<%BANNER%>

Physical Characteristics of a Seepage Irrigated Soil Profile in the Tri-County Agricultural Area, Northeast Florida

Permanent Link: http://ufdc.ufl.edu/UFE0022744/00001

Material Information

Title: Physical Characteristics of a Seepage Irrigated Soil Profile in the Tri-County Agricultural Area, Northeast Florida
Physical Description: 1 online resource (75 p.)
Language: english
Creator: Acharya, Subodh
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

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

Notes

Abstract: Physical properties significantly influence hydraulic characteristics of soils, which in turn determine the water and nutrient management practices and efficiencies of an agricultural field. In the Tri-County Agricultural Area of northeast Florida (TCAA), the sandy nature of surface soil overlying an impervious shallow hardpan has allowed controlled fluctuations of the perched water table levels for successfully irrigating crops. The irrigation system, referred to as ?seepage irrigation?, however, is inherently inefficient in terms of water use despite its popularity and manageability. The potential loss of nutrients from the agricultural areas of northeast Florida and the consequent nonpoint source (NPS) pollution in the Lower St. Johns River (LSJR) are attributed to the inefficient irrigation system. Potato best management practices (BMPs) in the area have been developed and implemented in order to increase the nutrient use efficiency and reduce environmental losses. However, potential nutrient loss from agricultural fields is determined to a great extent by physical characteristics of the soil profile. The data on physical properties of the soil profile and the water restrictive hardpan layer in the TCAA, on the other hand, is limited. Seventy soil core samples were collected at 22.5cm, 45cm, 67.5cm, 90cm and 120cm depths from a 4.2ha seepage irrigated field in 26m?26m grids. Particle-size distribution (PSD), saturated hydraulic conductivity (Ks), and bulk density (Db) at the five depths and moisture retention capacity at the first three depths were determined with the respective standard protocols. Horizontal and vertical variations in soil properties were analyzed with statistical and geostatistical techniques. The PSD showed > 90% of sand in most of the soil samples. Clay content (%clay) and Db increased with depth, the highest values being at 120cm depth. The first three sampling depths showed little change in Ks while it decreased sharply at 120cm depth, indicating the beginning of the hardpan. Moisture release curves determined for the first three soil depths indicated a low capacity for moisture retention as the soils progressively dried out. Variability of Db and Ks was least in the soil samples collected from the hardpan (120cm depth). Percentage clay on the other hand, was least variable at 22.5cm followed by 120cm depth. Highest variation of Ks and %clay was observed at 90cm while for Db, highest variation was found at 67.5cm followed by 90cm. The spatial variation in soil properties differed greatly with depth at the sampling interval of 26meters. A large nugget effect was prevalent in most of the cases while pure nugget effect was also observed in some of the property-depth combinations. The Db values were significantly higher and the Ks values were significantly lower within the hardpan with little variability, indicating that vertical movement of water and solute below this compact layer was minimal. An evaluation of potential water loss from the experimental field under seepage irrigation showed that as high as 87% of water received by the field as irrigation and/or rainfall could be lost by surface and subsurface drainage. Overall results of the study indicated that the nature of the soil profile in the field could encourage a steady state subsurface lateral flow of water (SLFW) through the soil profile thereby increasing the potential of substantial amounts of water and nutrient loss from the field.
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 Subodh Acharya.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Mylavarapu, Sambasiva R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022744:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022744/00001

Material Information

Title: Physical Characteristics of a Seepage Irrigated Soil Profile in the Tri-County Agricultural Area, Northeast Florida
Physical Description: 1 online resource (75 p.)
Language: english
Creator: Acharya, Subodh
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

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

Notes

Abstract: Physical properties significantly influence hydraulic characteristics of soils, which in turn determine the water and nutrient management practices and efficiencies of an agricultural field. In the Tri-County Agricultural Area of northeast Florida (TCAA), the sandy nature of surface soil overlying an impervious shallow hardpan has allowed controlled fluctuations of the perched water table levels for successfully irrigating crops. The irrigation system, referred to as ?seepage irrigation?, however, is inherently inefficient in terms of water use despite its popularity and manageability. The potential loss of nutrients from the agricultural areas of northeast Florida and the consequent nonpoint source (NPS) pollution in the Lower St. Johns River (LSJR) are attributed to the inefficient irrigation system. Potato best management practices (BMPs) in the area have been developed and implemented in order to increase the nutrient use efficiency and reduce environmental losses. However, potential nutrient loss from agricultural fields is determined to a great extent by physical characteristics of the soil profile. The data on physical properties of the soil profile and the water restrictive hardpan layer in the TCAA, on the other hand, is limited. Seventy soil core samples were collected at 22.5cm, 45cm, 67.5cm, 90cm and 120cm depths from a 4.2ha seepage irrigated field in 26m?26m grids. Particle-size distribution (PSD), saturated hydraulic conductivity (Ks), and bulk density (Db) at the five depths and moisture retention capacity at the first three depths were determined with the respective standard protocols. Horizontal and vertical variations in soil properties were analyzed with statistical and geostatistical techniques. The PSD showed > 90% of sand in most of the soil samples. Clay content (%clay) and Db increased with depth, the highest values being at 120cm depth. The first three sampling depths showed little change in Ks while it decreased sharply at 120cm depth, indicating the beginning of the hardpan. Moisture release curves determined for the first three soil depths indicated a low capacity for moisture retention as the soils progressively dried out. Variability of Db and Ks was least in the soil samples collected from the hardpan (120cm depth). Percentage clay on the other hand, was least variable at 22.5cm followed by 120cm depth. Highest variation of Ks and %clay was observed at 90cm while for Db, highest variation was found at 67.5cm followed by 90cm. The spatial variation in soil properties differed greatly with depth at the sampling interval of 26meters. A large nugget effect was prevalent in most of the cases while pure nugget effect was also observed in some of the property-depth combinations. The Db values were significantly higher and the Ks values were significantly lower within the hardpan with little variability, indicating that vertical movement of water and solute below this compact layer was minimal. An evaluation of potential water loss from the experimental field under seepage irrigation showed that as high as 87% of water received by the field as irrigation and/or rainfall could be lost by surface and subsurface drainage. Overall results of the study indicated that the nature of the soil profile in the field could encourage a steady state subsurface lateral flow of water (SLFW) through the soil profile thereby increasing the potential of substantial amounts of water and nutrient loss from the field.
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 Subodh Acharya.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Mylavarapu, Sambasiva R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022744:00001


This item has the following downloads:


Full Text

PAGE 1

PHYSICAL CHARACTERISTICS OF A SEEPAGE IRRIGATED SOIL PROFILE IN THE TRI-COUNTY AGRICULTURAL AR EA, NORTHEAST FLORIDA By SUBODH ACHARYA A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Subodh Acharya

PAGE 3

3 To my wife, my parents, and my sisters

PAGE 4

4 ACKNOWLEDGMENTS I express m y sincere gratitude to my adviso r, Dr. Rao S. Mylavarapu for his valuable guidance and support. I also thank my committee members (Dr. Chad Hutchinson, Dr. Yuncong Li, and Dr. Ying Ouyang) for their suggestions a nd corrections. I am grateful to Kelley Hines, Dipika Hindupur, Jordan Pugh, and Ananta Raj Ac harya for their help in sampling, lab-works, and data preparation throughout my research. I thank the farm crew at Hastings Research Unit farm, whose cooperation helped me in several ways to complete my study in time. Finally, I thank my parents, my sisters, and my lovely wife, whose selfless love has always helped me reach my goals.

PAGE 5

5 TABLE OF CONTENTS Page ACKNOWLEDGMENTS...............................................................................................................4LIST OF TABLES................................................................................................................. ..........7LIST OF FIGURES.........................................................................................................................8ABSTRACT...................................................................................................................................10 CHAP TER 1 INTRODUCTION..................................................................................................................12TriCounty Agricultural Area................................................................................................ 12Current Management Practices in the TCAA.................................................................. 13Fertilizer Management.....................................................................................................13Irrigation Management....................................................................................................13Environmental Impacts.................................................................................................... 14Objectives...............................................................................................................................152 LITERATURE REVIEW.......................................................................................................17Seepage Irrigation............................................................................................................. ......17Subsurface Lateral Water Flow.............................................................................................. 18Nutrient Problem in LSJR......................................................................................................19Soil Physical Properties..........................................................................................................20Moisture Retention Characteristics.................................................................................20Saturated Hydraulic Conductivity (Ks)............................................................................21Spatial Variation in Soil Properties........................................................................................ 223 MATERIALS AND METHODS........................................................................................... 25Site Description......................................................................................................................25Sampling Design.....................................................................................................................25Sample Collection...................................................................................................................27Sample Analyses.....................................................................................................................27Particle-Size Analysis......................................................................................................28Bulk Density (Db) and Saturated Hydraulic Conductivity (Ks).......................................28Moisture Retention..........................................................................................................29Statistical Analysis........................................................................................................... .......31Geostatistical Analysis............................................................................................................31Potential Water Loss from the Soil Profile............................................................................. 32

PAGE 6

6 4 RESULTS AND DISCUSSION............................................................................................. 36Statistical Analysis........................................................................................................... .......36Particle-Size Distribution................................................................................................36Bulk Density....................................................................................................................37Saturated Hydraulic Conductivity................................................................................... 38Moisture Retention..........................................................................................................38Spatial Analysis............................................................................................................... .......40Potential Water Loss from the Soil Profile............................................................................. 435 SUMMARY AND CONCLUSION....................................................................................... 65REFERENCES..............................................................................................................................68BIOGRAPHICAL SKETCH.........................................................................................................75

PAGE 7

7 LIST OF TABLES Table page 4-1 Summary statistics of PSD at the five sam pling depths.................................................... 464-2 Summary statistics of the v, Db, and Ks at the five sampling depths................................474-3 ANOVA and mean comparison results for %clay, Db and Ks...........................................484-4 Semivariogram parameters of the soil properties at differe nt sampling depths.................484-5 Calculated ET and net rainfall (ET RF) values for years 2006 and 2006 potato seasons...............................................................................................................................494-6 Calculated amounts of irrigation and dr ainage from a 4.7ha field during the potato seasons of 2006 and 2007..................................................................................................494-7 Average water discharge and ir rigation rate into the field................................................. 50

PAGE 8

8 LIST OF FIGURES Figure page 1-1 Tri-county agricultural area............................................................................................... 161-2 Water table control/drainage channel................................................................................ 163-1 Potato crop planted in beds separated by a water furrow..................................................343-2 Experimental field......................................................................................................... .....343-3 Soil sampling locations in the field.................................................................................... 353-4 Schematic presentation of Ks determination in the lab...................................................... 354-1 Average PSD as a function of depth: (a) %clay; (b) %silt; (c) % sand............................. 514-2 Coefficient variation of %cla y at the five sampling depths............................................... 524-3 Frequency distributions of %clay at the five sampling depths: (a) 22.5cm; (b) 45cm; (c) 67.5cm; (d) 90cm; (e) 120cm....................................................................................... 534-4 Bulk density as a function of depth....................................................................................544-5 Distribution of Db at the five sampling depths................................................................... 544-6 Coefficient of Variation of Db as a function of depth........................................................ 554-7 Coefficient of Variation of Ks at the five sampling depths................................................ 554-8 Saturated hydrau lic conductivity (Ks) as a function of depth............................................ 564-9 Distribution of Ks values at the five sampling depths........................................................ 564-10 Frequency dist ributions of Ks (cmhr-1) at the five sampling depths. (a) 22.5cm; (b) 45cm; (c) 67.5cm; (d) 90cm; (e) 120cm............................................................................ 574-11 Distributions of v in the soil samples at four different pressure levels. (a) 0kPa; (b) -33kPa; (c) -500kPa; (d) -1500kPa.................................................................................... 584-12 Soil moisture release curves of the soil samples at three depths....................................... 594-13 Semivariogram of the log-transformed %clay. (a) 90cm; (b) 120cm................................ 604-14 Semivariograms of the log-transformed Ks values at the five sampling depths. (a) 22.5cm; (b) 45cm; (c) 67.5cm; (d) 90cm; (e) 120cm........................................................ 61

