<%BANNER%>

Plant Analysis and Soil Testing As Tools to Develop a Sulfur Recommendation for Fresh Market Tomato (Solanum lycopersicum L.)

Permanent Link: http://ufdc.ufl.edu/UFE0041550/00001

Material Information

Title: Plant Analysis and Soil Testing As Tools to Develop a Sulfur Recommendation for Fresh Market Tomato (Solanum lycopersicum L.)
Physical Description: 1 online resource (133 p.)
Language: english
Creator: ESMEL,CAMILLE ELIZABETH
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: ANALYSIS -- IRRIGATION -- METHODOLOGY -- SULFUR -- TOMATO
Horticultural Science -- Dissertations, Academic -- UF
Genre: Horticultural Science thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Sulfur is one of the essential elements for plant growth and development. Current trend in fertilizer formulations is to use high analysis fertilizer sources, which exclude sulfur (S). In the past, anthropologic and environmental sources have met plant?s S requirement, but implementation of clean air referenda have reduced atmospheric S. Therefore, studies were conducted to determine, if a S response in tomato exists, 1) differences in sources of S and rates of S with nitrogen (N) on tomato production 2) if irrigation regime and S rates could influence tomato production, 3) the appropriate soil testing method for S and sulfate-sulfur (SO4-S), and 4) if a relationship could be established between tomato yield, tissue S, and soil testing results. Two field studies and one laboratory experiment were conducted at the Gulf Coast Research and Education Center in Balm, Florida to ascertain these objectives. The first field study used ammonium nitrate ((NH4)2NO3), potassium sulfate (K2SO4), ammonium sulfate nitrate (NH4NO3 + (NH4)2SO4), ammonium sulfate ((NH4)2SO4) as fertilizer sources. Muriate of potash was used to balance total potassium amounts. Tomato yields increased with the inclusion of S into the fertilizer program for Fall 2006 season. Sulfur sources were found to be similar and different from the non-treated control and NH4NO3 treatment for Fall 2006 season. The second field study utilized a split-plot design with three irrigation regimes (5,406, 8,109, and 10,812 L?ha-1?d-1) and six S rates (0, 28, 56, 112, 168, and 224 kg S?ha-1). Irrigation regimes chosen for this study did not influence tomato production. Only early tomato harvest was influenced by S rate for both seasons. The rate of 28 kg S?ha-1 increased yield over the non-treated control and beyond this level early yield did not increase. For the laboratory study, soil and tissue samples were collected and extracted for S or SO4-S. A suitable soil test was not found to relate soil S or SO4-S to crop S on sandy soils. It is recommended when S deficiency is diagnosed based upon tissue testing. An application rate of 28 to 37 kg S?ha-1 should be used in tomato.
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 CAMILLE ELIZABETH ESMEL.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Santos, Bielinski M.
Local: Co-adviser: Simonne, Eric H.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0041550:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041550/00001

Material Information

Title: Plant Analysis and Soil Testing As Tools to Develop a Sulfur Recommendation for Fresh Market Tomato (Solanum lycopersicum L.)
Physical Description: 1 online resource (133 p.)
Language: english
Creator: ESMEL,CAMILLE ELIZABETH
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: ANALYSIS -- IRRIGATION -- METHODOLOGY -- SULFUR -- TOMATO
Horticultural Science -- Dissertations, Academic -- UF
Genre: Horticultural Science thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Sulfur is one of the essential elements for plant growth and development. Current trend in fertilizer formulations is to use high analysis fertilizer sources, which exclude sulfur (S). In the past, anthropologic and environmental sources have met plant?s S requirement, but implementation of clean air referenda have reduced atmospheric S. Therefore, studies were conducted to determine, if a S response in tomato exists, 1) differences in sources of S and rates of S with nitrogen (N) on tomato production 2) if irrigation regime and S rates could influence tomato production, 3) the appropriate soil testing method for S and sulfate-sulfur (SO4-S), and 4) if a relationship could be established between tomato yield, tissue S, and soil testing results. Two field studies and one laboratory experiment were conducted at the Gulf Coast Research and Education Center in Balm, Florida to ascertain these objectives. The first field study used ammonium nitrate ((NH4)2NO3), potassium sulfate (K2SO4), ammonium sulfate nitrate (NH4NO3 + (NH4)2SO4), ammonium sulfate ((NH4)2SO4) as fertilizer sources. Muriate of potash was used to balance total potassium amounts. Tomato yields increased with the inclusion of S into the fertilizer program for Fall 2006 season. Sulfur sources were found to be similar and different from the non-treated control and NH4NO3 treatment for Fall 2006 season. The second field study utilized a split-plot design with three irrigation regimes (5,406, 8,109, and 10,812 L?ha-1?d-1) and six S rates (0, 28, 56, 112, 168, and 224 kg S?ha-1). Irrigation regimes chosen for this study did not influence tomato production. Only early tomato harvest was influenced by S rate for both seasons. The rate of 28 kg S?ha-1 increased yield over the non-treated control and beyond this level early yield did not increase. For the laboratory study, soil and tissue samples were collected and extracted for S or SO4-S. A suitable soil test was not found to relate soil S or SO4-S to crop S on sandy soils. It is recommended when S deficiency is diagnosed based upon tissue testing. An application rate of 28 to 37 kg S?ha-1 should be used in tomato.
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 CAMILLE ELIZABETH ESMEL.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Santos, Bielinski M.
Local: Co-adviser: Simonne, Eric H.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0041550:00001


This item has the following downloads:


Full Text

PAGE 1

1 PLANT ANALYSIS AND S OIL TESTING AS TOOLS TO DEVELOP A SULFUR RECOMMENDATION FOR FRESH MARKET TOMAT O ( Solanum lycopersicum L ) By CAMILLE ELIZABETH ESMEL A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Camille Elizabeth Esmel

PAGE 3

3 To my family and friends for their unwavering support during this gritty and sometimes crazy process

PAGE 4

4 ACKNOWLEDGMENTS M any people assisted me during my graduate studies. I would like to express my appreciation to them in this dissertation. First, I wish to express my appreciation to the chair of my supervisory committee, Dr. Bielinski M. Sant os f or his support and suggestion in the pursuit of this research topic. I also recognize my co chair, Dr. Eric H. Simonne for his extraordinary patience and literary advice. I wish to express my deepest appreciation to both Dr. Joseph W. Noling and Dr. Jack E. Rechcigl. Dr. Noling for his support and sticking with me through the changes in research topics and Dr. Rechcigl for his suggestion of pursuing analytical methodologies comparison as part of my research. I would also like to thank Dr. James P. G ilreath for seeing potential in me For technical support and laboratory assistance at the Gulf Coast Research and Education Center in Wimauma, FL, I would like to thank Elizabeth Golden, Bu tch Bradley, Silvia Slamova, Victor Alifonso, Humberto Moratinos, Jose Moreno, Tim Davis, Gitta Shurberg and Myriam Siham. I wish to express my thanks to Dr. Gurpal Toor, and Dr. Amy Shober for their expert advice in soil science. Their assistance in com pleting the analytical methodology portion of this dissertation can not be understated. Finally, I would like to thank my family and friends, and especially my parents, George and Lydia Fortier, Theodore McAvoy, Aparna Gazula, Joshua Adkins, Jennifer Bonin a Noseworthy, Kristine Hoffman, and Elizabeth Thomas.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 13 Sulfur the Forgotten Essential Plant Nutrient ................................ .......................... 13 Why is Sulfur so Important? ................................ ................................ .................... 14 Difficulty of Predicting Crop S Nu trient Status Using Soil S Levels ......................... 15 2 SULFUR IN CROP PRODUCTION AND SOIL DETERMINATION ........................ 17 Introduction ................................ ................................ ................................ ............. 17 Timing and Placement ................................ ................................ ............................ 17 Source and Rate ................................ ................................ ................................ ..... 19 Fertilizer Sources and Rates ................................ ................................ ............ 19 Irrigation Water ................................ ................................ ................................ 21 Soil ................................ ................................ ................................ ................... 21 Soil Testing and Plant Analysis ................................ ................................ ............... 22 Analytical Instruments ................................ ................................ ...................... 22 Nutrient Extraction ................................ ................................ ............................ 27 Calibration of Soil Testin g ................................ ................................ ................. 33 Rationale and Justification ................................ ................................ ...................... 36 3 INFLUENCE OF NITROGEN AND SULFUR SOURCES AND APPLICATION RATES ON FRESH MARKET TOMATO ................................ ................................ 39 Introduction ................................ ................................ ................................ ............. 39 Materials and Methods ................................ ................................ ............................ 43 Results and Discussion ................................ ................................ ........................... 46 4 SULFUR FERTILIZATION RATES AND IRRIGATION PROGRAMS ON TOMATO GROWTH AND YIELD ................................ ................................ ........... 53 Introduction ................................ ................................ ................................ ............. 53 Materials and Methods ................................ ................................ ............................ 54 Results and Discussion ................................ ................................ ........................... 56

PAGE 6

6 5 COMPARISON OF ANALYTICAL METHODS FOR D ETERMINE S AND SO 4 S PRESENT IN FLORIDA SANDY SOILS ................................ ................................ 64 Introduction ................................ ................................ ................................ ............. 64 Materials and Methods ................................ ................................ ............................ 69 Results and Discussion ................................ ................................ ........................... 70 6 COMPARISON OF EXTRACTANTS FOR DETERMINING SULFUR IN SOIL AND PLANT TISSUE ................................ ................................ .............................. 84 Introduction ................................ ................................ ................................ ............. 84 Materials and Methods ................................ ................................ ............................ 90 Results and Discussion ................................ ................................ ........................... 95 7 CALIBRATION AND DIAGNOSTIC TOOLS OF SOIL AND TOMATO TISSUE SULFUR FOR PRODUCTION ON FLORIDA SANDY SOILS .............................. 105 Introduction ................................ ................................ ................................ ........... 105 Materials and Methods ................................ ................................ .......................... 107 Results and Discussion ................................ ................................ ......................... 111 8 SUMMMARY AND CONCLUSIONS ................................ ................................ ..... 120 LIST OF REFERENCES ................................ ................................ ............................. 122 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 133

PAGE 7

7 LIST OF TABLES Table page 2 1 S ulfur containing fertilizers, formulas, and percent S and companion nutrients for materials used in crop production. ................................ ................................ 37 2 2 Chemical reagents and concentrations for extracting available S from soil and references in which extractants were used. ................................ ................. 38 3 1 Nitrogen, K, and S sources, rates applied and placement used to create Education Center in Balm, FL. ................................ ................................ ............ 49 3 2 Influence of N and S sources and applied rates on total marketable yield of tomato for Fall 2006 and Spri ng 2007 at Gulf Coast Research and Education Center in Balm, FL. ................................ ................................ ............................. 50 3 3 Marketable yield quality as influenced by N and S fertilization sources and applied rates for Fall 2006 and Spring 2007 at Gulf Coast Research and Education Center Balm ................................ ................................ ...................... 51 3 4 Effects of N and S sources and applied rates on tomato foliar S concentration as percent S for Fall 2006 and Spring 2007 at Gulf Coast Research and Education Center in Balm, FL. ................................ ................................ ............ 52 4 1 Plant height and vigor ratings for tomatoes grown under differing irrigation regimes and applied S rates during Spring and Fall 2008 at Gulf Coast Research and Education Center in B alm, FL. ................................ .................... 59 4 2 Estimated chlorophyll content for tomatoes grown under differing irrigation regimes and applied S rates f or Spring and Fall 2008 at Gulf Coast Research and Education Center in Balm, FL. ................................ ................................ ..... 60 4 3 Mid season and end of season soi l pH and soil S content for soil samples collected from tomato production under differing irrigation regimes and applied S rates during Spring and Fall 2008 at Gulf Coast Research and Education Center in Balm, FL. ................................ ................................ ............ 61 4 4 Influence of irrigation regime and applied elemental S rates on tomato leaf S concentrations for seasonal pooled samples for Spring and Fall 2008 at Gulf Coa st Research and Education Center in Balm, FL. ................................ .......... 62 5 1 Advantages and disadvantages of selected analytical instruments for determining S or SO 4 S. ................................ ................................ ..................... 76

PAGE 8

8 5 2 Instrumental details for comparison between sulfur and sulfate determination for selected samples obtained at the Gulf Coast Research and Education Center in Balm, FL. ................................ ................................ ............................. 77 5 3 Salts and their solubility as potential interfering ions when determining SO 4 S by turbidmetric methods. ................................ ................................ .................... 78 5 4 Potential interfering ions for determining SO 4 S by turbidmetric methods, their reported thresholds in solution, and reference ................................ ............ 79

PAGE 9

9 LIST OF FIGURES Figure page 4 1 Influence of rates of applied elemental S on f irst harvest of fresh market tomato for Spring and Fall 2008 at Gulf Coast Research and Education Center in Balm, FL. ................................ ................................ ............................. 63 5 1 Comparison between dry combustion total S and 0.025 M KCl turbidimetric SO 4 S for selected soil samples from fields at the Gulf Coast Research and Education Center in Balm, FL ................................ ................................ ............. 80 5 2 Comparison between dry combustion procedure for total S and de ionized water turbidimetric SO 4 S for selected soil samples from the fields at the Gulf Coast Research and Education Center in Balm, FL. ................................ .......... 81 5 3 Comparison between dry combustion procedure for total S and 0.025 M KCl Inductively Coupled Plasma tot al S for selected soil samples obtained from the fields at the Gulf Coast Research and Education Center in Balm, FL .......... 82 5 4 Comparison and regression analysis of dry combustion procedure for total S and Inductively Coupled Plasma de ionized water total S for soil samples obtained from the fields at the Gulf Coast Research and Education Center in Balm, FL ................................ ................................ ................................ ............. 83 6 1 Diagram of Gulf Coast Research and Education Center buildings, field layout, and nomenclature system for identifying production fields ..................... 99 6 2 Comparison of Modified Blair SO 4 S with Standard Reference (H 2 O) SO 4 S for selected soil samples from the Gulf C oast Research and Education Center in Balm, FL. ................................ ................................ ........................... 100 6 3 Comparison of Mehlich 3 S with de ionized water S for selected soil samples from the Gulf Coast Research and Education Center in Balm, FL. ................... 101 6 4 Comparison of Mehlich 3 S with Modified Blair S for selected soil samples collected at the Gulf Coast Research and Education Center in Balm, FL. ........ 102 6 5 Comparison of total plant tissue S by wet oxidation and by dry combustion total S procedure for selected soil sa mples collected at the Gulf Coast Research and Education Center in Balm, FL. ................................ .................. 103 6 6 Distribution of the deviation betwee n total plant tissue S by wet oxidation and dry combustion total S for selected soil samples collected at the Gulf Coast Research and Education Center in Balm, FL ................................ ................... 104

PAGE 10

10 7 1 Relationship between dry combustion total plant S and tomato yield for selected tissue samples from the Gulf Coast Research and Education Center in Balm, FL ................................ ................................ ................................ ....... 113 7 2 Relationship between total plant S and tomato yield for selected tissue samples from the Gulf Coast Research and Education Center in Balm, FL. .... 114 7 3 Relationship between dry combustion total soil S and tomato yield for selected soil samples from the Gulf Coast Research and Education Center in Balm, FL. ................................ ................................ ................................ .......... 115 7 4 Relationship between Modified Blair (0.025 M KCl) soil S and tomato yi eld for selected soil samples from Gulf Coast Research and Education Center in Balm, FL. ................................ ................................ ................................ .......... 116 7 5 Relationship between de ionized water soil S and tomato yield for selected soil samples from Gulf Coast Research and Education Center in Balm, FL. .... 117 7 6 Relationship between Mehlich 3 soil S and tomato yield for selected soil samples from Gulf Coast Research and Education Center in Balm, FL. .......... 118 7 7 Possible main fractions of soil S and analytical methods for S determination. 119

PAGE 11

11 Abstract of Dissertation Presented to the Graduate School of the U niversity of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PLANT ANALYSIS AND S OIL TESTING AS TOOLS TO DEVELOP A SULFUR RECOMMENDATION FOR FRESH MARKET TOMAT O ( Solanum lycopersicum L ) By Camill e Elizabeth Esmel May 2011 Chair: Bielinski M Santos Cochair: Eric H. Simonne Major: Horticultural Sciences Sulfur is one of the essential elements for plant growth and development. Current trend in fertilizer formulations is to use high analysis ferti lizer sources, which exclude sulfur (S). requirement but i mplement ation of clean air referenda have reduced atmospheric S. Therefore, s tud ies were conducted to determine if a S res ponse in tomato exists 1 ) differences in sources of S and rates of S with nitrogen (N) on tomato production 2 ) if irrigation regime and S rates coul d influence tomato production, 3 ) the appropriate soil testing method for S and sulfate sulfur (SO 4 S), and 4 ) if a relationship could be established between tomato yield, tissue S, and soil testing results Two field studies and one laboratory experiment were conducted at the Gulf Coast Research and Education Center in Balm, Florida to ascertain these objecti ves. The first field study used ammonium nitrate ((NH 4 ) 2 NO 3 ), potassium sulfate (K 2 SO 4 ), ammonium sulfate nitrate ( NH 4 NO 3 + (NH 4 ) 2 SO 4 ) ammonium sulfate ( (NH 4 ) 2 SO 4 ) as fertilizer sources. Muriate of potash was used to balan ce total

PAGE 12

12 potassium amounts T omato yields increased with the inclusion of S into the fertilizer program for Fall 2006 season Sulfur sources were found to be similar and different from the non treated control and NH 4 NO 3 treatment for Fall 2006 season T he second field study utilized a split plot design with three irrigation regime s ( 5,406, 8,109, and 10,812 Lha 1 d 1 ) and six S rate s ( 0, 28, 56, 112, 168, and 224 kg Sha 1 ) Irrigation regimes chosen for this study did not influence tomato production. Only e arly tomato harvest was influenced by S rate for both seasons T he rate of 28 kg S ha 1 increased yield over the non treated control and beyond this level early yield did not increase For the laboratory study, s oil and tissue samples were collected and extracted for S or S O 4 S A s uitable soil test w a s not found to relate soil S or SO 4 S to crop S on sandy soils. I t is recommended when S deficiency is diagnosed based upon tissue testing. An application rate of 28 to 37 kg S ha 1 should be used in tomato

PAGE 13

13 CHAPTER 1 INTRODUC TION Sulfur the Forgotten Essential Plant N utrient Sulfur (S) is the 16th with concentrations ranging from 0.06% to 0.10% (Ceccotti et al., 1998; Havlin et al., 2005), is essential for plant production of S conta ining amino acids cysteine, and methionine (Leustek, 2002). Although animals are unable to reduce sulfate ( SO 4 ) they rely upon t heir diet to attain S through S containing proteins. Therefore, if deficiencies exist in forage, grain, and vegetable crops t his reduced level of S then carrie s over to both animals and humans Deficiencies have begun to be widely reported in various crops (such as Brassica spp.) and crop by products (e.g. flour) (Duke and Reisenauer, 1986; Hagel, 2005 ; Sawyer and Baker, 2002 ). It has been suggested that deficiencies are the result of decreased S being put into the environme nt and the increased use of low S fertilizers. Clean air referenda across the world have caused reduction on world sulfur dioxide ( SO 2 ) emissions with est imates ranging from 55.2 to 68 Tgyr 1 S (De Kok et al., 2007; Smith et al., 2001; Stern, 2005). It is predicted that future plant available S from atmospheric depositions will be less than 10 kg S ha 1 per yr (Dgmmgen et al., 1998). Sulfur deficiency h as become recognized as a limiting factor for crop production (Schere, 2001) and more apparent where increased agricultural intensification of arable land and the use of low S fertilizer has occurred (Ceccotti et al., 1998). Low S fertilizers, commonly re ferred to as h igh analysis fertilizers, are a common fertilizer choice in vegetable production.

PAGE 14

14 Why is Sulfur so Important? Unlike animals that have a dietary requirement for S amino acids, plants are able to assimilate inorganic sulfate S ( SO 4 S ) which is reduced to sulfide (H 2 S) and is then incorporated into cysteine (Leu stek et al., 2000). A s humans w e rely directly or indirectly upon plants for S based amino acids for our diet. Sulfur is the least abundant of the macro elements found in plants, with an approximate concentration of 1,000 mg S kg 1 of dry matter (Leustek et al., 2000). Additionally, the S cycle is a biol ogical and chemically mediated cycle. T he majority of soil S is in organic S f orm and not readily available for plant uptake. Thus, th e process of mineralization is an important consideration when choosing fertilizer sources (i.e. elemental S versus SO 4 salts). In addition to these issues, SO 4 S, the form plants uptake from the soil, is an anion. As an anion, SO 4 S can be leached from the soil profile and thi s is influenced by the presence of other anions (Eriksen and Askegaard, 2000). Adsorption refers to the ability of an object to attract and hold particles to its surface. All soils especially clays and organic matter, have the ab ility to adsorb water, nutrients, and other chemicals. The order of adsorption of strength of anions in soils is as follows: phosphate > sulfate > nitrate > chloride, where as phosphates are the primary anion adsorbed (Havlin et al., 2005). Eriksen and Askegaard (2000) found an average of 20 kg S ha 1 was leached from a crop rotation over three years in sandy soils This amounted to approximately 60% of the total S input into the system. I t is possible ; because of the sandy soils dominate in Florida, to leach SO 4 S out of the s oil profile a nd into waterways H o wever, there is a lack of scientific information support ing t his hypothesis The potential of SO 4 S to leach may cause future pollution problems as S under anoxic conditions (aquatic systems) ha s been suggested to enhance the methylation of

PAGE 15

15 mercury (Ullrich et al., 2001; Ba tes et al., 2002). When Hg has been methylated, it may be bio a ccumulated by fish and stored in lipid tissues (Ullirch et al., 2001). Mercury accumulation in fish poses a hea lth risk for humans as well as other invertebrates when consumed. In Florida, it has been re ported that P and S run off from the E verglades Agricultural Area (EAA) have the potential to shift native vegetation population species as well as the microbial p opulation in the Everglades (Brix et al., 2010; Castro et al., 2005; Li et al., 2009). Therefore, the effects of excess nutrients in balance with the knowledge of proper fertilizer management strategies should be developed for S in Florida to maximize cro p growth and yield while maintain environmental quality. Difficulty of Predicting Crop S Nutrient Status Using Soil S Levels To develop fertilizer recommendations for any crop a robust relationship between crop response and soil testing needs to be estab lished. This association is the back bone of interpretation of a soil test, the requirement of an established relationship between soil, plant tissue content and relative yield (Bloem et al., 2001). Crop response s to S and SO 4 S applicat ions have been e stablished world wide (Scherer, 2001; Tabatabai, 1984). In Florida Susila (2001 ) found that timing and form of S did not influence tomato ( Solanum lycopersicum L.), but increased yields occurred with the inclusion of S into the fertilizer program It is possible given that a response to S has been found for tomato to establish a relationship between crop response and soil S. Currently, a relationship between crop response and soil testing has not been conclusively established, because of the factors infl uencing the soil S cycle. Seasonal fluctuation of inorganic S concentration in the soil solution occurs because of changes in the balance between microbial activity, soil drying, leaching, surface runoff, atmospheric inputs, plant uptake, plant senescence and fertilizer inputs (Blair et al., 1993; Williams,

PAGE 16

16 1968). These factors add to the difficulty of determining a relationship between soil S, tissue S and plant yield. Furthermore, S and SO 4 S soil testing lacks a universal reference method making it apart of routine soil testing, and many extracting solutions have been employed for testing soil S and SO 4 S (Lewis, 1999; Lisle et al., 1994; Matula, 1999; Rao and Sharma, 1997). A wide range of soil analytical methods will still be employed for S and SO 4 S determination until an established relationship between soi l S and crop yield or S content

