|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 PORE WATER TRANSPORT MECHANISMS AND SOIL DIAGENESIS IN AN IMPACTED ISOLATED WETLAND WITHIN THE LAKE OKEECHOBEE DRAINAGE BASIN: IMPLICATIONS FOR INTERNAL NUTRIENT LOADING By JEHANGIR H. BHADHA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009
2 2009 Jehangir H. Bhadha
3 To my dearest parents I d edicate this degree to you both; t hank you for all that you have done for me my entire life
4 ACKNOWLEDGMENTS I wish to thank my advisor Dr. James W. Jawitz for his commitment towards my research and friendship during the process. I would like to tha nk the members on my committee: Dr. Willie Harris, Dr. Mark Clark, Dr. John Jaeger, and Dr. Mark Brenner for their help and support towards the completion on this research. I wish to thank my funding agency: The Florida Department of Agriculture and Consum er Services for supporting the research. I wish to thank the University of Florida Land Use and Environmental Change Institute Laboratory (Mr. William Kenney), Soil Pedology Laboratory, Environmental Hydrology Laboratory, Wetland Biogeochemistry Laboratory (Dr. Yu Wang), and the Florida Institute of Paleoenvironmental Research Laboratory for letting me use their facilities for anal yses. I would like to thank my D epartment Chair (Dr. Ramesh K. Reddy), faculty and staff for their assistance and guidance whene ver I needed it. I would also like to thank my colleagues (past and present) for their encouragement and help with field wor k. I wish to thank my good friends Miss Me ghan ODonoghue, and Dr. Melroy Borges for their constant support and many stimulating dis cussions regarding my research. Last but not the least I wish to thank my parents back home in India for their prayers and blessings.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES ................................................................................................................................ 7 LIST OF FIGURES .............................................................................................................................. 8 ABSTRACT ........................................................................................................................................ 10 CHAPTER 1 INTRODUCTION ....................................................................................................................... 12 Background and Motivation ....................................................................................................... 12 Internal Loading of Nutrients ..................................................................................................... 15 Hypotheses and Objectives ......................................................................................................... 17 Study Area ................................................................................................................................... 18 Site Description .................................................................................................................... 19 Hydrology ............................................................................................................................. 19 2 SOIL CHARACTERIZATION BASED ON PHYSICAL AND CHEMICAL PROPERTIES .............................................................................................................................. 22 Introduction ................................................................................................................................. 22 Sampling and Analytical Methods ............................................................................................. 25 Results and Discussion ............................................................................................................... 27 Soil De scription and Classification .................................................................................... 27 Soil Total Phosphorus and Organic Matter Content .......................................................... 30 Summary and Conclusions ......................................................................................................... 32 3 PHOSPHORUS FLUXES AND MASS BALANCE CALCULATIONS BASED ON SURFACE AND GROUND WATER CHEMISTRY ............................................................. 42 Introduction ................................................................................................................................. 42 Sampling and Analytical Methods ............................................................................................. 44 Pressure Transducers ........................................................................................................... 45 Multi level Samplers ........................................................................................................... 45 Pore Water Equilibrators (peepers) .................................................................................... 46 Flux Calculations ................................................................................................................. 47 Using the Damkhle r Number to Evaluate Pore Water Concentration (Cpw) .................. 48 Using Peclet Number to Evaluate Magnitude of Diffusive Versus Advective Flux ....... 49 Phosphorus Budget .............................................................................................................. 49 Results and Discussion ............................................................................................................... 50 Wetland Hydrography and Hydroperiod ............................................................................ 50 Ground Water Chemistry .................................................................................................... 52
6 Diffusive Fluxes of P Based on Ficks First Law .............................................................. 53 Advective Fluxes of P Based on Ground Water Fluctuations .......................................... 55 Phosphorus Budget Estimation From May 2005 March 2006 ....................................... 56 Summary and Conclusions ......................................................................................................... 57 4 SOIL PHOSPHORUS RELEASE AND STORAGE CAPACITY BASED ON EXPERIMENTAL AND FIELD DATA................................................................................... 70 Introduction ................................................................................................................................. 70 Sampling and Analytical Methods ............................................................................................. 73 Batch Incubation and EPC0 Calculations ........................................................................... 74 Effect of Soil:S olution Ratio on EPC0 ............................................................................... 75 Oxalate Extraction, Analyses, and SPSC Calculations ..................................................... 76 Results and Discussion ............................................................................................................... 77 Phosphorus Sorption and Equilibrium Concentrations ..................................................... 77 Effect of Soil:Solution Ratio on EPC0 Values and Pore Water Concentrations .............. 80 Soil P Storage Capacity ....................................................................................................... 81 Summary and Conclusions ......................................................................................................... 83 5 PRACTICAL IMPLICATIONS AND RECOMMENDATIONS ........................................... 97 Soil Erosion and Stratification ................................................................................................... 97 Scaling up to Basin-wide P Loads Based on Using Local Flux Measurements ..................... 98 Basin -wide P Loads from Isolated Wetlands ..................................................................... 98 Basin -wide P Loads from Ditches .................................................................................... 100 Kinetic Control of Dissolved P in Aquatic Systems ............................................................... 101 Role of pH, Soil and Solution Composition in Evaluating P Mobilization ................... 104 Strategy to Reduce P Loads ...................................................................................................... 106 LIST OF REFERENCES ................................................................................................................. 109 BIOGRAPHICAL SKETCH ........................................................................................................... 120
7 LIST OF TABLES Table page 3 1 Diffusive flux of soluble P from sediments to overlying water in selected aquatic systems .................................................................................................................................... 59 3 2 A summ ary of wetland hydrologic parameters .................................................................... 60 4 2 Effect of soil:soilution ratio on EPC0 values and pore water concentration ...................... 90 5 1 Basin wide estimate of P load from isolated wetlands associated with diffusion, advection, ditch flow, and ground water ............................................................................. 107
8 LIST OF FIGURES Figure page 1 1 Map of study area showing the two wetlands ...................................................................... 21 1 2 Contrast in the appearance of wetland hydrology and vegetation-cover between wet and dry seasons. ...................................................................................................................... 21 2 1 Location of four soil cores collected from Wetland 1 ......................................................... 34 2 2 Soil description and charecteristics ...................................................................................... 35 2 3 Particle size distribution and soil texture based on the soil texture triangle for Core 2 (wetland) and Core 4 (upland). .............................................................................................. 36 2 4 Changes in s oil bulk density and porosity with depth. ........................................................ 37 2 5 Sequences of X ray diffraction patterns for soil clays from Core 2 at four different de pth intervals ....................................................................................................................... 38 2 6 Changes in soil or ganic matter content with depth .............................................................. 39 2 7 Changes in soil total phosphorus with depth. ....................................................................... 40 2 8 Amount of T P (kg ha1) estimated within various soil horizons in the upland and wetland. ................................................................................................................................... 41 2 9 Correlation between soil TP (mg kg1) versus % OM. ......................................................... 41 3 1 Conceptual model of the wetland P -budget .......................................................................... 61 3 2 Map of ground water sampling locations ............................................................................. 61 3 3 Pore water sampling devices. A) Multisampler B) Pore water equilibrator (pee per) ....... 62 3 4 Rainfall and water table (stage) fluctuation during the monitoring period between May 2005 and March 2006. ................................................................................................... 63 3 5 (top) Relationship between U.S.G.S. stage data (Cypress slough, near Basinger, FL) (bottom) Wetland hydroperiod (%) estimated for five years from 2004 to 2008 .............. 64 3 6 Pore water concentration of soluble reactive phosphorus (SRP), measured a t three separate locations within Wetland 1 ..................................................................................... 65 3 7 Pore water concentrations of SRP meas ured using peepers from Wetlands 1 and 2 in August 2008. ........................................................................................................................... 65
9 3 8 Diffusive flux of SRP calculated for the three sampling locations and their co rresponding cumulative load over time. ............................................................................ 66 3 9 Diffusive flux of SRP calculated using peepers from Wetland 1 and 2 during August 2008. ........................................................................................................................................ 67 3 10 Ad vective flux of SRP calculated from Wetland 1 and c orresponding cumulative load. ......................................................................................................................................... 68 3 11 Cumulative load of P exiting the wetland via ditch flow, and entering it via surface runoff and atmospheric deposition during the monito ring period. ..................................... 69 4 1 Lan gm u i r plots for P sorption of all four cores .................................................................... 92 4 2 Relationship between (i) oxalate -extractable P versus oxalate -extractable (Fe+Al), and (ii) P sorption maxima (Smax) versus oxalate -extractable (Fe+Al) for all four cores ........................................................................................................................................ 93 4 3 Surface hori zon EPC0 values for cores 1, 2, 3, and 4 .......................................................... 93 4 4 Depth distribution of EPC0 values compared with seasonally variable por e water concentrations and soil total P from wetland cores 1, 2 and 3 ........................................... 94 4 5 Relationship between EPC0 at field condition versus those measured using 1:20 soil:solution ratio for all four soil cores ................................................................................ 94 4 6 Relationship be tween the water soluble P and the degree of P saturation (DPS) calculated using an oxalate extraction from surface and subsurface horizons of all four cores. ............................................................................................................................... 95 4 7 Soil P storage capacity for all four cores, calculated using the PSR values and oxalate extractable Fe Al concentrations in the soil ......................................................................... 96 5 1 Dam structure constructed to reduce the amount of water e xiting wetland 1 via ditch flow ....................................................................................................................................... 108
10 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PORE WATER TRANSPORT MECHANISMS AND SOIL DIAGENESIS IN AN IMPACTED ISOLATED WETLAND WITHIN THE LAKE OKEECHOBEE DRAINAGE BASIN: IMPLICATIONS FOR INTERNAL NUTRIENT LOADING By Jehangir H. Bhadha August 2009 Chair: James W. Jawitz Major: Soil and Water Science The deteriorating water quality of Lake Okeechobee due to excess phosphorus (P) associated with beef cattle and dairy operations from surrounding basins have been an environmental concern for many decades. During this period, isolated wetlands within these basins have been extensively ditched and drained to improve farmlands and facilitate agriculture. Restoring the wetlands has been proposed as one of the strategies to reduce overland P loss by retaining it in the soils. Phosphorus fluxes associated with soils and shallow pore water from agricultural land s can contribute to internal loading which could undermine the effective use of these wetlands. Soil properties of deeper horizons within impacted isolated wetlands are rarely investigated due to the assumption that these wetlands are flooded, and that biogeochemical interactions occur at the soil -water interface. However, prolonged droughts and periodic draining can leave the wetlands dried for most of the year, and water tables can drop over a meter below the ground surface. As part of my doctoral researc h, I investigated soil properties of surface as well as deeper horizons from an impacted isolated wetland located on a beef farm and its surrounding upland pasture. The upland soils (Alfisols) showed pedologically formed horizons compared to the
11 stratified wetland soils (Entisols). The soils were sandy up to a depth of 100 cm, and showed increasing clay content further down core (>100 cm). This can have implications for higher P sorption capacity, perched water tables and slower hydraulic conductivity. Net P reserves were highest at the surface 28 kg ha1 (46%), however, nontrivial quantities were also found in the deeper horizons associated with the clay horizons at 15 kg ha1 (25%). Due to the transient hydrology, one of the hypotheses tested was that int ernal loading of P from diffusive and advective processes was a significant component of the P budget to the wetland. Diffusive fluxes of SRP to the wetland water column were estimated at < 1 mg m2 d1, while advective fluxes associated to water table flu ctuations were 8 (10) mg m2 d1. The wetlands also showed very low annual hydroperiods with multiple drawdown events, this can have a negative effect on the P loading to the water column due to fast changes in soil redox conditions. In addition, P releas e and storage capacity of the soils was evaluated as a function of depth, based on batch equilibrium experiments, and amorphous Fe and Al content. There was a good correlation between Smax and Fe+Al (r = 0.84), and between P and Fe+Al (r = 0.91), suggesting that Fe and Al can be used as a good indicator for P sorption capacity and P content. The EPC0 values at the wetland and upland surface (0 2 cm) was 1.8 (1.3) mg L1 and 8.9 mg L1 respectively, suggesting that the surface horizons could behave as a pot ential source of P to the wetland. Based on the soil P storage capacity experiments both the wetland and the upland soils showed positive storage capacities. Results from this study adds to our overall understanding of P dynamics within impacted wetland so ils in the Okeechobee basin, which will help develop sound P management strategies in the future.