PAGE 9

9 4-15 Semivariograms of the log-transformed Db at the five sampling depths. (a) 22.5cm; (b) 45cm; (c) 67.5cm; (d) 90cm; (e) 120cm...................................................................... 624-16 Average water table depth in the experime ntal field during sixweek crop periods of 2006 and 2007....................................................................................................................634-17 Water balance in a typical ag ricultural field in the TCAA................................................ 64

PAGE 10

10 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science PHYSICAL CHARACTERISTICS OF A SEEPAGE IRRIGATED SOIL PROFILE IN THE TRI-COUNTY AGRICULTURAL AR EA, NORTHEAST FLORIDA By Subodh Acharya August, 2008 Chair: Rao S. Mylavarapu Major: Soil and Water Science Physical properties significan tly influence hydraulic characte ristics of soils, which in turn determine the water and nutrient management practices and efficiencies of an agricultural field. In the Tri-County Agricu ltural Area of northeast Florid a (TCAA), the sandy nature of surface soil overlying an impervious shallow har dpan has allowed controlled fluctuations of the perched water table levels for successfully irriga ting crops. The irrigation system, referred to as seepage irrigation, however, is inherently inefficient in te rms of water use despite its popularity and manageability. The po tential loss of nutrients from the agricultural areas of northeast Florida and the conse quent nonpoint source (NPS) pollu tion in the Lower St. Johns River (LSJR) are attributed to the inefficien t irrigation system. Potato best management practices (BMPs) in the area have been developed and implemented in order to increase the nutrient use efficiency and reduce environmental losses. However, potenti al nutrient loss from agricultural fields is de termined to a great extent by physical characteristics of the soil profile. The data on physical properties of the soil profile and the water restrictive hardpan layer in the TCAA, on the other hand, is limited. Seventy soil core samples were collected at 22.5cm, 45cm, 67.5cm, 90cm and 120cm depths from a 4.2ha seep age irrigated field in 26mm grids. Particle-size distribut ion (PSD), saturated hydraulic conductivity (Ks), and bulk density (Db) at

PAGE 11

11 the five depths and moisture retention capacity at the first three depths were determined with the respective standard protocols. Hori zontal and vertical variations in soil properties were analyzed with statistical and geostatistical techniques. The PSD showed >90% of sand in most of the soil samples. Clay content (%clay) and Db increased with depth, the hi ghest values being at 120cm depth. The first three sampling dept hs showed little change in Ks while it decreased sharply at 120cm depth, indicating the beginning of the har dpan. Moisture release curves determined for the first three soil depths indicated a low cap acity for moisture retention as the soils progressively dried out. Variability of Db and Ks was least in the soil samples collected from the hardpan (120cm depth). Percentage clay on th e other hand, was least variab le at 22.5cm followed by 120cm depth. Highest variation of Ks and %clay was observed at 90cm while for Db, highest variation was found at 67.5cm followed by 90cm. The spatial variation in soil properties differed greatly with depth at the sampling interval of 26meters. A large nugget effect was prevalent in most of the cases while pure nugget effect was also observed in some of the property-depth combinations. The Db values were significantly higher and the Ks values were significantly lower within the hardpan with little variability, indicating that vertical movement of water and solute below this compact layer was minimal. An evaluation of potentia l water loss from the experimental field under seepage irrigation showed that as high as 87% of water received by the field as irrigation and/or rain fall could be lost by surface a nd subsurface drainage. Overall results of the study indicated that the nature of the soil profile in the field could encourage a steady state subsurface lateral flow of water (S LFW) through the soil profile thereby increasing the potential of substantial amounts of wa ter and nutrient loss from the field.

PAGE 12

12 CHAPTER 1 INTRODUCTION TriCounty Agricultural Area Tri-County Agricultural Area (TCAA) locat ed along the no rtheastern coastline of Florida covers approximately 15,000 ha cropland of St. Johns, Putnam and Flagler counties. The TCAA is the most important potato producing area in the state, accounting approximately 60% of total potato production (USDA, 2008). In 2007, the total production of spring potato in Florida was approximately 354,120 metric tons harvested from 11,000ha. Approximately 59% (209,420 metric tons) of the total production wa s contributed by Hasti ngs area of the TCAA alone (USDA, 2008). Both area and production were higher in 2008 than in the previous year with approximately 223,620 metric tons (61%) production from 6880ha Hastings area potato farms. Besides potato, other important crops in the TCAA are cabbage and cole crops. A typical one year crop rotation in the TCAA includes spring potato followed by a cover crop in the summer. Sorghum and corn are the most common cover crops in the area. Due to sandy nature of the soil, nutrient and water holding capacities ar e generally low. This area is characterized by a naturally occurring shallow hardpan at approximately one meter depth, with low Ks and high Db. The subsurface hardpan restricts vertical move ment of water and a resultant perched water table is built up, as the influx of water through surface irrigation and/or rainfall typically far exceeds the Ks of the hardpan. The level of the perched water table can be appropriately altered by controlling the lateral drainage through adjusting the riser boards in the discharge channels at the edges of the fields.

PAGE 13

13 Current Management Practices in the TCAA Much of the area in TCAA is planted with chipping variety of potato (var. Atlantic ) during the spring season (February to May) follo wed by summ er cover-crops such as corn or sorghum. A typical crop field in TCAA is divided into severa l crop beds separated by water furrows. Each crop bed is further divided in to 16-20 crop rows with 101cm center-to-center distance from one another and raised approximately 25cm above the alleys. Fertilizer Management Successful p otato production in TCAA has been accomplished for several years by the application of large amount of fe rtilizers. Nitrogen, being the mo st limiting nutrient, is typically applied in excess of the recommended amount. Average nitrogen application rate by producers in TCAA potato fields is 285 kg/ha (Hutchinson, 2005) whereas the recommended rates of fertilization in the ar ea are 224:168:67 kg N: P2O5:K2O per ha under soil test results of medium soil phosphorus content (Hutchin son et al., 2007). Although P2O5 is recommended only when the soil test value is Medium or Low, gr owers may prefer to apply additional amount of phosphorus to their crop as an insurance against crop failure. Fertilizer applications are typically in two installments. Complete dose of P2O5 and half the recommended dose of N and K2O are applied at the time of planting. The remaining N and K2O amounts are applied 35-40 days after planting (Hutchinson et al., 2007). Due the low nutr ient holding capacity of the soil, soluble N can leach into the perched water table and become unavailable to the crops. Application of fertilizers in split doses increases the bioavailability of N and K2O by decreasing leaching and ensures better production even unde r the conditions of leaching. Irrigation Management In seepage irrigation, the water table is m aintained at 20-25cm below the furrows between crop rows (Campbell et al., 1978) duri ng the entire cropping period. Water is supplied

PAGE 14

14 to the water furrows instead of applying to individual crop rows. As the water flows along the furrows, it infiltrates down and raises the perched water table. Water then moves into crop root zone by capillarity and subsurf ace lateral flow (Smajstrla et al., 2000). Any undesirable rise in the perched water table is cont rolled by means of a check-gate c onstructed at end of the ditch. The subsurface seepage irrigation practiced in TCAA is inherently an inefficient system because a large quantity of wate r is needed to maintain the water table at desired depth. According to Sjmastrla (1991), efficiency of seepage irrigation system can vary from 20-70%. The magnitude of water-loss from the system is both timeand site-specific because it depends on the permeability and continuity of the restrictive layer, and the management practices of the surrounding fields. Environmental Impacts Successful agricultural production requires a bundant use of fert ilizers coupled with adequate supply of water to th e crops. On the other hand, these two im portant management considerations increase the pot ential of nutrient loss, especi ally in high rainfall areas. The average annual precipitation in the area is approximately 125cm (FAWN, 2008). Summer rainfall is predominantly through short duration thunderstorms due to which a high potential for leaching exists. Application of N fertilizer can often exceed the recommended amounts as insurance against farmers percei ved risk of crop failure (Munoz et al., 2008). Due to fluctuating shallow water table, fertilizers can be lost ra pidly from crop root z one (Bonczek and McNeal, 1996). Nutrient loading into surface water bodies promotes rapid algal growth and accelerates the process of eutrophication (S JRWMD, 1996). Nitrogen contam ination of LSJR has been increased rapidly in recent years and it is estimated that 5 to 20% of the total nutrient pollution in LSJR can be attributed to the row crops of TCAA (SJRWMD, 1996). With a goal to increase fertilizer application efficiency and reduce leach ing loss, BMPs have been implemented in the

PAGE 15

15 TCAA to comply with the Total Daily Maximu m Load (TMDL) set under the Clean Water Act (WCA). Physical characteristics determine how much water soils can hold and can supply to crop roots, affecting the efficiency of a water management system. Permeability of underlying soil horizons has a significant effect on soil hydraulic properties (Blume et al., 1987). According to Ezeaku et al (2006), the potential for ground water contamination and the use of perched water for agricultural production in an area depends largely on the phys ical properties of the soil profile and the underlying hardpan la yer. It is therefore important to characterize the soil profile with respect to its physical prope rties. The knowledge on soil propert ies is also important for the assessment of alternative water and nutrient mana gement practices in the area. Soil properties vary significantly both in horizontal and vertical direction. Field scale sp atial variability study in these properties is therefore impor tant in developing site specif ic farming practices which can reduce the potential of NPS pollution in LSJR. Objectives The objectives of this study were to1. determ ine the physical and hydraulic properties of soils in the TCAA 2. study the field scale variation in soil physical properties bot h horizontally and vertically 3. evaluate the potential of water loss from the agricultural fields of TCAA under the conventional seepage irrigation system

PAGE 16

16 Figure 1-1. Tri-county agricultural area Figure 1-2. Water table c ontrol/drainage channel

PAGE 17

17 CHAPTER 2 LITERATURE REVIEW Seepage Irrigation Most of the gravity-flow irrigated farm s in Florida are under seepage irrigation system, as surface irrigation is rarely us ed in the state (Sjmastrla and Haman, 1998). Se epage irrigation or simply subirrigation is preferred because of its cost effectiveness and low maintenance requirement (Haman et al., 1989) while being simple to operate and effective for crop production (Bonczek and McNeal, 1996). Seepage irri gation system also has another advantage; the open ditches commonly used for irrigation also function as drainage systems (Sjmastrla et al., 1984). Due to a significant increase in the us e of water resource as well as the cost of installation and operation of irrigation systems, the design of efficient water management system has become more critical than before (Rosa, 2000). Under such a situation, conventional seepage irrigation may not be sustainable as it consumes a large amount of water. Clark and Stanley (1992) compared the conventional ditc h conveyance seepage irrigation with a fully enclosed subirrigation system a nd reported that the fully enclosed system was more efficient in uniform water distribution as well as drainage in subirrigated fields. Sjmastrla et al (1984) studied whether the water use and energy cons umption efficiency of conventional seepage irrigation could be increased by controlled water applications using an automatic float controller. Although thei r study showed only a small increase (8%) in the efficiency, it was pointed out that a proper and timely water table control system could have had better efficiency over the conventional system. Sjmastrla et al (2000) conducte d another experiment to study the feasibility of using automatic subsurface drip irrigation (SDI) system in a potato field in the TCAA. They reported that SDI system controlled the water table mo re accurately at the desired level and used

PAGE 18

18 approximately 36% less water than the conventi onal seepage system, but consumed 70% more energy, producing a similar yield as the seepage system. Cost effectiveness of the conventional seepage system and the lack of significant incr ease in the yield under SDI system therefore did not provide an economic incentive to growers towards changing the irrigation method to SDI. Subsurface Lateral Water Flow Subsurface lateral flow of water (SLFW ) is a process by which water is transmitted laterally through the soil profile. This flow is usually ignored or misinterpreted while studying the water balance of an ar ea (Styles and Burt, 1999). Howe ver, SLFW from fertilized agricultural lands is a major m echanism for moving nitrates a nd other agrochemicals to ground and surface water (Starr et al., 2005). Shaw et al (2001), from a tracer study in the upper coastal plain of Georgia, reported the la teral movement of an applied tr acer, which had occurred due to reduced hydraulic conductivity of lower horizons. Subsurface horizons having reduced hydraulic conductivity impeded the downward flow thus encouraging lateral movement. Occurrence of lateral flow though the permeable horizons overlying such water restrictive horizons has been re ported in different studies (e.g Blume et al., 1987; Bosch et al., 1999). According to Reuter et al (1998), rapid solute transport vi a perched water table occurs in the permeable horizons overlying hydraulically rest rictive fragipan or argillic horizon. This may hasten the offsite carriage of agrochemicals th ereby impairing the wate r quality of surface and subsurface water bodies. In a re cent tracer study performed in the TCAA (Mylavarapu et al., 2008), it was revealed that a tracer injected into a 4.7ha field was quickly removed by subsurface lateral pathways. In a typical seepag e irrigated field of TCAA, where a constant depth of perched water table was maintain ed throughout the crop seasons, the SLFW is typically governed by a steady state darcian flow given as follows.