PAGE 17

17 CHAPTER 2 SULFUR IN CROP PRODU CTION AND SOIL DETER MINATION Introduction I f S deficiencies are diagnosed for vegetable crops grown in Florida, it is recommende d to apply 28 to 37 kg S ha 1 (Simonne and Hochmuth, 2009). Yet, this recommendation for S in crop production has often been empirical in nature than based upon previous research. And positive yield and growth results have been found when S has been appl ied to vegetable crops (Pavlista, 2005; Rhoads and Olson, 2000; Susila, 2001). The objectives of the of this review are to review (1) existing S fertilizer research on vegetable crops for developing S fertilizer recommendations and (2) review the methodol ogy currently used to measure S fractions in plant and soil samples. A fertilizer recommendation for any nutrient consists of four fundamental components: timing, placement, rate, and source of the fertilizer. Timing refers to at what point in the crop g rowth cycle the fertilizer is applied (i.e. pre plant or at planting). Placement refers to where the fertilizer is placed (i.e. banded or broadcasted). Rate refers to the amount of nutrient applied per unit of land. Source refers to the particular fert ilizer salt used to supply the nutrient to the crop. Each component is essential when developing a fertilizer recommendation. Timing and Placement Timing of S application depends largely on whe ther the S fertilizer is a sulf ate form that is immediately us able or an elemental S that must be first oxidized before the plant can utilize it, and also the formulation (solution, suspension, granules, or prills) (Malhi et al., 2005). In the case of elemental S as granules or prills, soil moisture can be the domin ant factor in dispersion/oxidation. Soil temperature also factors into the

PAGE 18

18 dispersion/oxidation of elemental S via soil microbial action. However, Nuttall et al. (1990) found that external force was required to dispersion of elemental S prills. Elementa l S fertilizers applied annually for several years are expected to produce a cumulative residual effect on plant available S in soil and subsequently on seed yield and quality of canola ( Brassica napus L. ) (Malhi et al., 2005). Further research by Malhi e t al. (2009) found that timing did not influence wheat ( Triticum aestivum L. ) yield parameters in a canola application did improved wheat seed yield into 2006 and 2007 seasons These authors su ggested that fall applications of sulfur in this system were superior to spring applications. In contrast, Solberg et al. (2003) found that early broadcast applications of S fertilizers resulted in an increased recovery by soil testing for SO 4 S from fall ow fields. Very little information exists on only timin g of S fertilizer applications for vegetable production. Since in vegetable production often the timing of fertilizer applications is directly linked to the p lacement (i.e. pre plant broadcast, pre plant banded, season long drip injections of fertilizer ). It is difficult to separate the two topics and often they are discussed together as method of application. Sus il a (2001) fo und that methods of S application (pre plant, split application, and drip injection) did not influence yield or growth parameters of tomato, but did positively influence leaf S concentrations 6 weeks after transplanting. Drip injection of S had the highest level of leaf S content at 6 weeks after field planting. Other studies by Susila (2001) revealed that pepper ( Capsicum annuum L.) was not influenced by the method of S fertilizer application. Yet, studies conducted on cabbage ( Brassica olera cea L.) by Susila (2001) found method of S

PAGE 19

19 fertilizer application as a drip injectio n increased cabbage yield. This is not a surprising result for cabbage, since this vegetable crop has a high S requirement (Scherer, 2001). Source and Rate Crop plants fulfill their S requirement from a num ber of sources; including crop res idues, manures, rainfall, the atmosphere, soil amendments, fertilizers irrigation water, and soil (Tabatabai, 1984). In vegetable cropping systems, the number one source of nutrients is fertilizer, but often environmental sources are not accounted (i.e. irrigation wate r and soil levels ). The goal of this section is to discuss not only fertilizer sourc es and rates, but also irrigation water and soil levels of S available for plant uptake. Fertilizer Sources and Rates Elemental S is the least expensive S fertilizer sourc e when compared to other so urces of S that contain N ( (NH 4 ) 2 SO 4 ) ammonium thiosulfate (NH 4 ) 2 S 2 O 3 ) and NH 4 NO 3 + (NH 4 ) 2 SO 4 ) Elemental S can be an ideal S source for crop growth and yield because it is 1) high analysis of S and insoluble in water; 2) stab le in damp and humid conditions 3) has to be oxidized to SO 4 S for plant uptake (Solberg et al., 2003). The slow release of elemental S via oxidation to SO 4 would be considered a disadvantage in certain cropping systems, such as annual vegetable productio n. Information on elemental S as a S source in vegetable production is needed Other sources of S that contain N or K have been studied Susila (2001) found tomato yield respons e to S sources ((NH 4 ) 2 SO 4 versus (NH 4 ) 2 S 2 O 3 ) was similar, regardless of applic ation method (broadcast versus drip injected). Sulfur sources ( (NH 4 ) 2 SO 4 NH 4 NO 3 K 2 SO 4 CaSO 4 7H 2 O K 2 SO 4 2MgSO 4 ) have been found not to influence cabbage yield (Rhoads and Olson, 2000). Heeb et al. (2006) found that the

PAGE 20

20 inclusion of S into the fertilize r regime for greenhouse grown tomato increased yield and growth than when compared to fertilizer regi me without S This further supports Rhoads and Olson (2000), in which cabbage yield did not respond to treat ments where S was not included. Fertilizer appl ication rates are an important factor in creating a fertilizer recommendation in any cropping system and should be on the basis of available soil S and cro p nutrient requirements (Scherer, 2001). Crop nutrient requirement is defined as the total amount (k g ha 1 ) of an element required by the crop to produce optimal yield (Simonne and Hochmuth, 2009). Crop nutrient requirements are determined by studies conducted with the inclusion of nutrient into the fertilizer program with different increasing rates of the nutrient on land in which there is no history of S applications or S present Limited information is available on vegetable crop S requirements but some studies have explored S rates and the inclusion of S into the fertilizer program S uslia (2001) f ound linear increases in tomato yield when S fertilizer rates increased from 0 to 102 kg S ha 1 but no one rate was recommended for tomato production. Susila (2001) also found that 34 kg S ha 1 was the critical level for S applications on cabbage. After this rate, cabbage yield (in heads) was reduced. These results are in agreement with Rhoads a nd Olson (2001) in which 45 kg S ha 1 was found to be excessive and consistent higher yields resulted fro m 22 kg S ha 1 While fertilizer may be the primary sou rce of nutrients for crops, other sources of nutrients must be explored and accounted for when creating a fertilizer recommendation. Irrigation water and soil may also be sources of S in crop production.

PAGE 21

21 Irrigation Water Field sites without access to mine ral bound S sources may have insufficient S dissolved in soil water and shallow ground water bodies to move available S to plant roots (Eriksen et al. 1998). The concentration of S in irrigation waters used throughout North American is unknown, and little research is available related to the S content of irrigation waters to S requirement in a fertilizer program (Olson and Rehm, 1986). Studies on high sulfate irrigation waters utilized water with 175, 646, 862, and 1 743 mg SO 4 L 1 (Drost et al., 1997) an d 15.0 and 38.8 mol SO 4 L 1 (Papadopoulos, 1986) have found mixed results in improving yield of broccoli and tomato. In Florida, SO 4 levels in drinking water wells can range from 0.2 to 1400 mg SO 4 L 1 with higher concentrations coming from deeper wells ( Sacks, 1996 ). While these values cover drinking water, the levels of SO 4 or S within irrigation water in Florida are unknown It has been suggested that water pumped from the aquifer as well as surface waters, has a rotten egg smel l commonly associated w ith sulfide (H 2 S) levels This may often leads to the belief that S in the irrigation water is an S source for crop production. Little scientific information exists to support this claim. Soil Most Florida surface soils are low in plant available S, and additional sources are necessary to maintain optimum production despite the lack of actual cases of S deficiencies among crops (Mitchell and Blue, 1981 ). These authors also suggested that 4 kg S ha 1 as SO 4 derived from SO 2 from summer rainfall based upon calculations and estimations by Brezonik et al. (1980) would remain in the surface soil for plant uptake. They also suggested that 0.8 kg S ha 1 of the initial 4.8 kg S ha 1 estimate wo uld be leached or immobilized.

PAGE 22

22 Mitchell and Blue (1981 ) stated that S podosols and Entisols two of the major soil orders in Florida, lack large reserves of available S in the subsurface horizons Ultisols another predominate soil order in Florida, may provide S via the argillic horizon in the subsurface Florida sandy so ils (Spodosols and Entisols) were found to not have enough available S reserves for tomato crop growth (Susila, 2001). Both surface and subsoil horizons were inefficient in supplying enough S for six weeks of growth without the addition of pre plant appli cation of S as (NH 4 ) 2 SO 4 Although, this study utilized an Arredondo fine sand soil (loamy, siliceous, hyperther imic, Grossarenic Paleudult), a n Ultisol, growth of tomato on this soil required S applications. These results are in conflict with those of M itchell and Blue (1981 ) stating that Ultisols may have enough reserve S in the subsoil for plant growth. It may be possible that crops with large deep root systems would benefit from reserve S in the subsoil. Vegetable crops, such as tomato and cabbage, often have shallow root systems in compared to row crops (i.e corn ( Zea m ays L.) and alfalfa ( Medicago sativa L.) ). Soil Testing and Plant Analysis Analytical Instruments Current assessment of S and SO 4 S in agricultural soils for the development of fertil izer recommendations has been fraught with difficulties. The major difficulties that researchers confront are inherent variability in analytical results, lack of consistent associations between soil S, plant S, and crop yield, and S in soils and plant tis sues is often a separate method adding time and costs in an analytical laboratory. Ideal characteristics for S determination would be low cost, rapid, accurate, precise, with low interferen ce, and potentially offer multiple element determination opti ons Soil chemical analysis is the cheapest and most widely used

PAGE 23

23 chemical analysis employed in agriculture (Van Raij et al., 2002) and is often under utilized by many (Jones, 1993). It is primarily the cost of labor to operate analytical inst rument that determines the underl ying cost of each soil sample. In terms of speed, ma n y instruments (i.e. elemental analyzer inductively coupled plasma atomic emission spectrometer or inductively coupled plasma optical emission spectrometer (ICP AES/ICP O ES), ion exchange chromatograph (IC), and discrete analyzer) have been automated to allow sample loading time to be reduced. This can essentially cut labor and time for the laborat ory worker spent supervising instrument status. In the past, S and SO 4 S d etermination were labor intensive and required reagents have a finite shelf life for use. Such example is methylene blue, which is the color reagent for estimating the reduction of sulfate to sulfide (Tabatabai, 1996). The color of this reagent is import ant to the determination of S by spectrometric means; if kept away from direct sunlight the color can remain stable for 24 h or more (Tabatabai, 1996). This would be considered to be a time limitation or shelf life restriction for this method. Not proper ly understanding and identifying limitations in a method can lead to data quality issues. Quality control and assurance plans are implemented to maintain result quality. Quality control consists of using the following: proper calibration, blanks, duplica tion of samples, spiked samples and the use of certified standards ( Shampine, 1993 ). Quality assurance usually c onsists of validation of methods (characteristics and limits). Both of these are extremely important to S determination, because of the inhere nt variability in determining S and SO 4 S. It has been reported by Crosland et al. (2001) that the coefficient of variation from an inter laboratory study on soil S was between 36 to 45%.

PAGE 24

24 This means up to 45% of the variation in soil S values was due to random error and the decision to apply fertilizer based upon this data could be faulty. This random error can be attributed to operator error, loss of calibration during analysis or interferences from oth er elements within the aliquot. Interferences from other elem ents or ions are the most encountered error when determining S or SO 4 S. With the advent of newer technologies, such as ICP OES, claims that less interference would be encountered were publicized (Kola et al., 2002). Multi element determination s are limited to those instruments with the ability to determine multiple elements. These include, but are not limited to the following: elemental analyzers, ICP OES/ICP AES, and IC. Instruments with multiple element determination capabilities are viewe d as a versatile economic investment. These types of instruments reduce labor, because more than one element can be an alyzed within a single sample. Currently seven analytical instruments are capable of determining total S or SO 4 S. These instruments are ICP AES or ICP OES (S), inductively coupled plasma mass spectrophotometer (ICP MS) (S), elemental analyzer (S) by infrared radiation cell/spectrometer (IR) or titration of SO 2 IC (SO 4 S), and turbidimeter (SO 4 S). Of these six instruments, the three mos t commonly used to determine S or SO 4 S are ICP AES/ICP OES, elemental analyzer via IR or titration, and colorimetrically with a spectrometer (turbidimeter). The principle behind ICP AES/ICP OES is based in chemistry and physics. More specifically, it inv olves particle collision, excitement of out shell electrons, adsorption/emission of a wavelength, and separation/detection of sp ectral lines of

PAGE 25

25 interest. A comprehensive discussion of the theory will not be provided here, but several reviews and sources e xplain in detail such theory and mechanics of ICP AES/ICP OE S (Boss and Fredeen, 1997; Laju ne n and Permki, 2004; Manning and Grow, 1997; Welz et al., 2009). Several advantages for determining S with ICP AES/ICP OES are after initial costs, it is cost ef fective, rapid sample analysis, multiple element, low level of interferences, wide linear dynamic range and detection limits at the 1 level (Kola et al., 2002 Mermet and Poussel, 1995). One disadvantaged for determining S with ICP AES/ICP OES is tha t a vacuum line is required within the instrument as S is considered a non metal element. This is required, because emission lines for S lie in the ultraviolet range so that the lines are not absorbed by O 2 (Soltanpour et al., 1996). A range of emissions are used for S on ICP AES/ICP OES. In general, the range of wavelengths used is from 180.04 to 182.63 nm, and the three strongest emission lines of S in the vacuum ultraviolet spectrum are at 180.73, 182.04, and 182.63 nm (Wu et al., 2009). Two signific ant spectral overlaps occur around the 180.73 nm S line; 180.361 nm Ca and 180.74 nm Mg lines (Eames and Cosstick, 1992). Therefore, corrections or adjustments must be made to ensure interfering spectral lines are not distorting results for S. The princip le behind an elemental analyzer, often called by brand name LECO or by the method name Dumas, is titration via colorimetric determination or infrared radiation (IR) determination of total S. The D umas method is the combustion of samples a high temperature s (<1000 C) and the evolution of S as SO 2 and subsequent capture for determination. Earlier Dumas models use a photocell with SO 2 collected by HCl containing KI, starch, and a trace amount of KIO 3 approximately 0.22 M KIO 3

PAGE 26

26 (Tabatabai, 1996; Tiedemann and Anderson, 1971) or 1N NaOH (Tabatabai, 1996) with colorimetrically determining total S by methylene blue similar to Johnson and Nishita (1952). For IR determination of S, after combustion, a stream of gas is directed into a ballast tank, homogenized befo re a sample is trapped for infrared detection of S (Kowalenko, 2001). The infrared source consists of nichrome wire, which is resistance heated to 850 C, SO 2 absorbs infrared energy at 3.98 and 7. with a band pass filter to prev ent further infrared energy reaching the IR detector (LECO, 2006). The advantages of an elemental analyzer are that it is designed for simpli city, speed, convenience, carbon (C), N can also be determined by some models (Tabatabai, 1996), and capable of mea suring total S in soils including those with high organic matter content is possible unlike with wet digestion methods such as NaBrO (Kowalenko, 2001). Some disadvantages of elemental analyzers are the following: initial cost of the instrument, the built in flow conditions for S restrict the scope for optimizing the combustion cy cle for organic and total C dete rminations when analyzing C and S simultaneously (Wright and Bailey, 2001), and without an automated sample loader this process consumes time and l abor to load each sample individually. The principle behind SO 4 S determination by turbidity is the Beer which a reduction in the intensity of light passing through a solution containing suspended particles of barium sulfate (BaSO 4 ) (Beato n et al., 1968). The advantages of determining SO 4 S by turbidimetry are less time consuming, in some cases sensitive, and the most commonly used method for determining SO 4 S. Therefore many modifications are available to adjust for its short comings inc luding exchange resins,

PAGE 27

27 addition of charcoal filters, and flow injection analysis (Morais et al., 2006). The disadvantages to turbidimetric measurement of SO 4 S are insensitivity at low concentrations (less than 10 mgL 1 ), multiple cation interferences ( Na, K, Ca, Mg, NO 3 PO 4 and SiO 2 ), cati on co precipitations with Ba high organic matter co precipitation, and difficulty of reproducible BaSO 4 suspensions under uniform precipitating conditions (Beaton et al., 1968; Tabatabai, 199 6). Nutrient Extraction Assessing relationships between a newly proposed soil or plant extractant and a standard extractant is a useful first step technique in the evaluation a new extractant for the development of a soil testing program (Van Erp and Van Beusichem, 1998). This u seful first step technique seems to be lacking for soil S and SO 4 extractants in many soil testing programs that have been developed. Many studies have been conducted upon comparing determination methods of S (Ajwa and Tabatabai, 1993; Crosland et al. 200 1; Palomino et al., 2005; Reisman et al., 2007; Soon et al., 1996), but only few separate extraction solution comparison from the determination step. This may be a mistake made by many investigators trying to determine optimal soil extraction solutions fo r the development of a soil testing program. Soil tests for N and S, as well as for the micronutrients, are either limited in their applications to a narrow range of soil crop conditions or are nonexistent (Jones, 1985). The application of a soil test for S may be narrow in range, because soil S is in constant flux due to environmental influences via microbial activity on soil S (Spencer and Glendinning, 1980). Environmental influences such as soil disturbance, rainfall, temperature, and organic matter in put could all cause fluxes in soil S. Therefore, land history would be an important factor in the amount of soil S available to plants.

PAGE 28

28 Some of the most widely used solutions to extract soil S to predict plant availability and fertilizer applications ar e water (H 2 O) calcium chloride (CaCl 2 ) and phosphate based solutions (Anderson et al., 1992; Tan et al., 1994a). Water extractable S is primarily soil S readily leached out of the soil and is mostly likely in the inorganic SO 4 S form. Several solutions are often used and labeled when they are truly slightly saline solutions These solutions include, but are not limited to sodium (Na) calcium (Ca) or lithium chloride (LiCl) sol utions (Palomino et al., 2005). T hese solutions ex tract the soluble fraction of sulfate in soils (Sant oso et al., 1995). The calculated difference be tween water extractable S and a phosphate solution is related to the amount of sulfate adsorbed by the soil, but this does not always hold true (Santoso e t al., 1995). Often and depending upon soil organic matter and exchange sites, water extractable S as H 2 O can significantly dissolute more SO 4 S than salt solutions (Maynard et al., 1987; Santoso et al., 1995; Tan et al., 1994a ) or in some cases similar amounts (Anderson et al., 1998; Tan et al., 1994b; Zhao and McGrath, 1994). Many studies employ monocalcium phosphate (Ca(H 2 PO 4 ) 2 ) extraction solution at 0.01 M or 500 mgL 1 P This is one of the recommended soil SO 4 S extraction solutions for northeaste rn US (Singh et al., 1995) and extracts more soil SO 4 S than water via dissolution of SO 4 S off exchange sites on soil particles. Yet, this solution can cause interferences when using turbidimetric measurement of SO 4 S when 0.04 M solution is used for som e soils (Searle, 1998). The other issue with Ca based extraction solutions is that Ca may cause spectral interference of S when using ICP AES (Zhao and McGrath, 1994). In addition, various studies using Ca(H 2 PO 4 ) 2 have had variable

PAGE 29

29 results for extracted S correlating significantly to other parameters ( Anderson et al., 1998a; Anderson et al., 1998b; Blair et al., 1991; Blair et al., 1993; Chinoim et al., 1997; Palomino et al., 2005; Pandey and Girish, 2006; Spencer and Gle ndinning, 1980; Ven drell et al., 1 990 ). Blair et al. (1991) developed a soil extraction solution of 0.25 M potassium chloride ( KCl ) as a potential replacement to Ca(H 2 PO 4 ) 2 extraction solution as a response to poor performance of Ca(H 2 PO 4 ) 2 as a soil extractant. Potassium chloride was ori ginally investigated by Gianello and Bremner (1986) to measure inorganic N and potential mineralized N. Yet, concerns raised for this potential replacement have been centered on the parameters involved in the methodology. This method has a long incubatio n time (3h ) and requires the maintenance of 40 C for this time period while shaking. M odifications of this procedure have been made to improve time and detection limits. Bloem et al. (2002) examined six modifications of the 0.25M KCl extracting solution for SO 4 S. This study found that using 0.025 M KCl, 1 to 5 ratio of soil to extracting solution, shaking for 3 hr at room temperature was the fastest, precise method for SO 4 S extraction of agricultural soils (Bloem et al., 2002). While this optimizes s oil testing for SO 4 S alone, another possibility is the inclusion of soil SO 4 and S determination in extraction solutions already accepted by laboratories. An example is the Mehlich 3 extracting solution. The Mehlich 3 extraction solution excludes H 2 SO 4 which is a component of the Mehlich 1 extraction solution, therefore makes it ideal chemically for the determination of S. Mehlich 3 extractant was designed to replace the Mehlich 1 and Bray 1 P extractant (Jones, 1998) and is suitable for use on neutral to acidic soils. The benefit of

PAGE 30

30 using an extracting solution such as Mehlich 3 is that it allows for S to be determined as part of a multi element (including micronutrients) extraction step. This saves time, labor, and money, because separate chemicals a nd analysis do not need to be performed as has been described in a diagram by Jones et al. (1991). Few studies have examined Mehlich 3 for S extraction in soils, because the general belief that S is analyzed separately from all other nutrients. This is e ither due to the determination of SO 4 S by turibidimetry or ion chromatography in stead of using ICP technology. Rao and Sharma (1997) found strong relationships between Mehlich 3 and other extracting solutions commonly used to determine SO 4 S in soils. Th is was the first reported effort of using Mehlich 3 as an extractant for S in soils. Zbral (1999) found Mehlich 3 extraction correlated well to water extraction of SO 4 S. Also Zbral and elem ent determination in Czech agricultural soils. Further support of Mehlich 3 being used to include S in its multiple element extraction was by Pandey and Girish (2006). Yet, this study did not compare ability of each extracting solution to dissolute SO 4 S or S from the soil samples. Instead, the study focused upon correlation between extractants chosen and various plant gro wth and total S within tissue. Total S in plant tissues are most commonly extracted by two methods: dry combustion (also known as dry ash procedure) or wet oxidation (also known as acid digestion). While each has their own advantages and disadvantages, the focus here is on which ones commonly used in S research and suitability for S determination. For dry combustion, two methods are o ften used. The first is the Association of Analytical Communities (AOAC) Official Methods of Analysis for sulfur in plants in which organic

PAGE 31

31 matter can be destroyed by two diff erent chemicals, sodium peroxide (Na 2 O 2 ) or magnesium nitrate (MgNO 3 ) ( AOAC OM 9 20.10 or AOAC OM 923.01 ) ( AOAC, 2000 ). The AOAC Official Methods of Analysis are considered to be the reference methods in the analytical chemistry domain, but often these methods are not used by research scientists in agricultural fields. This is primar ily due to the chemical hazards, time, and labor involved in AOAC Official Methods of Analysis. The other dry combustion technique for total S in plant tissues is an automated procedure commonly termed LECO This procedure has two steps within on e piece of equipment: organic matter destruction and total S determination step. The LECO procedure combusts samples at temperatures up to 1400 C and measures S as SO 2 evolved by titration or infra red cell. In some cases, an accelerator or catalyst is added t o the sample such as tungsten trioxide (WO 3 ) or vanadium pent oxide (V 2 O 5 ) Kowalenko (2001) suggested that tungsten trioxide be added when simultaneously measuring C, N, and S with a LECO automated dry combustion instrument. Whether or not a catalyst was used (Etheridge et al., 1998) or one case the type of catalysis used (David et al., 1989) have not been reported when LECO automated dry combustion instrument is used for total S. While this may seem just a minor error, the lack of this information creat es difficulties in repeating the analytical research. In contrast the automated dry ash procedure, wet oxidation for the extraction of S from pl ant tissues is time consuming. The most common wet oxidation procedure for total S is nitric acid with perchlo ric acid (HClO 4 ) The number one difficulty with acid mixture for wet oxidation is the t endency of hot HClO 4 in the presence of easily oxidizable organic matter to be explosive (Jones, 1985). This in conjunction with the need for a special hood when