12 CHAPTER 1 INTRODUCTION Background and Motivation Intensive agricultural practices in the United States have maintained a surplus of nutrients including phosphorus (P) within agricultural watersheds (Sharpley et al., 2001; Sims et al., 2005), and surf ace water export from such areas has resulted in eutrophication of surrounding fresh water bodies (Dougherty et al., 2004). In the state of Florida the problem of e utrophication in Lake Okeechobee has been largely attributed to exce ss P in runoff from agr icultural lands within the drainage basin area (Flaig and Havens, 1995). The Lake Okeechobee Protection Act of 2000 set a total maximum daily load (TMDL) target of 140 Mg P yr1 for the watershed to be met by 2015 (SFWMD et al., 2004). Phosphorus management programs implemented within the watershed during the 1980s resulted in a decline in the P loading rate to the lake from a high of 2380 Mg P yr1 in 1985 to 1717 M g P yr1 by 1990 (Hiscock et al., 2003). However, P loading to Lake Okeechobee has continued to remain approximately 100 Mg P yr1 over the legally mandated target primarily because of discharge from basins that have active beef cattle and dairy operations within the watershed (Havens et al., 1996; Hiscock et al., 2003). Such operations account for 12% of the land area in the Ok eechobee basin, yet contribute a disproportionate amount of P (35%) to the Lake (Flaig and Reddy, 1995). In response to these issue s, the agricultural industry is mandated to achie ve P load reductions, and ranches used for dairy and beef cattle within the watershed are being targeted for a portion of these reductions. Concentrations of P in surface runoff tend to be lower from pasture s than f r om holding areas or high intensity areas yet the total P load from pastures is substantial due to the below ground storage over vast acreage (Capece et al., 2007).
13 Over the past 50 years the hydroscape of South Florida has been severely altered by the construction of ditches and canals (Steinman and Rosen, 2000) that connect over 800 historically drained isolated wetlands within the Lake Okeechobee basin with major lotic systems such as the Kissimmee River and Taylor Creek which ultimately drain into Lake Okeechobee. Among the P control and mitigation strategies being considered to achieve the TMDL target is the physical restoration of historically isolated wetlands within the watershed that were drained by ditching. These drained isolated wetlan ds cover nearly 7% of the approximately 5160 km2 Lake Okeechobee drainage basin area (Hiscock et al., 2003; McKee, 2005). The practical implications of restoring isolated wetlands would be to increase the ir flooding periods (hydroperiod) which would allow for greater P storage and retention within the soils. Increasing wetland hydroperiod would also involve an increase in wetland area, and this could facilitate greater P storage in above ground plant biomass. It has been suggested that hydrologic restorati on of these wetlands may help retain P by up to 13 kg P ha1 yr1 (Dunne et al., 2007a ). The low nutrient retention capacity and hydrological setting of the agricultural soils of subtropical Florida can pose a nutrient management challenge. Categorized as the most widely distributed soil order in the state of Florida, Spodosols of this region form in sandy parent material on flatwood landforms that are usually poorly or somewhat -poorly drained (Nair et al., 1998). The sandy nature of the soi ls in the Lake Okeechobee basin and the near ly level landscape (0 2% slope) aids in the rapid infiltration of water, resulting in significant subsurface flow (Capece, 1994). The P retention capacity within the A and E horizon in Spodosols is almost non existent (Harris e t al., 1996; Nair et al., 1998). The limited retention capacity of these soils is an impor tant factor contributing to the high P load entering Lake Okeechobee. These poorly -drained soils can accumulate ground water that emerges as artificially drained surf ace water in
14 depression basins (Hiscock et al., 2003). Campbell et al. (1995) observed that in such low relief landscapes, subsurface drainage is a significant mechanism for P transport. In a 120 ha field study in the Lake Okeechobee basin, Capece (1994) f ound no surface runoff, due to the combination of low topography (1.3% slope) and high infiltration, and estimated that all P export (averaging 25 kg P ha1 yr1) was via subsurface flow -paths. An estimated 60 70% of the P imported to the Okeechobee waters hed is stored within the Bh horizon in upland soils (Reddy et al., 1998). Although these soils could still have up to 75% of their maximum P storage capacity remaining, Capece et al. (2007) suggested that they are already a significant source of non -point P, despite their total storage capacity. A different set of challenges for P management in these soils is the presence of lower permeability argillic horizons (or lenses), composed of finer, P sorbing phyllosilicate minerals and quartz between 60 and 200 cm below ground surface (Villapondo and Greatz, 2001b). Argillic horizons with thicknesses varying between 20 to Alfisols, another common soil order in subtropical Florida The finer textured (loamy to clayey) horizons have the ability to sorb P, similar to the spodic -Bh horizon se en in Spodosols (Harris et al., 1996). Dissolved P moving downwards from the A horizon through the E generally accumulates in the Bh. However, the presence of loamy or clayey horizons below the Bh can impede the movement of water, thus promoting transport of P via surface runoff and lateral seepage to streams and canals, minimally sorbed and mitigated through A and E horizons (Allen, 1987). Further, alternating flooded (anaerobic) and drained (aerobic) conditions can significantly affect the distribution of P within impacted soils. Villapando and Graetz (2001b) found that up to 50% of total P associated with clay horizons was readily leachable and released to pore water
15 under drained conditions. They concluded that the movement of P from such soils would be gre ater from rainfall events following the dry season than during the flooded wet season. Many previous studies related to P retention in wetland soils within the Okeechobee drainage basin area have focused only on near -surface soil horizons (<30 cm) (for example, Richardson and Vaithiyanathan, 1995; Pant et al., 2002; Pant and Reddy, 2003; Dunne et al., 2006, 2007a ), and did not evaluate the effect of P -enriched horizons such as the Bh and Bt that usually exist below 30 cm. However, wetlands in the basin a re hydrologically linked to deeper soil horizons by fluctuating water tables. In a three year study of four isolated wetlands in the Okeechobee basin, Perkins (2007) recorded frequent water table fluctuations of 80 cm below ground surface to 50 cm of stand ing water in the wetlands. When the water table rises and the wetland re -floods following dry periods, there is the possibility of advection of P from the groundwater into the wetland (internal loading). In a laboratory column study, Villapando and Graetz (2001a) showed significant P leaching in impacted spodosols from the Lake Okeechobee basin due to a fluctuating water table. Maintaining the water table above (flooded) and below (drained) the Bh horizon resulted in substantial movement of P in the soil columns. Thus, it is suggested that in cases of fluctuating water tables, it may be important to assess the distribution of P deeper into a soil profile than the commonly studied surface horizons. Internal Loading of Nutrients The internal loading of nutrients from sediments (and/or soils) has been identified as a critical component of the nutrient budget, and can control the trophic conditions of aquatic systems suc h as lakes and wetlands (Berner, 1980; Hkanson 2004). Within the Lake Okeechobee drainage ba sin, this internal release of nutrients from shallow aquatic systems can occur via two different mechanisms: (i) release at or below the sediment -water interface during hypoxic or anoxic conditions, and the subsequent diffusion of dissolved nutrients to th e
16 overlying water column; and (ii) advection as a result of a fluctuating water table, bioturbation, and wind induced resuspension at the sediment surface (Moore et al., 1998; Fisher and Reddy 2001; Wu et al., 2001). The ability to quantify the nutrient fl ux within a system (from the ground to the water column) can be useful for ecosystem managers in determining the importance of internal versus external nutrient loading, thereby implementing strategies to mitigate them. The implementation of P control programs (PCPs) caused a decline in the loading to Lake Okeechobee from early 1980s to 1990 (Havens, e t al., 1996), however P loads to the La ke have still remained over the legally mandated target primarily because of discharge from t ributary basins such as Ta ylor C reek that has active beef cattle and dairy operations within its watershed. External nutrient loading from agricultural and cattle areas can be addressed through the implementation of Best Management Practices (BMPs), and interception strategies such as construction of storm -water treatment areas (STAs) and constructed wetlands in the state of Florida (Rice et al., 2002); h owever, internal loading from impacted soils and sediments has prevented the rapid improvement of water quality of numerous waters heds (Fisher and Reddy, 2001; Pant and Reddy, 2003). For example, studies conducted by Moore and Reddy (1994), and Reddy et al. (1995) in Lake Okeechobee estimated that mean P flux (mg m2 d1) as a result of internal P l oading from bottom sediments is app roxima tely equal to the rate of external P loading (~ 1.0 mg m2 d1). Based on their conclusions the primary source of nutrients that constitutes an internal load is the organic matter within the upper sediments. With frequent drought conditions experien ced in the state of Florida o ver the past decade most of the wetlands within the Okeechobee drainage basin remain dry for 8 9 months through out the year, and water levels are below the surface for prolonged periods (Perkins, 2007). This would allow for shallow ground
17 water to act as a potential source of nutrients that can affect internal loading as the water table fluctua tes above and below the soil -water interface. Besides drainage and other water management practices such as ditching, wetlands located o n ranches are also affected by cattle activity (bioturbation) (Tollner et al., 1990; Sigua, 2003; Sigua and Coleman, 2006; Tweel and Bohlen, 2008). Graz ing animals have been shown to a ffect the movement and uptake of nutrients through the soil and vegetati on either directly or indirectly, thus affecting the chemical and physical characteristics of pasture soils. Standing cattle can facilitate soil compaction and the formation of vertical preferential flow paths (PFP) up to 50 cm within the soil profile (Dre ccer and Lavado, 1993). Nutrients derived directly from manure can potentially seep through these PFPs quickly altering the chemical equilibrium between the soil and surrounding po re water. Mulholland and Fullen (1991) showed that the magnitude of stress caused by animal hooves mainly depends on the soil water content. When dry soils are trampled upon, the main process is compression beneath the hoof, collapsing the larger soil pores by mechanical disruption of aggregates. However, when wet soils are tramp led upon there is a plastic flow being generated around the hoof (Scholefield et al., 1985). Repeated treading under these conditions can produce deep hoof prints that coincide with the presence of surface water ponding up to depths of 5 20 cm (Taboada and Lavado, 1993). It is believe d that because of factors such as rapid changes in hydrology over short periods of time (months or even days), and the random movement of cattle in and out of the wetlands that thes e wetlands need to be perceived as very dynami c, transient system s especially within the top meter of the soil profile Hypotheses and Objectives The main theme of my dissertation deals with the potential for the wetlan d to act as a sink for nutrients under present hydrologic and manag ement conditions on a cattle ranch. Through
18 detailed field sampling, laboratory experiments, analyses and modeling I tested the following three hypotheses: 1 Within a soil profile c ertain horizon s may behave as hot -spots that can act as a potential source of nutrients 2 Nutrient fluxes across the soil -water interface associated with advective processes such as water table fluctuations are significant compared to molecular diffusion. 3 Due to the transient hydrology, f ast adsorption-desorption reaction s between soil and wate r can only partially explain in situ pore water P concentration Three primary objectives of my research were to: (1) investigate the physical and chemical characteristics of surface and subsurface soil horizons (up to 160 cm) and their influence on the su rrounding pore water P concentrations ; (2) estimate nutrient fluxes across the soil -water interface within the wetland associated with internal loading based on advective and diffusive processes; (3) conduct laboratory experiments t o evaluate soil P leachi ng and sorption capacities In the following three Chapters (2, 3, and 4) specific goals and set objectives will be discussed in detail, along with field (or experim ental) results that were used to test the above hypotheses. S tudy Area The two wetlands sel ected for this study are historically isolated emergent marshland s located on an active beef ranch within the S 154 priority sub-basin of the Lake Okeechobee watershed, which has been identified as contributing relatively higher loads of nutrients to Lake Okeechobee, compared to the surrounding lake sub-basins. (Figure 1 1). The se wetlands are located o n the Dixie -Larson Ranch (27.24 N, 80.83 W) and are at the head of ditch es that connect to the regional ditch network that ultimately drains in to Lake Okee chobee The two wetlands, Wetland 1 (or Larson East) and Wetland 2 (or Larson West) are about 1.1 and 2.2 ha respectively.