PAGE 19

19 Q = Ks i A (2-1) where, Q = Rate of subsurface lateral flow, Ks = Saturated hydraulic conductivity of the soil i = Hydraulic gradient of the field given as dl hh dl dh i 21 (2-2) where, h1 and h2 are the observed head at the start a nd end points of the fields respectively. The head is the column of perched water above the impervious layer. A = Cross-sectional area perpendicu lar to the direction of the flow The application of Darcys equation implies that Ks of the soil profile and hydraulic gradient of the field have a great influe nce on the total amount of SLFW from a field. According to Wilson et al (2002) the subsurface water movement is highly variable and site specific. Water moves rapidly form certain areas while little or no subsurface flow occurs in other areas, which is primarily due to the high variability in Ks of the soil profile. Nutrient Problem in LSJR Irrigated agriculture no tably increases crop productivity, but consum es high volumes of water and may induce off-site pollution of receiving water bodies (Causap et al, 2004). According to USEPA (2004), agricultural NPS pollu tion is the leading sour ce of water quality impacts to surveyed rivers and lakes, the th ird largest source of im pairments to surveyed estuaries and a major contribu tor to ground water contamination and wetlands degradation. According to the St. Johns River Water Management District (SJRWMD) (1996), approximately 5 to 20% of the nutrient polluti on in LSJR is attributed to the row crop agriculture of TCAA. The most severe algal blooms in the LSJR have been observed in the freshwater reach near Palatka which coincides wi th the peak runoffs from the potato fields in the TCAA (SJRWMD, 1996).

PAGE 20

20 In the past, there have been a few studi es regarding nutrient leaching from the subsurface seepage irrigated fields of TCAA. Wi th the growing concerns of NPS pollution due to agricultural activities in the area, researches have been directed to wards reducing the leaching losses and consequent NPS pollution. The main goal s of these studies have been to develop the BMPs for seepage irrigated vegetable and pot atoes in TCAA thereby reducing potential of nutrient losses. These BMPs are promoted in th e area through a cost share program developed by SJRWMD (Livingston-Way, 2000), which provi des the growers an economic incentive to voluntarily adopt BMPs without incurring any loss (Hutchin son et al., 2002). The BMPs implemented in the area include a defined set of fertilizer and water management practices aimed in increasing nutrient application effi ciency while reducing leaching losses. Other practices such as monitori ng water table depth through obs ervation wells and proper crop rotation are also parts of the BMPs. Soil Physical Properties Moisture Retention Characteristics Determ ination of soil moisture release characteristic curves is one of the most important measurements for characterizing soil hydraulic pr operties as it will allo w for assessment of water storage capacity of soils (Townend et al., 2001). According to Pachepsky et al (2001), soil moisture retention data is used in simulati ons of water and chemical transport in soil, estimations of water holding capacity and irri gation requirements, and assessment of water sorptivity to predict infiltration rates. Unger (1975) indicated that soil moisture storage capacity had a great influence in the management of irrigation water and precipitation for crop production. According to Hanks (1992) soil moisture characteristic curves are very nonlinear in nature. This means that for a gi ven change of water content ( ) at one value, the change of matric potential ( m) will be different than at another va lue. According to Hillel (1971), the

PAGE 21

21 amount of water retained at low tensions depe nds primarily on the pore-size distribution and soil structure whereas water retention at the hi gher pressures is due to adsorption and is influenced by soil texture and the surface area of clay fraction. Sandy soils contain large pores therefore most of the water is retained at low suctions. Clayey soils, on the other hand, release small amount of water at low suctions and retain a large proportion of water even at higher suctions. Jamison and Kroth (1958 ) indicated that, the order of available moisture storage capacity of particle-size fractions should be coa rse silt>fine silt>clay>fine sand>coarse sand from theoretical considerations. Hence, an in crease or decrease in any textural components tends to change available st orage capacity by its influen ce upon pore size distribution. Compaction of a soil reduces the total pore spaces, particularly macropore spaces that retain water at low suctions. According to Hillel (1998), saturated water content and the initial decrease of water content with the application of low suction are diminished due to compaction. Reeve et al (1973) reported that Db exerted a profound influence on water retention properties of soils but the effect varied betw een texture groups and horizons. Saturated Hydraulic Conductivity (Ks) Saturated hydraulic cond uctivity of soils is the ability to transmit water (Klute and Dirksen, 1986) when all the soil pores are filled with wate r. This is a very important soil water property from the standpoint of irrigati on and drainage as well as envi ronmental pollution. Information on Ks of surface soils is critical for agronomic a nd water management strategies including the design of irrigation systems. On the other hand, the knowledge on Ks of the shallow hardpan present in a soil profile is important from both agronomic and environmental perspectives as it influences the extent of solute and chemical movement through the soils. Surface water run-off, erosion, and deep percolation are directly affected by Ks of the soil (Suleiman and Ritchie, 2001).

PAGE 22

22 Although Ks is a soil property that is difficult to obtain (Ahuja et al., 1993), it is needed to run the water balance of many crop simu lation models (Suleiman and Ritchie, 2001). Saturated hydraulic cond uctivity of soils is determined by the PSD and the level of compaction. Soils dominated by sand particles have high Ks due to the presence of large number of macropores while in clay dominated soils, the opposite is usually true. According to Nakano and Miyazaki (2005), Ks is intrinsically related to the macropores in the soils and not necessarily related to the aver age soil porosity or average Db. Because of the influence of macropores, the variation of Ks is mostly higher than other physical properties. Mohanty et al (1994) observed different variability for the Ks values measured with different methods and indicated that such variability had occurred due to the pres ence and absence of open ended macropores in the soil as well as due to variab le soil compaction during core extraction. Some scientists have also indicated that the variability of Ks tends to decrease as the sampling volume increases (e.g. Anderson and Bouma, 1973; Bouma 1982). The number of continuous macropores in large samples is likely to be less than in small samples which can produce uniform Ks observations. Spatial Variation in Soil Properties Physical p roperties of the soils such as PSD, Ks, Db and, water retention vary from one location to another within an area which tends to be correlated in space, both horizontally and vertically (Warrick et al, 1986). According to Iqbal et al (2005), spat ial variability of soil physical properties within or am ong agricultural fields is inherent in nature due to geologic and pedologic soil forming processes bu t some of the variability may be induced by tillage and other management practices. Due to within field vari ation, uniform management of fields may often result in overapplication of inputs in low yielding areas and underapplic ation of inputs in highyielding areas (Ferguson et al., 200 2). Fine scale information on the spatial variability of soil is

PAGE 23

23 therefore necessary for implementing site speci fic management approa ches (McBratney and Pringle, 1999). Spatial distributi ons of soil properties at field a nd watershed scales may affect yield potential, hydrologic proces ses, and transport of nutrients and chemicals to surface or ground water (Cambardella et al., 1994). According to Russo (1986), it is important to analyze the spatial distribution of soil pr operties in the field in orde r to suggest improved irrigation management schemes. Gupta et al (2006) indi cated that an understanding of spatial and temporal variation in hydraulic properties of so ils was crucial for char acterizing the rate of water flow and solute transport through the soil profile. Accord ing to Mulla (1988), field scale modeling of attributes like inf iltration and solute transport re quires the knowledge of spatial patterns in soil water content, Ks and evaporation. Camberdella et al (1994) studied the spatial variability of im portant soil properties and classified their spatial dependence as weak, moderate, and strong based on the ratio of nugget to sill expressed as the percentage. They found a moderate spatial dependence of the Db along with several other soil properties. Gaston et al (2001) reported that the clay content of a surface soil in Mississippi delta was described by spherical semivariogram model. The range of the %clay was 220m and the nugget variance was approximately 30% of the total semivariance. Study of the spatial variation of soil properties over a large area many not always be feasible due to various constraints. Due to this reason, several experiments have been conducted in field scales to study the spat ial variation of soil properties. For example, Jung et al (2006) studied the spatial variability of soil properties in a 4ha field using 30m 30m grid sampling and reported that clay and silt c ontents at 15-30cm depth of th e soil profile were spatially autocorrelated at a separation distance of 40m. Duffera et al (2006) studi ed the spatial variation of texture, SWC at different pressures, Ks, and Db in a 12ha field in North Carolina. They found

PAGE 24

24 moderate to strong spatial autoco rrelation in the soil properties sa mpled in 60m grid intervals at five depths up to 72cm. The range s of clay content and Db varied from 63m to 411m while the range was not obtained for the Ks at 60m sampling distance. Sp atial dependence was strong at the first 4 depths while it was moderate at the last sampling de pth (72m). Sahandeh et al (2005) studied the spatial and temporal variation of various soil nitrog en parameters including clay content with the help of 63 samples collected at 8m m from a 2.5ha corn field. They reported that the spatial characteristics of soil te xture and other related pr operties varied greatly with depth and landscape position. Mulla (1988) intensively sampled two 660m long transects at 20cm spacing and analyzed the spatial variability of texture, SWC, and surface soil temperature using spherical semivariograms. Although different ranges of spatial variation in sand, clay, and SWC were obtained in the two transects, all the values were between 60m and 100m. No nugget effect was obtained for the water content in both transects. However, the nugget effects for clay and sand contents ranged from <3% to 26% and 11% to 13% respectively, indicating that the strength of spatial association in the properties was variab le in the two transects. Cemek et al (2007) studied the Ks, PH, exchangeable sodium percentage (ESP) and cation exchange capacity (CEC) of soil samples collected from 60 sampling si tes at four depths. They reported that Ks had the strongest spatial dependence while it was more variable at the top soil layer These literatures indicate that the inform ation on physical properties of soil profile characterized by high wate r table and a shallow ha rdpan layer is still lacking. Therefore, the objective of this experiment was to study th e spatial variation in soil texture, Db, and Ks of the soil profile and the shallow ha rdpan of a typical seepage i rrigated field in the TCAA.