PAGE 32

32 usin g HClO 4 rules it out for many laboratories when choosing a wet oxidation method. Therefore, it is not surprising that many studies have been conducted on comparing acid mixtures which include or exclude HClO 4 for the suitability as a matrix for metals and non metals (Havlin and Soltanpour, 1980; Hoenig et al., 1998; Huang et al., 2004; Ippolito and Barbarick, 2000; Oliva et al., 2003; Pyki et al., 2000; Soon et al., 1996). Two studies in particular (Pyki et al ., 2000; Soon et al., 1996) compared metho ds for t he determination of total S in plant tissues. Pyki et al ., (2000) compared five different wet oxidation methods and dry combustion for total S and six other elements (Al, Ca, K, Mg, Mn, and Zn) against the certified value. This study found that out of all wet oxidation methods nitric acid ( HNO 3 ) + hydrogen peroxide ( H 2 O 2 ) or HNO 3 alone were the most suitable for S determination using ICP. Also the HNO 3 + HClO 4 digestion procedure in almost every case gave the lowest results compared to other ac id digestion procedures (Pyki et al., 2000). The authors also stated that the combustion technique combined with infrared detection (LECO ) also gave relative good reproducibility for sulfur determination. In contrast, Soon et al. (1996) compared six to tal S methods with reference values. These methods were dry ashing with sodium bi carbonate ( NaHCO 3 ) + silver oxide ( Ag 2 O ) oxygen flask combustion with determination on two different an alytical equipment, LECO microwave digestion with HNO 3 + HClO 4 and m icrowave digestion with HNO 3 + H 2 O 2 + hydrochloric acid ( HCl ) B oth dry ashing with NaHCO 3 + Ag 2 O, and oxygen flask combustion can be time consuming for regular laboratory use when co mpared to microwave digestion. In addition, precipitation of AgCl 2 may occur in the case of dry ashing a s well as poor recovery of S from certain plant tissues Microwave digestion, while rapid, has a high

PAGE 33

33 initial cost with the investment of the equipment and may not be in used in commercial laboratories. Despite these conc erns, Soon et al (1996) suggested that HNO 3 + H 2 O 2 + HCl was suitable for multi element analysis using microwave digestion in conjunction with ICP spectrometry and that LECO results for total S in plant tissues were similar to this method. Calibration o f Soil Testing The results of soil tests can be used to diagnose and guide the correction of nutrient deficiencies, to diagnose and avoid nutrient and non nutrient toxicities, to assess the need for soil amendments and to monitor the effects through time o f fertilizer rates and other management practices (Helyar and Price, 1999). Interpretation of soil test results requires a correlation between yield and soil test values. Typically, the interpretation has been established for a crop, soil type, and area. In the case of S and tomato, this relationship has not been established. While i t is important to note that seldom does interpretation become established based upon a single carefully designed experiment; rather it requires hundreds of fertility trials and resulting in thousands of plant analyses (Melsted et al., 1969 ). An initial study must be conducted to establish a nutrient calibration curve. Nutrient calibration based upon mineral concentration in dry matter versus plant growth or yield has been de scribed by Smith (1962). This nutrient calibration curve is the basis of plant nutrition. It is a relationship often attempted to be established for many nutrients for interpretation of nutrient research results. Both total and relative yield can be used but it is not recommended to use relative yield as a variable Using relative yield can lead to erroneous conclusions due to the standardization of yield over several locations or seasons Different locations and multiple seasons often have variability in

PAGE 34

34 practices and environmental factors making standardization (i.e. dividing all yields by one yield from a certain location or an average across all) misleading. Standardization does not accurately reflect the variability in yield response. It has been suggested by Jackson (2000) that thorough evaluations of SO 4 S extraction procedures, along with more S response data are needed to create an S fertilization recommendation. Strong relationships between soil S and plant S are difficult to achieve and are the current focus of many S researchers throughout the world. Only calibration of soil test values against yield, S concentrations in plants or S uptake by plants make them suitable for judgment about the S nutritional status of a site (Schnug and Hanekla us, 1998). Since an official method of analysis is lacking (due to variability) for soil S, a standard reference method for determining SO 4 S or total S in soil is used. Examining the literature about SO 4 S or total S determination in soil resulted in m any studies recommending different soil extraction and determination methods (Table 2 2). Beaton and Burns (1968) cite two methods for determining extractable SO 4 from soils: by calcium phosphate (CaHPO 4 ) and LiCl The first determines SO 4 by turbidimet ry and the second by methylene blue. Jones and Jones (2001) have the determination of extractable SO 4 S as Ca(H 2 PO 4 ) 2 0.5 M ammonium acetate ( CH 3 COONH 4 ) 0.25 M acetic acid ( CH 3 COOH ) and 0.01 M CaCl 2 Blair et al. (1993) developed an ex tr action method f or SO 4 S for humid extractants as predictors of S status of pastures, they found that 0.25 M KCl extraction solutions shaken at 40 C had the highest significant coefficient of determination (Blair

PAGE 35

35 et al., 1991). More recently, Bloem et al. (2002) investigated this method for extracting procedures can be used in conjunction with turbidimetry, induct ively coupled pl asma emission spectrometry, and ion chromatography An important theme throughout S research is the wide spread variability for different methods when determining S and even laboratory to laboratory methodological differences in determining S and SO 4 S. An inter laboratory comparison of S in plant tissue and SO 4 S in soil found that variability [higher than 36% coefficient of variation (CV)] was large and that determination of S is a complex and difficult procedure (Crosland et al., 2001). This study made several recommendations that are applicable beyond the United Kingdom, where th e study was conducted. Some suggestions advanced by Crosland et al., (2001) were 1) laboratories should standardize extraction and digestion procedures for S, 2) reference materials should be utilized (none currently exist for soils), 3) analytical methods must be capable of determining low concentrations of S in soil extracts, and 4) methods must be calibrated for diagnostic purposes so that results obtained ar e comparable with those from other methods. The first recommendation is important, because until a standardization of extractions and digestions is created M any more studies will be conducted with different extraction and digestion procedures and variab ility will continue t o be an issue with S research. Yield responses to S application were not related to soil S test values for samples 0 to 30 cm deep in the soil profile (Jackson, 2000). Nitrogen and S sources for this study were as follows: urea, ammo nium sulfate, and potassium sulfate and were all applied as a pre plant broadcast. Jackson (2000) recommended that for canola about

PAGE 36

36 20 kg S ha 1 acetate acetic acid ext ractable S soil test results are <70 kg S ha 1 S oil testing f or evaluating S of soils has been suggested as a tool for determining if S fertilizers should be recommended (Lewis, 1999). Rationale and Justification Tomato growers typically use high analysis fertilizers wh ich are low in S content. This in combination with reduced S emissions from combustion has lowered S input to agricultural systems. Preliminary studies conducted suggest that S is a limiting factor in fresh market tomato yield. Therefore, unless the cha racterization of S in the crop production system is identified, an increasing trend in S deficiency in a ll crops will become apparent. The overall objective of this study is to explore, determine and apply soil SO 4 S extraction methods in Florida sandy soi ls to achieve optimal tomato growth and develo pment. Therefore the specific objectives are: to determine whether, a response to different sources and rates of SO 4 on tomato plant growth, development tomato yield, and S content exists. to determine the influence of irrigation rate and rate of applied S on tomato growth, development, and yield. to compare extraction methods and analytical methods for SO 4 and S determination in soil samples. to establish and interpret the relationship between soil SO 4 S, tomato tissue S concentrations or tomato yield

PAGE 37

37 Table 2 1. Sulfur containing fertilizers, formulas, and percent S and companion nutrients for materials used in crop production Material z Formula Plant Nutrient Content (%) S Other Ammonium nitrate s ulfate NH 4 NO 3 + (NH 4 ) 2 SO 4 14 26% N Ammonium polysulfide NH 4 S x 45 20% N Ammonium phosphate sulfate [(NH 4 ) 2 HPO 4 + (NH 4 )H 2 PO 4 ] + (NH 4 ) 2 SO 4 12 17 11 16% N, 13 48% P 2 O 5 Ammonium sulfate (NH 4 ) 2 SO 4 24 21 N Ammonium thiosulfate (NH 4 ) 2 S 2 O 3 26 12 N Calcium poly sulfide CaS x 22 6 Ca Calcium sulfate CaSO 4 16 23 29 Ca Calcium thiosulfate CaS 2 O 3 10 6 Ca Elemental S S or S 0 100 Ferric sulfate Fe 2 (SO 4 ) 3 9 H 2 O 18 30 20 35 Fe Ferrous sulfate Fe 2 (SO 4 ) 3 7H 2 O 16 30 20 35 Fe Gypsum CaSO 4 2H 2 O 19 24 Ca Magnesium sulfat e MgSO 4 13 10 Mg Potassium polysulfide KS x 23 22 K 2 O Potassium Mg sulfate or Sulfate potash of magnesia K 2 SO 4 MgSO 4 22 22 26 K 2 O, 11 Mg Potassium sulfate K 2 SO 4 18 50 K 2 O Potassium thiosulfate K 2 S 2 O 3 17 25 K 2 O Sulfuric acid H 2 SO 4 33 S coated triple s uperphosphate Ca(H 2 PO 4 ) 2 CaSO 4 2H 2 O + (NH 4 ) 2 SO 4 + S 0 10 20 38 43 P 2 O 5 14 16 Ca Single super phosphate Ca(H 2 PO 4 ) 2 CaSO 4 2H 2 O 14 20 P 2 O 5 18 20 Ca Triple super phosphate Ca(H 2 PO 4 ) 2 CaSO 4 2H 2 O 1.5 46 P 2 O 5 15 16 Ca Urea S CO(NH 2 ) 2 + S 10 20 38 N Urea sul furic acid CO(NH 2 ) 2 + H 2 SO 4 9 18 10 28 N Zinc sulfate ZnSO 4 H 2 O 18 36 Zn z Adapted from Bixby and Beaton (1970) and Maynard and Hochmuth (2007)

PAGE 38

38 Table 2 2. Chemical reagents and concentrations for extracting available S from soil and references in which extractants were used Reagent Reference z H 2 O Bansal et al., 1979 HCl (0.001 M) Bansal et al., 1979 HCl (0.05 M) Islam and Ponnamperuma, 1982 LiCl (0.1 M) Islam and Ponnamperuma, 1982 LiCl (0.01 M) Maynard et al., 1987 NaCl (10 g L 1 ) Bansal et al. 1979 KCl (0.01 M) Maynard et al., 1987 KCl (0.25 M) Blair et al., 1993 KCl (0.025 M) Bloem et al., 2002 CaCl 2 (0.01 M) Yli Halla, 1987 NH 4 OAc(0.5 M) Islam and Ponnamperuma, 1982 NH 4 OAc (0.003 M) Maynard et al., 1987 NH 4 Cl (0.003 M) Maynard et al., 1987 NH 4 Cl (0.01 M) Maynard et al., 1987 NaHCO 3 (0.5 M NaHCO 3 ) Tiwari et al., 1983 NaOAc (0.073 M) + HOAc (0.52 M) + DTPA (to pH 4.8) Wolf, 1982 NaOAc (1.0 M) + HOAc (to pH 4.8) Bansal and Pal, 1987 NH 4 OAc (0.5 M) + HOAc (0.5 M) Yli Halla, 1987 KH 2 P O 4 (500 mg L 1 P) Bansal et al, 1983 Ca(H 2 PO 4 ) 2 (500 mg L 1 P) Islam and Bhuiyan, 1988 Ca(H 2 PO 4 ) 2 (500 mg L 1 P + 2 M HOAc) Hoeft et al., 1973 NaCl (0.1 M) + KCl (0.25 M) + MgCl 2 (1 M) + CaCl 2 (2 M) KH 2 PO 4 (0.01M) Baker, 1973 z adapted from Johnson an d Fixen (1990)

PAGE 39

39 CHAPTER 3 INFLUENCE OF NITROGE N AND SULFUR SOURCES AND APPLICATION RATES ON FRESH MARKET TOMA TO Introduction Excess nutrients, particularly nitrate nitrogen (NO 3 N) and phosphate (PO 4 ), are subject to environmental losses through surface r unoff and leaching (Sato et al., 20 09). In vegetable production in sandy soils, NO 3 N is of concern because it has the potential to leach out of the soil profile and into waterways. Best Management Practices (BMP ) are practices or compilation of practice s which are scientific based, field tested, reviewed, voluntary, a nd incentive driven to improve water quality in agricultural and urban discharges (Bartnick et al., 2005). Sulfur has also been subjected to legislative control, because of acid rain devel opment from SO 2 emissions. a term generally applied to describe micronutrient malnutrition in humans (Welch and Graham, 1999), but can also be used to describe nut rient levels below optimum in plants (Bould, 1968; Dwivedi and Randhawa, 1 974; Smith, 1962). This is an old term used to describe sub adequate levels in which nutrient deficiency systems may not be evident, yet growth and yield are below the optimum. Positive yield an d growth results have been shown when S has been applied to vegetable crops (Pavlista, 2005; Rhoads and Olson, 2000; Susila, 2001). Sulfur deficiencies are often confused with N deficiency in plants as N and S assimilation are closely coordinated (Schere, 2001; Tabatabai, 1984; Zhao et al., 1997) and yellowing of plant tissues is generally upon first observation considered to be N deficiency. In addition, NO 3 reduction or protein synthesis is inhibited by S deprivation (Duke and Reisenauer, 1986) and this can be considered an antagonistic relationship between exce ss N and deficient S (Eriksen et al., 2001; Eriksen and Mortensen, 2002). It has been stated by Haneklaus

PAGE 40

40 et al. (2008) for a 1 kg shortfall of S, 15 kg of N are loss to the environment due to volatilization and leaching, because N utilization efficiency has been reduced (Schnug, 1999). Eriksen et al. (2001) found that increasing N application at low S supply caused S deficiency symptoms to become more pronounced and resulted in a serious retardation of plant growth, suggesting that high N levels decrease d the redistribution of S from older leaves. Yet, no connection to barley ( Hordeum vulgare L.) yield was made in this study to determine if S deficiency resulted in lower yield of barley. Eriksen and Mortensen (2002) also examined the impact of N and S o n barley yields. An interaction with N application and the relative yield increase with S applications were much higher at high N levels than at the lower N application rates. The later the S application in time with high levels of N, the decrease in yie ld was more pronounced for grain than yield of S amino acids (Erik sen and Mortensen (2002). T hese studies did not focus upon the N to S ratio which is often used as an index for methio nine and cystine concentrations, as well as for general plant growth (R adford et al., 1977; Sexton et al., 1998) The N to S ratio has been used as an indicator of S deficiency, but it has some problems with the determination of each element being separately and not concurrent (Blake Klaff et al., 2001). Although, this ratio has been stated as being less variable during plant growth and development (Maynard et al., 1983; Spencer and Freney, 1980), and like all ratios, the magnitude of the individual elements must be taken into account. An N to S ratio threshold of 15:1 has b een suggested by Mills and Jones (1996). Yet, even this ratio has been subjected to some debate. For winter wheat ( Triticale aestivum L.) it has been indicated that lower than 19:1 would be suitable for some growth stages (Spencer and Freney, 1980). Ca lvo et al. (2008) showed that only

PAGE 41

41 early growth stages of red wheat had N to S ratios below the 15:1 threshold level, but advanced growth stages were above this N to S ratio threshold. The authors suggested that N to S ratio should be used in advanced sta ges of wheat growth to correct S deficiencies. Th us using N to S ratio c ould be a tool for diagnosing S deficiencies in strawberry ( Fragaria ananassa Duch.) As strawberries production season is approximately 6 months long Most Florida surface soils ar e low in plant available S, and additional sources are necessary to maintain optimum production despite the lack of actual cases of S deficiencies among crops (Mitchell and Blue, 1981). These authors also suggested that 4 kg S ha 1 as SO 4 derived from SO 2 from summer rainfall based upon calculations and estimations by Brezonik et al. (19 80) would remain in the soil surface for plant uptake. They also suggested that 0.8 kg S ha 1 of the initial 4.8 kg S ha 1 estimate would be either leached or immobilized. Mitchell and Blue (1981) also stated that Spodosols and Entisols lack large reserves of available S in the subsurface horizons in contrast with th at provided by the agrillic horizon of the Ultic soils. The lack of available S reserves for crop growth in Florida sandy soils was further supported by S usila (2001). Both surface and subsoil horizons were inefficient in supplying enough S for six weeks of gro wth without the addition of pre plant application of S as (NH 4 ) 2 SO 4 Although, this study utilized an Arredondo fine sand soil (loamy, siliceous, hyperther imic, Grossarenic Paleudult), a n Ultisol, growth of tomato on this soil required S applications. These results are in conflict with Mitchell and Blue (1981) who stat ed that Ultisols may have enough res erve S in the subsoil for plant growth.

PAGE 42

42 It is recommended that if S deficiencies are diagnosed for vegetable crops grown in Florida, applications of 28 to 37 kg S ha 1 should be provided (Simonne and Hochmuth, 2009). Yet, recommendations for S in crop pro duction have often been empirical in nature rather than based upon previous research. These recommendations might not accurately reflect the current S status of the soil since the reduction in atmospheric S h as occurred over the past 20 year s This has b een a recurring problem in S research, namely soil recommendations not accurately reflecting the current status of S in soils and therefore crops. Li mited information exists on tomato response to concurrent N and S applications as either different sources or rates. Information does exist for N applications as different sources or rates as NO 3 movement in soils is an important issue for Florida tomato growers. Hochmuth and Cordasco (2000) summarized that yield of tomato occasionally increase d when N is app lied over 224 kg N ha 1 and yield responses were similar when using drip or subsurface irrigation in Florida. Regardless of the differences in soil type within the major tomato growing ar eas of Florida, this finding le d to the implementation of a recommen dation of 224 kg N ha 1 on tomato Supplemental N application of 34 kg N ha 1 is recommen ded 1) after a leaching rain, (7.3 cm in 3 d or 9.8 cm in 7 d ), 2) measured (Olson et al., 2009). Ozores Hampton et al. (2006) suggested that the actual rate of N needs to be adjusted based upon planting date and irrigation method This adjustment of the current recommendation is based upon the differences in fertilizer salt movement with temperature (planti ng date) and water application (drip versus subsurface irrigation).

PAGE 43

43 Similarly to concurrent N and S research, S leaf sufficiency range, source and rates studies in field grown tomato are lacking. Those studies conducted involve greenhouse grown tomato and hydroponic culture, which does not involve the soil S cycle, but solution pH. Susila (2001) found that tomato yield response s to different S sources ( (NH 4 ) 2 SO 4 versus NH 4 S 2 O 3 ) were similar, regardless of application method (broadcast versus drip injected ). Heeb et al. (2006) found that the inclusion of S into the fertilizer program for greenhouse grown tomato increased yield and growth in comparison with fertilizer program without S This study also found that tomatoes with S in the fertilizer program h ad preferred taste when compared to similar trea tment program s with out S. For field grown tomato, the sufficiency range within leaf tissue is 3,000 and 8 000 mg S kg 1 on a dry weight basis (Olson et al., 2009). Despite previous studies finding positive r esults for tomato within this range (Susila, 20 01) ; it is possible that this range of S sufficiency is too broad for field g rown tomato. Therefore, the objective of this study was to compare sources and rates of N and S fertilizers on field grown fresh mar ket tomato and tomato leaf S concentrations. Materials and Methods Two separate fie ld trials were conducted in the fall 2006 and spring 2007 at the University of Florida (UF) Gulf Coast Research and Education Center (GCREC), Institute of Food and Agricultu re Sciences (IFAS), University of Florida in Balm, FL on a sandy, siliceous, hyperthermic Oxyaquic Alorthod soil with 1.5% organic matter and a soil pH of 7.3. The soil was very low in SO 4 content (<30 mg S kg 1 ) as revealed by BaSO 4 turbidimetry and soil test interpretation performed 4 weeks before transplanting by a commercial laboratory. Fertilizer sources for treatment applications were

PAGE 44

44 ammonium nitrate (NH 4 NO 3 ), potassium sulfate (K 2 SO 4 ), ammonium sulfate ((NH 4 ) 2 SO 4 ), and ammonium sulfate nitrate ( NH 4 NO 3 + (NH 4 ) 2 SO 4 ) at rates 0 and 186 kg N ha 1 with 0 and 213 kg S ha 1 Fertilizer sources were chosen based upon S refers to fertilizer that is banded on top of a raised bed. This fertilizer blend has a high irrigation water in a subsurface irrigation system. Therefore, the fertilizers (salts) are diluted via subsurface irrigation. Muriate of potash (KCl) (60% K) was used to balance total K amounts in those treatments using NH 4 NO 3 or (NH 4 ) 2 SO 4 as a fertilizer source to ensure that this nutrient was under similar non limiting conditions (Table 3 1). The plots were arrange d in a randomized complete block design with four replications. Experimental plots were 8.8 m long with a 2.9 m long non treated buffer zone at the end of each plot. Fertilizer treatments were applied 21 d before transplanting within two 7.6 cm deep bands separated 34 cm apart centered on the bed cm wide at the top, 19.6 cm high, and spaced 1.47 m apart on centers. Pressed beds were fumigated with methyl bromide plus chloropicrin (67:33 v/ v) at a rate of 196 kgha 1 to eliminate soilborne diseases, nematodes and weeds on 1 Aug. 2006 and 23 Feb. 2007. Simultaneously, plan ting beds were covered with 1.5 the center of each bed on 17 Aug. 2006 and for the second experiment on 12 Mar. 2007. Irrigation was supplied as subsurface irrigation at an ap proximate rate of 61,983 Lha 1 per day in Fall 2006. Irrigation was supplied as a hybrid irrigation system in Spring

PAGE 45

45 2007. A hybrid irrigation system uses both subsurface and drip irrigation and employed by those growers transitioning from all subsurface to drip irrigation. Irrigation supplied as subsurface was at the approximate rate of 21,035 Lha 1 per day. Drip irrigation tubing (T Tape Systems International, San Diego, CA) was buried to a depth of 2.5 cm in the center of the raised bed with a flow rate of 0.058 Lm 1 min 1 The irrigation rate using drip system was at an approximate rate of 39,095 Lha 1 per d Therefore, the total water applied via irrigation for Spring 2007 was 60,130 Lha 1 per d The water table was maintained between 46 and 61 cm deep and constantly monitored with piezometers located in the fields. Phosphorus (P) was not added, because soil analysis revealed 108 mg P kg 1 and this is considered high production (Simonne and Hochmuth, 2009). Tom ato plants were s taked and tied as described by Csizinszky et al. (2005) and weekly spray applications to maintain pest and disease control based upon weekly field scouting results. Newly mature leaves were randomly collected from each plot at 12 wk after transplanting (WAT) to determine foliar S concentration in Fall 2006 and Spring 2007. These samples were dried at 109 C and ground to pass through a 0.853 mm sieve aperture ( 20 mesh screen ) Samples were sent to a commercial la b for total S determinati o n by I CP OES Tomato fruit were harvested twice (8 and 10 WAT in Fall 2006, 10 and 12 WAT in Spring 2007) and graded as marketable and non marketable based upon size. Marketable tomato fruits were graded according to the current industry standards for si ze categories (Brown, 2009; Sargent and Moretti, 2004). The sizes were 5X6 (7.0 cm) 6X6 (7.06 to 6.32 cm) 6X7 (6.43 to 5.72 cm), and cull ( < 5.4 cm). Cull fruit were those smaller than 5.4 cm in diameter which is below the minimum

PAGE 46

46 diameter for a 6x7 to mato fruit. All data were analyzed using ANOVA to determine treatments effects (P<0.05) and treatment means were sep arated using Waller Duncan mean separation procedure and orthogonal contrasts (SAS Institute, 2000). Single degree of freedom orthogonal co ntrasts were used to compare individual treatments. All contrasts were checked to ascertain that they were stochastically independent, and coefficients equal to zero. Results and Discussion Total marketable yield and quality of tomato were both significan tly influenced by S applications (Table 3 2 and Table 3 3) for Fall 2006 only In Fall 2006, both NH 4 NO 3 + K 2 SO 4 + KCl and NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 had higher total yield than the non treated control, NH 4 NO 3 + KCl and (NH 4 ) 2 SO 4 + KCl with an average tot al yield of 42,099 kgha 1 Any treatment that included S as either K 2 SO 4 or NH 4 NO 3 + (NH 4 ) 2 SO 4 were significant when compared to NH 4 NO 3 + KCl. Also the comparison between (NH 4 ) 2 SO 4 + KCl against NH 4 NO 3 + K 2 SO 4 + KCl and NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 were si gnificant. For Spring 2007, all treatments had higher yields when compared to the non treated control with a total average marketable yield of 72,741 kgha 1 None of the single degree of freedom orthogonal contra sts were significant suggest ing that tempe rature, rainfall, and possibly the hybrid irrigation system used in the second season influenced yield response. The differences between total marketable yield and quality of yield could be due to the inherent season temperature and rainfall patterns that occur in Florida. The average rainfall and temperatures for each season were 27.9 and 17.9 cm with minimum temperatures of 18 and 14.5 C, and maximum temperatures 30.9, and 28.7 C, Fall 2006 and Spring 2007, respectively. It is also