19 Site D escription The majority of P export to Lake Okeechobee comes from four priority sub-basins (Hiscock et al., 2003). Approximate ly 21% of the total catchment area is comprised of depressional wetlands and other water features (streams, canals and ditches), with only 15% represented by wetlands alone (Boggess, et al., 1995). Nearly 45% of the wetlands today have been hydrologically managed by a network of ditches and canals that drain the landscape and divert the water off -site into rivers and streams that ultimately flow into Lake Okeechobee The wetlands are typically less than one m eter deep and exhibit a pan shape d morphology. They are remnant s of historic sinkhole collapse features. They occur at a relatively small scale in comparison to the resolution of soil mapping units defined by the USDA Soil Survey for Okeechobee County (Lewis et al., 1997). Typical vegetation at these wetlands includes: Juncus effuses, Panicum sp., Pontedaria cordata var. lancifolia, Ludwigia repens, Hydrocotyle ranunculoides and Andropogon glomeratus (Dunne et al., 2006) Hydrology In the mid 1950s, an extensive network of ditches and canals was created t hroughout the Okeechobee drainage basin in an attempt to lower the regional water table and partially drain depressio nal wetlands so that more land could be used for grazing pasture (Flaig and Reddy, 1995). These ditches do not completely drain the depress ional wetlands, as they were excavated to depths above the b ottom of the wetlands. As a result, the water in the wet land must reach a critical height before water can exit via the ditch. Howeve r, with low topographic gra dients it is not uncommon for ditch flow rates to drop t o undetectable levels. By limiting the water depth in the wetland throughout the year, the se wetland s and the ditches serve as cooling ponds for the cattle Such a strategy is advantageous for cattle productivity; however, the wetlands and ditches
20 that are hydrologically connected to the lake can create a rapid conduit for P rich water to be transported off the pastures The effect of ditching on soils and shallow ground water is two -fold. First, it can create a recharge/discharge poin t for ground water in the landscape. Second, the ditch serves to drain the wetland quicker t han natural conditions, thus producing a greater frequency of water table fluctuations. Ground water from the landscape surrounding the ditches can flow towards the ditch carrying w ater -soluble nutrients with it; conversely, water from the ditch can be transported laterally along the edges of the ditches to the surrounding wate rshed. Wetlands located within the Okeechobee basin remain dry for prolonged periods, and w ater levels are below the surface for months primarily due to periodic drought conditions experienced in the state of Florida (McKee, 2005; Perkins, 2007), and due to draining as a result of ditching. Prolonged dryness can lead to change in vegetation cove r and erosion of the topsoil. Figure 1 2 shows the contrast in appearance of the wetland between the two extreme wet and dry periods.
21 Wetland 2 Wetland 1 NLarger ditch transporting water off the ranch Ditch Okeechobee Co., Florida Isolated Wetlands 100 m Figure 1 1. Map of study area showing the two wetlands. Dotted line represents the outer boundary of the wetlands. Wat er from these wetlands dr ains off into ditches that flow into a larger ditch transporting water off the ranch Figure 1 2. Contrast in the appearance of w etland hydrology and vegetation cove r between wet and dry seasons
22 CHA PTER 2 SOIL C HARACTERIZATION BASED ON PHYSICAL AND CHEMICAL PROPERTIES Introduction Over the past 50 years the Lake Okeechobee basin has been impacted as a result of ditching and cattle activity from beef and dairy farming. These operations account for onl y 12% of the land area in the Okeechobee basin, yet contribute a disproportionate 35% of the phosphorus (P) load to the Lake. Within this landscape historically isolated wetlands are known to store water and nutrients such as phosphorus (P), and exist alo ng the landscape continuum between the upland and the l ake (Braskerud, 2000; Hiscock et al., 2003). At present, nearly 45% of the wetlands within the drainage basin are ditched and drained, surrounded by cow -calf pastures located on active beef or dairy fa cilities (SFWMD, 2004). Draining the isolated wetlands via ditches and canals was done to provide improved cow -calf grazing pastures, which have left the basins hydroscape severely altered. One of the strategies recently proposed (Lake Okeechobee Protecti on Plan, 2004) to help mitigate P loss from improved pastures and reduce P loads to Lake Okeechobee was to restore the hydroperiod of these isolated wetlands located on beef and dairy farms, so as to facilitate water storage and P retention (Johns and ODe ll, 2004). An understanding of so il characteristics can help in planning future r estoration work and implement sound management strategies. Studies of wetland ecosystem functions have shown that soils are the most important longterm P storage compartment compared to litter or plant biomass (Dolan et al., 1981; Graham et al., 2005), and that their properties can influence wetland water quality. Most of the P stored in wetlands is in organic form derived from dead vegetation (Graham et al., 2005, Dunne et al ., 2007 a ), typically concentrated in the surface horizons. Under flooded conditions biogeochemical cycling of P occurs within a few centimeters at or near the soil -water interface (Fisher and
23 Reddy, 2001; Debusk and Reddy, 2003), therefore previous studies have focused primarily on soil properties that relate to P storage capacity of surface soils (Reddy et al. 1998; Pant et al., 2002; Dunne et al., 2007a ). However, due to the transient hydrologic conditions experienced by these wetlands the re is a need to investigate the deeper horizons. It is common for water tables to drop below a meter in the ground during the dry season (Perkins, 2007) potentially transporting nutrients to the deeper horizons, in addition to causing sudden shifts in the soils redox con ditions. Studies have shown that the P retention capacity of saturated soils that have undergone periodic drawdown diminish on re -flooding compared to continually flooded soils (Watts, 2000; Klotz and Linn, 2001). Bostic and White (2006) evaluated the beha vior of periodically saturated marsh soils and found that up to 6% of total P (TP) from the soils was released from a single drawdown and reflooding event. The reason for observing this type of behavior is the sudden change in redox conditions which affect s the iron oxide minerals in the soils, and its ability to retain P. Such episodic flooding and drying cycles are a common occurrence within the Okeechobee basin and could mobilize P from deeper horizons, releasing P to the water column on re -flooding (int ernal loading). Within so il profiles clays typically accumulate in deeper horizons, below the E and within the B (Bh or Bt) horizons (Zhou, et al., 1997). In addition to the hydrologic impact of clay horizons, clay particles provide surface area for miner al to -mineral and mineral -to -organic interactions (Rengasamy et al., 1984). Capece et al. (2007) reported clay layers below 60 cm near Lake Placid, FL., about 40 km west of our study site. They concluded that the accumulation of P in deep soils from histor ic fertilization had an overriding influence on P loads in surface runoff. Consequently, net P imports from deeper horizons would also need to be addressed in order to achieve mandated water quality targets in the long term.
24 Ditches and canals created arti ficially to drain the landscape can also have an impact on soil properties. Soil morphology has been used to evaluate a ditchs effectiveness, because morphology has been found to change when the hydrology is altered (Hayes and Vepraskas, 2000). For exampl e, Daniels and Gamble (1967) showed that soils (ultisols) along stream dissected edges of the Middle Coastal Plains (North Carolina) had redder colored Bt horizons than soils further from the edge. This red-edge effect was believed to have occurred after the Coastal Plain had been incised and drained by streams. Prior to the stream incision, restricted surface and subsurface drainage caused reduced conditions to persist for long periods. Stream incision resulting from lowered water tables and the subseque nt oxidation of ironaluminum minerals produced red and yellow soil horizons in areas of improved drainage. Such changes in redox conditions can have an overriding effect on P solubility and sorption mechanisms regardless of the soil/sediment mineralogy (R eddy et al. 1998b ; Hayes and Vepraskas, 2000). Isolated wetlands cover nearly 7% of the land within the Okeechobee basin, and nearly as much is covered by ditches and canals (Hiscock et al., 2003; Gathumbi et al., 2005). With open access to wetlands and d itches, cattle can increase nutrient pollution by depositing manure into surface waters (Bottcher et al., 1999; Alloush et al., 2003), or by stimulating nutrient release from sediments stirred up by cattle activity (Line et al., 1998). Hea vy stocking can a lso lead to reduction in vegetation cover and an increase in soil erosion (Capece et al., 2007). Cattle trampling can increase soil bulk density and decrease soil porosity (Tollner et al., 1990; Mulholland and Fullen, 1991; Bezkorowajnyj et al., 1993). Mat ure cattle can exert a static ground pressure of approximately 1.7 kg cm2 of hoof -bearing area, which can affect bulk density up to 100 cm deep within a wet saturated soil profile (Rhoades et al., 1964; Bezkorowajnyj et al., 1993). In addition, cattle act ivity can also increase the physical breakdown of litter, stimulating
25 enhanced decomposition of organic matter (OM) at the surface (Klemmedson, and Tiedemann, 1995). In order to evaluate the effectivenes s of wetland restoration a thorough understanding of pre restoration soil characteristics is important Soil ch aracteristics of horizons deeper than a meter are not routinely investigated in impacted wetlands within the Lake Okeechobee drainage basin, and soil s urvey data for Okeechobee County only provide limited information because i solated wetlands in the Okeechobee drainage basin occur at a relatively small scale compared to the resolution of s oil mapping units The objectives of this study were to (i) determine differences in soils characteristics as a function of depth between an impacted isolated wetland and the surrounding upland, and (ii) determine the P status within deeper horizons that may be affected by water -table fluctuations. The physical properties descri bed included color, texture (based on particle size distribution), bulk density, porosity and s aturated hydraulic conductivity. Chemical analyses included soil organic matter (OM) content, total P (TP) concent ration and clay mineralogy. Sampling and Analytical Methods Four intact soil cores w ere collected in July 2006 using the vibracoring technique. Vibracoring works on the principle of liquefaction in fine -grained sediments by displacing sediment to allow passage of the pipe and retrieval of long (up to 10 m), continuous sections from uncons olidated saturated sediments (Smith, 1984). Vibracoring works best in saturated organic sediments, clays, silty clays, silts and fine sands, but is inefficient in firm clays and medium to coarse sands. Cores 1, 2, and 3 were collected from within the wetla nd basin, while core 4 was collected from a nearby upland site within 20 m from the ditch (Figure 2 1). These cores were 140, 144, 129 and 161 cm long, respectively. The water table was 6 to 10 cm below the ground surface at the time of coring, so the core s were virtually saturated T o ensure
26 saturation they were completely flooded with de ionized water. The cores were transported to the laboratory where they were split lengthwise, described and photographed. Soil description was conducted on wet soil using USDA protocols (Soil Survey Staff, 2007). One section of the split core was used to measure soil bulk density (BD) by gamma ray attenuation (Geotek Standard Multi -sensor Core Logger, MSCL -S) at 0.5 cm increments, and capture highresolution digital images (Geoscan II, 40 pixels cm1) of the entire core lengthwise. The accuracy of the bulk density (r2 > 0.99) was determined using a standard aluminum density calibration piece. Particle size distribution (PSD) was measured only on selected soil horizons extra cted from core 2 (wetland) and core 4 (upland) based on morphological changes. The pipette dispersion method was used to measure PSD (Gee and Bauder, 1986). Minerals in the clay fraction were identified by X ray diffraction (XRD) using a computer -controlle d X ray diffractometer equipped with stepping motor and graphite -crystal monochromator. Oriented mounts were prepared by sedimentation on unglazed ceramic tiles under suction. Diagnostic cation saturation with MgCl and glycerol were performed to aid phyllosilicate identific ation (Harris and White, 2008). The remaining halves of all four split soil cores were sectioned at 2 -cm intervals, weighed, freeze -dried, and re -weighed prior to chemical analyses. Freezedrying the samples was preferred over conventiona l oven-drying (2 h at 105 C), because the process sublimates the moisture content, with minimal physical or chemical alteration (McClymont et al., 2007). The sam ples were stored in air tight containers to preve nt any rehydration Wet and dry soil weights were used to determine water content for each 2 -cm section. Soil porosity was calculated using measured BD and a fixed particle density of 2.65 g cm3 (quartz sand).