PAGE 25

25 CHAPTER 3 MATERIALS AND METHODS Site Description The study was perform ed in a 4.7ha (approx.260 m 182m) seepage irrigated field of UF/IFAS Research and Demonstration Site lo cated at Hastings, Florida. The soil at the experiment site is classified as Ellzey fine sand (sandy, siliceous, hyperthermic, arenic ochraqualf; 90-95% sand, <2.5 % clay, <5% silt) (USDA, 1983). The field is planted to spring potato, which is the main crop, followed by sorghu m or corn during summer as the cover crops. The field resembled typical TCAA agricultura l fields and consisted 14 crop beds each approximately 17.2m wide. Each crop bed wa s further divided into 16 rows raised approximately 25cm above the alleys (furrows between each crop row). The irrigation furrow separating each crop bed was approximately 1.01m wide. The distance from the first crop row of one bed and the last row of the adjacent be d was approximately 3m which allowed for the movement of farm vehicles, part icularly during the cropping season. The irrigation of the experimental field was through seepage system as described previously. During crop seasons, the water furrows con tinuously conveyed water along the bedrows towards a drainage ditch constructed perpe ndicular to the field in the east-west direction. A check gate installed at the western end of the di tch controlled the level of water. The level of water in the drainage ditch in turn controlled the perched wate r table depth under the crop beds as determined. A few inches below the crop root zone, the soil profile was saturated throughout the cropping season and water was supplied to the root zone by upward capillary movement. Sampling Design Representative soil sam pling is critical to th e success of any fiel d scale study. According to Petersen and Calvin (1986), th e purpose of sampling is to estim ate the various parameters of

PAGE 26

26 a population with accuracy. A sampling plan that fails to resemble the entire population in any one of the attributes of interest will result in inconsistent results. Therefore, a best suited sampling design that falls within the research bounds should always be employed, depending on the nature and variability of the attributes. Since soils are highly heterogeneous and characterized by high variation, it is necessary to consider the nature of the variation during sample collection for the proper representation of a particular soil population (Petersen and Calvin, 1986). Soil samples can be collected either randomly from the study area or in a systematic way. Random sampling is the easiest method in which soil samples are collected at randomly selected locations throughout the study area. Random sampling is unbiased and hence every point within the study area has an equal chance of being sampled. However, in several studies random sampling may not be appropriate especia lly when some of th e soil properties under consideration are not distributed uniformly over the entire study area. Under such a condition, random sampling is likely to miss th e true variation of soil proper ties and therefore a different sampling schemes needs to be adopted in order to address the issue. In our case, regular grid samp ling scheme was used to collect soil samples from the soil profile. Grid sampling is usually used when intensive sampling is necessary from a relatively smaller area so that the extent of spatial varia tion in the properties can be represented in the field. The experimental field was first divided into 70 square gr ids of size 26m 26m followed by locating the sample spots at the center of eac h grid. The width of the crop beds in the TCAA fields vary from 18 to 24m (16-20 crop rows). Therefore, the sampling scale of 26m was considered suitable to analyze whether any spatial autocorrelation in soil properties existed among the crop beds.

PAGE 27

27 The sampling points in the field were determined with the help of measuring tapes. In order to locate the sampling spots, first a point at the southeast corner of the field was selected as the reference point. The boundary lines of the field were then determined by the right-angled triangle method using measuring tapes. On bot h the boundary lines, a point at 13m from the corner was flagged since the first set of sampling-points was located at 13m from the boundaries (Figure 3-3). Using th e measuring tapes, a perpendicular line passing through the flagged point on one of the boundary line was locate d. The sample spots on the line were then located by stretching the tape along this perpendi cular line. The process was repeated for all other sampling locations at every 26m. Sample Collection Soil sam ples were collected from the soil prof ile up to 1.2m at five different depths: 22.5cm, 45cm, 67.5cm, 90cm, and 120cm. At the fi rst three depths, undisturbed soil cores were collected using a core sampler along with bulk soil samples. Ho wever, undisturbed soil samples could not be collected at the last two depths due to excessive mois ture content in the soil profile. Only disturbed soil samples were therefore collected at these two depths. In order to collect the bulk soil samples from 90cm and 120cm depths, fi rst a hole was drille d down to the sampling depth in the soil profile with the help of a tractor-mounted augur and the sample was taken using a hand augur. After sample collection, the actual coordinates of the samplingpoints in the field were recorded using a GPS device (Trimb le Inc.). The recorded coordinates were then differentially corrected in or der to achieve high precision. Sample Analyses The soil samples were analyzed for their physical properties at the Soil Moisture Laboratory of Soil and Water Science Departm e nt, University of Florida. The measured physical properties included the texture, Db, Ks and soil moisture characteristic curves.

PAGE 28

28 Particle-Size Analysis The PSD of the soil sam ples was determined for all the five depths using the hydrometer method (Gee and Bauder, 1986).Dry and sieved soil samples were first saturated with water in beakers and 27% hydrogen peroxide solution was added to the soils for digestion of organic matter, which took approximately 2-3 days. After digestion, the samples were oven dried and subject to dispersion by adding sodium hexameta phosphate solution. The content of the cylinder was then stirred for 3 minutes to disperse the so il particles. The final volume of the suspension was brought to one liter by adding water. The cylinder was then closed and shaken with hands to ensure proper dispersion. A hydrometer was dropped gently into the suspension and the particles were allowed to settle down for 40 seconds This was the time when all the particles >5 mm were settled down at the bot tom of the cylinder. The dens ity of the suspension at 40 seconds was measured from which the combin ed silt and clay content was obtained. After 40 seconds the cylinders were shaken again, hydrometers were dropped into each suspension, and left for two hours. By this time most of the silt partic les present in the soil would be settled down at the bottom and the dens ity of the suspension wo uld solely be due to the clay particles suspended in water. From this hydrometer reading, %clay was calculated. Sand content (%sand) and silt content (%silt) we re then obtained by using the two observations. Bulk Density (Db) and Saturated Hydraulic Conductivity (Ks) Bulk density of the intact soil cores collected at the first three depths was determined as described in (Blake and Hartge, 1986). The cylinders containing intact soil cores were oven dried after recording the combined weight of the cylinder and the soil. The oven dried cylinders with soil were weighed in order to calculate the mass of oven dry soil, which was then used to calculate Db.

PAGE 29

29 Saturated hydraulic conductiv ity was determined by constant head method (Klute and Dirksen, 1986) which is described in the Darcy s equation. The cylinders with soil were covered by cheese cloth at one of the ends and saturated overnight. A constant head 3cm was maintained for 5 minutes at the top of each saturated soil core and the volume of discharge from the lower end was measured. The Ks values were calculated using th e following expression as given in Klute and Dirksen (1986). HAt QL Ks (3-1) where, Q = volume of discharge collected L = length of the soil core A = cross-sectional area of the core t = time H = hydraulic head difference across the soil core The Db and Ks of the soil samples at 90cm and 120c m depth were determined from the reconstructed samples as undisturbed cores could not be collected due to field conditions. The bulk samples were reconstructed by careful tramp ling with hands in the laboratory. In order to achieve high structural integrity, the bulk samples collected in the field were maintained as intact as possible. While reconstr ucting a core was inserted into th e soil and pressed gently from both the sides to maintain the integrity of th e core. Care was taken not to apply very high pressure so that the total porosity of th e samples would not decrease significantly. Moisture Retention The soil m oisture retention by the undisturbe d soil cores at different pressures ranging from 0kPa to -1500kPa were determined by usin g tempe cells for low pressures (0kPa to 33kPa) and, pressure chambers for high pressures (-500kPa and -1500kPa). The method comprised of repetitive weighing of the tempe cell assembly (which included the soil core with

PAGE 30

30 the cylinder, a porous ceramic plate, two rubber rings and the two tempe cell caps) before and after applying a particular pressure to the soil. Tempe cell method used to determine the moisture retention is based on the principle that when the pressure inside the tempe cell is increased above the atmosp heric level, it forces the soil water to move out through the pores pres ent in the ceramic plate. The ceramic plate was boiled for several minutes to expel any air remain ing in the pores prior to assembling the tempe cells. After assembling the cells, they were satura ted over night in order to bring the soil water potential to zero. As the positive pressure was applied, water was extracted from the pores. Larger pores release water earlier than the smaller pores as micropores hold water with greater tenacity. Continuity of the pores between the soil sample and the ceramic plate was maintained as the pressures are increased gradually starting at saturation point. Positive pressures were created by increasing the water level in the columns as appropriately as required. After a through extraction of moisture and reaching equilibrium at each pressure, the soil sample assembly was wei ghed and then exposed to next higher pressure. After reaching the equilibrium at -33kPa pre ssure, the assembly was weighed and then disassembled in order to record the combined we ight of the soil core and cylinder. The rings were then resaturated and expos ed to higher pressures in the pressure chamber followed by weighing in the end of each step. Finally, all the samples were oven dried in order to get the weight of the oven dry soil. Bulk dens ity and the volumetric water content ( v) at respective pressures were determined from the data collected. Soil moisture release curves were determined only for the first three depths (22.5cm, 45cm, and 67cm). Due to shallow water tabl e depths maintained during potato season, undisturbed samples could not be collected from 90cm and 120cm depths. Twenty one soil

PAGE 31

31 cores were randomly selected from each set of sa mples at each depth for a total of 63 samples for determination of moisture retention characteristics. Statistical Analysis Com parison of depth wise measurements of PSD, Db, Ks and moisture retention was performed using SAS (SAS Inc., 2005) and R (R Development Core Team, 2008) statistical software. The distribution and variation in th e soil properties was ev aluated using general statistical tools. Depth wise comparison of the %clay, Db, and Ks was performed by ANOVA followed by Tukeys test. Geostatistical Analysis Classical s tatistical approaches assume that the measurements of soil properties are independent of space, that is, they are random ly distributed. Since th e soil properties are spatially autocorrelated, these techniques may not be able to explain the underlying variation in the attributes. Spatial autocorrelation of soil properties renders classical statistical tools inadequate to explain the varia tion in those properties. The vari ation of soil properties with respect to space can be better explained by geosta tistical techniques. Geostatistical study of an attribute assumes that the observations in sp ace have some connection or continuity among themselves (Gutjahr, 1985). Geostatistical techni ques are based on the theory of regionalized variables and stationarity which provide theoretical basis for analysis of spatial dependence of the attribute under study using sp atial autocorrelation or semivariograms (Trangmar et al., 1985). Spatial dependence of physical properties of the soil profile including hardpan was studied using geostatisti cal methods. Spatial dependence in the soil properties are commonly described with a semivariogram, which study sp atial structure in th e data distribution (Kravchenko, 2003). Semivariogram is the plot of semivariance against the separation distance

PAGE 32

32 between the samples, also called separation distance or Lag. Semivari ance is defined as the half of the estimated squared differences betw een the sample observations at a given Lag (Trangmar et al., 1985). The estimated semivariance at a given lag is given by the following equation. )]()([2)(2 1 )(hxiAXiAhN h (3-2) where, A(Xi) and A(Xi+ h) are the observations of the regionalized variable A at locations X and X+h respectively, h is the lag and N is the number of pair of observations separated by the lag h. Semivariograms, thus, illustrate the change in the semivariance of the soil property data with changing lag distance thus giving an idea of the exte nt of spatial autocorrelation in the measurements. Experimental semivariograms were created at all the five depths for Db and Ks while for %clay, semivariograms were created only at 90cm and 120cm. Semivariograms of %clay at 22.6cm, 45cm and 67.5cm could not be created due to severe skewness in the data. Prior to producing the semivariograms, the datasets were tested for normality using Shapiro and Wilk Test (Shapiro and Wilk, 1965). Any variable that deviated significantly from normality was logtransformed in order to bring th e variable close to normality. The experimental semivariograms produced fo r each property were fitted with a best fitting model using wherever possible using Gamma Design Software in order to obtain the semivariogram parameters nugget, sill and range. Potential Water Loss from the Soil Profile Due to the nature of the water m anagement system, a large amount of water applied to the fields of TCAA is lost without being u tilized. A significant por tion of drainage is contributed by subsurface lateral flow because there is no vertical movement due to the

PAGE 33

33 occurrence of the hardpan and the shallow water table. Since system reaches a steady state soon after initiating water app lication to the crop, the total amount of water ap plied is equal to the total amount drained plus the total amount lost by evapotranspiration (ET). Potential loss of water from the experimental field during potato seasons was therefore calculated using the ET, total rainfall (RF) and the total amount of irri gation applied during the seasons. Rainfall data of Hastings area was obtained from Florida Au tomated Weather Network (FAWN, 2008) for the required time periods. The data was then us ed to calculate average daily rainfall for approximately four months time period. Evapot ranspiration rates for potato was calculated using the potato crop-coefficient (Kc) reported by Singleton ( 1990) where different Kc values for February, March, April, a nd May months corresponding to potato growth stages were reported. The ET values for the four months crop growth period were therefore calculated using the corresponding Kc and ET0 values which were then used to calculate the average daily ET rates in mm per day (mmd-1). At steady state, there is no vert ical movement of water which implies that the drainage is only in the lateral direction. Unde r seepage irrigation drainage is considered to be predominated by subsurface lateral flow. Therefore the tota l amount of drainage fr om the field can be expressed as followsQd = Qi [ET RF] (3-3) where, Qd = Rate of drainage (mm d-1) Qi = Rate of irrigation (mm d-1) ET RF = Net rainfall (+/-, mm d-1) Average daily loss of water calculated for th e experimental field therefore included the average daily precipitation receiv ed during the potato seasons.