PAGE 47

47 possible that even though the irrigation water amount was similar in both seasons. The differences in the irrigation system s could have influenced the increased total yield in Spring 2007. For Fall 2006, both 5x6 and 6x6 grade tomato followed the same pattern as total mark etable yield for the season (Table 3 3 ). Both NH 4 NO 3 + K 2 SO 4 +KCl and NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 had higher quality yield than those treatments receiving no S and (NH 4 ) 2 SO 4 + KCl. The yield by tomato size (grade) for 5x6 and 6x6 were 17,647 and 13,223 kg ha 1 for NH 4 NO 3 + K 2 SO 4 +KCl and NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 respectively. Tomato yield for 6x7 size was similar to both 5x6 and 6x6 in respect to mean separation but NH 4 NO 3 + KCl was also similar to NH 4 NO 3 + K 2 SO 4 + KCl for all sizes of tomato. T he si ngle degree of freedom orthogonal contrasts were significant for NH 4 NO 3 + KCl compared to NH 4 NO 3 + K 2 SO 4 + KCl and NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 for this size Plus (NH 4 ) 2 SO 4 + KCl compared to NH 4 NO 3 + K 2 SO 4 + KCl and NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 contrasts wer e significant. This suggested that those treatments receiving NH 4 NO 3 + K 2 SO 4 + KCl or NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 were superior in quality of tomato yield. For Spring 2007, only 5x6 and 6x6 grades were significant for yield of tomato size (grade) and then only 6x6 had significant single degree of freedom orthogonal contrasts. The non treated control was the lowest yielding treatment for 5x6 tomato es in Spring 2007. All other treatments responded similarly with a total average yield at 57,437 kgha 1 Fo r 6x6 grade tomato in Spring 2007, (NH 4 ) 2 SO 4 + KCl and NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 were similar, yet NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 was not different from all the other treatments, including the non treated control. This was unexpected and the contrasts furthe r reveal that (NH 4 ) 2 SO 4 + KCl was the superior

PAGE 48

48 treatment when compared to NH 4 NO 3 + K 2 SO 4 + KCl, NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 and NH 4 NO 3 + KCl. Only Spring 2007 was foliar S concentrations within tomato leaves significant (Table 3 4). All treatments were sim ilar except for NH 4 NO 3 + KCl which had the lowest S leaf concentration at 6,500 mg S kg 1 This amount is still within the recommended range (3,000 to 8,000 mg S kg 1 ) as stated by Olson et al. (2009). Three single degree of freedom orthogonal contrasts were significant; NH 4 NO 3 + KCl versus (NH 4 ) 2 SO 4 + KCl, NH 4 NO 3 + KCl versus NH 4 NO 3 + K 2 SO 4 + KCl; NH 4 NO 3 + KCl versus NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 uptake of this nut rient only i n Spring 2007 Yet, the quality (i.e. size increase ) of tomato harvested was high er in Fall 2006 with the same treatments. This suggests that although S increases total yield and quality of tomato, seasonal impacts in terms of temperature and rainfall wi ll influence how much S will improve quality. In this case, positive tomato yield response was found as concen tration increased from about 6, 737 and 9,612 mgkg 1 in Fall 2006 and Spring 2007, respectively when S was included in the fertilizer regime. This demonstrated that application of S in tomato fertilization programs is essential to increase marketable yields. These results agree with those presented by Pavlista (2005), Rhoads and Olson (2000) and Susila (2001), in which S sources did not cause diffe rential responses to crop growth or development for cabbage, tomato, or pepper

PAGE 49

49 Table 3 1. Nitr ogen, K, and S sources, rates applied and placement used to create treatments for determining the r Fall 2006 and Spring 2007 at Gulf Coast Research and Education Center in Balm, FL. Treatment Sources Pre plant Drip Injected Total N K S N K S N K S kgha 1 1 Non treated control 0 270 0 0 0 0 0 270 0 2 NH 4 NO 3 + KCl 186 270 0 0 0 0 186 270 0 3 ( NH 4 ) 2 SO 4 + KCl 186 270 213 0 0 0 186 270 213 4 NH 4 NO 3 + K 2 SO 4 + KCl 186 270 213 0 0 0 186 270 213 5 NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 186 270 213 0 0 0 186 270 213

PAGE 50

50 Table 3 2. Influence of N and S sources and applied rates on total marketable yield of tomat o for Fall 2006 and Spring 2007 at G ulf Coast Research and Education Center in Balm FL Fertilizer sources N Rate K Rate S Rate Yield z Fall 2006 Spring 2007 (kgha 1 ) Non treated control 0 270 0 13,123b 42,384b NH 4 NO 3 + KCl 186 270 0 19,154b 68,773 a (NH 4 ) 2 SO 4 + KCl 186 270 213 15,079b 78,758a NH 4 NO 3 + K 2 SO 4 + KCl 186 270 213 44,096a 66,735a NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 186 270 213 40,102a 76,699a P Values <0.01 <0.01 CV 25 13 Single degree of freedom orthogonal contrasts P values NH 4 NO 3 + KCl v s (NH 4 ) 2 SO 4 + KCl 0.40 0.12 NH 4 NO 3 + KCl vs NH 4 NO 3 + K 2 SO 4 + KCl <0.01 0.74 NH 4 NO 3 + KCl vs NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 <0.01 0.21 (NH 4 ) 2 SO 4 + KCl vs NH 4 NO 3 + K 2 SO 4 + KCl <0.01 0.07 (NH 4 ) 2 SO 4 + KCl vs NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 <0.01 0.74 NH 4 NO 3 + K 2 SO 4 + KCl vs NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 0.41 0.12 z mean separation of total yield by Waller Duncan

PAGE 51

51 Table 3 3. Marketable yield quality as influenced by N and S fertilization sources and applied rates for Fall 2006 and Spring 2007 at Gu lf Coast Research and Education Center Balm Fertilizer sources N Rate K Rate S Rate Grade Distribution (kgha 1 ) z (kgha 1 ) 5 x 6 6 x 6 6 x 7 Fall 2006 Spring 2007 Fall 2006 Spring 2007 Fall 2006 Spring 2007 Non treated control 0 270 0 3,505b 29,7 91b 3,394b 9,598b 5,624c 2,995a NH 4 NO 3 + KCl 186 270 0 4,238b 55,731a 6,358b 10,881b 8,559bc 2,159a (NH 4 ) 2 SO 4 + KCl 186 270 213 4,535b 57,565a 4,238b 19,032a 6,276c 2,160a NH 4 NO 3 + K 2 SO 4 + KCl 186 270 213 19,562a 54,407a 13,530a 10,311 b 11,004ab 2,017a NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 186 270 213 15,731a 61,946a 12,715a 12,634ab 11,656a 2,119a P Values <0.01 <0.01 <0.01 0.03 <0.01 0.81 CV 52 14 30 32 23 56 Single degree of freedom orthogonal contrasts P values NH 4 NO 3 + KCl vs (NH 4 ) 2 SO 4 + KCl 0.92 0.72 0.24 0.01 0.12 0.99 NH 4 NO 3 + KCl vs NH 4 NO 3 + K 2 SO 4 + KCl <0.01 0.24 <0.01 0.54 0.04 0.97 NH 4 NO 3 + KCl vs NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 <0.01 0.24 <0.01 0.54 0.04 0.97 (NH 4 ) 2 SO 4 + KCl vs NH 4 NO 3 + K 2 SO 4 + KCl <0.01 0.54 <0.01 < 0.01 <0.01 0.88 (NH 4 ) 2 SO 4 + KCl vs NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 <0.01 0.40 <0.01 0.04 <0.01 0.96 NH 4 NO 3 + K 2 SO 4 + KCl vs NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 0.29 0.16 0.64 0.42 0.64 0.91 z mean separation of yield quality by Waller Duncan

PAGE 52

52 T able 3 4. Effects of N and S sources and applied rates on tomato foliar S concentration as percent S for Fall 2006 and Spring 2007 at Gulf Coas t Research and Education Center in Balm FL Fertilizer sources N Rate K R ate S Rate Foliar S concentration z (kgha 1 ) (mgkg 1 ) Fall 2006 Spring 2007 Non treated control 0 270 0 5,333a 8,725a NH 4 NO 3 + KCl 186 270 0 4,550a 6,500b (NH 4 ) 2 SO 4 + KCl 186 270 213 5,450a 9,400a NH 4 NO 3 + K 2 SO 4 + KCl 186 270 213 6,900a 8,625a NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 186 270 21 3 6,450a 9,725a P Values 0.16 0.02 CV 23 15 Single degree of freedom orthogonal contrasts P value NH 4 NO 3 + KCl vs (NH 4 ) 2 SO 4 + KCl 0.35 <0.01 NH 4 NO 3 + KCl vs NH 4 NO 3 + K 2 SO 4 + KCl 0.02 <0.01 NH 4 NO 3 + KCl vs NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 0.06 <0.01 (NH 4 ) 2 SO 4 + KCl vs NH 4 NO 3 + K 2 SO 4 + KCl 0.15 0.40 (NH 4 ) 2 SO 4 + KCl vs NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 0.30 0.72 NH 4 NO 3 + K 2 SO 4 + KCl vs NH 4 NO 3 + (NH 4 ) 2 SO 4 + K 2 SO 4 0.64 0.24 z Foliar S concentration data means separated with Waller Duncan

PAGE 53

53 CHAPTER 4 SULFUR FERTILIZATION RATES AND IRRIGATION PROGRAMS ON TOMATO GROWTH AND YIELD Introduction Crop plants obtain their S requirement from a number of sources; including soil, crop residues, manures, irrigation waters, rainfall, the atmosphere, and soil amendments and fertilizers (Tabatabai, 1984). Field sites without access to mineral bound S sources may have insufficient S dissolved in soil water and shallow ground water bodies to move available S to plant roots (Eriksen et al. 199 8). The concentration of S in irrigation waters used throughout North American is unknown, and little research is available related to the S content of irrigation waters to S requirement in a fertilizer program (Olson and Rehm, 1986). Studies on high sul fate irrigation waters utilized water with 175, 646, 862, and 1743 mgL 1 SO 4 (Drost et al., 1997) and 15.0 and 38.8 mol SO 4 L 1 (Papadopoulos, 1986) have found mixed results in improving yield of broccoli and tomato. In Florida, SO 4 levels in drinking wa ter wells can range from 0.2 to 1400 mgL 1 with higher concentrations coming from deeper wells (Sacks, 1996). While these values cover drinking water, the levels of SO 4 or S within irrigation water in Florida is unknown It has been suggested that water pumped from the aquifer as well as surface waters, has a rotten egg smell commonly associated with S. This may often leads to the belief that S in the irrigation water is an S source for crop production. Tomato production in Florida often occurs on deep Spodosols (fine sandy soils) with low organic matter (>2%) and therefore inherently low in organic and inorganic S. Previous studies on tomato and S has found that S sources ( (NH 4 ) 2 SO 4 NH 4 S 2 O 3 NH 4 NO 3 + (NH 4 ) 2 SO 4 K 2 SO 4 ) affected tomato yields similarly (Susila, 2001; Santos et al., 2007). Elemental S is a less expensive S fertilizer source when compared to other

PAGE 54

54 sources of S that contain N Yet, limited information ex ists on using elemental S with irrigation programs to impr ove tomato S content and yi eld in Florida. Elemental S can be an ideal S source for crop growth and yield because it is 1) high analysis of S and insoluble in water; 2) stable in damp and humid conditions 3) has to be oxidized to SO 4 S for plant uptake (Solberg et al., 2003). In o rder to achieve high yields and to minimize S leaching, rates of fertilizer S should be recommended on the basis of available soil S and crop requirement (Scherer, 2001). Therefore, the goal of this study was to investigate whether or not irrigation water in conjunction with S applications can supply the necessary S for to mato production. The objective of this study was to determine the influence of irrigation program and S rates on fresh m arket tomato growth and yield. Materials and Methods Two field stu dies were conducted at GCREC on a Myakka fine sand (sandy siliceous hyperthermic Aeric Alaquods) for two seasons, Spring 2008 (Mar. to July 2008) and Fall 2008 (Aug. to Dec. 2008). All crop management was conducted according to current recommendations for tomato (Olson et al., 2008). So il was fumigated with 50:50 (v: v) methyl bromide to chloropicrin mixture at 190 kgha 1 using seedlings were transplanted on 6 Mar. 2008 and 28 Aug. 2008, 21 d after soil fumigation into raised beds (19.6 cm high and 68.6 cm wide). The experimental design was a split plot with six replications. This design allows for more precision to be given to factors within sub plots. Main pl ots were drip irrig ation volumes of 5,406, 8,109, and 10,812 Lha 1 d 1 while sub plots were S rate at 0, 28, 56, 112, 168, and 224 kg S ha 1 Irrigation rates were based upon the historical Penman method for central Florida (Simonn e et al., 2009 ). Irrigation rates were cr eated by changing times and frequencies of drip irrigation

PAGE 55

55 events. Water levels of SO 4 S at GCREC was found to be 57.8 mgL 1 Sulfur rates were applied pre fumigation as elemental S (0 0 0 90S). All other nutrients were applied at recommended rates base on soil test results (Olson et al., 2008) using a 6 2 8 (N P 2 O 5 K 2 O ) liquid fertilizer. The following variables were collected: soil samples at 5 and 13 WAT in Spring 2008 and 4 and 15 WAT in Fall 2008 for soil S, and pH. Dried soil samples were submitte d to Waters Agricultural Laboratory in Camilla, GA for SO 4 determination by the turbidimetric method. Plant tissue samples for S were collected 5 and 13 WAT in Spring 2008 and 4 and 11 WAT in Fall Season 2008. Dried and ground tissue samples were also su bmitted to Waters Agricultural Laboratory in Camilla, GA for total S determination by ICP OES Soil samples were held at 4.4 C and tissue samples at 0 C until shi pment to the lab for analysis. Soil moisture was monitored every other week for 5 d using a time domain reflectometry probe (TDR) (FieldScout TDR 200 Soil Moisture Meter, Spectrum Technologies, Inc., Plainfield, IL). Plant vigor was assessed on 5, 8 and 12 WAT in Spring 2008. Plant vigor was a subjective visual assessment on a scale from 0 to 100; where 0 indicate s a dead plant and 100 was a very healthy green plant. Plant height was collected on 5 and 13 WAT in Spring 2008. For Fal l 2008, plant height was collected as an alternative to the subjective plant vigor rating at 5 WAT. Estimated chlorophyll was collected as leaf greenness using a Minolta SPAD 502 meter at 4, 6, and 11 WAT in Spring 2008 and at 7 and 11 WAT in Fall 2008. Tomato fruit were harvested during 12 and 13 WAT in Spring 2008 and 11 and 14 WAT in Fall 2008. Harvest ed frui t were then separated as marketable and non marketable based upon

PAGE 56

56 size. Marketable tomato fruits were graded according to the current standards for size categories (Brown, 2009; Sargent and Moretti, 2004). The sizes were 5X6 (7.0 cm) 6X6 (7.06 to 6.32 c m) 6X7 (6.43 to 5.72 cm) and cull ( < 5.4 cm). These sizes are based upon the number of horizontal rows and columns of fruit fitted into an 11.3 kg tomato box. Cull fruit were those smaller than 5.40 cm in diameter which were below the minimum diameter f or a 6X7 marketable tomato fruit. Data were subjected to an analysis of variance to determine treatment significance using a general linear model (SAS Institute, 2000). Regression models were also fit to significant data using SigmaPlot (Systat, 2008). T reatment means were separated by standard error s Results and Discussion No interactions were obs erved between irrigation volumes and S rates for any collected plant variable. Irrigation volumes were found to influence early season plant vigor in Spring 2 008, but was not evident at the end of the season An increase in vigor was observed when irrigation volumes were increased from the lowest rate to the highest rate, although a linear contrast was not significant (Table 4 1). Plant height was not influen ced by either irrigation or S rates in either growing season (Table 4 1). Estimated chlorophyll was also not influ ence by either irrigation volumes or S rates for Spring 2008 growing season (Table 4 2). However, late in Fall 2008, estimated chlorophyll w as influenced by applied S rates and both linear and quadratic contrasts were significant for this observation (Table 4 2). Mid season soil SO 4 S and soil pH were not influenced by either irrigation program or S rate for both Spring and Fall 2008 (Table 4 3). End of season soil SO 4 S and soil pH were not influenced by irrigation regime and applied S rates for Fall 2008. However, Spring 2008, end of season pH were influenced by applied S rates and a significant linear contrast was found. It is

PAGE 57

57 possible t hat season temperature and rainfall differences between cropping seasons may be responsible for the pH differences within the soil, because oxidation of S is highly dependent upon environmental factors. Solberg et al. (2003) has suggested elemental S fert ilizers require oxidation time to become plant available SO 4 S which may help explain the lack of differences in S. The oxidation of elemental S is highly dependent upon soil conditions particularly to those soil microbes involved in the S cycle. Unexpec tedly, an increase in irrigation rate did not influence soil SO 4 S or soil pH. Increased soil moisture is believed to increase elemental S oxidation in the soil. This suggests that soil C or soil substrates may be the limiting factor of S oxidation in Fl orida sandy soils. A seasonality effect of S applications on leaf S concentration was found to exist (Table 4 4). After separation of the two seasons, S leaf concentrations were no longer influenced by applied S rates For Spring 2008, S leaf concentrati ons ranged from 5,456 to 5,714 mg S kg 1 Despite these results for Spring 2008, irrigation was found to influence Fall 2008 S leaf concentration and a significant linear contrast was found. Leaf S concentration decreased with increasing irrigation regim e, possibly indicating that as applied S oxidized into SO 4 form it was leached from the root zone. All the leaf S concentrations were within acceptable range for leaf S for tomatoes (Olson et al., 2008; Susila, 2001), rega rdless of season. Early tomato yi eld was influenced by S rates (P < 0.05), but not irrigation (data not shown). Yields were o bserved to increase (Figure 4 1 ) in a linear plateau fashion described by the following equation (y = 10,851/ (1 + e ( (x+1.15)/1.16)))). This curve was also used by Blake Kalff et al. (2000) to describe S fertilization influence on yield of

PAGE 58

58 wheat. Increasing S applications from 0 to 28 kg S ha 1 improved early yield by over 3,500 kgha 1 Adding 28 kg Sha 1 is equivalent to increasing t omato yield by one hundre d 11.3 kg boxes per hectare. These results were similar to tomato yield incre ases reported by Susila (2001). Irrigation program did not influence soil S concentration, soil pH, leaf S concentrations or ea rly tomato yield. This suggested that irrigation w ater may not have met the S needs of tomatoes when grown with drip irrigation. It is possible that the form of S present within the irrigation water needs to be oxidized by microbes. An increase in tomato leaf S and early yield were found with the additi on of elemental S to the fertilizer regime. Based upon these results, at least a minimal amount of S should be incorporated into the fertilizer regime to maximize early yield of fresh market tomato in Florida.

PAGE 59

59 Table 4 1. Plant height and vigor ratings for tomatoes grown under differing irrigation regimes and applied S rates durin g Spring and Fall 2008 at G ulf Coast Research and Education Center in Balm FL Plant Height z (cm) Plant Vigor z (%) Spring 2008 Fall 2008 Spring 2008 5 WAT 13 WAT 5 WAT 5 W AT 8 WAT 12 WAT Irrigation Regime (Lha 1 d 1 ) 5,406 39a 79a 32a 52b 64a 64b 8,109 39a 77a 32a 58ab 65a 70a 10,812 39a 78a 32a 60a 60a 70a P Value 0.70 0.48 0.57 0.03 0.48 0.35 Linear Contrast Irrigation 0.61 0.31 0.45 0.6 5 0.16 0.90 Soil S (kgha 1 ) 0 38a 78a 32a 54a 58a 66a 28 39a 78a 32a 56a 60a 69a 56 39a 78a 31a 57a 63a 71a 112 39a 79a 31a 60a 66a 66a 168 39a 78a 33a 56a 66a 67a 224 38a 76a 33a 60a 6 5a 69a P Value 0.54 0.62 0.26 0.70 0.22 0.59 Orthogonal Contrasts Linear 0.20 0.24 0.11 0.22 0.01 0.76 Quadratic 0.96 0.24 0.41 0.66 0.28 0.63 Cubic 0.18 0.47 0.66 0.57 0.74 0.10 CV 6 6 8 22 20 14 z means separation by Waller Duncan

PAGE 60

60 Table 4 2. Estimated chlorophyll content for tomatoes grown under differing irrigatio n regimes and applied S rates for Spring and Fall 2008 at G ulf Coast Research and Education Center in Balm FL. Estimated Chlorophyll z (%) Spring 2008 F all 2008 4 WAT 6 WAT 11 WAT 7 WAT 11 WAT Irrigation Regime (Lha 1 d 1 ) 5,406 51a 48a 48a 45a 48a 8,109 51a 48a 44a 44a 43c 10,812 51a 46b 83a 44a 44b P Value 0.75 0.17 0.40 0.84 0.37 Linear Contrast Irrigation 0.38 0.03 0.22 0.93 0.04 Soil S (kgha 1 ) 0 50a 47a 46a 46a 45b 28 50a 48a 46a 45a 48a 56 51a 46a 44a 44a 45b 112 52a 48a 45a 42a 46ab 168 51a 47a 47a 44a 44bc 224 51a 47a 129a 46a 42c P Value 0.78 0.38 0.50 0.11 <0.01 Ort hogonal Contrasts Linear 0.34 0.32 0.16 0.68 <0.01 Quadratic 0.56 0.83 0.21 <0.01 0.02 Cubic 0.74 0.90 0.36 0.50 0.40 CV 15 17 517 09 15 z means separation by Waller Duncan with

PAGE 61

61 Table 4 3. Mid season and end of season soil pH and soil S content for soil samples collected from tomato production under differing irrigation regimes and applied S rates during Spring and Fall 2008 at G ulf Coast Research and Education Center in Balm, FL. Soil pH Soil S (kgh a 1 ) Spring Fall Spring Fall Irrigation Regime (Lha 1 d 1 ) 5 WAT 13 WAT 4 WAT 15 WAT 5 WAT 4 WAT 15 WAT 5,406 6.7a 6.3a 6.8a 6.7a 25.2a 34.3a 31.9a 8,109 6.6a 6.3a 6.7a 6.7a 25.6a 35.4a 33.1a 10,812 6.6a 6.1a 6.7a 6.8a 24.6a 34.2a 30.0a P Value 0.91 0.51 0.79 0.21 0.91 0.22 0.49 Linear Contrast Irrigation 0.73 0.33 0.62 0.48 0.60 0.49 0.37 Soil S (kgha 1 ) 0 6.7a 6.8a 6.7a 6.7a 22.6a 33.2a 34.9a 28 6.6a 6.6a 6.7a 6.8a 25.8a 36.0a 27.3a 56 6.6a 6.2b 6.7a 6.7a 22.9a 33.6a 30.8a 112 6.6a 6.2a 6.7a 6.8a 22.8a 34.8a 31.7a 168 6.6a 6.0bc 6.8a 6.8a 27.3a 35.3a 31.2a 224 6.6a 5.8c 6.8a 6.8a 29.5a 39.1a 34.8 P Value 0.27 <0.01 0.85 0.71 0.12 0.81 0.62 Orthogonal Contrasts Linear 0.12 <0.01 0.53 0.15 0.03 0.61 0.67 Quadratic 0.68 0.49 0.30 0.59 0.23 0.71 0.20 Cubic 0.59 0.50 0.89 0.97 0.39 0.70 0.48 CV 2 9 3 3 35 19 37 z means separation by Waller Duncan with