27 Soil samples were also analyzed for TP and organic matter (OM) content. Total P was analyz ed colorimetrically (Bran Lueb b e Autoanalyzer) using the acid persulfate digestion for soil TP (Nelson, 1987). Organic matter content was measured using the loss on ignition (LOI) technique (Dean, 1974) where 2.0 g of dry mass (freeze -dried) sample were ox idized at 550C in a muffle furnace for two hours. The weight loss on ignition represents the OM content of the sample. Note that the LOI method used in this study may overestimate the OM content in clayey material due to dehydroxylation reactions of secon dary phyllosilicates in the temperature range of OM combustion. Analytical precision for TP and OM was assessed using duplicates every tenth sample, and reported as relative percent difference (RPD, 3.3% for TP and 5.2% for OM content). Saturated hydraulic conductivity was measured using the slug test method (slugout technique) from five fully penetrating onsite wells extending approximately 1.5 m below the ground; one well located in the wetland, and four in the surrounding upland. The Hvorslev method (Fe tter, 1994) was used to analyze the slug test data for fully penetrating wells with well length more than eight times the well radius. Results and Discussion Soil Description and C lassification The soil textur e in all four cores was sandy down to a depth o f 100 cm (Figure 2 2). The top of upland core 4 showed a distinct Ap horizon with high OM, underlain by an eluvial E horizon (10 40 cm depth), a Bt horizon (40 80 cm), a sandy loam Btg1 horizon (80140 cm), and a sandy clay Btg2 horizon (140 161 cm). While the upland core comprised horizons Ap, E, Bt and Btg formed by pedogenesis, all three wetland cores showed stratified C horizons. Deposition of surface sediments eroded from surrounding uplands may have given rise to the stratified morphology observed in all three of the wetland cores. Below 100 cm there were finer textured
28 horizons (sandy loam and sandy clay), in which fine stratification was not as prominent as in the overlying horizons. In upland core 4 I observed what may be a result of the red-edge effect within the Bt horizon where a well defined yellow to red coloration (high chroma) is visible between 75 83 c m. Being close to the ditch, it is possible that followi ng ditch incision, lowering of the water table due to surface and subsurfa ce drain age may have caused oxidizing conditions, conducive for formation of iron oxides The upland soil is classified as an Alfisol, while the wetland soils are classified as Entisols by virtue of having > 50 cm of recent material (stratification). The genesis of Alfisols involves the translocation of clay from the upper to lower soil zones ultimately forming thick argil lic horizons. Such horizons have the ability to sorb P, similar to the spodic Bh horizons seen in Spodosols. Dissolved P moving vertically downw ards from the A and O horizon through the E generally accumulates in the B horizon. However, the presence of clay within the B horizon can impede the movement of water, thus promoting transport of P via surface runoff and lateral seepage to the ditches through A and E horizons. The texture of the wetland soils below the stratified horizon became loamier, gradually transitioning to clay at the core bottom. Particle size distributio n analysis of cores 2 and 4 showed that the sand fraction (0.052.0 mm) grad ually decreased with depth, while the clay fraction (<0.002 mm) increased, up to 36% in core 2 and up to 38% in core 4 (Figure 2 3). The sandy clay layer observed in all four cores will subsequently be referred to as clay horizons. The top of the clay horizon was encountered at approximately 120 cm in all three wetland cores. The clay thickness is unknown, but it extended all the way to the bottom of the soil cores. A similar 15 -cm clayey horizon was seen around 85 100 cm in core 2. Suc h clay horizons cou ld
29 establish a perched water table, and separate it from underlying aquifers, and reduce the loss of P through leaching. The porosity of all four cores was higher (0.49 0.05) within the top 20 cm (Figure 2 4). This is probably due to the presence of the loosely bound OM associated with plant roots within the top 20 cm (Justin and Armstrong, 1987). Below 20 cm, the porosity gradually decreased to less than 0.45 in the wetland cores. The small range in porosity ( 0.006) between 50160 cm in upland core 4 r eflects uniform grain size. Bulk density increased with depth in all four cores: in the wetland soils from 1.33 0.15 g cm3 in the top 20 cm to 1.56 0.03 g cm3 in the bottom 20 cm, and in the upland soil from 1.39 0.11 g cm3 in the top 20 cm to 1.70 0.01 g cm3 in the bottom 20 cm (Figure 2 4). Cattle poaching, which is the penetration of the soil surface by the hooves of grazing animals (Mulholland and Fullen, 1991), may be responsible for the increase in BD observed at 10 cm depth within the soi l profile. The sharp increase in BD from 1.45 to 1.70 g cm3 in the upland core 4 at the E -Bt horizon interface is because of the change in soil texture due to pores being filled during illuviation. Saturated hydraulic conductivity within the wetland base d on slug test measurements from the wetland well was 1.62 m d1, while the hydraulic conductivity in the surrounding upland pastures was 0.82 m d1 ( 0.21). These values are comparatively lower than those reported by Soil Survey Staff (2007) for the Basi nger and Valkaria fine sands of this region (3.7 to 12.2 m d1). The discrepancy in values may be because the 1.5 m long screened wells intercept the clay horizons that could lower the integrated c onductivity for the soil unit down to the well depth. The XRD analyses on four selected soil horizons from core 2 showed that smectite was the dominant mineral in the clay -size fraction, and that smectite enrichment gradually increased with depth, as evidenced by the smectite:kaolinite ratio (based on peak areas) of 1.16, 1.38, 3.07, 2.23
30 from samples collected at 45, 60, 92 and 144 cm depth respectively (Figure 2 5). Smectites are typically considered to have minimal P sorption capacity giv en their low edge to face ratio and permanent negative charge They impart a higher cation exchange capacity than kaolinite, suggesting less retention of species such as orthophosphate than kaolinite (Pant et al., 2002). Soil Total P hosphorus and Organic Matter C ontent Characterizing P storage in wetland soils is often accompani ed by OM characterization, as the TP pool within wetland ecosystems is often dominated by organic P, which can comprise > 50% of soil TP (Reddy et al., 1998b ). High concentrations of OM in the form of manure and decomposed vegetation can drastically increa se the availability of P within surface soils since organic anions formed by the decomposing OM can compete with P for the same adsorption sites such as clays and iron alumino minerals (El Dewiny et al., 2006). Except for core 1, OM concentrations in the c ores were greatest within the top 2 cm (Figure 2 6). The OM content within the top 6 cm was 38% ( 14%) and 48% ( 2%) for the wetland and upland soils respectively. The high OM content at the surface is believed to be associated with decomposing vegetatio n and cattle manure. Below 6 cm the OM concentrat io ns steadily dropped to < 2% at the top of the clay horizon. Due to the presence of finer clay material OM content in those zones may be slightly overestimated, because small quantities of finer inorganic minerals could have also been lost by the incineration process. An irregular decrease in OM with depth was consistent with the observed fine stratification within the wetland soils. Similar to OM, clays can also influence the P storage capacity with depth, because of the ability to sorb P. Except for core 1, soil TP concentrations were also highest within the top 2 cm of the soil profile (Figure 2 7). In the wetland soils these concentrations gradually decreased with depth to about 100 cm, but then increas ed again as the clay content increased. In comparison, the upland core showed a sharper increase in TP concentrations below 120 cm. This
31 increase in TP concentrations below a meter in the soil profile is also associated with the clay fraction within and be low the Btg1 horizon at those depths. Depth averaged soil TP concentratio ns between 0 20 cm were 62.8 mg kg1 ( 28), 182.6 mg kg1 ( 241), 166.1 mg kg1 ( 104), and 124.2 mg kg1 ( 172) for cores 1, 2, 3, and 4, respectively. The high degree of variabi lity is primarily beca use of the high concentration of TP within the top 2 cm of the soil profile. The increase in P concentration between 40 60 cm seen in core 4 corresponds to the affinity for iron -bearing minerals for P present within the Bt horizon in the upland. Based on the BD values and average concentration of soil P for a given thickness of soil horizon, the net P reserve within wetland and upland soil horizons was estimated (Figure 2 8). While the majority of the P is accumulated in the upper 10 c m (46 49%), up to 25% is also present in the deeper argillic horizons of the upland, and 17% in the deep, finer textured layers of the wetland. The increasing trend of TP observed at the bottom of core 2 and 4 indicates that there may still be more P in ho rizons deeper than the core bottom; this would further decrease the relative proportion of net TP reserves estimated in the upper 10 cm of the wetland and upland soils. T he low -permeability clay horizons could impede vertical movement thereby forcing the flow laterally, potentially transporting P to nearby ditches, a process previously addressed by Allen (1987). Under such conditions ground water could be short -circuited rapidly along preferential flow paths and emerge as surface water (Kadlec and Wallace, 2008). The TP concentrations within the soil profiles followed OM trends in all four cores (r = 0.8 9, Figure 2 9). The regression was dominated by the surface horizon (r = 0.90), because TP concentrations at the surface were nearly two orders of magnitude higher than in the rest of the soil profile. Correlations in other horizons were all much weaker: for example, root -zone (10 20 cm) r = 0.56, mixed zone (20 100) r = 0.65, and finer textured zone (>100 cm) r = 0.6 The
32 strong OM to P correlation within t he top 10 cm likely corresponds to P in cattle manure and decaying vegetation deposited at the soil surface. This is because both OM and P would tend to accumulate at the surface, whether they are chemically associated or not. The slope of the regression f or surface soils corresponded to 10.3 mg kg1 of TP for every 1% increase in OM. In comparison, Dunne et al. ( 2007a ), estimated an increase of 4.2 mg kg1 of TP for every 1% increase in OM (r2 = 0.86) in surface soils (0 10cm) from adjoining wetlands inclu ding this one located on the same pasture land. An inverse correlation between OM and P was observed between 10 20 cm within the root -zone (rhizosphere). This is not unusual, because at this depth C:P ratios are typically elevated, and virtually all the P is sequestered within the rooted macrophytes, responding to the increase in soil P concentrations (Davis, 1991). The P within the clay horizons is primarily associated with the increase in clay content, and its ability to retain P on reaction sites. This d eep pool of TP could be critical for understanding soil P retention dynamics associated with clayey horizons, especially if the water table intercepts those depths during the dry periods. Summary and Conclusions S oils recovered from the wetland are classif ied as Entisols by virtue of un -obliterated stratification seen throughout the entire soil profile possibly resulting from upland soil erosion. The upland soils are classified as Alfisols, because of a distinct argillic horizon and the Bh does not meet th e taxonomic criteria for a spodic diagnostic horizon. All four soil cores recovered from the site included clayey material below 120 cm in the soil profile. The two dominant phyllosilicate minerals present within the wetland soi ls were smectite and kaolini te. T hese minerals can lower hydraulic conductiv ity by impeding subsurface flow, however, kaolinite can increase P sorption capacity due to its lower cation exchange capacity and a higher edge to face ratio than smectite The high -chroma reddish yellow hor izon within the Bt -horizon in the upland
33 soil may have resulted from the altered hydrology as a result of ditching (the red -edge effect). The effect of cattle trampling is evident within the top 10 cm of the wetland soils causing an increase in soil BD. Organic matter and P within the top few centimeters of the soil profile were highest compared to the rest o f the soil profile; both were at least partially derived from manure inputs. Restricting the cattle from en tering the wetland could help control the high P concentrations seen at the surface. Although not supported by measured data, m oderation of cattle stocking ratios is probably the most effective means of decreasing the effect of overgrazing and erosion, in addition to reducing P inputs from cattle activity at the surface in the uplands. Significant amounts of P were also found within the deeper Btg1 and Btg2 horizons (upl and), and in the clay horizons be low 100 cm in the wetland The vertical translocation of P due to infiltration may result in fur ther accumulation of P at those depths. Net TP reserves within the deeper horizons are nontriv ial, and cannot be ignored when computing P budgets. T he clay horizon accounts for nearly 25% of TP ha1 of the measured soil profile in the upland, and 17% of TP ha1 of the measured soil profile in the w etland. F urther investigation of soil characteristics in the deeper horizo ns on neighboring wetlands is encouraged to improve our understanding of below -ground soil P dynamics and in delineating soil properties in response to ditching and surrounding landuse practice.