PAGE 34

34 Figure 3-1. Potato crop planted in beds separated by a water furrow Figure 3-2. Experimental field

PAGE 35

35 Figure 3-3. Soil sampling locations in the field Figure 3-4. Schematic presentation of Ks determination in the lab

PAGE 36

36 CHAPTER 4 RESULTS AND DISCUSSION Statistical Analysis Particle-Size Distribution The summary statistics (average, v ariance, maximum and minimum, and coefficient of variation) of the PSD at the five depths are pr esented in Table 4-1. At all sampling depths, the average sand content was >90%. However, the average %sand content decreased by 6% from 22.5cm to 120cm depth while the average clay content increased by >5%. The maximum amount of clay recorded was 11.64% at 120cm de pth. The average %silt was low (<2%) at all sampling depths with most of the samples having 0-1%. Higher clay content in subsurface soil layers was also reported by Bo sch et al (1999) in a sandy loam profile of Georgia, where a prominent layer of clay was found in the soil pr ofile from 1 to 4m depth which was responsible for low and high Increased clay content of subsurface soil layers was possibly due to gradual translocation and deposit ion of the clay particles cause d by fluctuating water table. Deposition of fine clay part icles tends to decrease Ks of a soil horizon because fine clay particles can occupy void spaces be tween coarse soil particles thus reducing the permeability of the horizon. Although the maximum variation of %clay was ob served at 120cm depth, the coefficient of variation (CV) was highest in the three intervening depths between 22.5cm and 120cm. The magnitude of CVs decreased in the following order90cm>67.5cm>45cm>120cm>22.5cm. The least CV and the least average %clay at 22.5cm depth indicated that the soil at the surface was predominantly sandy throughout the field. On the other hand, higher average %clay and low CV at 120cm suggested a uniform distribution of cl ay at the hardpan layer. High CVs of the

PAGE 37

37 intervening soil layers was probably because of the entrapment of clay particles during vertical translocation. The frequency distributions of %clay at the fi rst four depths were highly skewed to the right, showing a maximum frequency in the range 1 to 4%. On the other hand, the distribution at 120cm was relatively symmetrical The frequency of %clay obser vations at 120cm was highest in the range 6 to 11%, which further affirmed that the clay particles were uniformly distributed at the hardpan layer. Analysis of variance of PSD data showed th at the average %clay at upper three sampling depths were similar while the values at 90cm a nd 120cm were significantly higher than the first three depths (P-value < 0.001). At the same time, the average %clay values at 120cm depth were significantly higher than the values at 90cm. Bulk Density The summary statistics f or Db is presented in Table 4-2. Bulk density increased significantly with increasing depth (P-value < 0.05). However, the increase was highest at 120cm depth supporting the results of observed PSD data. Variation in Db values was relatively lower in the horizontal direction (CV = 3-7%) than in the vertical direction (CV = 10%). These values were similar to the CV values reported by Warrick and Nielsen (1980). Similar to %clay, Db exhibited highest variation in the two intervening soil laye rs (67.5cm and 90cm) while the least variation was observed in the 22.5cm and 120cm samples. Although the CV values of %clay and Db differed significantly, depth wise change in variability was similar in both properties (Figure 4-2 and Figure 4-5).

PAGE 38

38 Saturated Hydraulic Conductivity Summ ary statistics for Ks is presented in Table 4-2. Saturated hydrau lic conductivity showed a relatively larger variati on at the first four sampling depths while it was slightly lower at 120cm depth. The CV for the Ks ranged from 61 to 77%. Satura ted hydraulic conductivity of the soils is regarded as a highly variable soil property with possible CV at >100% (Warrick and Nielsen, 1980). Relatively lower CV for observed Ks values in the soil profile therefore indicated that the number of samples was ad equate to provide a good estimation of the Ks of the whole population. Despite the higher variation in Ks values within each of th e first four sampling depths, the average Ks among the depths was similar. At 120cm depth however the average Ks value was significantly lower. Lower Ks values at 22.5cm depth were possibly because of temporary compaction tractor wheels or because of the organic matter content present in the soil. Average Ks for 120cm was 3.8 cmhr-1 although several values ranged from 0.13cmhr-1 to 2cmhr-1. In contrast, samples at 90cm and above had significantly higher Ks values. The sudden increase in %clay, Db, combined with a decrease in Ks values at 120cm depth suggested that the hardpan layer in the soil profile began at a depth between 90cm 120cm from the surface. Moisture Retention Average v at field capacity (33kPa) was approximately 15% while at permanent wilting point (-1500kPa), it was close to 5% or less at all three depths. There was no significant difference among the three depths with respect to average v values at 0kPa, -33kPa, -500kPa and -1500kPa pressures. At 0kPa (saturation), v increased with depth wh ile at higher pressures (-500kPa and -1500kPa), the water c ontent decreased. At -33kPa (field capacity), on the other hand, the distribution of v at 67.5cm depth was quite dissimila r from the distribution at other

PAGE 39

39 three pressure levels. Although the median v at that depth was close to the median value at 22.5cm, greater number of observations lied within the lower 50 percentile of the distribution, indicating that despite th e observed difference in v distribution, the values tended to decrease in the subsurface at higher pressures. This phenomenon not only supported the notion of the presence of macropores in subsurface layers but also suggested that th e plough layer of the soil profile contained relatively highe r amount of organic matter which helped to retain more water at higher pressures. At higher pressure, the wa ter retained in the soil is mostly in the micropores, which is either due to %clay, %silt or organi c matter content. However, as revealed by the texture data, the soil at 22.5cm depth was ve ry low in %clay. Moistu re holding capacity of the surface soil layer was enhanced possibly due to orga nic matter presence but at the subsequent two depths, lack of organic matter and increased bulk density without a corresponding increase in silt or clay content resulted in low microporosity. Therefore, water retention by the soil samples at 22.5cm could be attributed to the organic matter content which was routinely incorporated in the field every year. Volumetric water content exhibited a high variability at all depths. The CV of v was highest at -1500kPa (permanent wilting point) while the CV was lowest at 0kPa (saturation). This high variability could be due to either a sm all number of samples or due to the differences among soil samples in the number of macropores caused by small sample-size. Moisture release curves for the three depths were characterized with the gradual decline in v at lower pressures (0kPa to -33kPa) followed by a steep fa ll between -33kPa and -500kPa, confirming the low water holding cap acity of the soils in the field. Moisture release curves at all the three depths were in close agreement with typical curves for highly sandy soils.

PAGE 40

40 Spatial Analysis Spatial variation analysis of the soil propert ies at each sampling depth was studied with semivariograms. Experimental semi variograms produced from the %clay, Db and Ks data were fitted with a best-fit model in order to dete rmine the parameters of the semivariograms. Semivariogram parameters were calculated by fitting the three common models which are given by the following formulae (Clark, 1979). 1. Linear model : This model fits a straight line th rough semivariogram points The linear model is given by (h) = C0 + m h for 0 h a (4-1) (h) = (h) = C0 + C1 for h a (4-2) where, m = slope of the best fit line h = lag distance C0 = nugget C1= sill a = range 2. Spherical Model : This model is given by the formula 3 102 1 2 3 )( a h a h CCh for 0 h a (4-3) (h) = (h) = C0 + C1 for h a 3. Exponential model : This model is given by the formula 0 1 0exp1 )( a h CCh for 0 h d (4-4) where, d is the maximum distance over which the variogram is defined and a0 = a/3 Semivariogram parameters nugget, sill and range provide valuable information on the spatial dependence of soil properties. The Nugget -tosill ratio can be used as a measure of the

PAGE 41

41 strength of spatial dependence while the range is helpful in designing sampling schemes for future studies. Experimental and fitted semivariogr ams for the log transformed %clay, Db, and Ks are presented in figures 4-13, 4-14, a nd 4-15 respectively. The parameters, nugget sill and range of the semivariograms are presented in Tabl e 4-4. The semivariograms of the %clay were calculated only for the 90cm and 120cm depth as the distribution of %clay in the upper three depths was severely skewed. More than 90% of the samples in the upper three depths had 3% or less of clay. The spatial varia tion of %clay at 90cm and 120c m was described by exponential and spherical models respectively. The range and nugget of the semivariograms were higher at 120cm than at 90cm depth whereas the sill was approximately equal at both sampling depths. On the other hand, the semivariogram was smoo ther at 120cm (R2 = 0.99) than at 90cm (R2 = 0.82). Values of nugget and range of %clay were similar to thos e obtained by Iqba l et al (2005) in a deep soil horizon (1m depth) in Mississippi delta. Spatial va riation of %clay at 90cm and 120cm depth could be attributed to the varying rates of illuviation in the profile. However, presence of large nugget effect at 120cm (44%) suggested that a denser sampling was necessary for explaining all the vari ation in %clay at 120cm. Saturated hydraulic conductiv ity did not show any spat ial autocorrelation at the sampling interval of 26m except at the 22.5cm depth (Figure 4-14). Even at 22.5cm, a high nugget effect was found while the range of spatial variation could not be defined. Semivariograms of the Ks values of lower sampling depths we re hard to fit to any model and could be described as pure nugget effects The Ks at these sampling depths behaved more randomly instead of showing any spatial autocorrelation. High nugget variance of Ks had also been reported in various the l iteratures (e.g., Gupta et al., 2006; Mallants et al., 1997; Bosch and

PAGE 42

42 West, 1998). It was thus found in our case that, for the Ks values, classical statistical tools were more suitable than the spatial statistical techniques at the sampling interval of 26m. Sample collection at distance closer than 26m would probably be necessary to obtain the spatial association in Ks values Semivariograms of Db for the five sampling depths ar e presented in Figure 4-15. There was no spatial association observed in Db at 22.5cm depth while a relatively better autocorrelation was observed at the subsequent three depths. Linear, exponential and linear semivariograms models were fitted, respectively, at 45cm, 67.5cm and 90cm sampling depths. The best correlation of the fitted model was found at 67.5cm depth (R2> 95%) with nugget and range values of 0.005 (35%) and 135m respectively. However, at 45cm and 90cm depths, nugget variances were approximately 54% and 62% respectively indicating the existence of a weak spatial autocorrelation. Also, the range of spatial dependence was larger than the length of the field at 45cm and 90cm depths. At 120cm, Db values exhibited a pure nugget effect or a random variation possibly because of coarser sampling resolution. In summary, spatial dependence of soil pr operties differed ba sed on the depth of sampling. At 120cm, there was no spatia l autocorrelation in both the Ks and Db values but the spatial variation of %clay was well described by a spherical model. The range of spatial autocorrelation was larger than the length of the field for some properties while the range was well defined for others. The nugget effect was prevalent in the semivariograms showing that the sampling interval of 26m was not adequate to ex plain the total spatial variation present in the properties. Sampling at a higher intensity was requ ired for adequately explaining the field scale spatial variations in the soil properties.