PAGE 62

62 Table 4 4. Influence of irrigation r eg ime and applied elemental S rates on tomato leaf S concentrations for seasonal pooled samples for Spring and Fall 2008 at G ulf Coast Research and Education Center in Balm FL. S Leaf Concentration (mgkg 1 ) z Spring 2008 Fall 2008 Irrigation Regime (L ha 1 d 1 ) 5,406 5,539a 6,079b 8,109 5,711a 6,642a 10,812 5,714a 5,900b P Value 0.19 0.02 Linear Contrast Irrigation 0.90 <0.01 Soil S (kgha 1 ) 0 5,456a 6,429a 28 5,622a 5,958b 56 5,589a 6,079ab 112 5,686a 6,308ab 168 5,584a 6,275ab 224 5,680a 6,192ab P Value 0.40 0.11 Orthogonal Contrasts Linear 0.08 0.99 Quadratic 0.46 0.24 Cubic 0.58 0.01 CV 10 7 z means separation by Waller Duncan

PAGE 63

63 Figure 4 1. Influence of rates of applied elemental S on first harvest of fresh market tomato for Spring and Fall 2008 at G ulf Coast Research and Education Center in Balm FL

PAGE 64

64 CHAPTER 5 COMPARISON OF ANAL YTICAL METHODS FOR D ETERMINE S AND SO 4 S PRESENT IN FLORIDA S ANDY SOILS Introduction Current assessment of S and SO 4 S in agricultural soils for the development of fertilizer recom mendations has been difficult The major difficulties that researche rs confront are inherent variability in analytical results, lack of consistent associations between crop yield, and S in soils or plant tissues. In addition to these difficulties, soil or plant S determination is often a separate method adding time and cos ts in an analytical laboratory. Ideal characteristics for S determination would be low cost, rapid, accurate, precise, with low interference, and potentially offer multi element determination options (Table 5 is the cheapest and most widely used chemical analysis employed in agriculture (Van Raij et al., 2002) and is often underutilized by many (Jones, 1993). It is primarily the cost of labor to operate analytical instrument that determines the underlying cos t of each soil sample. In terms of speed, ma n y instruments (i.e. elemental analyzer inductively coupled plasma atomic emission spectrometer or inductively coupled plasma optical emission spectrometer (ICP AES/ICP OES), ion exchange chromatograph (IC), and discrete analyzer) have been automated to allow sample loading time to be reduced. This can essentially cut labor and time for the laborat ory worker spent supervising instrument status. In the past, S and SO 4 S determination were labor intensive and req uired reagents have a finite shelf life for use. Such example is methylene blue, which is the color reagent for estimating the reduction of sulfate to sulfide (Tabatabai, 1996). The color of this reagent is important to the determination of S by spectrom etric means; if

PAGE 65

65 kept away from direct sunlight the color can remain stable for 24 h or more (Tabatabai, 1996). This would be considered to be a time limitation or shelf life restriction for this method. Not properly understanding and identifying limitati ons in a method can lead to data quality issues. Quality control and assurance plans are implemented to maintain result quality. Quality control consists of using the following: proper calibration, blanks, duplication of samples, spiked samples and the u se of certified standards ( Sham pine, 1993 ). Quality assurance usually consists of validation of methods (characteristics and limits). Both of these are extremely important to S determination, because of the inherent variability in determining S and SO 4 S It has been reported by Crosland et al. (2001) that the coefficient of variation from an inter laboratory study on soil S was between 36 to 45%. This means up to 45% of the variation in soil S values was due to random error and the decision to apply fe rtilizer based upon this data could be faulty. This random error can be attributed to operator error, loss of calibration during analysis or interferences from oth er elements within the aliquot. Interferences from other elements or ions are the most commo n error encountered when determining S or SO 4 S. With the advent of newer technologies, such as ICP OES, claims that less interference would be encountered were publicized (Kola et al., 2002). Multi element determinations are limited to those instruments with the ability to determine multiple elements. These include, but are not limited to the following: elemental analyzers, ICP AES/ICP OES, and IC. Instruments with multiple element determination capabilities are viewed as a versatile economic investme nt. These types

PAGE 66

66 of instruments reduce labor, because more than one element can be analyzed within a single sample. Currently five analytical instruments are capable of determining total S or SO 4 S. These instruments are ICP AES or ICP OES, inductively coupled plasma mass spectrophotometer (ICP MS), elemental analyzer by infrared radiation cell/spectrometer (IR) or titration of SO 2 IC, and turbidimeter. Of these five instruments, the three most commonly used to determine S or SO 4 S are ICP AES/ICP OES, elemental analyzer via IR or titration, and colorimetrically with a spectrometer (turbidimeter). The principle behind ICP AES/ICP OES is based in chemistry and physics. More specifically, it involves particle collision, excitement of outer shell electro ns, adsorption/emission of a wavelength, and separation/detection of spectral lines of interest. An ext ended review of the theory will not be presented here, but several reviews and sources explain in detail such theory and mechanics of ICP AES/ICP OES (B oss and Fredeen, 1997; Lajunen and Permki, 2004; Manning and Grow, 1997; Welz et al., 2009). Several advantages for determining S with ICP AES/ICP OES are after initial costs : cost effectiveness rapid sample analysis, multiple element, low level of int erferences, wide linear dynamic range and detection limits a 1 level (Kola et al., 2002 Mermet and Po ussel, 1995). One disadvantage for determining S with ICP AES/ICP OES is that a vacuum is required within the instrument as S is considered a non me tal element. This is required because emission lines fo r S lie in the ultraviolet range so that the lines are not absorbed by O 2 (Soltanpour et al., 1996). A range of emissions are used for S on ICP AES/ICP OES. In general, the range of wavelengths used is from 180.04 to 182.63 nm and the three strongest em ission lines of S in the

PAGE 67

67 vacuum ultraviolet spectrum are at 180.73, 182.04, and 182.63 nm (Wu et al., 2009). Two significant spectral overlaps occur around the 180.73 nm S line; 180.361 nm Ca and 180.74 nm Mg lines (Eames and Cosstick, 1992). Therefore, corrections or adjustments must be made to ensure interfering spectral lines are not distorting results for S. The principle behind an elemental analyzer, often called by brand name LECO or by the method name Dumas, is titration via colorimetric determinat ion or IR determination of total S. The D umas method involves the combustion of samples a t high temperatures (<1000 C) and the evolution of S as SO 2 and subsequent capture for determination. Earlier Dumas models use a photocell with SO 2 collected by HCl containing KI, starch, and a trace amount of KIO 3 approximately 0.22 M KIO 3 (Tabatabai, 1996; Tiedemann and Anderson, 1971) or 1N NaOH (Tabatabai, 1996) with colorimetrically determining total S by methylene blue similar to Johnson and Nishita (1952). Fo r IR determination of S, after combustion, a stream of gas is directed into a ballast tank, homogenized before a sample is trapped for infrared detection of S (Kowalenko, 2001). The infrared source con sists of nichrome wire, which was resistance heated to 850 C, SO 2 al., 2006), with a band pass filter to prevent further infrared energy reaching the IR detector (LECO, 2006). The advanta ges of an elemental analyzer were that it was designed for simpli ci ty, speed, convenience, C and N could also be determined with some models (Tabatabai, 1996), and capable of measuring total S in soils including those with high organic matter content is possible unlike with wet digestion methods such as sodium hypobromite

PAGE 68

68 ( NaBrO ) (Kowalenko, 2001). Some disadvantages of elemental analyzers are the following: initial cost of the instrument, the built in flow conditions for S restrict the scope for optimizing the combustion cy cle for organic and total C deter minations when analyzing C and S simultaneously (Wright and Bailey, 2001), and without an automated sample loader this process consumes time and labor to load each sample individually. The principle behind SO 4 S determination by turbidity was the Beer in which a reduction in the intensity of light passing through a solution containing suspended particles of BaSO 4 (Beaton et al., 1968). The Beer Lambert law was written equations states that absorbance (A) is directly proportional to molar solution wi thin the cuvette. Absorbance was used instead of perce nt transmission because absorbance follows a linear line when plotted against concentration. This relationship or law is not obeyed at higher concentrations. The advantages of determining SO 4 S by turbidimetry are that the determination is rapid, in some cases sensitive, most commonly used for determining SO 4 S in soils Therefore many modifications are available to adjust for its short comings including exchange resins, addition of charcoal filters, and flow injection analysis (Morais et al., 2006). Th e disadvantages to turbidimetric measurement of SO 4 S are insensitivity at low concentrations (less than 10 mgL 1 ), multiple cation interferences (Na, K, Ca, Mg, NO 3 PO 4 and SiO 2 ), cation co precipitations with Ba high organic matter co precipitation, and difficulty of reproducible BaSO 4 suspensions under uniform precipitating conditions (Beaton et al., 1968; Tabatabai, 1996).

PAGE 69

69 While these are the most commonly available methods in the US for the determination of S and SO 4 S, it is unknown if a relation ship can be established between them. The goal of this study was to find the ideal analytical equipment that would accurately determine S as either total S or SO 4 S The objective of this study was to compare analytical equipment used to determine soil S and SO 4 S for Florida sandy soils. Materi a ls and Methods Soil samples were collected on 16 May 2007 from tomato fields at G ulf Coast Research and Education Center in Balm, Florida Fifty two soil samples were randomly taken from within an 8.8 m linear pl ot of the raised bed ( 78 cm wide at base, 19.6 cm high an d 68.6 cm wide at the top, spaced 1.47 m apart ) between tomato plants and were approximately 15 to 20 cm deep of surface soil. After collection soil samples were air dried, sieved, and stored at 4. 4 C until analysis was conducted. Soil SO 4 S status was characterized by extractions with de ionized w ater, and 0.025 M KCl (Table 5 2 ) using a spectrophotometer (Spectronmic 100, Bausch & Lomb, Rochester, NY) on 16 Sept. 2009 at a commercial laboratory. The standard method for determining SO 4 in water by turbidimetry (AOAC OM 973.57, ASTM D 516 02) (AOAC, 2000; ASTM, 2006) was considered to be the standard reference method for SO 4 S determination and used for comparison with other methods at Waters Agri cultural Laboratory in Camilla, GA Total soil S status was characterized by de ionized water, and 0.025 M KCl, u sing ICP OES (Optima 2100 DV, Perkin Elmer, Waltham, MA) with Eschelle grading and under a constant vacuum (Table 5 2 ) at GCREC Analytical Lab in Balm, FL on 19, 20 and 21 Dec. 2008 8, 9 and 21 July 2009. The LECO ( LECO Corp., St. Joseph, MI) method with SO 2 measurement by infrared radiation (CNS 2000, LECO Corp. St. Joseph, MI )

PAGE 70

70 (Table 5 2 ) was conducted on 16 June 2009 at a MicroMacro Intern ational Laboratory in Athens, GA All data were an alyzed using a linear regression model (y ij i i x ij ij ) with test slope = 1 and test intercept = 0 statements with P<0.05 (SAS Institute, 2000). These parameters of the regression line were tested to determine if plotted regression lines were similar to the ideal relationship between analy tical instruments This ideal relationship is a regression line with a slope of 1 (1 rise to 1 run or 1:1) with an intercept of 0. The refore, if the test slope statement test H 0 i = 1 is not significant (P>0.05) then the slope is not significantly diff erent from 1 and if the test intercept = 0 statement test H 0 i = 0 is not significant (P>0.05) then the intercept is not significantly different from 0. This comparison against a 1:1 line is a similar relationship tested as Sikora et al. (2005) used for comparing methods for the determination of P The R 2 values have been reported to show the variability within the chosen regression model. Regression lines were plotted with significant data using Sigma Plot software program (Systat Software, 2008). R esults and Discussion The regression relationships between LECO total S and 0.025 M KCl SO 4 S and LECO total S and de ionized water total S were the only signif icant regression lines (P<0.01) The range in LECO total S values were 0.30 to 636.10 mg SO 4 kg 1 where as the range in 0.025 M KCl SO 4 S values were 12.41 to 276.99 mg S kg 1 The deviation between LECO total S and 0.025 M KCl SO 4 S was not normal ly distributed around 0 with a slight left skewediness (negative numbers) (data not shown). As both th e distribution and range in values depicts, 50% of the time LECO total S values were higher in concentration than 0.025 M KCl SO 4 S and the rest of time they were not

PAGE 71

71 higher. In addition, a large clumping of points between 0 and 100 mg SO 4 S kg 1 on the y axis represents points between 0 and 50 mg S kg 1 which indicates that LECO total S does not accurately predict turbidimetric SO 4 S. Further more, the calculated organic S content of the soil (total S SO 4 S) for samples extracted with 0.025M KCl are al l negative, except for four samples (data not shown). The regression line plotted for LECO total S and 0.025 M KCl SO 4 S was y = 73.2 + 0.29x with an R 2 of 0.26 (Figure 5 1 ). Both the test slope =1 and intercept = 0 were significant and therefore the te st hypotheses of H 0 i = 1, and H 0 i = 0 were rejected. This relationship does not meet the requirements of the ideal 1:1 relationship. Given that Florida sandy soils are low in organic matter (<2%), it is most likely that the relative insensitivity (i .e. interferences occurring) of the turbidimetric method to accurately determine SO 4 S. This interference could be from various other ions extracted, co precipitated, and present in the solution matrix as The second significant relationship between two methods was for LECO and de ionized water SO 4 S (Figure 5 2 ). This relationship has two major clusters close to 0 mgkg 1 The first cluster is between 0 and 50 mg S kg 1 total S and 0 and 100 mg SO 4 kg 1 de ionized water SO 4 S The second cluster is above the first cluster, located b etween 0 and 100 mg Skg 1 total S and 120 to 200 mg SO 4 kg 1 de ionized water SO 4 S The range in LECO values were 0.30 to 636.1 mg S kg 1 The range in de ionized water SO 4 S were 8.79 to 232.65 m g SO 4 S kg 1 In addition, the deviation between LECO and de ionized water SO 4 S was found to be not normally distributed around zero and skewed to the left (data not presented). The calculated organic S for de ionized water SO 4 S for all samples was ne gative, except for two (data not shown). This

PAGE 72

72 demonstrates that turbidimetric SO 4 S is not an accurate method for determining SO 4 S and should not be used without modifications. The regression line plotted between was y = 88.6 + 0.17x with an R 2 = 0.13. Both the test slope = 1 and test intercept = 0 were significant and the test hypotheses of H 0 : i = 1, and H 0 i = 0 were rejected. This means that both the slope and intercept are different from the ideal 1:1 relationship tested. Regardless of extraction solution, turbidimetric determination of SO 4 S appears to not accurately reflect SO 4 withi n the solution Interfering ions, co precipitation of other insoluble products and occlusions (Kolthoff and Sandell, 1952) during the extraction process and determination of SO 4 S are possible reason for this insensitivity (Table 5 3) It has been state d by Beaton et al. (1968) that interferences can occur with high concentrations of Na, K, Ca, Mg, PO 4 NO 3 and SiO 2 in solution when determining turbidimetric SO 4 S The 0.025 M KCl solution should be well below the threshold of interference at 0.013 M K. Also, Cl is below the 5000 mgL 1 threshold (Table 5 4) for causing negative interference at 0.012 M Cl. Calcium, PO 4 or SiO 2 could be interfering with the determination of SO 4 by forming other Ba salts (Table 5 4) Florida sandy soils tend to be high in Ca, but also are categorized as phosphatic soils. Yet, the dissolution of PO 4 from soil colloids is highly unlikely given the pH requirement for any of the ortho phosphates to be in solution (pH <4). Plus, polyphosphates as low as 1 mg L 1 will inhib it BaSO 4 precipitation causing a negative interference (Table 5 4) (ASTM, 2006). Nitrate and SiO 2 are also possible for interference when determining SO 4 turbimetircally, but NO 3 is readily leached from sandy soils as it is an anion. Silicates can possib ly interfere positively by co precipitating with BaSO 4 ion as it has a relatively low threshold,

PAGE 73

73 500 mgL 1 (Table 5 4) (ASTM, 2006). However, it is not known how much silicates were present in the extraction solutions for turbidimetric procedures in this study. It also possible for organic matter colloids interfere via co precipitation. This is often resolved by using cation exchange resins in conjunction with turbidimetric determination of SO 4 or by filtering, and treating the solution with activated carbon. These measures often add time and labor to the determination of SO 4 S by turbidimetric procedures. Both relationships between LECO total S and ICP total S were not significant. For the relationship plotted between LECO and 0.025 M KCl total S h ad only one large cluster of points between 0 and 100 mg S kg 1 of LECO and 0 and 40 mg S kg 1 0.025 M KCl ICP total S (Figure 5 3). Also the range in LECO was 0.30 to 636.1 mg S kg 1 and the range in 0.025 M KCl ICP total S was 0 to 173 mg S kg 1 For a large change in LECO the corresponding change in 0.025 M KCl ICP total S is small. The ICP procedure is not as precise compared to the LECO procedure. Therefore, it is not surprising that the deviation between LECO and 0.025 M KCl ICP total S was not n ormally distributed around zero and bimodal (data not shown). The regression relationship found was y = 25.9 0.02x with an R 2 of 0.007. As this regression relationship was not significant (P= 0.55) and the random error explaining most of the model plo tted (small R 2 ). Both the test slope = 1 and test intercept = 0 were significant and the test hypotheses of H 0 i = 1, and H 0 i = 0 were rejected at P<0.05. This relationship does not meet the requirements for the ideal 1:1 as both slope was not equal to 1 and intercept is not 0. For the plotted relationship between LECO and de ionized water ICP total S, a l arge cluster was located between 0 and 150 mg S kg 1 LECO total S and 0 and 20 mg

PAGE 74

74 S kg 1 de io nized water ICP total S This cluster is only a small portion of the total S from LECO The range in LECO total S was 0 to 636.1 mg S kg 1 where as de ionized water ICP total S ranged from 0 to 63.27 mg S kg 1 The regression line plotted for this data was y = 12.3 + 0.02x with an R 2 of 0.04 (P = 0.17). Only the test slope = 1 statement was significant, meaning i meaning i = 0 Therefore, this comparison between LECO total S and de ionized water relationship, but it met one of th e two requirements. Regardless of meeting one of the requirements, de ionized water ICP total S would not be recommended, because the R 2 value is small and random error explains a large extent of the regression model. An issue with ICP determination of S is that Ca may cause spectral interference of S when using inductively coupled plasma adsorption emission spectroscopy (ICP AES) (Zhao and McGrath, 1994). In Florida, sandy soils can be high in Ca (Simonne and Hochmuth, 2009) and previous research has f ound levels >400 mg Ca kg 1 in Florida Spodosols (Esmel, 2005). There are two suggested methods for correcting spectral interferences of Ca on S when using ICP OES: 1) calibrating with pure S standards, running a series of Ca standards as samples and fit ting a regression line using S concentration as a function of Ca concentration and 2) inter element correction method (IEC) (Kola et al., 2002). For IEC, the interfering element is measured (in this case Ca) at another wavelength and applying a predetermi ned correction factor to the results (Boss and Fredeen, 1997). Both of these correction methods require additional labor and time to correct the interference, but these methods do not correct matrix Ca interferences (Kola, et al., 2002). Without correcti on erroneous results will occur when

PAGE 75

75 high levels of Ca are present within the aqueous sample. It has been described by Knauthe et al. (2000) and supported by Knauthe and Otto (2001) that a weighted compromise conditions be used when determining S using I CP OES to maximize operating conditions. In addition, Kola et al. (2002) use d multiple linear regressions to correct for matrix interferences of Ca on S determination by ICP OES. This method reduces labor and time by eliminating 6 of the original 12 cali bration standards used to correct for Ca matrix interferences with an improvement in accuracy. Therefore, if no adjustments for Ca are conducted by any of the above s uggested means, the results should be similar to this study. If a choice is possible be tween all the analytical equipment used in this study, Ca levels within samples are suspected to be high and it is unknown if corrections for spectral or matrix interferences B ased upon the results from this study it is suggested that LECO dry combusti on total S be used for the determination of S for all samples. For this method, the possibility of interfering ions, co precipitation of other Ba salts or organic matter, and spectral interferences are virtually eliminated.

PAGE 76

76 Table 5 1. Advantages and disadvantages of selected analytical instruments for determining S or SO 4 S. Instrument ICP OES Elemental Analyzer Turbidimeter Cost (per sample) < $10 $20 $12 Speed (min./sample) 2 to 3 min. 4 min. 4 min. Precise/accurate Yes Yes No y Multi ple element Yes Yes No Hazardous waste No No Yes z y without modifications z Barium salts are classified as a health hazard

PAGE 77

77 Table 5 2 Instrumental details for comparison between sulfur and sulfate determination for selected samples obtained at the Gulf Coast Resea rch and Education Center in Balm, FL Analytical Equipment Specification and Settings ICP OES Instrumentation Perkin Elmer Optima 2100 DV Generator 750 to 1500 W Wavelength 181.975 nm Readout mgL 1 Frequency 40 MHz Forward power RF 1350 W Nebuliz er Gemtip cross flow Internal standard Yttrium Sample volume 10 mL Auxiliary coolant Liquid Argon Auxiliary coolant flow rate 15 Lmin 1 Auxiliary gas 0.2 Lmin 1 Liquid uptake rate 1.5 mLmin 1 Carrier gas rate 0.85 Lmin 1 Grating Eschelle Calib ration curve 5 point linear, 0, 0.1, 1.0, 10, and 50 mgL 1 S Optical view Axial Dilution ratio 1:10 Dry combustion S Instrumentation LECO CNS 2000 Wavelength Infrared Radiation Catalyst none Carrier gas O 2 Reaction time 4 to 6 min Standard r eference material Orchard Turbidimetric Spectrophotometer Bausch Lomb Spectromic 100 Wavelength 420 nm Conditioning reagent Glycerol, HCl, H 2 O, 95% ethyl alcohol, and NaCl Reagent BaCl Sample volume 15 mL Dilution ratio 1:7 Standard reference N a 2 SO 4 at 100 mgL 1 Reaction time 4 min

PAGE 78

78 Table 5 3 Salts and their solubility as potential interfering ions when determining SO 4 S by turbidmetric methods. Salt Solubility ( mgL 1 ) z Cold Water Hot Water Ba(NO 3 ) 2 8.7 X 10 24 3 42 000 X 10 10 0 BaPO 3 sl s -BaHPO 4 100 to 200 -Ba(H 2 PO 4 ) 2 d d Ba 3 (PO 4 ) 2 i i Ba 2 P 4 O 7 100 sl s Ba 3.75 X10 31 5 .9 X 10 105 BaSO 4 2.66 4.13 Ca(CO 3 ) 2 1.4 X 10 26 1.8 X 10 76 Ca(H 2 PO 4 ) 2 1.8 X 10 34 d CaSO 4 2H 2 O (Gypsum) 2.41 X 10 3 sl s CaSO 4 H 2 O (Plaster of Paris) sl s sl s CaSO 4 6 94 X 10 2 2.09X10 33 1 619 X 10 100 CaNO 3 1 .212 X 10 24 376 0000 X 10 10 0 CaCl 2 7.45 X 10 25 159 0000 X 10 100 KNO 3 2.8 X 10 6 413 0000 X 10 100 K 2 SO 4 3.63 X 10 5 121 6 000 X 10 100 MgSO 4 7H 2 O (Espsom salt) 1.0 X 10 37 2.0 X 10 80 z notation i = insoluble, sl = slightly soluble, s = soluble, d = decomposes

PAGE 79

79 Table 5 4 Potential int erfering ions for determining SO 4 S by turbidmetric methods, their reported thresholds in solution and reference Interfering ion threshold (mgL 1 ) Reference AsO 4 > 10 ppm Iwasaki et al., 1957 Na+ 200 ppm (no interference) van Staden and Taljaard, 1996 K + > 5000 ppm 200 ppm (no interference) Iwasaki et al., 1957 van Staden and Taljaard, 1996 Ca +2 > 100 ppm 50 ppm Baban et al., 1980 van Staden and Taljaard, 1996 Cl > 5000 mg/L 200 ppm ASTM, 2006 van Staden and Taljaard, 1996 CO 3 2 & HCO 3 > 100 ppm > 1000 mgL 1 Iwasaki et al, 1957 Morais et al., 2003 Mg +2 no interference 100 ppm Morais et al., 2003 van Staden and Taljaard, 1996 MoO 4 > 10 ppm Iwasaki et al., 1957 Mn +2 10 ppm van Staden and Taljaard, 1996 NH 4 + > 5000 ppm Iwasaki et al., 1957 NO 3 > 50ppm > 100 mg/L Iwasaki et al., 1957 Reisman et al., 2007 PO 4 H 3 PO 4 1 > 25 ppm ASTM, 2006 Dasgupta et al., 1978 S 2 O 3 > 10 ppm Iwasaki et al., 1957 SeO 4 > 10 ppm Iwasaki et al., 1957 SiO 2 > 500 mg/L ASTM, 2006 Zn +2 200 ppm van Staden and Taljaard, 1 996

PAGE 80

80 Figure 5 1. Comparison between dry combustion total S and 0.025 M KCl turbidimetric SO 4 S for selected soil samples from fields at the Gulf Coast Research and Education Center in Balm, FL

PAGE 81

81 Figure 5 2. Comparison between dry combustion procedure for total S and de ionized water turbidimetric SO 4 S for selected soil samples from the fields at the Gulf Coast Research and Education Center in Balm, FL.