34 Core 4 Core 1 Core 2 Core 3 LEGEND Wetland 1 area Soil core sampling location Ditches Okeechobee Co., Florida 14 5 2 11 Figure 2 1. Location of four soil cores collected from Wetland 1. S oil survey map units for the area:  Basinger fine sand;  Valkaria fine sand;  Imokalee fine sand;  Myakka fine sand
35 ApHigh OMESandy, 2.5Y 6/2; splotches of OMBtHigh chroma ; Fe oxideBtg1Sandy loam, 10R 5/1; masses associated with Fe redoxBtg2Sandy clay C1Mixed and stratifiedC2StratifiedC3Loamy sandC4 Sandy loam C1 C2StratifiedC3Sandy loamC4Loamy sandC5Sandy clay; gleyed 5/5BC1 Mixed and stratified C2 Stratified C3Loamy sand C4 Sandy loam 17 cm 5 cm 10 cmCore 4 Core 1 Core 2 Core 3Depth (cm) Rededge effect Figure 2 2. Soil description. Cores 1, 2, and 3 were collected from the wetland, while Core 4 was collected from the upland < 20 meters from the ditch. Relative difference in elevation fro m the upland shown on the top
36 10 0 8 0 6 0 4 0 2 0 0 p e r ce n t 16 0 1 4 0 12 0 10 0 8 0 6 0 4 0 2 0 0 C o r e 4 10 0 8 0 6 0 4 0 2 0 0 p e r ce n t 16 0 1 4 0 12 0 10 0 8 0 6 0 4 0 2 0 0D e p t h ( c m) C o r e 2Texture sand loamy sand sandy loam sandy loam loamy sand sandy clay loam sandy clay Texture sand sand sandy loam sandy clay loam sandy loam sandy loam sandy clay loam sandy clay Silt Sand Clay Figure 2 3. Particle size d istribution and soil texture based on the soil texture triangle for Core 2 (wetland) and Core 4 (upland).
37 Figure 2 4. Changes in soil bulk density (solid circles), and porosity (empty circles) with depth.
38 Figure 2 5. Sequences of X ray diffraction patterns for soil clays from Core 2 at four different depth intervals, and scanned at 25 C after cation -saturation of MgCl and glycerol smectites kaolinite quartz kaolinite quartz 2 Relative intensity
39 16 0 15 0 14 0 13 0 12 0 11 0 10 0 9 0 8 0 7 0 6 0 5 0 4 0 3 0 2 0 1 0D e p th ( c m ) 1 0 8 6 4 2 0 1 0 8 6 4 2 0 1 0 8 6 4 2 0 16 0 15 0 14 0 13 0 12 0 11 0 10 0 9 0 8 0 7 0 6 0 5 0 4 0 3 0 2 0 1 0 2 5 2 0 1 5 1 0 5 0 Core 1 Core 2 Core 3 Core 4 1 0 0 6 4 2 0 6 0 4 0 2 0 0 6 0 4 0 2 0 0 % O rg a n i c m a tt e r ( L O I ) 1 0 0 6 0 4 0 2 0 0 Figure 2 6. Changes in s oil organic matter content with depth.
40 1 0 0 6 0 4 0 2 0 0 8 0 0 60 0 40 0 20 0 0 40 0 30 0 20 0 10 0 0 T P ( m g/ k g ) 1 0 0 8 0 0 60 0 40 0 20 0 0 16 0 15 0 14 0 13 0 12 0 11 0 10 0 9 0 8 0 7 0 6 0 5 0 4 0 3 0 2 0 1 0D e p th ( c m ) 8 0 4 0 0 12 0 8 0 4 0 0 12 0 8 0 4 0 0 16 0 15 0 14 0 13 0 12 0 11 0 10 0 9 0 8 0 7 0 6 0 5 0 4 0 3 0 2 0 1 0 12 0 8 0 4 0 0 C ore 1 C ore 2 C ore 3 C ore 4 Figure 2 7. Changes in soil total phosphorus with depth.
41 24.2 kg P ha1(49%) 11.7 kg P ha1(24%) 5.2 kg P ha1(10%) 8.6 kg P ha1(17%)10 cm 20 cm 100 cm ~130 cm Ap E Bt Btg110 cm 40 cm 80 cm 140 cm ~160 cm27.9 kg P ha1(46%) 5.9 kg P ha1 (9%) 6.1 kg P ha1(10%) 6.1 kg P ha1(10%) 15.4 kg P ha1(25%) Btg2 Upland WetlandStratified horizon Figure 2 8. Amount of T P (kg ha1) estimated within various soil horizons in the upland and wetland. y = 9.6x + 7.2 R = 0.89 0 100 200 300 400 500 600 700 800 0 10 20 30 40 50 60 Top 6 cmOM %TP (mg kg1) Figure 2 9. Correl ation between soil TP (mg kg1) versus % OM compared using 280 samples from all four cores. Every 1% increase in OM corresponds to a 10.3 mg kg1 increase in TP within surface soils (0 10 cm).
42 CHAPTER 3 PHOSPHORUS FLUXES AND MASS BALA NCE CALCULATIONS BAS ED ON SURFACE AND GROUND WATER CHE MISTRY Introduction The internal loading of P from sediments and soils has been identified as a critical component of the nutrient budget that can control the trophic conditions of aquatic systems such as lakes and wetlands (Berner, 1980; Hkanson 2004). Within the Lake Okeechobee drainage basin, this internal release of P from shallow aquatic systems can occur via two different mechanisms: (i) release at or below the sediment -water interface during hypoxic or anoxic conditions, and the subsequent diffusion of dissolved P to the overlying water column (Moore et al., 1998; Fisher and Reddy, 2001; Pant and Reddy, 2003) ; and (ii) advection as a result of a fluctuating water table (Corstanje and Reddy, 2004). Studies conducted by Moore and Reddy (1994), and Reddy et al. (1995) in Lake Okeechobee have shown that the mean P flux (mg m2 d1) as a resul t of internal P l oading from bottom sediments is approximately equal to the rate of external P loading (~ 1.0 mg m2 d1). Based on their conclusions the primary source of P that constitutes an internal load is the organic matte r within the topmost sedimen ts. The implementation of P control programs in the Okeechobee basin resulted in a decline in P loading from early 1980s to 1990 (Havens, et al., 1996), however P loads to Lake Okeechobee still remain over the legally mandated target primarily because of d ischarge from tributary basins such as Nubin Slough and Taylor C reek that have active beef cattle and dairy operations within their watershed s External loading of P from agricultural and cattle areas has been addressed through the implementati on of best m anagement practices (Bottcher et al., 1997), and interception strategies that involve the restoration of isolated wetlands in the state of Florida (Rice et al., 2002). I nternal loading through impacted soils however, has prevented the rapid improvement of water quality within the basin. In addition, wetlands that are exposed to less than
43 100% hydroperiods can also be affected by drawdown events that increase soil oxidation and subsequent mobilization of P during reflooding ( Bostic and White, 2006). In the past P fluxes from wetland soils have been routinely estimated using a general procedure that involves exposing a soil core to a layer of P -free water (Olila et al., 1997; Fisher and Reddy, 2001; Pant and Reddy, 2003; Dunne et al., 2006); concentration gr adients between the water column and pore water were measured over time to generate a flux (mass area1 time1). However, these are typically abiotic experiments, in which plants are absent, and transport via groundwater flow is non-existent. Hence, such e xperiments place a limit on how much labi le P can be released under low flooding conditions, and it is dangerous to extrapolate the results to fu ll -scale field scenarios as they fail t o capture P fluxes associated with the advective mode of transport (Kadl ec and Wallace, 2008). Wetlands within the Okeechobee basin remain dry for prolong ed periods, and water levels can remain below the land surface for months due to periodi c drought conditions in Florida (McKee, 2005; Perkins, 2007), and draining as a result o f ditching. This would allow shallow ground water (pore water) to act as a potential source of P that can advect internally as the water table rises above the soil -water interface and mix with the water column. Zhang et al. (2006) show ed that up to a 20% P load reduction from beef cattle ranch facilities within the Okeechobee drainage basin could be achieved by maintaining a fixed surface water detention depth of 0.5 inches on the landscape. It is hypothesize d that under fluctuating water table conditions the internal load of P from an impacted isolated wetland may be si gnificant compared to diffusion. The ability t o estimate the load depends on measuring the flux accurately Figure 3 1 represents a conceptual model of surface and ground water hydrology b ased on a fluctuating water table. Quantifying diffusive and advective P fluxes from these
44 wetlands can be useful for ecosystem managers to determine the importance of internal versus external P loading, enabling them to develop strategies to mitigate such loads When dealing with a shallow -water system such as a wetland, a simple one -dimensional (depth) diagenetic model such as Ficks first law of diffusion is often used to estimate the vertical flux based on close interval estimation of pore water nutrien t concentration gradients (Berner, 1980; Harper et al., 1997). Such models assume microbial -mediated processes are uniform horizontally (American Society of Li mnology and Oceanography, 1998). N utrient gradients can be measured from pore water samples using equilibration devices such as peepers (Pardue and Jackson, 2001) or multisamplers (Martin et al., 2003). Quantifying nutrient fluxes via advective modes i s slightly more difficult, as they depend on transient processes that are hard er to calibrate and mea sure, such as wind effects, wave action, and fluctuating water table s Some of the techniques used in the past to quantify advective fluxes include chemical tracers (Rama and Moore, 1996), ground water modeling (Pandit and El -Khazen, 1990), and seepage met ers ( Micheal et al., 2003). In this chapter I present results from a field study that was done to estimate P fluxes in side and out side of the wet land The objectives of the study were to: (i) calculate diffusive and advective fluxes of soluble reactive P ( SRP) based on in -situ measured pore water concentrations; (ii) develop a P budget that will help evaluate the importance of diffusion and advection from wetlands affected by flooding and drying cycles. Sampling and Analytical Methods The location of sampli ng sites is shown in Figure 3 2, and the various sampling techniques and analytical methods used are described below.