PAGE 43

43 Potential Water Loss from the Soil Profile Above results confirm ed that the soil pr ofile of a typical field in TCAA was characterized by highly permeable sandy layers overlying an impermeable hardpan. The results also indicated that Ks was higher and more variable in th e surface layers than in the hardpan suggesting the possibility of occurrence of m acropore flow through the soil profile during storm events. Subsurface hardpan layer present in the soil profile of the TCAA is being used advantageously for crop irrigation by creating a perched water table for several decades. Although the occurrence of shallo w hardpan has been beneficial for easy and cost effective water management for crop production, the water management system itself however exposes a high potential of NPS pollution into nearby water bodi es. In seepage irrigation, soil profile immediately below the root zone is saturated with the perched water and the root-zone is wetted through capillary rise from the pe rched water table. Continuous application of water into the water furrows allows for raising the perched wate r table level, which in turn is controlled by a riser board installed in the drai nage channel. The height of th e riser board is adjusted to facilitate precise controls on the water table level based on the water flowing from the spigots at the north-end of the field. Excess water will dr ain over the riser board, thus creating a steady state lateral subsurface wa ter flow between the NE the SW e dges of the field. Subsurface lateral flow is therefore the single most significant mechanism of water loss from TCAA fields during a cropping season. In a recent study performed in the area by Myla varapu et al (2008), the movement of an injected tracer was monitored in a seepage irri gated with a network of 90 wells over a six weeks period. It was reported that the tracer injected at NE-corner of the experimental field was quickly removed from the field by lateral subsurf ace flow. The distribution of tracer in the field

PAGE 44

44 and its subsequent removal by subsurface late ral flow was more rapid during the summer seasons when there was frequent rainfall, sugge sting that the subsurface lateral movement was hastened by precipitation. High precipitation that occurs in TCAA duri ng the crop seasons can raise the perched water table up to the crop root zone, which potentially can flush a large amount of nutrient out of the crop root zone. Figure 4-16 shows the average water table de pth in a typical TCAA potato field during the spri ng and summer seasons of 2006 and 2007 respectively. Due to shallow water table, even a relatively small amo unt of rainfall can raise the perched water table up to the crop root-zone, which dissolves the nu trients applied to the cr op. When the water table settles down, the nutrients are transported belo w the root-zone and subsequently removed from the field by lateral transport. Inefficient use of irrigation water and potenti al nutrient leaching are therefore the two most important disadvantages of seepage irri gation. In a typical spring potato season of approximately 100 days, the total amount of wate r applied to a seepage irrigated field is approximately 58.32 million L or 583,200 L daily. Average daily ET (mmd-1), average RF (mmd-1), and average daily net-rainfall (ET-RF) (mmd-1) during the potato seasons of 2006 and 2007 are presented in table 4-5. Approximate amount of water drainage (surface and subsur face) for the two years (2006 and 2007) is presented in table 4-6. The crop coefficient for potato crop ranged fr om 0.41 in February to a high of 1.35 in April (flowering and tubering stag e) before reducing to 1.28 in May before harvest stage. The calculated ET based on the Kc values during the crop growth period ranged from 1.028 to 5.635 mm d-1. The average amount of water that was lo st through lateral drainage (surface and

PAGE 45

45 subsurface) was 509,184 L d-1 or 50.9 million L in 100 days of a typical potato crop produced under seepage irrigation system in the TCAA. Although a portion of drainage occurs as a surface flow, a significant amount of water drains as a lateral subsurface flow. The amount of water being lost through surface flow can possibly be determined by setting up measuring devi ces at the drainage end of each of the water furrows which can then be used to estimate the extent of subsurface lateral flow. Based on the above results, approximately 87% of applied water was lost via lateral flow (surface and subsurface), after meeting the ET demands during a typical potato season. This massive water loss from the fields also has the potential to remove plant nutrients from the crop root-zone. Because of the macroporous nature of surface soil layers, frequent storm events during the crop seasons can generate macropore flow thus hast ening the leaching process. Therefore, such losses may likely result in lowered nutrient use efficiencies and crop deficiencies rather than water quality concerns due to sign ificant dilution effect caused by the continuous application of large volume of irrigation water.

PAGE 46

46 Table 4-1. Summary statistics of PSD at the five sampling depths Soil Properties Depth of Sampling(cm) Average (%)Variance Max Min CV (%) 22.5 2.08 0.35 5.64 1.64 28.4 45 2.09 0.903 9.00 1.64 45.5 67.5 2.25 1.74 10.00 1.64 58.6 90 4.04 6.34 11.00 1.64 62.3 %clay 120 7.51 7.63 11.64 1.64 36.8 22.5 97.22 1.45 98.36 92.03 1.2 45 96.88 2.59 98.36 88.00 1.7 67.5 97.05 3.66 98.36 87.00 2.0 90 94.33 10.05 98.36 86 3.4 %sand 120 91.66 7.59 98.36 87.36 3.0 22.5 0.64 0.72 4.00 0.00 132.6 45 1.01 1.28 4.66 0.00 112.0 67.5 0.71 1.18 4.67 0.00 153.0 90 1.61 1.63 6.00 0.00 79.3 %silt 120 0.81 0.62 3.67 0.00 97.2

PAGE 47

47 Table 4-2. Summary statistics of the v, Db, and Ks at the five sampling depths Soil Properties Depth of Sampling (cm) Average Variance Max Min CV (%) 22.5 0.365 0.01512 0.6350 0.154 33.69 45 0.360 0.00365 0.5011 0.259 16.78 v 0kPa (cm3cm-3) 67.5 0.395 0.00485 0.5839 0.321 17.63 22.5 0.176 0.00955 0.3139 0.012 55.54 45 0.155 0.00715 0.3041 0.0256 54.56 v -33kPa (cm3cm-3) 67.5 0.157 0.01362 0.3325 0.0014 74.34 22.5 0.064 0.00044 0.1094 0.0333 32.78 45 0.0482 0.00063 0.1247 0.0102 51.99 v -500kPa (cm3cm-3) 67.5 0.056 0.00288 0.2158 0.0185 95.90 22.5 0.047 0.00059 0.0907 0.0041 51.55 45 0.036 0.00074 0.1136 0.0025 75.31 v 1500kPa (cm3cm-3) 67.5 0.047 0.00234 0.1845 0.0125 102.90 22.5 13.099 103.68 52.920 1.05 77.73 45 13.41 107.67 57.110 3.410 77.38 67.5 15.288 104.49 36.680 1.05 66.86 90 10.953 87.21 31.23 0.140 85.26 Ks (cmhr-1) 120 3.833 5.586 14.670 0.130 61.66 22.5 1.37 0.007 1.66 1.21 6.11 45 1.42 0.0092 1.69 1.22 6.75 67.5 1.52 0.0129 1.76 1.28 7.47 90 1.60 0.0122 1.81 1.41 6.90 Db (gmcm-3) 120 1.78 0.0032 1.94 1.62 3.18

PAGE 48

48 Table 4-3. ANOVA and mean comp arison results for %clay, Db and Ks Soil properties Sampling depth(cm) Average F-value P-value (ANOVA) 22.5 2.08* %clay 45 2.09* 67.5 2.25* 76.00 <.0001 90 4.04** 120 7.51** 22.5 1.37** Db 45 1.42** 67.5 1.52** 170.33 <.0001 90 1.60** 120 1.78** 22.5 13.09* Ks 45 13.41* 67.5 15.28* 9.65 <.0001 90 10.95* 120 3.833** Not significantly different; ** significantly different Table 4-4. Semivariogram parameters of the soil properties at diffe rent sampling depths Soil Propertie s Depth (cm) Model R2 Lag (m) Nugget Sill % Nugget Range (m) 90 Exponential 0.82 36 0.009 0.346 2.601 78 %clay 120 Spherical 0.99 36 0.135 0.304 44.57 164.5 45 Linear 0.84 31 0.0066 0.011 59.72 67.5 Exponential 0.98 29 0.0052 0.0145 35.58 135 Db 90 Linear 0.96 30 0.0094 0.0145 64.82 Ks 22.5 Linear 0.88 33 0.356 0.746

PAGE 49

49 Table 4-5. Calculated ET and net rainfall (ET RF) values for years 2006 and 2006 potato seasons 2006 2007 RF* Kc ET0 ET* ET-RF* RF Kc ET0 ET ET-RF Feb 5.4 0.41 2.28 1.03 -4.37 0.06 0.41 2.43 1.10 1.04 Mar 0.5 0.99 2.82 2.80 2.30 3.4 0.99 2.65 2.62 -0.73 Apr 0.8 1.35 3.86 5.21 4.40 1.04 1.35 3.44 4.64 3.60 May 0.3 1.28 4.4 5.63 5.37 1.02 1.28 3.92 5.01 1.10 *. Daily average values; unit: mmd-1. Table 4-6. Calculated amounts of irrigation and drainage from a 4.7ha field during the potato seasons of 2006 and 2007 2006 2007 Qi (100Ld-1) Qi (mmd-1) ET-RF (mmd-1) Qd (mmd-1) Qd (100Ld-1) Qi (100Ld-1) Qi (mmd-1) ET-RF (mmd-1) Qd (mmd-1) Qd (100Ld-1) Feb 5832 12.46 -4.37 16.9 7899.85832 12.46 1.04 11.4 5344.6 Mar 5832 12.46 2.30 10.2 4754.9 5832 12.46 -0.73 13.2 6172.9 Apr 5832 12.46 4.40 8.1 3772.15832 12.46 3.60 8.9 4146.5 May 5832 12.46 5.37 7.1 3318.1 5832 12.46 1.10 11.4 5316.5 Avg. 5832 12.46 -4.42 10.6 4937.45832 12.46 1.04 11.21 5246.3 *. Total drainage that also accounts for th e net rainfall (net rainfall = ET RF). Ld-1 = Liters per day

PAGE 50

50 Table 4-7. Average water discharge and irrigation rate into the field Reading Spigot 1 (Ls-1) Spigot 9 (Ls-1) Spigot 13 (Ls-1) Spigot 15 (Ls-1) Pump Meter (Ls-1) Qi (Ld-1) Qi (mmd-1) 1 0.48 0.31 0.48 0.53 7.14 2 0.45 0.3 0.5 0.53 7.28 583,200 12.46 3 0.43 0.31 0.5 0.56 7.28 AVG 0.46 0.31 0.49 0.54 7.23 Ls-1 = Liters per second

PAGE 51

51 (a) (b) (c) Figure 4-1. Average PSD as a function of depth: (a) %clay; (b) %silt; (c) % sand

PAGE 52

52 Figure 4-2. Coefficient va riation of %clay at th e five sampling depths

PAGE 53

53 (a) (b) (c) (d) (e) Figure 4-3. Frequency distributi ons of %clay at the five samp ling depths: (a) 22.5cm; (b) 45cm; (c) 67.5cm; (d) 90cm; (e) 120cm

PAGE 54

54 Figure 4-4. Bulk density as a function of depth Figure 4-5. Distribution of Db at the five sampling depths

PAGE 55

55 Figure 4-6. Coefficien t of Variation of Db as a function of depth Figure 4-7. Coefficien t of Variation of Ks at the five sampling depths