PAGE 82

82 Figure 5 3. Comparison between dry combustion procedure for total S and 0.025 M KCl I nductively Coupled Plasma total S for selected soil samples obtained from the fields at the Gulf Coast Research and Education Center in Balm, FL

PAGE 83

83 Figure 5 4. Comparison a nd regression analysis of dry combustion procedure for total S and Inductively Coupled Plasma de ionized water total S for soil samples obtained from the fields at the Gulf Coast Research and Educat ion Center in Balm, FL

PAGE 84

84 CHAPTER 6 COMPARISON OF EXTRAC TANTS FOR DETERMININ G S ULFUR IN SOIL AND PLANT TISSUE Introduction Assessing relationships between a newly proposed soil or plant extractant and a standard extractant is a first step technique in the evaluation a new extractant for the development of a soil testing program (Van Erp and Van Beusichem, 1998). This useful first step technique seems to be lacking for soil S and SO 4 extractants in many soil testing programs that have been developed. Many studies have been conducted comparing determination methods of S (Ajwa and Tabatabai, 1993; Crosland et al. 2001; Palomino et al., 2005; Reisman et al., 2007; Soon et al., 1996), but only a few separate extraction solution comparison from the determination step. Some of the most widely used solutions to extract soil S to predict plant availability and fertilizer applications are water, CaCl 2 and PO 4 solutions (Anderson et al., 1992; Tan et al., 1994a). Water extractable S is primarily soil S readily leach ed out of the soil and is mostly likely in the inorganic SO 4 S form. Several solutions are often used and s include Na, Ca or LiCl solutions (Palomino et al., 2005), because it is accepted that these soluti ons extract the soluble fracti on of SO 4 in soils (Santoso et al., 1995) as the solutions contain the chloride ion which can easily replace SO 4 ion in solution chemistry. It is believed that the difference between water extractable S and a phosphate soluti on is related to the amoun t of SO 4 adsorbed by the soil, but this does not always hold true (Santoso et al., 1995). Often and depending upon soil organic matter and exchange sites, water extractable S as H 2 O can significantly dissolute more SO 4 S than sal t solutions (Maynard et al., 1987; Santoso et al., 1995; Tan et al., 1994a ) or in some cases similar amounts (Anderson et

PAGE 85

85 al., 1998; Tan et al., 1994b; Zhao and McGrath, 1994). It may be possible just to use H 2 O as an extracting solution instead of a sal t solution if water soluble SO 4 was desired. Many studies employ Ca(H 2 PO 4 ) 2 extraction solu tion at 0.01M or 500 mg P L 1 This is one of the recommended soil SO 4 S extraction solutions for northeastern US (Singh et al., 1995) and extracts more soil SO 4 S than water via dissolution of SO 4 S off exchange sites on soil particles. The Ca(H 2 PO 4 ) 2 solution can cause interferences when using turbidimetric measurement of SO 4 S when 0.04 M solution is used for some soils (Searle, 1998). The other issue with Ca based extraction solutions is that Ca may cause spectral interference of S when usi ng ICP AES (Zhao and McGrath, 1994). In addition, various studies using Ca(H 2 PO 4 ) 2 have had variable results for extracted S correlating significantly to other parameters ( Anderson et al., 1998a; Anderson et al., 1998b; Blair et al., 1991; Blair et al., 1993; Chinoim et al., 1997; Palomino et al., 2005; Pandey and Girish, 2006; Sp encer and Glendinning, 1980; Ven drell et al., 1990). In response to the variability found with e xtracting S with Ca(H 2 PO 4 ) 2 and to the poor performance of soil S testing with Ca(H 2 PO 4 ) 2 to estimate soil SO 4 supplying cap acity from the organic S pool, Blair et al. (1991) developed a soil extraction solution of 0.25 M KCl as a potential replacement to Ca(H 2 PO 4 ) 2 extraction solution. Potassium chloride was originally investigated by Gianello and Bremner (1986) to measure inorganic N and potential mineralized N. Yet, concerns raised for this potential replacement have been centered on the parameters inv olved in the methodology. This method has a long incubation time (3h ) and requires the maintenance of 40 C for this time period while shaking. Therefore, it is not surprising that modifications to improve time and detection limits of this extraction sol ution have been made. Bloem et al.

PAGE 86

86 (2002) examined six modifications of the 0.25 M KCl extracting solution for SO 4 S. This study found that using 0.025 M KCl, 1 to 5 ratio of soil to extra cting solution, shaking for 3 h at room temperature was the fastes t, precise method for SO 4 S extraction of agricultural soils (Bloem et al., 2002). While this optimizes soil testing for SO 4 S alone, another possibility is the inclusion of soil SO 4 and S determination in extraction solutions already accepted by laborato ries. An example of a widely accepted soil extracting solution is the Mehlich 3. This extracting solution excludes H 2 SO 4 which is a component of the Mehlich 1 extraction solution Mehlich 3 extractant was designed to replace the Mehlich 1 and Bray 1 P extractants (Jones, 1998) and is suitable for use on neutral to acidic soils. The benefit of using an extracting solution such as Mehlich 3 is that it allows for S to be determined as part of a multi ple element (including micronutrients) extraction step This saves time, labor, and money, because separate chemicals and analysis do not need to be performed as has been described in a diagram by Jones et al. (1991). Few studies have examined Mehlich 3 for S extraction in soils, because the general belief that S is analyzed separately from all other nutrients. This is either due to the determination of SO 4 S by turibidimetry or ion chromatography instead of using ICP technology. Rao and Sharma (1997) found strong relationships between Mehlich 3 and other extracting solutions commonly used to determine SO 4 S in soils. This was the first reported effort of using Mehlich 3 as an extractant for S in soils. Zbral (1999) found Mehlich 3 extraction correlated well to water extraction of SO 4 S. Also Zbral an d element determination in Czech agricultural soils. Further support of Mehlich 3 being used to include S in its

PAGE 87

87 multiple element extraction was showed by Pandey and Girish (2006). Yet, this stu dy did not compare ability of each extracting solution to dissolute SO 4 S or S from the soil samples. Instead, the study focused on correlation between extractants chosen and various plant growth and total S within tissue as a way to calibrate the soil te st. Total S in plant tissues are most commonly quantified by two methods: dry combustion (also known as dry ash procedure) or wet oxidation (also known as acid digestion). While each has their own advantages and disadvantages, the focus here is on which ones commonly used in S research and their suitability for S determination. For dry combustion, two procedures are often used. The first is the Association of Analytical Chemists (AOAC) Officia l Methods of Analysis for S in plants in which organic matte r can be destroyed by two different chemicals, Na 2 O 2 or MgNO 3 ( AOAC OM 920.10 or AOAC OM 923.01) (AOAC, 2000) The AOAC Official Methods of Analysis are the reference methods in the analytical chemistry domain, but often these methods are not used by res earch scientists in agricultural fields. This is primarily due to the chemical hazards, time, and labor involved in AOAC Official Methods of Analysis. The other dry combustion technique for total S in plant tissues is an automated p rocedure commonly ter med LECO This procedure has two steps within on e piece of equipment: organic matter destruction and total S determination step. The LECO procedure combusts samples at temperatures up to 1400 C and measures S as SO 2 evolved by titration or infra red cel l. In some cases, an accelerator or catalyst WO 3 or V 2 O 5 is added to the sample. Kowalenko (2001) suggested that WO 3 be added when simultaneously measuring C, N, and S with a LECO automated dry combustion instrument. Whether or not a catalyst was used ( Etheridge et al., 1998) or one case the

PAGE 88

88 type of catalysis used (David et al., 1989) h ave not been reported when LECO automated dry combustion instrument is used for total S. While this may seem just a minor error, the lack of this information creates diff iculties in repeating the analytical research. In contrast the automated dry ash procedure, wet oxidation for the extraction of S from plant tissues is time consuming. The most common wet oxidation proc edure for total S is HNO 3 with HClO 4 The number o ne difficulty with acid mixture for wet oxidation is the tendency of hot HClO 4 in t he presence of easily oxidized organic matter to be explosive (Jones, 1985). This in conjunction with the need for a special hood when using HClO 4 rules it out for many lab oratories when choosing a wet oxidation method. Therefore, it is not surprising that many studies have been conducted on comparing acid mixtures which include or exclude HClO 4 for the suitability as a matrix for metals and non metals (Havlin and Soltanpou r, 1980; Hoenig et al., 1998; Huang et al., 2004; Ippolito and Barbarick, 2000; Oliva et al., 2003; Pyki et al., 2000; Soon et al., 1996). Two studies in particular (Pyki et al., 2000; Soon et al., 1996) compared methods for the determination of total sulfur in plant tissues. Pyki et al., (2000) compared five different wet oxidation methods and dry combustion for total S and six other elements (Al, Ca, K, Mg, Mn, and Zn) against the certified value. This study found that out of all wet oxidation me thods HNO 3 + H 2 O 2 or HNO 3 alone were the most suitable for S determination using ICP. Also the HNO 3 + HClO 4 digestion procedure in almost every case gave the lowest results compared to other acid digestion procedures (Pyki et al., 2000). The authors s tated that the combustion technique combined with infrared detection (also known as LECO ) provide d relative good reproducibility for sulfur determination. In contrast, Soon

PAGE 89

89 et al. (1996) compared six total S methods with reference values. These methods w ere dry ashing with NaHCO 3 + Ag 2 O, oxygen flask combustion with determination on two differ ent analytical equipment, LECO microwave digestion with HNO 3 + HClO 4 and microwave digestion with HNO 3 + H 2 O 2 + HCl. B oth dry ashing with NaHCO 3 + Ag 2 O, and oxyge n flask combustion can be time consuming for regular laboratory use when compared to microwave digestion. In addition, precipitation of AgCl 2 may occur in the case of dry ashing as well as poor recovery of S for certain plant tissues. Microwave digestion while rapid, has a high initial cost with the investment of the equipment and may not be in use d in commercial laboratories. Despite these concerns, Soon et al. (1996) suggested that HNO 3 + H 2 O 2 + HCl was suitable for multi element analysis and that LEC O gave equivalent total S determination as this method. Currently, there was no soil test in Florida for determining recommended applications of S for Florida vegetable production. The recommendation for an S application is empirical based and uses S ra nges within plant tissues. Previous research on tomato and S applications in Florida have found positive results in yield when plant tissue S are within and outside of the current S recommended range. Therefore, the goal of this study was to find an extr actant for soil S or SO 4 in Florida sandy soils to base recommended S applications, and a total S plant tissue method that represents total S in tomato tissue. The objectives of this study were to compare soil extractants (de ionized water, 0.025 M KCl, a nd Mehlich 3) for soil S and SO 4 compare plant analysis for total S, and determine the relationship between each extractant and total plant S digestion method.

PAGE 90

90 Materials and Methods Florida has a humid sub tropical climate with an average temperature high of 32 C, minimum of 0 C, and average temperature range of 16 to 28 C. Average annual rainfall is typically in the range of 130 to 150 cm. Florida tomato production occurs predominate ly in the soil order of Spodosol, which has an impermeable spodic h orizon capable of perching water table for subsurface irrigation. The soil types at the location tested were Myakka fine sand (sandy siliceous hyperthermic Aeric Haplaquods), and Zolfo fine sand (sandy siliceous hyperthermic Grossarenic Entic Haplohumods) The land was located at the Gulf Coast Research and Education Center (GCREC) IFAS, University of Florida, in Hillsborough County, Balm, FL. This land was a former citrus grove renovated to vegetable production in May 2005. Field renovations began i n October 2003 when the citrus trees were pushed over, piled up, and burned. The ashes and remnants of the burned citrus trees were then spread out over the fields. All fields were left in fallow until January 2005 in which some were put into vegetable a nd ornamental plant production. Vegetable production uses row and raised bed culture. These rows were orientated directly perpendicular to the orientation of the previous citrus grove rows in each field. T he t wo areas of land which were utilized for th is study are identified as R3 and R4 (Figure 6 1). These areas were chosen because of their individual land history differences after renovation and possible potential for differences in soil S. In addition, these areas were also both in tomato producti on at the time of sample collection. For the R3 parcel, okra ( Abelmoschus esculentus L. (Moench.)) 15 0 30 (N P 2 O 5 K 2 O ) dry fertilizer was applied as a broadcast rate of 112 kg N ha 1 After cover

PAGE 91

91 crop destruction, the R3 parcel was disked three times, harrowed, packed, leveled and irrigation ditches pulled before bed preparation. At raised bed preparation, the field was subsurface irrigated two weeks in advanc e of pre bedding, po lyethylene mulch, and fumigant applications. Raised beds (78 cm wide at base, 19.6 cm high, and 68.6 cm wide at the top) spaced 1.47 m apart were formed and fumigated with 1,3 dichloropropene (372 Lha 1 ), methyl bromide:chloropicrin ( 66:33, 393 kgha 1 ) or chloropicrin alone (168 kgha 1 ) as a shank application, or one of various fosthiazate rates (O ethyl S (1 methylpropyl)(2 oxo 3 thiazolidinyl)phosphonothioate) (0, 2.3, 3.5, and 5.6 Lha 1 ) with or with out metam sodium (701 Lha 1 ) as a drip application on 16 Feb. 2006. Land where cucurbits were grown were treated with propenal (449 kgha 1 ), propylene oxide (92 Lha 1 ) applied through the drip tape on 6 Mar. 2006 and methyl bromide: chloropicrin (66:33, 393 kgha 1 ) as a shank app lications on 16 Feb. 2006. variety was variety was direct seeded into raised beds on 21 Mar. 2006. Subsurface irrigation was utilized for tomato plant establishment (2 wk) and drip irrigation was used for the remainder of the season (12 wk) starting on 21 Mar. 2006 at (T Tape, 1.7 L min 1 340 m 1 1 tube) 7,358 Lha 1 per d A pre plant broadcast application of fertilizer at 45 kg N ha 1 was applied using 4 8 4 (N P 2 O 5 K 2 O) Banded applications of fertilizer 10 0 20 (N P 2 O 5 K 2 O ) at a rate of 336 kg N ha 1 were made. Okra was seeded as a cover crop on 18 Apr. 2006 into R3 in areas were tomato and cucurbits not being grown during spring season (approximately 2.5 ha). On 8 June 2006, okra was fertilized with 15 0 30 (N P 2 O 5 K 2 O ) at a rate of 89.8 kg N ha 1 In December 2006, the okra cover crop was incorporated into the soil.

PAGE 92

92 In January 2007, R3 was prepared for tomato production as described above. On 23 Feb. 2007, fertilizer was banded into the raised beds before fumigation. Raised beds were fumigated with methyl bromide plus chloropicrin (67:33 v/v) at a rate of 196 kgha 1 to eliminate soilborne diseases, nematodes and weeds. Fertilizer rates applied in the bands were 0, 186 or 279 kg N ha 1 270 or 404 kg K ha 1 and 0, 213, or 319 kg S ha 1 single rows into raised beds. The irrigation system was a hybrid system which included subsurface irrigati on and drip irrigation. Subsurface was applied at the approximate rate of 21,035 Lha 1 per d Drip irrigation tubing (T Tape, 1.7 L min 1 340 m 1 1 tube) was buried to a depth of 2.5 cm in the center of the raised bed. The irrigation rate using drip s ystem was at an approximate rate of 39,095 Lha 1 per d Therefore, the total water applied via irrigation for the tomato crop in Spring 2007 was 60,130 Lha 1 per d For the R4 parcel the land was in fallow (no fumigation, tillage, cover crop, fertili zer or irrigation applications) until January 2006. In 2006, R4 was tilled, harrowed, leveled and allowed to go fallow until January 2007. In January 2007, R4 was prepared as described above for tomato production. The soil was treated with and without fu migation prior to vegetable production (0 and 257 kgha 1 with fumigation) on Feb. was applied via fertigation at 0, 224, or 336 kg N ha 1 Drip irrigation was utilized in R3 for tomato crop production. Drip tubing ( T Tape, 1.7 L min 1 340 m 1 1 tube ) was buried to a depth of 2.5 cm in the center of the raised bed. The irrigation rate using drip system was at an approximate rate of 39,095 Lha 1 per d

PAGE 93

93 Soil and plant tissue samples were collected on 16 May 2007 from the described fields at GCREC. The area of tomato fields w h ere both soil and tomato plant tissue samples were marked so as to be clearly identified for sampling. Fifty two soil samples were randomly take n from within an 8.8 m linear plot of the raised bed, between tomato plants and were approximately 15 to 20 cm of surface soil. After collection soil samples were air dried, sieved, and stored at 4.4 C until analysis was conducted. Tomato plant tissue s amples were randomly collected within the same 8.8 m linear plot to be paired exactly with soil samples. Each tomato plant tissue sample consisted of 10 recently matured leaves adjacent to an inflorescence as described by Mills and Jones (1991). Tomato p lant tissue samples were dried at 43 C, ground with a Wiley mill to pass through 0.853 mm sieve aperture (20 mesh screen) and stored at 0 C until analysis was conducted. Paired sampling was used to establish the strongest relationship possible between t wo variables. Soil SO 4 S status was characterized by extractions with de ionized water, and 0.025 M KCl (Table 5 1) using a spectrophotometer (Spectronmic 100, Bausch & Lomb, Roche ster, NY) on 16 Sept. 2009 at Waters Agricultural Laboratory in Camilla, GA The standard method for determining SO 4 in water by turbidimetry (AOAC OM 973.57, ASTM D 516 02) (AOAC, 2000; ASTM, 2006) was considered to be the standard reference method for SO 4 S determination (ASTM D 516 02) and used for comparison with other method s. Total soil S status was characterized by de ionized water, 0.025 M KCl, and Mehlich 3 using ICP OES with Eschelle grading (Optima 2100 DV, Perkin Elmer, Waltham, MA) at GCREC Analytical Lab in Balm, FL on 19, 20 and 21 Dec. 2008, 8, 9

PAGE 94

94 and 21 July 2009 Mehlich 1 is the standard calibrated soil extractant for mineral soils in Florida. For this study, Mehlich 3 was chosen over Mehlich 1 as an extractant because Mehlich 3 excludes H 2 SO 4 which is a component of Mehlich 1 The H 2 SO 4 included in the Mehl ich 1 extracting solution could artificially and non uniformly increase S levels during extraction. In addition Mehlich 3 has been found to improved estimation of P, Cu, Mn, and B in Florida sandy soils over that of Mehlich 1 (Mylavarapu et al., 2002). LECO total S (CNS 2000, LECO, St. Joseph, MI) with SO 2 measurement by infrared radiation was conducted on 16 June 2009 at MicroMacro International Laboratory in Athens, GA Plant tissue samples were digested with HNO 3 and 30% H 2 O 2 ( HNO 3 + H 2 O 2 ) as describe d by Mills and Jones (1991). One gram of dried plant material was weighed directly into a 100 mL glass test tube. Eight milliliters of concentrated nitric acid were added, covered, and left overnight to predigest. Tubes were then placed into an aluminum heating block capable of holding 50 tubes at 120 C and held for 1 h at this temperature and allowed to cool. Eight milliliters of 30% H 2 O 2 were then added to each tube and the block was re heated to 95 C. Solutions were diluted to 20 mL with nano pure water after cooling to room temperature (approximately 22 C) and filtered with Whatman No. 44 filter paper (20 25 m particle retention) Only soil samples extracted with Mehlich 3 and tissue samples digested with HNO 3 plus 30% H 2 O 2 were diluted to 1:10 ratio for ICP OES to prevent salting of the plasma torch (which reduces quality control and assurance). All data were converted from % or mgL 1 to mgkg 1 for data analysis.

PAGE 95

95 All data were analyzed using SAS system using proc reg and linear regression hypothesis is false (P<0.05) then the slope is significantly different from 1. The test intercept is significantly different from 0. These parameters of the regression line were tested to determine regression lines were similar to the ideal r elationship between two soil extractants. This ideal relationship is a regression line with a slope of 1 (1 rise to 1 run or 1:1) with an intercept of 0. Therefore, for every incremental increase in one variable there is the same corresponding incrementa l increase in the second variable (i.e. 1 to 1) This comparison against a 1:1 line is a similar relationship tested as Sikora et al. (2005) used for comparing methods for the determination of P The R 2 values have been reported to show the variability. Regression lines were plotted with significant data using Sigma Plot software program (Systat Software, 2008). Results and Discussion The regression relationship between the standard reference method (SRM) for soil sulfate and modified Blair (0.025 M KC l) for soil sulfate was the only significant regression line (P<0.05) (Figure 6 2). The regression model found was y = 51.6 + 0.53x with a R 2 = 0.38. The test statistics for test slope = 1 and test intercept = 0 were both significant (P<0.05). The devia tion between SRM and 0.025 M KCl was plotted and found to be normally distributed around 0 (data not shown). Despite, the wide range of x values (12.41 to 276.99 mg SO 4 kg 1 ), y values (8.79 to 232.65 mg SO 4 kg 1 ), lack of large clustering of points; this relationship is not similar to the ideal 1:1 relationship and

PAGE 96

96 interpretation of results would have to occur if 0.025 M KCl was used in replacement of the SRM for soil SO 4 The regression relationship between de ionized water and Mehlich 3 for soil S was not significant (P=0.65) (Figure 6 3). The regression model found was y = 11.8 + 0.09x with a R 2 = 0.004. The test statistics for test slope = 1 and test intercept = 0 were both significant (P<0.05). For this relationship, the range of x was 12.02 to 6 0.99 mg S kg 1 and the range in y was 0 to 63.27 mg S kg 1 Therefore, it is not surprising that a large cluster is centered around 20 mg S kg 1 The deviation between de ionized water and Mehlich 3 for soil S was plotted and found not to be normally dis tributed or centered around 0 (data not shown). This relationship does not fit the ideal relationship between two soil extraction methods and meets none of the statistical requirements for use (i.e. slope = 1, intercept = 0, R 2 >0.70). These results are in direct contrast to Zbral (1999) in which water extraction and Mehlich 3 were found to have a significant linear relationship for total S (y = 6.94 + 0.742x, R 2 = 0.62). This study also found that 70% of soil S extracted was in the sulfate form. The regression relationship between 0.025 M KCl and Mehlich 3 for soil S was not significant (P=0.38) (Figure 6 4). The regression model was found to be y = 11.2 + 0.47x with a R 2 = 0.01. The test statistics for test slope =1 and test intercept = 0 were sign ificant (P<0.05). For this relationship, the range in x was 12.02 to 60.99 mg S kg 1 and the range in y was 0 to 173 mg S kg 1 It is not surprising based upon these ranges that a cluster exists between 20 and 40 mg S kg 1 for Mehlich 3. The deviation b etween modified Blair and Mehlich 3 was normally disturbed but not centered around 0 (data not shown). This relationship also does not fit the ideal relationship between two soil

PAGE 97

97 extraction methods and meets none of the statistical requirements for use (i .e. slope = 1, intercept = 0, R 2 >0.70). The regression relationship between total plant tissue S by wet digestion with HNO 3 + 30% H 2 O 2 and tota l plant tissue S by LECO was not significant (P=0.19) (Figure 6 5). The regression model was to be y = 3637.8 + 0.04x with a R 2 = 0.03. The test statistics for test slope = 1 and test intercept = 0 were significant (P<0.05). For this relationship, the range in x was 5274 to 11896 mg S kg 1 and the range in y was 3261 to 4965 mg S kg 1 Based upon this relations hip there is a large change in the value for HNO 3 + H 2 O 2 plant tissue S for a small change in LECO plant tissue S. Even though this relationship is not significant and does not meet the ideal statistical requirements (i.e. slope = 1, intercept = 0, R 2 >0.7 0). This comparison between total plant tissue S methods has a deviation relationship. The HNO 3 + H 2 O 2 plant tissue S was systematically higher than the LECO plant tissue S. This was evident by the plotting of the deviation between HNO 3 + H 2 O 2 plant ti ssue S and LECO plant tissue S (Figure 6 6). The deviation is normally disturbed, but not centered around 0. These differences may be due to the inherent errors with in the ICP OES or LECO combustion S because the CV was 9 for this comparison and intern al quality control and assurance was met for each extraction method. When comparing this study results to a study conducted Pyki et al. (2000) on comparison of dissolution methods. Pyki et al. (2000) had two out of five S analyses higher in S concent ration for HNO 3 + H 2 O 2 than LECO. One of these two was an unknown S content leaf sample. It becomes apparent that our results are possibly unique since all of the samples with wet oxidation by HNO 3 + H 2 O 2 had higher S concentrations.