45 Pressure T ransducers The wetland surface water elevation was monitored for a three year period between 20032006 (Perkins, 2007). For this study, an 11 -month period between May 2005 and March 2006 was selected for monitoring work, sampling, and testing water quality. An internally -logging pressure transducer (Mini -troll STP, In -si tu, Inc.) was deployed in a 0.032 m d iameter fully screened P VC well in the deepest location of Wetland 1. T he wetland monitoring well was 3.0 m deep below ground surface, and the above -ground portion of the monitoring well was housed inside a la rger diameter PVC casing (0.051 m diameter) to provide structural prot ection. Wetland water sur face elevations were recorded at half -hour intervals and were subsequently averaged on daily time intervals. A simple input/output hydrologic budget was used to estimate daily c hanges in stage (cm) in wetland water elevation that was later used to calculate the velocity (v ) at which the ground water fluctuate d with time (cm d1) (Perkins, 2007) Multi -level S amplers Surface and pore water sampling was conducted on six occasions from Wetland 1 during the monitoring period (July05, September 05, October 05, January06, Febuary 06, March06). On each of these occasions surface water samples were collected by the simple dip and grab method in 100 ml polypropylene bottles. Depth variable groundwater (pore water) samples were collected u sing a mu ltisampler, which is simply a multilevel piezometer (Figure 3 3 A) Three such multisamplers (MS1, MS2, and MS3) were deployed in the wetland 60 meters apart. A peristaltic pump was used to extract water from sampling ports located at 10 cm depth increments up to 110 cm below the soil -water interface. Approximately 100 ml of pore water was purged prior to sampling, so as to discard any residual water remaining in the tubes from previous sampling. Water samples were acidified, preserved and transpo rted to the University of Florida where they were analyzed for soluble reactive phosphorus (SRP) using a Technicon
46 AAII autoanalyzer Standard EPA method 365.3 was used for analyzing SRP. Analytical precision was calculated by analyzing duplicates every te nth sample, and calculating relative percent difference (RPD) for each sampling event (RDP < 5%). Pore Water E quilibrators (peepers) In addition to the multisamplers, in August 2008 three pore water equilibrators (peepers) were deployed in Wetland 1 and 2 to capture vertical changes in solute concentrations at every 1 cm depth interval near the soil water interface. Using peepers helps identify the precise depth ( 1 cm) below the soil -water interface where ground water concentration is highest. In contrast to the multisampler, peepers can provide a much finer resolu tion in concentration gradients but limit the pore water volume to 8 ml, only enough for analysis of a single analyte. For this study I was only interested in SRP concentration The three peepers in each wetland were deployed strategically, one near the head of the wetland (or entrance of the ditch), the second in the cent er of the wetland, and the third at the bou ndary of the wetland. This provide s insight into the spatial variability of the of p ore water concentrations within the wetland. The peepers used in this study were similar in design to the device described by Hesslein (1976). It consisted of a 2 10 60-cm laths of acrylic into which are milled 8 -cm3 cells that were spaced vertically 1 cm apart (Figure 3 3 B). The cells were filled with deionized water, covered by a sheet of 0.2 m poly ethersulfone membrane with a 300 m sheet of silk netting placed over it. An acrylic cover was tightly screwed on to the lath, and the entire set up was placed in containers, sealed, and purged of O2 using a N2 gas line for 48 hours prior to field deployment. On site, the peep ers were pushed into the wetland soil so that 10 cells (10 cm) were in the water column, and the remaining cells were buried below g round. Equilibrium studies have shown that between 2 3 weeks is sufficient time for the solution inside the cells to
47 equilibrate with pore water P (Carignan et al., 1985). The equilibration period for this study lasted for 19 days, and during t his time tro pical storm Fay passed directly over the wetlands. After the equilibration period, the peepers were retrieved from the soils, and the water samples drawn out of individual cells using a syringe, acidified, and stored at 4 C until analysis. Flux C alculatio ns Diffusive fluxes were calculated using Ficks first law (Berner, 1980): JD = Ds z C 8.64 105 (3 1) w here JD is the vertical diffusive flux (mg m2 d1), is the porosity (cm3 cm3), Ds is the bulk soil diffusion coefficien t ( c m2 s1), and 8.64 105 is a units conversion. Porosity of soil for each location was depth averaged from an intact core study up to 10 cm depth. Li and Gregory (1974) reported the diffusion coeff icients of HPO4 2and H2PO4 in pure water as 7.34 106 and 8.46 106 cm2 s1, respectively. An average of the two values was used for our calculations (7.9 106 cm2 s1). The diffusion coefficient was modified for the restrictive effect of soil stru cture by dividing it by the square of the soil tortuosity ( Where 2 = 0.47 + 1.91 (Berner, 1980; Sweerts et al., 1991). Ground water advective fluxes were estimated using the hydraulic loading rate (q) of ground water moving in and out of the wetland (cm d1) deriv ed from hydrologic budget calculation (Perkins 2007), and multiplying it by either the pore water concentration (Cpw) of SRP for fluxes into the wetland, or multiplying by the surface water concentration (Csw) of SRP for fluxes out of the wetland. Jin = q Cpw (3 2 ) Jout = -qCs w (3 3 ) This would suggest that at times the advective flux (Jin) would be positive (+) and move up from
48 the pore water into the water column, and at other times the flux (Jout) would be negative ( -), i.e., when the w ater table drops below the soil -water interface carrying dissolved P with it. Advective fluxes (+/ ) were computed for times when the water table drove solutes in and out of the wetland between May 05 to March 06. Diffusive fluxes were calculated for all s ix sampling periods While diffusion can occur at all times when the wetland is flooded, advection from groundwater to the wetland is limited only to times when the water table rises across the soil water interface, based on the hydraulic loading rate (q) The process can be considered as a single event -based process that occurred five times during the monitoring period. The downward advection from the water column back into groundwater was calculated for times when the wetland stage drops due to infiltrati on into the ground. Using the Damkhler Number to Evaluate Pore Water C oncentration (Cpw) The Damkohler number (Da) is a unit less parameter that has been used to evaluate the rate of change by a reaction over the rate of transport by advection (Kadlec, 2000; Carleton, 2002). Defined as: Da = kvL/u (3 4) w here kv is the volumetric rate constant (s1) used to evaluate the sorption reaction between the soil and pore water as the front moves upwards within the soil profile; L is the soil length; and u is the pore water velocity (m s1). Chen et al. (1996) estimated rate coefficients for a two -site kinetic model reaction through columns of dry sandy spodic soils (Myakka) ranging from 1.2 103 to 4.4 103 s1. Using this range of kv and different ve locities of the wetting front (ranging between 3 and 30 cm d1) I was able to estimate Da for different scenarios of high and low velocities. The use of pore water concentrations measured at 10 cm depth for all times when calculating Jin was justified by a large D a (> 10) ranging between 351294, meaning that the soil
49 and pore water were in equilibrium based on the residence time of the slow -moving wetting front. Using Peclet Number to Evaluate Magnitude of Diffusive Versus Advective Flux The Peclet Number i s a dimensionless number (Pe), which works out to be the product of system length (L) and fluid velocity (U), divided by diffusivity (D); such that, Pe = LU/D. It is the measure of the relative importance of advection to diffusion processes in the system. For any given system, the higher the Peclet number (Pe>>1), the more important is advection. Similarly, smaller the Peclet number (Pe<<1), diffusion become the more important mode of solute transport in the system. This approach can be used to evaluate whe ther P fluxes across the soil -water interface into the wetland water column are dominated by advective of diffusive processes. Clearly, the bigger the L the higher the Pe. Hence it is important to set spatial and temporal boundary conditions because Pe wil l vary with varying boundary conditions. To calculate the Pe within a 1 m2 soil domain, we assumed steady state conditions with a velocity of 3 cm d1. The diffusivity of diffusion coefficient was 10 6 cm2 s1 (0.0864 cm2 d1, Li and Gregory, 1974). Base d on the above values Pe = 3472, suggesting that advection would be the dominant mode of solute transport. If we were to reduce the soil domain to just 10 cm, the Pe would still be >> 1, and advection would still be dominant. Using this simple calculation we can come up with a good first estimate that advective processes are more dominant than simple molecular diffusion at the wetland sites. Phosphorus B udget The P budget was based on a simple input versus output equation, comprising parameters whose cumula tive P mass loads (kg) were estimated over the eleven -month monitoring duration: C = A + R + GWin GWout D (3 5 ) w here C is the amount of P in the wetland generated from input parameters such as atmospheric
50 loading (A), surface runoff (R), and gro und water loading into the wetland water column associated with diffusive and advective fluxes (GWin), and output parameters such as ground water load out of the wetland water column re lated to the advective flux (GWout), and load exiting the wetland via d itch flow (D). Ahn and James (2001) estimated atmospheric P concentration in the Okeechobee drainage basin area to vary between 6.8 to 32.6 g L1 depending on wet and dry seasons respectively, with an annual average of 0.112 mg P m2 d1. Using this estim ate we calculated atmospheric loading to the wetland based on the periodicity of rain events. Concentration of P from pasture runoff was not measured as part of this study; however, Capece et al. (2007) reported concentrations of P from pasture runoff rang ing between 0.15 0.63 mg L1 based on a six year average (19982003) from a nearby ranch (Buck Island, Florida). For this study runoff P concentration of 0.6 to 1.0 mg L1 (reported in Reddy et al 1999) was used as a range of best estimate values based on review of water quality data provided by the South Florida Water Management District related to a beef cattle density of 1.2 cow ha1 and improved pasture as the land type. Phosphorus load exiting the wetland via the ditch was calculated using the volumet ric loss of water via the ditch times the concentration of surface water in the wetland for periods when the wetland stage is greater than the bottom of the ditch elevation (0.3 m). Results and Discussion Wetland H ydrography and H ydroperiod Both wetlands 1 and 2 are situated at the head of a ditch, with no ditch inflows, thus sources of inflow to the wetland are only from rainfall, surface runoff and ground water input mechanisms. These inflows should accou nt for surface water storage, evapo transpiration ( ET) ditch flow out of the w etland, and groundwater discharge. Based on the stage data from wetland 1, the wetland remained flooded for 230 days out of the 335 days of monitoring ; the remaining
51 105 days the water table remained be low the ground surface (Fi gure 3 4 ). One of the key controls on ground water nutrient cycling processes in seasonally flooded wetlands is the alter n ation between flooded and dry conditio ns During the monitoring period wetland 1 exhibited this type of season al variation on six occa sions: thrice when the wetland was dry and the water table rose up past the soil surface and flooded the wetland, and thrice when the stage of water in the wetland g radually dropped below the soil surface. The maximum water table fluctuation was between 56 cm above ground and 88 cm below g round surface. The a verage yearly rainfall in Okeechobee County is between 1.3 and 1.6 m (Lewis et al., 2001). During the monitoring period precipitation was just below the annual average at 1.16 m. Severe drought conditi ons rendered these wetlands dry for prolonged periods and since April 2006 the w etland remained dry until July 2008, except for a few days between July and O ctober 2007. In July 2008, the w etlands were floode d once again and the water level in the wetl and was 120 cm above ground after tropical storm Fay (08/19/2008) The U.S. Geological Survey (U.S.G.S.) has been involved in monitoring real -time surface and ground water data for over two decades The data are accessible to the public (http://waterdata.usg s.gov/nwis/rt ). I used five -year stage data from one of the U.S.G.S. monitoring site s (Cypress slough, near Basinger, FL), which is within 30 km from the wetland site s to estimate the yearly hydroperiod of our study site. The calibr ation was based on the s ame 335day monitoring period of our study. The relationship between our wetland measurement and the U.S.G.S. stage data was fitted using a sixth order polynomial curve (r2 = 0.58) (Figure 3 5, top). Using the equation of the line represented by the fit be tween the U.S.G.S. stage data and the wetland I was able to estimate the average yearly hydroperiod of the wetland from the beginning