PAGE 56

56 Figure 4-8. Saturated hydraulic conductivity (Ks) as a function of depth Figure 4-9. Distribution of Ks values at the five sampling depths

PAGE 57

57 (a) (b) (c) (d) (e) Figure 4-10. Frequenc y distributions of Ks (cmhr-1) at the five sampling depths. (a) 22.5cm; (b) 45cm; (c) 67.5cm; (d) 90cm; (e) 120cm

PAGE 58

58 (a) (b) (b) (d) Figure 4-11. Distributions of v in the soil samples at four different pressure levels. (a) 0kPa; (b) -33kPa; (c) -500kPa; (d) -1500kPa

PAGE 59

59 Figure 4-12. Soil moisture release curves of the soil samples at three depths

PAGE 60

60 (a) (b) Figure 4-13. Semivariogram of the l og-transformed %clay. (a) 90cm; (b) 120cm

PAGE 61

61 (a) (b) (c) (d) (e) Figure 4-14. Semivariograms of the log-transformed Ks values at the five sampling depths. (a) 22.5cm; (b) 45cm; (c) 67.5cm; (d) 90cm; (e) 120cm

PAGE 62

62 (a) (b) (c) (d) (e) Figure 4-15. Semivariograms of the log-transformed Db at the five sampling depths. (a) 22.5cm; (b) 45cm; (c) 67.5cm; (d) 90cm; (e) 120cm

PAGE 63

63 Figure 4-16. Average water table depth in the experimental fiel d during six-week crop periods of 2006 and 2007

PAGE 64

64 Figure 4-17. Water balance in a t ypical agricultural field in the TCAA

PAGE 65

65 CHAPTER 5 SUMMARY AND CONCLUSION Result of this soil profile study thus provided valuable inform ation on the characteristics of soil properties at different depths in a seepag e irrigated field in the TCAA. Clay content of the soil profile increased with depth, while the %sand decreased. The difference in clay percent was insignificant at the first four depths but was significant at 120cm indicating deposition of clay particles in the hardpan. Vert ical translocation of fine clay particles and their deposition in the lower horizons is common in the high wate r table soils. The TCAA soils are continuously wetted by the perched water table at an average depth of 40-50cm during most parts of the year. Continuous flushing of the profile by fluctuating water table e nhanced illuviation process and deposition of clay particles in the hardpan. The av erage silt percentage in the soil was very low at all sampling depths. Bulk density also exhib ited a similar increase pattern with the sampling depth; the increase was abrupt and steep in the impervious zone. Saturated hydraulic conductivity was not significantly different at th e first three depths but it dropped significantly at 90cm and further reduced at 120cm thus confirming the results of %clay and Db. Results indicated that creation of perche d water table above th e hardpan occurred rapidly after the field received water from irrigation and/ or rainfall, as the infiltration rate of overlying layers is very high. However as the hardpan was not complete ly impervious, it was indicated that some amount of water also moved vert ically down through the hardpan. Moisture retention characteri stics determined through seque ntial increase in pressures from 0kPa to -1500kPa indicated lo w water retention in the soil pr ofile, particularly at higher pressures. Average water content at 0kPa was found to be near 37% which dropped to about 16% at -33kPa. This implied that >50% of the total water retained in the soil at saturation was contributed by macropores. The aver age percentage of water held at -1500kPa was 4% or less.

PAGE 66

66 Variability in measured soil properties wa s observed in both horizontal and vertical direction. The CV was high for %clay, %silt, Ks, and v, while it was low for Db and %sand, which supported the variations reported in lite rature (Warrick and Nielsen, 1985). Overall variation in the vertical dire ction was high in %clay, %silt, v at -1500kPa, and Ks while the variation was comparatively lower in Db, %sand, and v at 0kPa and -33kPa. Spatial variation analysis pe rformed with the semivariogram s indicated the prevalence of large nugget effects in all the soil properties studi ed. The variation of %clay, which was studied only in 90cm and 120cm, wa s well described by exponential (R2 = 0.82) and spherical (R2 = 0.99) models re spectively. The nugget variance in 90cm was only 2.6% of the total semivariance while it was 44% at 120cm. The range was 78m and 164m respectively in the two depths thus indicating strong and moderate spatial variation of clay percentage. Variation in Ks did not show any spatial autocorrel ation at depths below 22.5cm (pure nugget effect). At 22.5cm the variation was best described by a linear model (R2 = 0.88) with a nugget of 47%, while the range was greater than the total length of the experimental field, indicating a weak spatial autocorrelation in Ks at the sampling distance of 26m. The nugget effect, which occurred either due to measuremen t error or due to spatial autocorrelation at distance shorter than the sampling interval, mo st likely would be reduced if the sampling resolution could be increased. Unlike Ks, Db did not show any spatial autocorr elation at 22.5cm depth while the autocorrelation could be described by linear, exponential and linear models at 45cm, 67.5cm and 90cm respectively. The Nugget variance ranged from 35% to 63% indicating moderate to weak spatial variation at these depths At the hardpan layer, however, pure nugget effect was observed in the semivariograms of Db as well.

PAGE 67

67 Potential water loss from the experiment field was calculated using the ET and RF data which showed that at steady state, as high as 87% of the total water receive d by the field (including total rainfall) was lost in the potato seasons years 2006 a nd 2007. The result indicated the potential of a substantial loss of not only the irrigation water but al so the nutrients applied to the crops In conclusion, the soil profile of the experiment field was characterized with increased %clay, increased Db and reduced Ks at the water restrictive horizon (120cm) indicating the effect of continuously standing high water table in the area. Th e crop root-zone had very high and variable Ks and a low moisture holding capacity show ing the possible existence of vertical macropore flow during precipitation. There was moderate to very w eak spatial variation in the measured soil properties. Occurrence of high nugget effects suggested that the sampling distance was higher than that w ould be required for capturing the total spatial variability. The results, in conclusion, indicated that the nature of the soil pr ofile in the TCAA could encourage subsurface lateral transport which could further be intensified by the nature of irrigation system and high rainfall. Massive water loss from the pr ofile thus suggested that the seepage irrigated fields of the TCAA were also prone to reduced nutrient use efficiencies while potentially increasing nonpoint source pollution in the Lower St. Johns River. Therefore, there is an immediate need for developing alternate ways of water management in the TCAA that can effectively reduce water and nutrient loss wh ile maintaining the level of production.

PAGE 68

68 REFERENCES Ahuja, L. R., O. W endroth, and D. R. Nielsen. 1993. Relationship between initial drainage of surface soil and average profile saturated conductivity. Soil Sci. Soc. Am. J. 57 (1): 1925. Anderson, J. L., and J. Bouma. 1973. Relationships between saturated hydraulic conductivity and morphometric data of an argillic horizon. Soil Sci. Soc. Am. J.37 (3): 408-413. Beckett, P. H. T., and R. Webster. 1971. Soil va riability a review. So ils and Fert. 34(1): 15. Blake, G. R., and K. H. Hartge. 1986. Bulk density. p. 363-375. In A. Klute (ed.) Methods of soil analysis, Part I. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI. Blume, L. J., H. F. Perkins, and R. K. Hubbard, R. K. 1987. Subsurface water movement in an upland coastal plain soil as influenced by plinthite. Soil Sci. Soc. Am. J. 51: 774. Bonczek, J. L., and B. L. McNeal. 1996. Specifi c gravity effects on fertilizer leaching from surface sources to shallow water tables. Soil Sci. Soc. Am J. 60: 978-985. Bosch, D. D., R. K. Hubbard, R. A. Leonar d, and D.W. Hicks. 1999. Tracer studies of subsurface flow patterns in a sandy loam profile. Trans. ASAE. 42(2): 337-349. Bosch, D., and L. West. 1998. Hydraulic conductivity variability for two sandy soils. Soil Sci. Soc. Am. J. 62 (1): 90-98. Bouma, J. 1982. Measuring the hydraulic c onductivity of soil horiz ons with continuous macropores. Soil Sci. Soc. Am. J. 46 (2): 438-441. Byrd, C. W., and D. K. Cassel.1980. The effect of sand content upon cone index and selected physical properties. Soil Science 129: 197. Cambardella, C. A., T. B. Moorman, T. B. Parkin, D.L. Karlen, J. M. Novak, and R. J. Turco. 1994. Field-scale variability of so il properties in central Iowa soils. Soil Sci. Soc. Am. J. 58 (5): 1501-1511. Campbell, K. L., J. S. Rogers, and D. R. Hensel. 1978. Water table control for potatoes in Florida. Trans. ASAE. 21 (3): 701-705. Causap, J., D. Qulez, and R. Arags. 2004. Assessment of irrigation and environmental quality at the hydrolog ical basin level: I. Irrigation quality. Agri Water Manage. 70 (3):195-209. Cemek, B., M. Gler, K. Kili, Y. Demir, and H. Arslan. 2007. Assessment of spatial variability in some soil properties as related to soil salinity and alkalinity in Bafra plain in northern Turkey. Environ Monit. Assess. 124 (1): 223-234.

PAGE 69

69 Clark, G. A., and C. D. Stanley. 1992. Subirri gation by microirrigation. Applied Eng. Agric.8 (5): 647-652. Clark, I. 1979. Practical geostatis tics. Applied Sci. Publ., London. Duffera, M., J. G. White, and R. Weisz. 2007. Spa tial variability of southeastern U.S. coastal plain soil physical properties: implications for site-specific management. Geoderma. 137 (3/4): 327-339. FAWN. 2008. Florida Automated Weather Network. http://fawn.ifas.ufl.edu Ferguson, R. B., G. W. Hergert, J. S Schepers C. A. Gotway, J. E. Cahoon, and T. A. Peterson. 2002. Site-specific nitrogen management of irrigated maize: yiel d and soil residual nitrate effects. Soil Sci. Soc. Am. J. 66: 544. Franzen, D. W., and T. R. Peck. 1995. Field Soil sa mpling density for variable rate fertilization. J. Prod. Agric. 8:568. Gaston, L. A., M. A. Locke, R. M. Zablotowic z, and K. N. Reddy. 2001. Spatial variability of soil properties and weed populations in the Mississippi de lta. Soil Sci. Soc. Am. J. 65: 449. Gee, G.W., and J. W. Bauder. 1986. Particle-size Analysis. p. 383-411. In A. Klute (ed.) Methods of soil analysis. Part I. 2nd. Ed. Agron. Monogr 9. ASA and SSSA, Madison, WI Geypens, M., L. Vanongeval, N. Vogels, and J. Meykens. 1999. Spa tial variability of agricultural soil fertility parameters in a gl eyic podzol of Belgium. Precision Agric. 1 (3): 319-326. GS+ for Windows. 2001. Gamma Design Soft ware: Version 5.1.1, Plainwell, MI Gupta, S. D., B. P. Mohanty, and J. M. K hne. 2006. Soil hydraulic c onductivities and their spatial and temporal variations in a vert isol. Soil Sci. Soc. Am. J. 70 (6):1872-1881. Gutjahr, A. 1985. Spatial variability : geostatistical methods. p. 9-34. In (Nielsen and Bouma (ed.) Soil spatial variability, Proc. Works hop ISSS and SSSA, Las Vegas, USA, 30 Nov1 Dec. Haman, D. J., A. G. Smajstrla, F. S. Zazuet a, and D. R. Hensel. 1989. Water recycling in seepage irrigation of potatoes. Report Subm itted to St. Johns River Water Management District, Palatka Florida. Hanks, R. J. 1992. Applied soil physics: soil wa ter and temperature applications. SpringerVerlag, New York. Hillel, D. 1998. Environmental soil physics. Academic Press, San Diego, CA.