PAGE 98

98 These results sugge st ed that the predictability of any soil extractant with tomato plant tissue S or yield is low, thereby rendering a soil test based on these extractions unpractical. While HNO 3 + H 2 O 2 consistently had higher S concentrations than LECO without further info rmation within this study on spectral interferences, the true reason can not be elucidated. Hence, the current Florida recommendation for S using plant analysis should continue to be utilized, regardless of current S responses being shown with the additio n of S to the fertilizer regime. It is also recommended that total S in plant tissues be determined by LECO as this is the absolute total S within plant tissues.

PAGE 99

99 Figure 6 1. Diagram of Gulf Coast Research and Education Ce nter buildings, field layout, and nomenclature system for identifying production fields

PAGE 100

100 Figure 6 2. Comparison of Modified Blair SO 4 S with Standard Reference (H 2 O) SO 4 S for selected soil samples from the Gulf Coa st Research and Education Center in Balm, FL.

PAGE 101

101 Figure 6 3. Comparison of Mehlich 3 S with de ionized water S for selected soil samples from the Gulf Coast Research and Education Center in Balm, FL.

PAGE 102

102 Figure 6 4. Comparison of Mehlich 3 S with Modified Blair S for selected soil samples collected at the Gulf Coast Research and Education Center in Balm, FL.

PAGE 103

103 Figure 6 5. Comparison of t otal plant t issue S by wet oxidation and by dry combustion total S procedure for selected soil samples collected at the Gulf Coast Research and Education Center in Balm, FL.

PAGE 104

104 Figure 6 6. Distribution of the deviati on between total plant tis sue S by wet oxidation and dry combustion total S for selected soil samples collected at the Gulf Coast Research and Education Center in Balm, FL

PAGE 105

105 CHAPTER 7 CALIBRATION AND DIAG NOSTIC TOOLS OF SOIL AND TOMATO TISSUE S ULFUR FOR P RODUCTION ON FLORIDA SANDY SOILS Introduction The results of soil tests can be used to diagnose and guide the correction of nutrient deficiencies, to diagnose and avoid nutrient and non nutrient toxicities, to assess the need for soil amendments and to mon itor the effects through time of fertilizer rates and other management practices (Helyar and Price, 1999). Interpretation of soil test results requires the use of correlation between yield and soil test values. Typically, the interpretation has been esta blished for a crop, soil type, and area. In the case of S and tomato, this relationship has not been established. This study is attempting to determine if a relationship exists between tomato yield, soil S and tissue S. While i t is important to note tha t seldom does interpretation become established based upon a single carefully designed experiment; rather it requires hundreds of fertility trials and resulting in thousands of plant analyses (Melsted et al., 1969 ). An initial study must be conducted to e stablish a nutrient calibration curve. Nutrient calibration based upon mineral concentration in dry matter versus plant growth or yield has been described by Smith (1962). This nutrient calibration curve is the basis of plant nutrition. It is a relation ship often attempted to be established for many nutrients for interpretation of nutrient research results. Both total and relative yield can be used to establish nutrient calibration curves. I t is not recommended to use relative yield as a variable beca use the standardization of yield can lead to erroneous conclusions It has been suggested by Jackson (2000) that thorough evaluations of SO 4 S extraction procedures, along with more S response data are needed to create an S

PAGE 106

106 fertilization recommendation. S trong relationships between soil S and plant S are difficult to achieve and are the current focus of many S researchers throughout the world. Only calibration of soil test values against yield, S concentrations in plants or S uptake by plants make them su itable for judgment about the S nutritional status of a site (Schnug and Haneklaus, 1998). Since an official method of analysis is lacking for soil S, a standard reference method for determining SO 4 S or total S in soil is used. Examining the literature about SO 4 S or total S determination in soil resulted in many studies recommending different soil extraction and determination methods (Table 2 2). Beaton and Burns (1968) cite two methods for determining extractable SO 4 from soils: by CaHPO 4 and LiCl The first determines SO 4 by turbidimetry and the second by methylene blue. Jones and Jones (2001) have the determination of extractable SO 4 S as Ca(H 2 PO 4 ) 2 0.5 M CH 3 COONH 4 0.25 M CH3COOH and 0.01 M CaCl 2 Blair et al. (1993) developed an exaction met hod for SO 4 S for humid extractants as predictors of S status of pastures, they found that 0.25 M KCl extraction solutions shaken at 40 C had the highest coefficient of determ ination (Blair et al., 1991). More recently, Bloem et al. (2002) investigated this method for extracting soil S can be used in conjunction with turbidimetry, ICP emission spectrometry and IC An important theme throughout S research is the wide spread variability for different methods when determining S and even laboratory to laboratory methodological differences in determining S and SO 4 S. A multiple laboratory compari son of S in plant

PAGE 107

107 tissue and SO 4 S in soil found that variability ( higher than 36% CV) was large and that determination of S is a complex and difficult procedure (Crosland et al., 2001). This study made s everal recommendations that are applicable beyond t he United Kingdom, where the study was conducted. Some of these suggestions by Crosland et al., (2001) were 1) laboratories should standardize extraction and digestion procedures for S, 2) reference materials should be utilized (none currently exist for s oils), 3) analytical methods must be capable of determining low concentrations of S in soil extracts, and 4) methods must be calibrated for diagnostic purposes so that results obtained are comparable with those from other methods. The first recommendation is important, because until a standardization of extractions and digestions is created many more studies will be conducted with different extraction and digestion procedures and variability will continue to be an issue with S research. Sulfur research n eeds to include reference material in analytical studies. This would establish the consistency in measurements, which in turn would aid in interpretation and quantification. Currently and most importantly, no reference materials have been developed and t his has to be determined and established for S research as a whole. The determination of low concentrations of S allows for establishment of lower limits of determination. The calibration of methods will assist in increasing the consistency and continuit y of S research world wide. The object of this study was to determine if a relationship exists between soil S or tomato leaf S with tomato yield. Materials and Methods A study was conducted at the Gulf Coast Research and Education Center (GCREC) in Balm, FL. The soil was a sandy, siliceous, hyperthermic Oxyaquic Alorthod soil with 1.5% organic matter and a water pH of 7.3. Plant nutrients, other than N and S,

PAGE 108

108 were supplied under non limiting conditions through drip irrigation following current UF/IFAS rec ommendations (O lson et al., 2008). Phosphorus was not added, because soil analysis revealed 108 mg P kg 1 and this is considered high by interpretations (Simonne and Hochmuth, 2009). Pressed beds were fumigated with methyl bromide plus chloropicrin (67:3 3 v/v) at a rate of 196 kgha 1 to eliminate soilborne diseases, nematodes and weeds on 23 Feb. 2007. Simultaneously, planting beds were covered with 1. Experimental plots were 8.8 m long with a 2.9 m long non treated buffer zone at the end of each plot. Fertilizer sources for N and S appl ications were NH 4 NO 3 K 2 SO 4 (NH 4 ) 2 SO 4 and NH 4 NO 3 + (NH 4 ) 2 SO 4 at rates 0 and 186 kg N ha 1 with 0 and 213 kg S ha 1 These fertilizer sources and rates were applied pre fumigation in grooves located on the top of the bed. These grooves were two 7.6 cm deep bands separated ds were 78 cm wide at the base, 68.6 cm wide at the top, 19.6 cm high, and spaced 1.47 m apart on centers. Fertilizer sources were chosen based upon S content and likelihood of being used fertilizer that is banded on irrigation system. Therefore, the fertilizers (salts) are diluted via subsurface irrigation. Muriate of potash (KCl) (60% K) was used to balance total K amounts in those treatments using NH 4 NO 3 or (NH 4 ) 2 SO 4 as a ferti lizer source to ensure that K was under similar and non limiting conditions.

PAGE 109

109 A hybri d system uses both subsurface and drip irrigation and employed by those growers transitioning from all subsurface to drip irrigation. Irrigation supplied as subsurface was at the approximate rate of 21,035 Lha 1 per d Drip irrigation tubing (T Tape Sys tems International, San Diego, CA) was buried to a depth of 2.5 cm in the center of the raised bed with a flow rate of 0.058 Lm 1 min 1 The irrigation rate using drip system was at an approximate rate of 39,095 Lha 1 per day. Therefore, the total wate r applied via irrigation was 60,130 Lha 1 per day. The water table was maintained between 46 and 61 cm deep and constantly monitored with piezometers located in the fields. c enter of each bed on 12 Mar. 2007. Tomato plants were staked and tied as described by (Csizinszky et al. 2005) and weekly spray applications to maintain pest and disease control based upon weekly field scouting results. Tomato fruits were harvested twice on 10 and 12 WAT, and graded as marketable and non marketable based upon size. Marketable tomato fruits were graded according to the Florida Tomato Committee regulations for size categories (Brown, 2009; Sargent and Moretti, 2004). The sizes were 5X6 (7 .0 cm) 6X6 (7.06 to 6.32 cm) 6X7 (6.43 to 5.72 cm), and cull ( < 5.4 cm). Cull fruits were those smaller than 5.4 cm in diameter which below the minimum di ameter for a 6x7 tomato fruit. Twenty four tomato leaf and soil samples were randomly taken from wit hin an 8.8 m linear plot of the raised bed. Ten leaf samples were taken from the most recently matured leaf adjacent to an inflorescence as described by Mills and Jones (1991) within a plot. Leaf samples were dried at 43 C, ground with a Wiley mill to p ass through at 0.853 mm sieve aperture ( 20 mesh screen ) and stored at 0 C until analysis was

PAGE 110

110 conducted. Soil samples collected from between tomato plants and were approximately 15 to 20 cm of surface soil within plots After collection soil samples were air dried, sieved, and stored at 4.4 C until analysis was conducted. Plant tissue samples were digested were HNO 3 and 30% H 2 O 2 as described by Mills and Jones (1991) One gram of dried plant material was weighed directly into a 100 mL glass test tube. Eight milliliters of concentrated nitric acid was added, covered, and left overnight to predigest. Tubes were then placed into an aluminum heating block capable of holding 50 tubes. The heating block was ramped to 120 C and held for 1 h at this temper ature and allowed to cool. Eight milliliters of 30% H 2 O 2 was then added and the block re heated to 95 C. Solutions were diluted to 20 mL with nano pure water after cooled to room temperature (approximately 22 C) and filtered with Whatman No. 44 filter paper (20 25 m particle retention) Soil SO 4 S status was characterized by extractions with de ionized water and 0.025 M KCl using a spectrophotometer (Spectronmic 100, Bausch & Lomb, Rochester, NY) on 16 Sept. 2009 at a commercial laboratory. The standard method for determining SO 4 in water by turbidimetry (AOAC OM 973.57, ASTM D 516 02) (AOAC, 2000; ASTM, 2006) was considered to be the standard reference method for SO 4 S determination and used for comparison with other methods. Total soil S status was characteri zed by de ionized water, and 0.025 M KCl, us ing ICP OES (Optima 2100 DV, Perkin Elmer, Waltham, MA) with Eschelle grading and under a constant vacuum at GCREC Analytical Lab in Balm, FL on 19, 20 and 21 Dec. 2008, 8, 9 and 21 July 2009. Only soil samples extracted with Mehlich 3 and tissue samples digested with HNO 3 plus 30% H 2 O 2 were diluted to 1:10 ratio for ICP OES. The Dumas method of SO 2

PAGE 111

111 measurement by infrared radiation (CNS 2000, LECO, St. Joseph, MI) was conducted on 16 June 2009 at a commercial l aboratory. All data was converted from % or mgL 1 to mgkg 1 for data analysis. The parameters of the regression line were tested to determine if plotted regression lines were similar to the ideal relationship between extractants. This ideal relationsh ip is a regression line with a slope of 1 (1 rise to 1 run or 1:1) with an intercept of 0. This comparison against a 1:1 line is a similar relationship tested as Sikora et al. (2005) used for comparing methods for the determination of P If the ideal 1:1 relationship is met, then the extractant comparisons do not require interpretation for over or under estimation of soil S or SO 4 S. Data was analyzed using SAS system using proc reg and linear regression model (y ij i i x ij ij ) with test slope = 1 and test intercept = 0 statements with P<0.05 (SAS Institute, 2000). The test slope statement test H 0 i = 1, if this test hypothesis is not significant (P>0.05) then the slope is not significantly different from 1. The test intercept = 0 statement test H 0 i = 0, if this test hypothesis is not significant (P>0.05) then the intercept is not significantly different from 0. The R 2 values have been reported to show the variability within the chosen regression model For relationships between extractants and tomato yield, regression lines were fitted to the data to determine if each soil test could be calibrated to tomato yield. Regression lines were plotted with significant data using Sigma Plot software program (S ystat Software, 2008). Results and Discussion None of the plant tissue testing methods for to tal plant tissue S (acid digestion and LECO ) had satisfactory relationships with total tomato yield (Fig ures 7 1 and 7 2 ) None of the soil S or SO 4 S methods rel ated well to total tomato yield as no linear regression re lationship could be established (Fig ures 7 3 7 4, 7 5, and 7 6 ). While

PAGE 112

112 presenting correlation coefficients may be possible; they may lead to erroneous conclusions about the relationship between tom ato yield and soil or tissue S. It is important to note that unlike previous studies, which have used dry matter (Matula, 1999; Pandey and Girish, 2006), leaf tissue S (Anderson et al., 1998a; Rao and Sharma, 1997), or relative yield (Anderson et al., 19 92; Chinoim et al., 1997; Spencer and Glendinning, 1980; Zhao and McGrath, 1994; this study used actual yield values for relationship determinations. This being stated and given the lack of an ideal relationship found between the three soil extraction met hods used for soil S and SO 4 S in this study. It is possible that the fractions extracted do not relate well to plant yield (Fig ure 7 7) Therefore, it is suggested that other soil extractants be investigated if a soil test based recommendation for S is desired for vegetable production in Florida. Based upon these results, the current recommendation of application of 28 to 37 kg S ha 1 based upon tissue testing should be followed until an appropriate soil test for S can be found.

PAGE 113

113 Figure 7 1 R elationship between dry combustion total p lant S and tomato yield for selected tissue samples from the Gulf Coast Research and Education Center in Balm, FL

PAGE 114

114 Figure 7 2 R elationship between total p lant S and tomato yield for selected tissue samples from the Gulf Coast Research and Educa tion Center in Balm, FL

PAGE 115

115 Figure 7 3 R elationship between dry combustion total soil S and tomato yield for selected soil samples from the Gulf Coast Research and Educ ation Center in Balm, FL

PAGE 116

116 Figure 7 4 R elationship between Modified Blair (0.025 M KCl) soil S and tomato yield for selected soil samples from Gulf Coast R esearch and Education Center in Balm, FL

PAGE 117

117 Figure 7 5 R elationship between de ionized water soil S and tomato yield for selected soil samples from Gulf Coast Research and Educa tion Center in Balm, FL

PAGE 118

118 Figure 7 6 R elationship between Mehlich 3 soil S and tomato yield for selected soil samples from Gulf Coast Research and Education Center in Balm, FL

PAGE 119

119 Figure 7 7 Possible m ain fractions of soil S a nd analytical methods f or S determination.

PAGE 120

120 CHAPTER 8 SUMMMARY AND CONCLUS IONS These studies on S fertilization provided information on S source, rate, and irrigation programs on fresh market tomato growth, development and yield. Sulfur applications, regardless of source ( (NH 4 ) 2 SO 4 K 2 SO 4 NH 4 NO 3 + (NH 4 ) 2 SO 4 ) incr eased tomato yields when compared to the non treated control. Therefore, source of S was not contributed to the yield increase. Elemental S, when applied as the only source of S, increased early fresh market tomato yi eld. At approximately 5.6 kg Sha 1 of applied S, A yield increase of 2,919 kgha 1 over the non treated control was determined. Irrigation program did not influence soil S concentration, soil pH, S concentration within tomato leaves or early yield. Thi s suggests that irrigation water may not meet the S needs of tomatoes when grown with drip irrigation. It is possible that the form of S present within the irrigation water needs to be oxidized by microbes. An increase in tomato leaf S and early yield we re found with the addition of elemental S to the fertilizer regime. The laboratory study provided information on S soil testing for Florida sandy soils, which is predominate soil type that tomatoes are grown on in Florida. Based upon the results, soil S O 4 S and S are highly variable in determination which is supported by previous studies and difficult to completely determine a relationship between soil SO 4 S, soil S, plant tissue S an d tomato yield. All regression relationships between soil S and SO 4 S had low R 2 values ( < 0.70). The relationship between soil SO 4 S and plant S had a negative slope. These were not ideal statistically to determine the relationship between these variables. An increase in early tomato yiel ds were found throughout the studi es when S was included in the fertilizer regime This suggests that S should be included in the fertilizer

PAGE 121

121 regime for fresh market tomatoes grown on Florida sandy soils. Further research to re evaluate the S sufficiency rates for tomato production would be helpful on determining the appropriate leaf sufficiency ran ge for tomato a s total S i n tomato tissue ranged from 4550 mg S kg 1 to 9400 mg S kg 1 with little correlation to yield response It is also apparent that irrigation water may not contribute to the total S necessary to grow a tomato crop in Florida. From this study further support of the variability in soil testing for S was determined. For Florida sandy soils, no one method can be recommended for determining SO 4 S or S. Therefore, when S def icien cy is suspected or tissue S is shown to be in the low range (<3, 000 mg S kg 1 ) in a vegetable crop 28 to 38 kg S ha 1 should be applied as part of the fertilizer regime.

PAGE 122

122 LIST OF REFERENCES Anderson, G.C., G.J. Bl air, and R.D.B. Lefroy. 1998a. So il extractable sulfur and pastur e response to applied sulfur : Studie s under glasshouse conditions. Aust ral J. Exp. Ag. 38:567 574. Anderson, G.C., G.J. Bl air, and R.D.B. Lefroy. 1998b. Soil extractable sulfur and pastu re response to applied sulfur: Season al variation in soil sulfur tests and sulfur response by pa stures under field conditions. Aust ral J. Exp Ag. 38:575 582. Anderson, G. C., R. Lefroy, N. Chinoim, and G. Blair. 1992. Soil sulphur testing. Sulphur Ag. 16:6 14 American Society of the Testing of M aterials (ASTM) International. 2006. ASTM D 516 02: Standard test method for sulfate ion in water. In: Annual Book of ASTM Standards. Volume 19, Section 5. ASTM International, West Conshohocken, PA Association of Analytical Chemis ts International ( AOAC). 2000. Chapter 11 Water and Salt: AOAC Official Method 973.57 Sulfate in Water. Turbidimetric Method. p. 23. In: W. Horwitz (ed.). Official Methods of Analysis of the Association of Anal ytical Chemists International. Association of Analytical Chem ists International (AOAC). Arlington, VA. Ajwa, H.A. and M. A. Tabatabai. 1993. Comparison of some methods for dete rmination of sulfate in soils. Commun. Soil Sci. Plant Anal. 24:1817 1832. Baban, S., D. Beetlestone, D. Betteridge and P. Sweet. 1980. The determination of sulphate by flow injection analysis with exploitation of pH gradients and EDTA. Anal. Chim. Acta 114:319 323. Baker, D.E. 1973. A new approach to soil testing: Ionic e quilibria involving H, K, Ca, M g, Mn, Fe, Cu, Zn, Na, P and S. Soil Sc i. Soc. Amer. Proc. 27:537 541. Bansal, K.N., D.P. Motiramani, and A.R. Pal. 1983. Studies on sulfur in vertisols: Soil and plant tests for diagnosing sulfur deficiency in soybean (Glycine max (L) Merr.) Plant Soil. 70:133 140. Bansal, K.N and A.R. Pal. 1987. Evaluation of a soil test method and plant analysis for determining the sulfur status o f alluvial soils. Plant Soil. 98:331 336. Bansal K.N., D.N. Sharma, D. Singh. 1979. Evaluation of some soil testing methods for measuring available sulfur in all uvial soils of Madhya Pradesh, India J. Ind. Soc. Soil Sci. 27:308 313 Bates, A.L., W.H. Orem, J. W. Harvey, E.C. Spiker. 2002. Tracing sources of sulfur in the Florida Everglades. J. Environ. Qual. 31:287 299.

PAGE 123

123 Beaton, J.D., G.R. Burns and J. Platou 196 8. Determination of sulfur in soils and plant material. Tech. Bul. No. 14 Sulfur Inst. Washington, DC. Brix H., B. Lorenzen, I.A. Men delssohn, K. L. McKee, S. Miao. 2010. Can differences in phosphorus uptake kinetics explain the distribution of cattail a nd sawgrass in the Flori da Everglades? BMC Plant Bio. 10:23 36. Bixby D.W. and J.D. Beaton. 1970. Sulfur Containing Fertilizers: properties and applications. Tech. Bul. No. 17. Sulfur Inst. Washington, D.C. Blair, G.J., N. C hinoim, R.D.B. Lefroy, G. C. An derson, and G.J. Crocker. 1991. A soil sulfur test for pastures and crops. Aust ral J. Soil Res. 29:619 629. Blair, G.J., R.D.B. Lefroy, N. Chin oim, and G.C. Anderson. 1993. S ulfur soil testing. Plant Soil 155& 156:383 386. Blake Kalff, M.M.A., M.J. H awkesf ord F.J. Zhao, and S.P. McGrath. 2000. Diagnosing sulfur deficiency in field grown oilseed rape ( Brassica napus L.) and wheat ( Triticum aestivum L.). Plant Soil 225:95 107. Blanchar, R. W. 1986. Measurement of sulfur in soils and plant s. p. 455 490. In: M A. Tabatabai (ed.). Sulfur in Agriculture. ASA CSSA SSSA Madison, WI Bloem, E., S. Haneklau s, G. Sparovek, and E. Schnug. 2001. Spatial and temporal variability of sulfate concentration in soil. Commun. So il Sci. Plant Anal. 32:1391 1403. Bloem, E., S Ha neklaus, and E. Schnug. 2002. Optimization of a metho d for soil sulfur extraction. Commun. Soil Sci. Plant Anal. 33:41 51. Boss, C B. and K. J. Fredeen (eds.). 1997. Correcting for spectral interferences in ICP OES. p. 4 16 to 4 22. In: Concepts, In strumentation, and Techniques in Inductively Coupled Plasma Optical Emission Spectrometry. PerkinElmer. Waltham, MA Brezonik, P.L., E.S. Edgerton, an d C.D. Hendry. 1980. Acid precipitation and sulfate deposition in Florida. Sci ence 208:1027 1029. Brown, R.L. 2009. Florida Tomato Committe e: Regulatory Bulletin No.1. 20 August 2009. < http://www.floridatomatoes.org/domestic.html >. Castro, H., S. Newman, K. R. Reddy, and A. Ogram. 2005. Distribution and stability of sulfate reducing prokaryotic and hydrogenotr ophic methanogenic assemblages in nutrient impacted regi ons of the Florida Everglades. Appl. Environ. Microbio. 71:2695 2704.