52 of 2004 to the end of 2008. Results show that from 2004 to 2008 the wetlands have only been inundated for short periods ( Figure 3 5, bottom). Ground Water C hemistry Ground water chemistry was evaluated based on the pore water samples collected using multisamplers (during the 20052006 monitoring period), and peepers (in August 2008). During the 20052006 monitoring period, w etland 1 pore water concentrations of SRP showed both spatial and temporal variability (Figure 3 6). Concentration of SRP was highest at 10 cm depth for all six sampling perio ds; the mean values were 3.9 mg L1 ( 2.1) at MS1, 3.4 mg L1 ( 1.4) at MS2, an d 2.2 mg L1 ( 0.4) at MS3. It is not unusual for pore water P concentrations to be high in the surface horizons due to inputs from plant associated P during senescence and manure inputs from cattle activity. With depth SRP concentration drops to < 0.02 mg L1 below 80 cm depth at MS2 and MS3, however, SRP concentrations increased below 70 cm at MS1. This could be a result of the upward diffusion of P from the deeper horizons that could have transported water from high concentration zones to get intercept ed at those depths. At all three sampling locations the lowest concentrations at the surface (10 cm depth) were measured during the February (winter) sampling event, while the highest concentrations were observed between JulyOctober (hur ricane season) sam pling events. Pore water SRP concentrations were also measured at three locations from wetlands 1 and 2 during a one time sampling event using peepers in August 2008 (Figure 3 -7). Results show that in wetland 1 pore water concentration of SRP at P1 is highest at 11 cm below the surface (12.5 mg L1). At P2 and P3 SRP concentrations are highest at 9 cm (6.7 mg L1) and 8 cm (6.2 mg L1), respectively. In wetland 2 pore water concentrations of SRP at P4 is highest at 17 cm (7.2 mg L1) below the surface. At P 5 and P6 SRP concentrations are highest at 8 cm (4.6 mg L1) and 7 cm (7.0 mg L1), respectively. The concentration of pore water SRP analyzed from the peepers
53 nearest to the ditch were relatively higher than those from either the center of the wetland or its boundary. Regardless of the sampling locations SRP pore water concentrations measured using the peepers in August 2008 were significantly ( p <0.01) greater than those measured using multisamplers during t he 20052006 sampling events. It is believe d tha t this increase in pore water concentrations may be directly related to an event -based phenomenon associated to tropical storm Fay. While it is also possible that P loading to the wetland in 2008 was greater than in 20052006, Olila et al. (1997) showed th at dissolved reactive P from drained soils th at were rapidly reflooded were four times higher compared to controlled, continuously flooded soils. While the shallow ground water P concentrations are useful in calculating diffusive fluxes to the w etland wat er column, concentrations deeper in the profile provide insight into the below ground dissolved P repository that could potentially advect to the surface driven by a fluctuating water table. Diffusive F luxes of P B ased on Ficks First Law Based on the conc entration gradients between the pore water and the surface water the direction of the diffusive fluxes was always from the ground (pore water) to the water column (positive). The upward flux of SRP from the soil into wetland 1 was estimated for all six sam pling events, with highest rates observed at MS1 (Figure 3 8). Out of the six sampling events the highest rate of SRP flux to the water column was 0.97 mg m2 d1 recorded during the September 2005 sampling period, while the lowest flux was observed for the month of February 2006 from all three sites (MS1, MS2, and MS3). The high degree of variability in measured fluxes is reflective of the spatial distribution of the multisamplers, and also the temporal varia tions between surface and pore water concentrati ons because diffusive fluxes are highly redox sensitive. Malecki et al. (2004) and Sen et al. (2007) were able to show that soluble
54 reactive P flux released from sediments under anaerobic conditions are significantly greater than the flux under aerobic con ditions. For our study, the wetlands are only seasonally flooded, hence ground water concentrations (and therefore the fluxes) are highly influenced by changing redox conditions. Nearly 78% (18) flux reduction was observed between September (high-flux) an d February (low -flux) month s from sites MS1, MS2, and MS3. Diffusive fluxes of SRP were also calculated using the pore water concentration gradients collected from all six peepers that were deployed in wetlands 1 and 2 (Figure 3 9). The diffusive fluxes of SRP calculated using peepers in 2008 were only slightly greater than the ones calculated using multisamplers during 20052006, even though the concentration in the shallow pore waters was significantly higher during the August 2008 sampling event. This is primarily because there is relatively little change in the concentration gradients between the two sampling events. The average SRP fluxes calculated using peepers from both wetlands 1 and 2 was 0.84 mg m2 d1 ( 0.31). Several studies have been performe d on P fluxes in a variety of aquatic systems using intact soil incubation (Table 3 1). While the diffusive fluxes calculated from this study are lower than those estimated from the peat soils within the Everglades Water Conservation Area (WCA) 2A (Fisher and Reddy, 2001), they are slightly higher than those from within the sandy zones from within Lake Okeechobee (Moore e t al., 1998). The concentration gradient is greater between pore water and surface water when the water column is underlain by P rich peat soils as compared to sandy sediments. It is evident that while the accretion of P -rich material at the soil -water interface could increase the diffusive fluxes, an increase in surface water concentrations from runoff inputs can reduce ground water fluxes from shallow soils (Dunne et al., 2006). However, diffusive fluxes of SRP alone do not account for the total fluxes from shallow soils in the wetland associated with internal loading. Advection plays an integral
55 role in driving soluble P in andout of the wetland, and needs to be accounted for when computing net internal loads. Advective Fluxes of P Based on Ground Water F luctuations During the eleven-month monitoring period the stage of the water in the wetland fluctuated almost daily. Thrice the water le vel dropped below the soil surface, rendering the wetland completely dry, and then rose, filling the wetland once again. Each fluctuation is indicative of the head difference between the water level in the wetland and the upland. T his corresponds to a puls e of P driven in or out of the wetland controlled by the velocity (hydraulic loading rate) of water across the soil -water interface. Once the front crosses the interface ground water can still advect in and out of the wetland carrying with it soluble P (i. e., even under flooded conditions advection via ground water can occur based on the hydraulic loading rate). The net volumetric rate of ground water into and out of the wetland was evaluated based on a hydrologic budget that accounts for changes in precipi tation, ET, and surface runoff (Perkins, 2007). Based on the hydrologic budget calculations, during the eleven month monitoring period ground water into the wetland occurs five times for a few of days per month, generating a pos itive (+) advective flux. D u ring the same period water from the wetland wa s drawn down into the ground (ground water out) on eleven occasions (every month), generating a negative ( ) advective flux associated with seepage/infiltration (Figure 3 10). T he advective fluxes of P to the wetland via ground water were greater than the diffusive fluxes for all sampling events. Positive adv ective fluxes were greatest during the month of May 2005, when the water table rose > 80 cm in just two days resulting from ~13 cm of rainfall. Ground water flux of SRP to the wetland for that month was 27.4 mg m2 d1, primarily driven by the higher hydraulic loading rate of 1.37 cm d1 (Table 3 2). Negative fluxes were greatest during September 2005 at 45.9 mg m2 d1. During the eleven-month monitoring per iod, the net
56 advective P fluxes leaving the wetland were far greater than those e ntering it via ground water. Nevertheless the po sitive advective flux of SRP were nontrivial and indicate that while diffusion temporally dominates the wetlands hydrologic r egime (230 days out 335), it accounts for smaller total fluxes than advection via ground water movement. P hosphorus Budget E stimation F rom May 2005 March 2006 The internal loading of P to the wetland water column is primarily associated with the two mode s of transport diffusion and advection. Based on diffusion alone the cumulative P load to the wetland water column was 5.2 kg ( 1.4) accounting for 52% of the total P load entering the wetland. In comparison, the cumulative P load associated with adve ction to the wetland was 0.93 kg accounting for 9% of the total load. The cumulative P load entering the wetland via runoff was estimated at 3.8 kg, accounting for 38% of the total load to the wetland (Figure 3 11). During the monitoring period the cumula tive load of P from the atmosphere (dry + wet) was 0.15 kg, accounting for only 1% of the total P loads entering the wetland water column and was by far the smallest input. The cumulative load of P exiting out of the wetland water column via infiltration was 10.6 kg; by far the domi nant component of the P budget, 81% of total P load leaving the wetland water column The cumulative P load exiting the wetland via the ditch flow was estimated at 2.5 kg during the monitoring period, accounting for only 19% of the P load out of the wetland. It is important to note that wetland vegetation is a vital component of the P budget that can sequester dissolved P by incorporating it into non labile above and below ground biomass (Dunne et al. 2007a ; Kadlec and Wallace, 2008). Conversely, plant colon ization during dry periods in P -enriched soils is a significant mechanism for P release from the soil at the time of flooding (Bostic and White, 2006). This component of the wetland P budget was not evaluat ed in this study. M y calculations suggest that about 3 kg of P wa s contained within the wetla nd water column over the eleven -month monitoring period.
57 Summary and Conclusions Both advective and diffusive proc esses need to be evaluated when calculating total P fluxes from wetl ands that are affected by flooding and drying cycles. Results indicate tha t advective fluxes associated with ground water fluctuations are greater than diffusive fluxes based on concentration gradients between surface and pore water. During the mon itoring period between May 2005 and March 2006 the wetland hydroperiod was 230 days, and diffusion occurred for the entire period generating a maximum flux of 0.97 mg m2 d1 recorded in September 2005 at sites MS1 and MS2. During the same monitoring period advect ion occurred for only 21 days, yet generated a maximum flux of 27 mg m2 d1 in the month of May 2005. This pulse of high P flux was a direct effect of a single two -day rain event that caused the ground water to rise rapidly across the soil -water interface Similar upward movement of ground water was observed on four other occasions during the monitoring period, however, the hydraulic loading rate s associated with those events were much slower hence the fluxes associated with th em were much smaller (2.2 4. 0 mg m2 d1). Using the online U.S.G.S stage data we were able to calibrate the stage of the water in our wetland, allowing us to extrapolate its hydroperiod over five years. Results indicated that the wetland has been fairly dry for prolonged periods ov er the last five years. It also shows the number of times the wetlands undergo drawdown events ; this could have a negative effect on nutrient loading, and also on the ability of the soil to retain P. In the future the wetland stage can be monitored using a 31.5 ft threshold value for the U.S.G.S. site above which the wetland starts to fill assuming the input of water remains the same, and the l ocal topography does not change Based on the P budget calculations the internal loading of P to the wetland wate r column via advection and diffusion accounted for nearly 61% of the total P entering the wetland. By far the greatest sink of P within the wetland was the GWout component via advection as surface
58 water seeps into the soil. This type of advective transport accounted for nearly 81% of P loss to the wetland. The P budge t calculations for this study do not consider all possible sources and sinks, and do not include functions such as plant uptake/release, or the effect of cattle on wetland hydrodynamics. From a management stand point maintaining prolonged flooded conditions by increasing the hydroperiod in the wetland may be considered as a strategy to r educe advection and suppress nutrient loading associated with it. Dunne et al. (2007a ) proposed a strategy wher eby increasing the wetland area by 20% could potentially increase wetland P storage by up to 13 kg ha1 within plant biomass. Both these strategies could be implemented in the future if wetland restoration work involves either deepening the wetland basin, raising the ditch elevation, or simply damming it.