PAGE 70

70 Hillel, D. 1971. Soil and water: physical principles and proces ses. Academic Press, New York. Hutchinson, C.M., E. H. Simmone, G. J. Hochmuth, W. M. Stall, S. M. Olson, S. E. Webb, T.G. Tylor, and S. A. Smith. 2007. Potato production in Florida. HS 733. Gainesville, Florida. University of Florida. Hutchinson, C. M. 2005. Influence of a controlled release nitrogen fertilizer program on potato ( Solanum tuberosum L.) tuber yield and qualit y. Acta Hort. ISHS. 684:99-102. Hutchinson, C. M., W. A. Tilton, P. K. Li vingstonWay, G. J. Hochmuth. 2002. Best management practices for potato production in northeast Florida. HS 877. Gainesville, Florida. University of Florida. Iqbal, J., J. A. Thomasson, J. N. Jenkins, P. R. Owens, and F. D. Whisler. 2005. Spatial variability analysis of soil phys ical properties of alluvial so ils. Soil Sci. Soc. Am. J. 69 (4): 1338-1350. Jamison, V. C., and E. M. Kroth. 1958. Availabl e moisture storage capacity in relation to textural composition and organic matter conten t of several Missouri soils. Soil Sci. Soc. Am. Proc. 22 (3):189-192. Jung, W. K., N.R. Kitchen, K. A. Sudduth, and S. H. Anderson. 2006. Spatial characteristics of claypan soil properties in an agricultural field. Soil Sci. Soc. Am. J. 70 (4). 1387-1397. Klute, A., and C. Dirksen. 1986. Hydraulic cond uctivity and diffusivity: Laboratory methods. p. 687-734. In A. Klute (ed), Methods of soil anal ysis, Part 1.2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI. Kravchenko, A. N. 2003. Influence of spatial structure on accuracy of interpolation methods. Soil Sci. Soc. Am. J. 67 (5). 1564-1571. Larson, W. E., and P. C. Robert 1991. Farming by soil. p. 103-112. In R. Lal and F. J. Pierce (ed.) Soil management for agriculture. SWCS. Ankeny, IW. Lin, H., D. Wheeler, J. Bell, and L. Wilding. 2005. Assessment of soil sp atial variability at multiple scales. Ecol. Model.182 (3-4): 271-290. Livingston-Way, P. 2000. Tri-county agricultura l area water quality protection cost share program: applicants handbook. St. Johns River Water Management District, Palatka, FL. Mallants, D., B. P. Mohanty, A. Vervoort, a nd J. Feyen. 1997. Spatial analysis of saturated hydraulic conductivity in a soil with macropores. Soil Tech. 10 (2): 115-131. McBratney, A. B., and M. J. Pringle. 1999. Esti mating average and proportional variograms of soil properties and their potential use in precision agriculture. Prec. Agric 1: 125-152.

PAGE 71

71 Meirvenne, M. V. 2003. Is the soil variability within the small fi elds of Flanders structured enough to allow precision agricultu re? Prec. Agric. 4 (2): 193-201. Mohamed, S. B., E. J. Evans, and R. S. Shiel. 1996. Mapping techniques and intensity of soil sampling for precision farming. p. 217-226. In Robert et al. (ed.) Proc. 3rd international Conference on Precision Agriculture. Minneapolis, MN. Mohanty, B. P., R. S. Kanwar, and C. J. Ev erts. 1994. Comparison of saturated hydraulic conductivity measurement methods for a glacialtill soil. Soil Sci. Soc. Am. J. 58 (3): 672-677. Mulla, D. J., 1988. Estimating spatial patterns of water content, matric suction and hydraulic conductivity. Soil Sci. Soc. Am. J. 52: 1547-1553. MunozArboleda, F., R. Mylavarapu, C. Hutc hinson, and K. Portier. 2008. Nitrate-nitrogen concentrations in the perched ground water under seepageirrigated potato cropping systems. J. Environ. Qual. 37: 387-394. Mylavarapu, R. S., C. M. Hutchinson, Y. C. Li, and S. Acharya. 2008. Measurement of lateral water flow in a subsurface seepage irrigation potato production system in the tri-county agricultural area. A Report Submitted to the Lower St. Johns. River Water Management District, Palatka, Florida. Nakano, K., T. Miyazaki. 2005. Predicting the satu rated hydraulic conductivity of compacted subsoils using the non-similar media con cept. Soil Till. Res. 84 (2): 145-153. Newman, B. D., A. R. Campbell, and B. P. W ilcox. 1998. Lateral subsurface flow pathways in a semiarid ponderosa pine hill slope. Wa ter Res. Research. 34 (12): 3485-3496. Nielsen, D., J. Biggar, and K. Erh .1973. Spatial variability of field-measured soil-water properties. Hilgardia. 42 (7): 215-259. Ovalles, F. A., and M. E. Collins. 1988. Evaluati on of soil variability in northwest Florida using geostatistics. Soil Sci. Soc. Am. J. 52: 1702. Pachepsky, Y., W. J. Rawls, and D. Gimenez. 2001. Comparison of soil wate r retention at field and laboratory scales. Soil Sci. Soc. Am. J. 65 (2): 460-462. Perrone, J., and C. A. Madramootoo. 1994. Ch aracterizing bulk density and hydraulic conductivity changes in a potato cropped field. Soil Tech..7 (3): 261-268. Petersen, R. G., and L. D. Calvin. 1986. Sampling. p.33-51. In A. Klute, (ed.) Methods of soil analysis. Part I. 2nd.ed. Agron. Monogr. 9. ASA and SSSA Madison, WI. Pettitt, A. N., and A. B. McBratney. 1993. Samp ling designs for estimating spatial variance components. J. Appl. Stat.42 (1):185-209.

PAGE 72

72 R Development Core Team. 2008. R: A language a nd environment for statistical computing. R Foundation for Statistical Computing, Vi enna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org Reuter, R. J., P. A. McDaniel, J. E. Ha mmel, and A. L. Falen. 1998. Solute transport in seasonal perched water tables in loess-derived soils capes. Soil Sc. Soc. of Am. J. 162 (4):977983. Rosa, J. A. 2000. Modeling water table response to subirrigation with a buried microirrigation line source for potato production. PhD Disse rtation, University of Florida. Russo, D. 1986. A stochastic approach to th e crop yield-irrigation relationships in heterogeneous soils: I. Analysis of the field spatial variability. Soil Sci. Soc. Am. J. 50 (3): 736-745. Saldaa, A., A. Stein, and J. A. Zinck. 1998. Spa tial variability of soil properties at different scales within three terraces of the Henares River (Spain). Catena, 33 ( 3-4): 139-153. SAS Institute. 2005. SAS system for windows. Release 9.1.3. SAS. Inst. Cary, NC. Scott, H. D. 2000. Soil physics: agricultural and environm ental applications, Iowa state university press, Ames, Iowa. Shahandeh, H., A. L. Wright, F. M. Hons, a nd R. J. Lascano. 2005. Spatial and temporal variation of soil nitrog en parameters related to soil texture and corn yield. Agron. J. 97: 772-782. Shapiro, S. S., and M. B. Wilk. 1965. An analysis of variance test for normality. Biometrika. 52: 725-731. Shaw, J. N., D. D. Bosch, L. T. West, C. C. Truman, and D. E. Radcliffe. 2001. Lateral flow in loamy to sandy kandiudults of th e upper coastal plain of Ge orgia (USA). Geoderma. 99: 1-25. Shaw, J.N., West, L. T., C. C. Truman, and D. R. Radcliffe. 1997. Morphologic and hydraulic properties of soils with water restrictive horizons in the Georgia coastal plain. Soil Sci. 162:875. Singleton, V. D. 1990. Investigation of potato wate r use in the Tri-count y agricultural area of Putnam, St. Johns, and Flagler Counties, Florida. SJ 90-13.Tech. Publ. Johns River Water Management District. Palatka, FL. Smajstrla, A. G., B. J. Boman, G. A. Clark, D. Z. Haman, D. S. Harrison, F. T. Izuno, D. J. Pitts, and F. S. Zazueta.1991.Efficiencies of Florida agricultural irrigation systems. Bul. 247.Gainesville, Florida: University of Florida.

PAGE 73

73 Sjmastrla, A. G., D. R. Hensel, D. S. Ha rrison, and F. S. Zazueta.1984.Improved seepage irrigation efficiency by controlled water ap plications. Proc. Fla. State. Hort. Soc. 97:181-187. Starr, J. L., A.M. Sadeghi, and Y. A. Pachepsky.2005. Monitoring and modeling lateral transport through a large in situ cham ber. Soil Sci. Soc. Am. J. 69:1871-1880. St. Johns River Water Management District (SJRWMD).1996.Nutrient po llution how much is a lot? Tri-county crop talk. Newsle tter Summer/Autumn. Palatka, FL. St. Johns River Water Management District (SJRWMD).1996.Can less mean more when fertilizing? Tri-county crop talk. Newsletter, Fall. Palatka, FL. Styles, S.W., and C. M. Hurt. 1999. Subsurface flows water balance components for irrigation districts in San Joaquin valley. Confer ence on benchmarking irrigation system performance using water measurement and water balances. ITRC. Paper. 99-002.San Luis Obispo, CA. Suleiman, A. A., J. T. Ritchie. 2001. Estimati ng saturated hydraulic conductivity from soil porosity. Trans of ASAE. 44 (2): 235. Townend, J., M. J. Reeve, and A. Carter. 2001. Water release characteristics. p. 95-140. In Smith, K. A, and C. Mullins (ed.). Soil and environmental analysis: physical methods. New York. Trangmar, B. B., R. S. Yost, and G. Uehara. 1985. Application of geostatistics to spatial studies of soil properties. Adv.Agron.38:45-93. Unger, P. W. 1975. Water retention by core and siev ed soil samples. Soil Sci. Sc. Am. J. 3 (6): 1197-1200. USDA. 2008. National Agricultural Sta tistics Service, Quick Stats. USDA. 1983. Soil survey of St. Johns County, Florida. Soil Conservation Service : 196. USEPA. 2004. Nonpoint source pollution: the na tion's largest water quality problem. Pointer No. 1. EPA841-F-96-004A. http://www.epa.gov/owow/nps/facts/point1.htm Warrick, A. W., D. E. M yers, and D. R. Nielsen. 1986. Geostatis tical methods applied to soil science. p. 53-82. In A. Klute (ed.) Methods of Soil Analysis. Part I. 2nd. ed. Agron. Monogr. 9. ASA and SSSA. Madison, WI. Warrick, A. W. and Nielsen, D. R., 1980. Spatial variability of soil physic al properties in the field. p. 319-344. In : D. Hillel (ed) Applic ations of soil physics. Academic Press, NY. Wilson, G. V., Jardine, P.M., Luxmoore, R.J., Jones, J.R., 1990. Hydrology of a forested hillslope during storm events. Geoderma. 46:119.

PAGE 74

74 Wilson, R. C., R. S. Freeland., J. B. Wilkers on, and R.E. Yoder.2002. Imaging lateral migration of subsurface moisture using electromagnetic induction. ASAE Paper No.023070.2002 ASAE Annual Internatio nal Meeting, Chicago, IL, USA, July 2031. Zotarelli, L., J. M. Scholberg, M. D. Dukes, and R. Munoz-Carpena. 2007. Monitoring of nitrate leaching in sandy soils: comparison of three methods. J. Environ. Qual. 36 (4): 953-962.

PAGE 75

BIOGRAPHICAL SKETCH I was born in Kavrepalanchok district, Nepal, on January 25, 1980. In 1996, I finished my high school from the Kavre Secondary School at Banepa. I completed my undergraduate in Agriculture (B. Sc. Ag.) from the Institute of Ag ricultural and Animal Sciences at the Tribhuvan University, Nepal. In August 2006, I was matriculat ed to the Soil and Water Science Department at the University of Florida as a Masters stude nt. After completing my Masters Degree, I plan to pursue the Degree of Doctor of Philosophy at the same department.