PAGE 124

124 Ceccotti, S.P, R.J. Morris, and D.L. Messick. 1998. A global ove rview of the sulfur situation: ds, and commercial aspe cts of sulfur fertilizers. p. 175 202. In: E. Schnug (ed.). Sulfur in Agroecosystems. K l uwer Academic Publisher s. Chinoim, N., R.D.B. Le froy, and G.J. Blair. 1997. Effect of crop duration and soil type on the ability of soil sulfur t ests to pre dict plant response to sulfur. Aust ral J. Soil Res. 35:1131 1141. Crosland, A.R., F.J. Zhao, and S.P. McGrath. 2001. Inter laboratory comparison of sulfur and nitrogen analysis in plants and soils. Commun. Soil Sci. Plant Anal. 32:685 695. Cs izinszky, A.A., D.J. Schuster, J.B. Jones, and J.C. van Lenteren. 2005. Crop Production. p. 237 256. In: E. Heuvelink (ed.) Tomatoes. CABI Publishing. Dasgupta, P.K., L.G. Hangley Jr., and P.W. West. 1978. Spectrophotometric determination of trace sulfur in water. Anal. Chem. 50:1793 1795. David, M. B., M.J. Mitchell, D. Aldcorn, and R. B. Harrison. 1989. Analysis of sulfur in soil, plant, and sediment materials: Sample handling and use of an automated analyzer. Soil Biol. Biochem. 21:119 123. Dmmgen, U., K. Walker L. Grnhage, and H J. Jger. 1998. The atmospheric sulfur cycle. p. 75 114. In: E. Schnug (ed.). Sulfur in Agr oecosystems. Kluwer Academic Pub lishers. Netherlands De Kok, L. J., M. Durenkamp, L. Yang, and I. Stulen. 2007. Atmospheric s ulfur. p 91 106. In: M.J. Hawkesford and L.J. De Kok (eds.). Sulfur in Plants: A n Eco logical Perspective. Springer. Dordrecht, The Netherlands Drost, D.T., J.W. MacAdam L.M. Dudley, and N. Soltani. 1997. Response of bean and broccoli to high sulfate irrigation water. HortTech nology 7:429 434. Duke, S. H. and H.M. Reisenauer. 1986. Roles and requirements of sulfur in plant nutrition. p. 123 168. In: M.A. Tabatabai (ed.). Sulfur in Agriculture. ASA CSSA SSSA, Madison, WI Eames, J.C. and R. J. Cosstick. 1992. Dete rmination of forms of sulfur in colas and related materials by Eschka digestion and inductively coupled plasma atomic em ission spectrometry. Analyst. 117:1581 1584. Eriks en, J. and M. Askegaard. 2000. Sulfate leaching in an organic crop rotation on sandy s oil in Denmark Ag. Ecosystem Environ. 78:107 114. Eriksen J. and J.V. Mortensen. 2002. Effects of timing of sulfur application on yield, S uptake, and quality of barley. Plant Soil. 283:283 289.

PAGE 125

125 Eriksen, J ., M.D. Murphy, and E. Schnug. 1998 The soil s ulfur cycle. p. 39 73. In: E.Schnug (ed.). Sulfur in Agroecosystems. Kluwer Academic Publishers, Netherlands Esmel, C. E. 2005. Influence of calcium on postharvest quality and yield of strawberry fruit ( Fragaria ananassa Duch.). Univ. FL Thesis. Etherig e, R.D., G. M. Pesti, and E.H. Foster. 1998. A comparison of nitrogen values obtained utilizing the Kjeldahl nitrogen and Dumas combustion methodologies (Leco CNS 2000) on samples typical of an animal nutrition analytical labo ratory. Anim. Feed Sci. Tech. 73: 21 28. Gianello, G. and Bremner J. M. 1986. A simple chemical method of assessing potentially avai lable organic nitrogen in soil. Commun. Soil Sci. Plant Anal. 17:195 214. Gong, R.K, L. Chen, and S. P. Nian. 2006. A new method to improve the precision of SO 2 measurement in atmosphere. J. Phys. 48:1162 1166. Hagel, I. 2005. Sulfur and baking quality of bread making wheat. Landbauforschung Vlkenro de (FAL Agricultural Research). 283:23 36. Havlin, J.L, J.D. Beaton, S.L. Tisdale, and W.L. Nelson (eds.) 200 5. Sulfur. p. 219 233. In: S oil Fertility and Fertilizers: An introd uction to nutrient management. Pearson Prentice Hall, Upper Saddle River, NJ Havlin, J. L., and P. N. Soltanpour. 1980. A nitric acid plant tissue digst method for use with inductivel y cou pled plasma spectrometry. Commun. Soil Sci. Plant Anal. 11:969 980. Heeb, A., B. Lundegr dh, G. Savage, and T. Ericsson. 2006. Impact of organic and inorganic fertilizers on yield, taste, and nutritional quality of tomatoes. J. Plant Nutr. Soil Sci. 169:53 5 541. Helyar, K.R. and G. H. Price. 1999. Making recommendations based on soil tests. p. 331 357. In: K.I. Peverill, L.A. Sparrow, and D.J. Reuter (eds.). Soil Analysis : an Interpretation Manual. CSIRO Pub. Australia. Hoeft, R.G, L.M. Walsh, and D.R. Keen ey. 1973. Evaluation of various extracta nts for available soil sulfur. Soil Sci. Soc. Am er J. 37:401 404 Hoenig, M., H. Baeten, S. Vanhentenrijk, E. V assileva, and Ph. Quevauviller. 1998. Critical discussion on the need for an efficient mineralization pr ocedure for the analysis of plant material by atomic spectrometric methods. Anal. Chim. Acta. 358:85 94. Huang, L., R. W. Bell, and J. Woodward. 2004. Rapid nitric acid digestion of plant material with an open vessel microwave sys tem. 35:427 440.

PAGE 126

126 Ippolito, J.A. and K. A. Barbarick. 2000. Modified nitric ac id plant tissue digest method. Commun. Soil Sci. Plant Anal. 31:2473 2482. Islam, M.M. and N.I. Bhuiyan. 1988. Evaluation of various extractants for available sulfur in wetland rice (Oryza sativa) soils in Bangladesh Ind. J. Agr Sci. 58:603 606 Islam, M.M. and F.N. Ponnamperuma. 1982. Soil and plant tests for available sulfur in wetland rice soils. Plant Soil. 68:97 113. Iwasaki, I., S. Utsumi, K. Hagino, T. Tarutani, and T. Ozawa. 1957. Spectrophotomet ric method for the determination of small amounts of sulfate ions. Bull. Chem. Soc. Japan 30:847 851. Jackson, G.D. 2000. Effects of nitrogen and sulfur on can ola yield and nutrient uptake. Agron. J. 92: 44 49. Johnson, C. M. and H. Nishita. 1952. Microest imation of sulfur. Anal. Chem. 24:736 742. J ones, J.B. Jr. and J.B. Jones. 2001. Laboratory guide for conducting soil tests and plant analysis. CRC Press. Boca Raton, FL. Jones, J. B. Jr., 1998. Soil test methods: Past, present, and future use of soil extra ctants. Commun. Soil Sci. Plant Anal. 29:1543 1552. Jones, J. B. Jr. 1993. Modern interpretation systems for soil and plant analysis in the United States of America. Austral. J. Exp. Ag. 33:1039 1043. Jones, J. B Jr, 1985. So il testing and plant analysis: G uides to fertil ization of horticultural crops. p. 1 68. In: J. Janick (ed.). Horticul tural Reviews. AVI Pub. Westport, CT. Jones, J. B. Jr, B. Wolf, and H. A. Mills. 1991. Plant Analysis Handbook: a practical sampling, preparation, analysis and interpretati on guide. Micro Macro Pub. Athens, GA Knauthe, B. and M. Otto. 2001. Nonmetals in the argon inductively coupled plasma optical emissi on spectrometry: Phosphorus, sulfur and car bon. Chem. 371:1052 1056. Knauth e, B., M. Otto, and F. Mart in. 2000. A systematic approach to optimum working conditions with inductively coupled plasma atomic emission spectrometry. 367:679 685. Kola, H., P. Permki, and I. Vlinki. 2002. Correction of spectral interference of cal calc ium in sulfur determination by inductively coupled plasma optical emission spectrometry us ing multiple linear regression. J. Anal. At. Spectrom. 17:104 108.

PAGE 127

127 Kolthoff, I.M. and E.B. Sandell (eds.). 1952. Determination of sulfur as barium sulfate. p. 322 336 In: Textbook of Quantitative Inorganic Analysis, 3 rd Edition. Macmillia n Pub. New York, NY. Kowalenko, G. 2001. Assessment of LECO CNS 2000 analyzer for simultaneously measuring total carbon, nitrogen and sulphur in soil. Commun. Soil Sci. Plant Anal. 32:2065 2078. La j unen, L.H. J. and P. Per m ki (eds.) 2004 Plasm a atomic emission spectrometry. p. 203 256. In: Spectrochemicla Analysis by Atomic Absorption and Emission 2 nd Edition. Royal Soc. Chem. Cambridge, UK. LECO Corp. 2006. TruSpec C N Carbon/ Nitrogen Determinator: Instruction Manual. St. Joseph, MI. p 7 11. Li, S., I.A. Mendelssohn, H. Chen, and W. H Orem. 2009. Does sulphate enrichment promote expansion of Typha domingesis (cattail ) in the Flor ida Everglades? Freshwater Bio. 54:1909 1923. L isle, L., R. Lefroy G. Anderson and G. Blair, G. 1994. Methods for measuring sulfur in plants and soil. Sulfur in Ag riculture 18:45 54 Leustek, T, M.N. Marti n, J. A. Bick and J.P. Davies. 2000. Pathways and regulation of sulfur metabolism revealed thro ugh molecular and genetic studies. Annu. Rev. Plant Physiol. Plant Mol. Biol. 51:141 165. Leustek, T. 2002. Sulfate metabolism. Arabidopsis Book. 42:1 16. Lewis, D.C. 1999. Sulfur. p. 221 228. In: K.I. Peverill, L.A. Sparrow, and D.J. Reuter. (eds.). Soil Analysis: An Interpretation Manual. CSIRO Pub Australia Malhi, S.S., J.J. Shoenau, and C. L. Vena. 2009. Influence of six successive annual applications of sulphur fertilizers on wheat in a canola wheat rotatio n on a sulphur deficient soil. Canadian J. Pl ant Sci. 89:629 644. Malhi, S.S., J.J. Schoenau, and C.A. Grant. 2005. A review of sulfur fertilizer management for optimum yield and quality of canol a in the Canadian Great Plains. Can. J. Plant Sci. 85:297 307 Manning, T.J. and W. R. Grow. 1997 Inductiv ely coupled plasma: A tomic em ission spectrometry. Chem. Ed. 2:1 19. Matula, J. 1999. Use of multinutrient soil tests for sulfur determination. Commun. Soil Sci. Plant Anal. 30:1733 1746. Maynard, D.G., Y.P. Kalra, and F.G. Radford. 1987. Extraction and det ermination of sulfur in or ganic horizons of forest soils. Soil Sci. Soc. Am er J. 51:801 805.

PAGE 128

128 M aynard, D.N. and G.J. Hochmuth. 2007. table Growers. Wiley N J Melsted, S. W., H. L. M otto, and T. R. Peck. 1969. Critical plant nutrient composition values useful in interpreting pla nt analysis data. Agron. J. 61: 17 20. Mermet, J M. and E. Poussel. 1995. ICP emission spectrometers: 1995 analytical figu res of merit. Appl. Spectrosc. 49:12A 18A Mills, H.A. and J. B. Jones. 1991. Plant Analy sis Handbook II. MicroMacro Publishing Athens GA Mitchel l, C.C. and W.G. Blue. 1981 The sulfur fertility status of Florida soils: An evaluation of subs oil sulfur on plant nutrition. Proc. Fla. Soil Crop Sci. Soc. 40:77 82. Morais, I. P. A., M.R. S. Souto, T.L.M. S. Lopes, and A.O.S. S. Rangel. 2003. Use of single air segment to minimize dispersion and improve mixing in sequential injection: turbidimetric determ ination of sulphate in waters. Water Res. 37:4243 4249. Morais, I.P. A, I.V. Tth, A.O.S.S. Rangel. 2 006. Turbidimetric an d nephelometric flow analysis: concepts and appli cations. Spectrosc. Letters. 39:547 579. Mylavarapu, R.S, J.F. Sanchez J.H. Nguyen, and J.M. Bartos. 2002. Evaluation of Mehlich 1 and Mehlich 3 extraction procedures for plant nutrient s in acid mineral soils of Florida Commun. Soil Sci. Plant Anal. 33:807 820. Olson, R.A. and G.W. Rehm. 1986. Sulfur in precipitation and irrigation waters and its eff ects on soils and plants. p. 279 293. In: M.A. Tabatabai. (ed.). Sulfur in Agriculture. ASA CSSA SSSA. Madison, WI Olson, S.M., W.M. Stall, G.E. Vallad, S.E. Webb, T.G. Taylor, S.A. Smith, E.H. Simonne, E. McAvoy and B.M. Santos 2009 Tomato production in Florida. p. 291 312. In: S.M. Olson and E.H. Simonne (eds.). Vegetable Production Hand book for Florida. Vance Publishing. Lenexa, KS. Oliva, S.R., H. Raitio and M. D. Mingorance. 2003. Comparison of two wet digestion procedures for multi elem ent analysis of plant samples. Commun. Soil Sci. Plant Anal. 34:2913 2923. Palomino, L M., Ch. Sie be, and A. Herre. 2005. Air drying affects extractable sulfate in soils of v ariable charge: T est of two extraction and two quantification methods. Commun. Soil Sci. Plant Anal. 36:2513 2528. Pandey, R.N. and B. H. Girish. 2006. Evaluation of some extractant s for predicting soil sulphur availability to cowpea ( Vigna unguiculata ). Ind. J. Ag. Sci. 76:200 202.

PAGE 129

129 Papadopoulos, I. 1986. Effect of high sulfate irrigation water s on soil salinity and yields. Agron. J. 78:429 432. Papadopoulos, I. 1984. Effect of sulfa te waters on soil salinity, growth and yield of tomatoes. Plant Soil. 81:353 361. Pavlista, A.D. 2005. Early season applications of sulfur fertilizers increase potato y ield and reduce tuber defects. Agron. J. 97:599 603. Pyki, R., H. Torvela, P. Permki T Kuokkanen, and H. Rkkmki. 2000. Comparison of dissolution methods for multi element analysis of some plant materials used as bio indicator of sulphur and heavy metal deposition determined b y ICP AES and IPC MS. Analusis. 28:850 854. Rao, T.N., and P.K. Sharma. 1997. Evaluation of Mehlich 3 as an extract ant for available soil sulfur. Commun. Soil Sci. Plant Anal. 28:1033 1046. Rhoads, F.M. and S.M. Olson. 2000. Cabbage response to sulfur source and nitrog en rate. Soil Crop Sci. Soc. Fla 60:37 40. Reisman, D. J.V. Sundaram, S. R. Al Abed, a nd D. Allen. 2007. Statistical validation of sulfate quantification methods used for a nalysis of acid mine drainage. Talanta. 71:303 311. Sacks, L.A. 1996. Geochemical and isotopic composition of ground water with emphasis on sources of sulfate in the Upper Floridian aquifer in parts of Marion, Sumter, and Citrus counties, Florida: US Geological Survey Tallahasse, FL. Water Resources Investigations Report 95 4251 < http://fl.water.usgs.gov/publications/Abstracts/wri95_4251_sacks.html > Santoso, D., R.D.B. Lefroy, and G. J. Blair. 1995. A comparison of sulfur extract ants for weathered acid soils. Austral. J. Soil Res. 33:125 133. Sarg ent, S.A. and C.L. Moretti. 2004. Tomato. In: The commercial storage of fruits, vegetables florists, and nursery stock. In: K.C. Gros s, C.Y. Wang, and M. Saltveit (e ds.). Ag. Handbook No. 66. USDA ARS. April 2004. < http://www.ba.ars.usda.gov/hb66/contents.html > Sawyer, J.E. and D.W. Barker. 2002. Corn and soybean response to sulfur application on Iowa soils. p. 157 163. I n: Proceedings of the 32nd North Central Extension Industry Soil Fertility Conference; 20 21 November 2002; Des Moines, IA. Potash and Phosphate Inst. Brookings, SD Scherer, H.W. 2001. Sulfur in crop production: I nvited paper. Euro. J. Agron. 14:81 111.

PAGE 130

130 Sc hnug, E. and S. Haneklaus. 1998. Diagnosis of sulfur nutrition. p. 1 38. In: E. Schnug (ed.). Sulfur in Ag roecosystems. Kluwer Academic Publishers The Netherlands Searle, P. L. 1988. The determination of phosphate extractable sulphate in soil with an anio n exchange membrane. Commun Soil Sci. Plant Anal. 19:1477 1493. Sexton, P.J., W.D. Batchelor, and R. Shibles. 1997. Sulfur availability, Rubisco content and photosynthetic rate of soybean. Crop Sci. 37:1801 1806. Shampine, W.J. 1993. Quality assurance a nd quality control in monitoring programs. Environ. Monitor. Assess. 26:143 151. Sikora, F.J., P.S. Howe, L.E. Hill, D. C. Reid, and D. E. Harover. 2005. Comparison of colorimetric and ICP determination of phosphorus in Mehlich 3 soil extracts. Commun. Soil Sci. Plant Anal. 36:875 887. Singh, R., D.K. Bhumbla, and R. F. Keefer (eds.). 1995. Rec ommended soil sulfate S tests. p. 46 50. In: Recommended Soil Testing Procedures for the Northeastern U.S. 2 nd edition. Northeastern Publication No. 493. Simonne, E.H., M.D. Dukes, and D.Z. Haman. 2009. Principles and practices of irrigation management for vegetables. p. 33 39. In: S.M. Olson and E.H. Simonne (eds.). Vegetable Production Handbook for Florida. Vance Publ ishing. Lenexa, KS. Simonne, E .H. and G. J. Hochmuth. 2009. Soil and fertilizer management for vegetable production in Florida p. 3 15. In: S.M. Olson and E.H. Simonne (eds.). Vegetable Production Handbook for Florida Vance Publishing. Lenexa, KS. Smith, P. F. 1962. Mineral analysis of plant tissues. Ann. Rev. Plant Phys. 13:81 108. Smith, S.J., H. Pitcher, T.M. L. Wigley. 2001. Global and anthropogenic sulfur dioxide emissions Global Planet. Change 29:99 119. Solberg, E.D., S.S. Malhi, M. Nyborg and K.S. Gill. 2003. Fertilizer type, tillage, and application time effects on recovery of SO 4 S from elemental sulfur fertilizers in fallow field soils. Comm. Soil Sci. Plant Anal. 34:815 8 30. Soon, Y.K., Y.P. Kalra, and S. A. Abboud. 1996. Comparison of some metho ds for the determination of total sulfur in plant tissues. Commun. Soil Sci. Plant Anal. 27:809 818. Spencer, K and J.R. Freney. 1980. Assessing the sulfur status of field grown whea t by plant analysis. Agron. J. 72:469 472 Spencer, K. and J.S Glendinnin g. 1980. Critical soil test values for predicting the phosphorus and sulfur status of sub humid temperate pastures. Austral. J. Soil Res. 18:435 45.

PAGE 131

1 31 Stern, D. I. 2005. Global sulfur emissions from 1850 to 20 00. Chemosphere. 58:163 175. Susila, A.D. 2001. Su lfur nutrition requirement for drip irrigat ed tomato, pepper and cabbage. Univ. Fla. Gainesville, Ph.D. Diss. UMI 3027001 Systat Software. 2008. Tabatabai, M.A. 1984. Importance of sulfur in crop produ ction. Biogeochem istry 1:45 62. Tabata bai. M.A. 1996. Sulfur. p. 921 960. In: D.L. Sparks (ed.). Methods of Soil A nalysis : Part 3. Chemical Met hods. SSSA ASA. Madison, WI Tan, Z., R.G. McLaren, and K. C. Cameron. 1994a. Forms of sulfur extracted from soil s after different methods of sample preparation. Austral. J. Soil Res. 32:823 834. Tan, Z., R.G. McLaren, and K. C. Cameron. 1994b. Season variations in forms of extractable sulfur in some New Zealand soils. Austral. J. Soil Res. 32:985 993. Tiedemann, A.R and T. D. Anderson. 1971. Rapid analysis of total sulphur in soils and plant materials. Plant Soil 35:197 200. Tiwari, K.N. V. Nigam, and A.N. Pathak. 1983. Evaluation of some soil test methods for diagnosing sulfur deficiency in rice in al luvial soils of Ultar Pardesh. J. Ind. Soc. Soil Sci. 31:245 249. Ullrich, S.M., T.W. Tanton, S.A. Abdrashitova. 2001. Merc ury in the aquatic environment: A review of factors affecting mercury methylation. Crit. Rev. Environ. Sci. Tech. 31:241 293. Van Erp. P.J. and M. L. Van Beusichem. 1998. Soil and plant testing programs as a tool for optimizing fertilizer strategies. p. 53 80. In: Z. Rengel.(ed.). Nu trient Use in Crop Production. Haworth Press. Binghamton, NY Van Raij, B., H. Cantarella, and J. A. Quaggio. 2002. Rationa le o f the economy of soil testing. Commun. Soil Sci. Plant Anal. 33:2521 2536. Vendrell, P. F., K. Frank, and J Denning. 1990. Determination of soil sulfur by inductivel y coupled plasma spectroscopy. Commun. Soil Sci. Plant Anal. 21:1695 1703 Welz, B., F. Lepri, R. Araugo, S. Ferriera, M D. Huang, M. Okruss, and, H. Becker Ross. 2009. Determination of phosphorus, sulfur, and the halogens using high temperature molecular absorption spect rometry in flames and furnaces: A review. Anal. Chim. Acta. 647:137 148

PAGE 132

132 Willams, C.H. 1968. Seasonal fluctuations in mineral sulfur under subterranean clover pasture in southern New South Wales Aust ral J. Soil Res. 6:131 139 Wolf, B. 1982. An improved universal extracting solution and its use for diagnosing soil fertility Comm un Soil Sci. Plant Anal. 13:1005 1033 Wright, A.F., and J. S. Bailey. 2001. Organic carbon, total carbon, and total nitrogen determinations in soils of variable calcium carbonate contents using a LECO C N 2000 dry combustion analyzer. Commun. Soil Sc i. Plant Anal. 32:3243 3258. Wu, P., X. Wu, X. Hou, C. G. Young, and B. Jones. 2009. Inductively coupled plasma optical emission spectrometry in the vacuum ultraviolet region. Applied Spectro. Rev. 44:507 533. Yli Halla, M. 1987. Assessment of extraction an d analysis methods in estimating the amount of plant available sulfur in the soil. Acta Agr Scand. 37:419 425 Zbral, J. 1999. Comparison of some extraction methods for determination of sulphur in soils of the Czech Repbulic. Rostlinn Vroba. 45:439 444 Integrating Mehlich 3 extractant i nto Czech soil testing scheme. Commun. Soil Sci. Plant Anal. 31:2171 2182. Zhao, F. and S. P. McGrath. 1994. Extractable sulphate and organic sulphur in soils and their availability to plants Plant Soil 164:243 250.

PAGE 133

133 BIOGRAPHICAL SKETCH Camille grew up in New Hampshire (in the heart of scenic New England), surrounded by the majestic White Mountains National Forest, the historic Robert Frost Farm, and the quaint colonial New England sea coast She developed an appreciation for nature and the outdoors from both of her parents. Her father, (Curtis R. Esmel, Sr.), is an avid outdoorsman. Her mother, (Lydia E. Fortier), is a gardener and naturalist. Camille attended Pinkerton Academy in Derry, New Hampshire, where she studied horticulture through high school vocational classes and participation in the National F uture Farmers of America organization. Camille remained active in the National F uture Farmers of America organization while attending the U niversity of New Hampshire. She graduated from the University of New Hampshire in May 2002 with a Bachelor of Science in environmental h orticulture. Camille graduated from the University of Florida in May 2005 with a Master of Science in Horticultural Sciences.