59 Table 3 1. Diffusive flux of soluble P from sediments to overlying water in selected aquatic systems Aquatic System Diffusive Flux (mg P m 2 d 1 ) Reference Wetland 1 (multisamplers) Wetland 1 (peepers) W etland 2 (peepers) 0.06 0.97 0.61-1.1 0.53 1.2 This study (2005 2006) This study (2008) This study (2008) Lake Apopka, Florida 1.0 5.3 Reddy et al., 1996 Lake Apopka, Florida 0.3 0.7 Olila et al., 1997 Lake Okeechobee, Florida Moore et al., 1998 Mud zone 0.1 1.9 Peat zone 0.2 2.2 Sand zone 0.1 0.5 Littoral zone 0.6 1.5 Indian River Lagoon, Florida 1.6 Reddy at al., 1999 Lower St. Johns River Estuary, Florida 0.13 4.55 Malecki et al., 2004 Everglades (Water Conservation Area 2A), Florida 1.5 6.5 Fisher and Reddy, 2001 Beaver Reservoir, Arkansas 0.01 1.77 Sen et al., 2007 Lake Eucha, Oklahoma 1.03 4.4 Haggard et al., 2005
60 Table 3 2. A summary of wetland hydroperiod, duration of events, ground water flow, wetland size, and water chemist ry on a monthly basis during the monitoring period (May 2005 -March 2006). NA = not applicable Wetland hydroperiod Number of days out of 335 GW flow in GW flow out Wetted wetland area Surface water conc. Pore water conc. (d) Rained Ditch flow Runoff GW in GW out (cm d 1 ) (cm d 1 ) (m 2 ) (mg SRP L1) (mg SRP L1) May 6 8 0 2 2 2 1.37 6.06 831 0.40 2.0 Jun 30 19 27 8 8 4 0.15 0.13 18044 0.40 2.0 Jul 31 14 16 2 0 15 NA 0.69 15176 0.49 3.1 Aug 31 15 0 8 0 31 NA 1.53 5529 0.40 2.0 Sep 13 27 0 7 0 12 NA 8.45 1065 0.54 4.6 Oct 16 17 7 5 4 15 0.05 2.07 12618 0.72 4.3 Nov 30 8 14 5 4 26 0.2 0.70 13725 0.40 2.0 Dec 31 6 0 6 0 31 NA 1.06 8728 0.40 2.0 Jan 6 4 0 2 0 6 NA 5.56 2281 0.62 3.2 Feb 25 6 2 4 3 22 0.25 0.83 13021 0.61 1.3 Mar s 11 3 0 2 0 10 NA 1.53 6669 0.32 2.6 Total 230 127 66 51 21 174
61 3. 2. 1. Fluctuating water table UPLAND WETLAND DITCH Figure 3 1. Conceptu al model of the wetland P -budget shown by arrows representing 1. external load entering the wetland, 2. internal loading (diffusive and advective fluxes) within the wetland, 3. total load exiting the wetland via ditch. Due to seasonal fluctuation of the water table, groundwater is an important source of P to the wetland. Wetland 2 Wetland 1 MS1, P1 MS2, P2 MS3, P3 P4 P5 P6 NLarger ditch transporting water off the ranch Ditch 100 m Figure 3 2. Map of ground water sampling locations. Three multisamplers (MS1, MS2, and MS3) and three peepers (P1 P2, and P3) used to sample pore water from Wetland 1, and three peepers (P4, P 5, and P6) used to sample pore water from Wetland 2.
62 Water Soil 10 cm 20 cm 30 cm 40 cm 50 cm 60 cm 70 cm 80 cm 90 cm 110 cm Tubing Ports at every 10 cm increments A B Ports at every 1 cm depth increments Water Soil 10 cm 20 cm 30 cm 40 cm 50 cm 60 cm 70 cm 80 cm 90 cm 110 cm Tubing Ports at every 10 cm increments A B Ports at every 1 cm depth increments Figure 3 3. Pore water sa mpling devices. A) Multisampler consi sting of sampling ports at 10 cm depth increments. B) P ore water equilibrator (peeper) consisting of sa mpling port at 1 -cm depth increments.
63 May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Ditch elevation Ground surface Figure 3 4. Rainfall represented in grey bars and water table (stage) fluctuation in solid black line during the monitoring period between May 2005 and March 2006.
64 -240 -180 -120 -60 0 60 2004 2005 2006 2007 2008Wetland water fluctuation (cm)(39%) (69%) (17%) (29%) (32%)Hydroperiod -140 -120 -100 -80 -60 -40 -20 0 20 40 60 80 30 31 32 33 34 35 36 U.S.G.S. stage data for Cypress slough (ft)Measured wetland stage (cm)R2= 0.59 n = 335 Figu re 3 5. (top) Relationship between U.S.G.S. stage data (Cypress slough, near Basinger, FL) and measured wetland stage data for 335 days. Trendline represents a polynomial fit used as a calibration for extrapolating wetland hydroperiod. (bottom) Wetland hy droperiod (%) estimated for five years from 2004 to 2008. The two shaded areas represent the monitoring duration used for this study.
65 Figure 3 6 Pore water concentration of soluble reactive phosphorus (SRP), measured at three separate locations (MS1, M S2, and MS3) within Wetland 1 using a multisampler. Conc entrations were measured for times throughout the monitoring period. Dates are month.day.year. Figure 3 7 Pore water concentrations of SRP measured using peepers from Wetlands 1 and 2 in August 2 008.
66 Wetland 1 2005 2006 Figure 3 8 Diffusive flux (mg m2 d1) of SRP calculated for the three sampling locations represented by the vertical bars, and their corresponding cumulative load (kg) over time (May 2005March 2006) represented by the line with tiny square boxes
67 Figure 3 9 Diffusive flux (mg m2 d1) of SRP calculated using peepers from Wetlands 1 and 2 during August 2008.
68 Figure 3 10. Advective flux (mg m2 d1) of SRP represented by vertical grey bars calculated for times when ground water moved in (+) and out ( ) of wetland 1 between May 2005 and March 2006 monitoring period. Corresponding cumulative load (kg) represented by the line with square boxes.
69 Figure 3 11. Cumulative load (kg) of P exitin g the wetland via ditch flow, and entering it via surface runoff and atmospheric deposition during the monitoring period May 2005March 2006). Average wetted wetland area for individual months represented by vertical grey bars. Runoff at 1 mg L1Runoff at 0.6 mg L1
70 CHAPTER 4 SOIL PHOSPHORUS R ELEASE AND STORAGE C APACITY BASED ON EXPERIMENTAL AND FIE LD DATA Introduction Years of intensive agricultural practices associated with the beef and dairy industries have maintained a surplus of P within the Lake Okeechobee watershed (Boggess, 1995). The l ateral transport of P from manure impacted surface soils to adjacent streams is facilitated by seasonally high water tables, and this has been attributed as a major source of P loading to the Lake (Havens et al., 1996). Historically isolated wetlands withi n the landscape have been used to intercept surface runoff thereby increasing water storage and P retention (Braskerud, 2002). However, ditching the landscape to provide greater expanse s of improved pasturelands ha s left the wetland soils drained and expos ed for extended periods. Generally, at the soil -water int erface when inorganic P is present at concentrations considerably greater than the interstitial (pore) water P is sorbed to th e soil, whereas at low P concentrations wetlands can release P (Pant et al., 2001; Dunne et al., 2006). However, changes in redox conditions associated with drying events can significantly alter P solubility and sorption mechanisms. Investigations have found that more P is released from dried soils upon reflooding than from pe rpetually flooded soils and sediments (Baldwin et al., 2000; Watts, 2000; Klotz and Linn, 2001). This could pose a serious concern for managing P from wetlands within the Okeechobee basin that are exposed to prolonged dry periods pri marily due to droughts and partly due to ditching practices. Soils within the Lake Okeechobee drainage basin are typically sandy, and retain less P than finer textured soils because of a lack of mineral components such as metal oxides and phyllosilicates that have surface affin ity for orthophosphate (o -P) (Sample et al., 1980). In addition, desiccation and consequent oxidation have been shown to cause significant changes to soil (and sediment) mineralogy (Sah et al., 1989; De Groot and Van Wijck, 1993; Klotz and
71 Linn, 2001), whi ch in turn can affect P solubility and sorption capacity of the soil. Repeated drying and wetting cycles have been shown to cause an irreversible aging of the mineral phases in the soil (Baldwin et al. 2000). Dried mineral phases become more crystalline making them difficult to reduce (Lovely and Phillips, 1986). Over time the increase in crystallinity reduces the soils ability to sequester o -P or release it under both oxidizing and reducing conditions. For example, De Groot a nd Van Wijck (1993) show ed that when anoxic wetland soils were exposed to air the ferrous sulfide previously present was rapidly oxidized to amorphous ferric oxyhydroxide. These ferric (oxy)hydroxides have both a large surface area and high affinity for o -P; however, extended period s of desiccation lead to a significant reduction of the o P binding capacity of ferric hydroxides in both flooded and dra ined soils (Sah et al., 1989). Phosphorus sorption by aluminum (Al) does not involve oxidationreduction reactions, and thus is not aff ected by aerobic or anaerobic conditions (Reddy et al., 1998a ). The P sorption by Al compounds is regulated by pH and organic matter. The availability of o -P in solution is controlled by the magnitude of organic acid production derived from detrital organi c matter within the soil profile (Traina et al., 1986). The chemical basis for this observation appears in the hydrolysis of exchangeable aluminum (Al3+) and the subsequent reaction of the newly formed surface hydrolytic species with o P from the soil solu tion. The presence of organic acids such as citric, formic, and tartaric can inhibit the Al3+ hydrolysis reaction. Characterizing P storage in wetland soils is often accompanied by orga nic matter (OM) measurement because the total phosphorus (TP) pool wit hin wetland ecosystems is often dominated by organic P, which can comprise >50% of soil TP (Reddy et al., 1998b ). High concentrations of OM in the form of manure and decomposed vegetation can drastically increase the availability of P within surface soils since organic anions formed by the decomposing OM can compete with P for the same
72 adsorption sites such as clays and iron aluminum (Fe -Al) minerals (El Dewiny et al., 2006). Goh and Huang (1984) show ed that elevated concentrations of citric acid in solutio n can displace Al3+ in favor of the formation of soluble Al3+-citric acid compl exes. M etal -complexing organic acids tend to reduce exchangeable Al3+ by inhibiting hydrolysis, which in turn increases o-P solubility. Standard soil test methods have been use d to evaluate the environmental risks of P loss from agricultural soils (Maguire and Sims, 2002). As concerns of excess P in soils arose the concept of saturation index was introduced to assess the potential risk of soluble forms of P. The degree of P s aturation (DPS) was expressed as the molar ratio extractable P to the sum of extractable Fe and Al, expressed as a percent and incorporating a factor which made it approximate the percent of total P sorption capacity (Bree uwsma et al., 1995). The simple mo lar ratio of extractable P to extractable Fe plus Al is referred to as the phosphorus saturation ratio (PSR) (Nair and Harris, 2004). The relationship between measures of P des orption (e.g., water soluble P WSP) and the DPS or PSR show a change point at which water extractable P starts to increase. Pautler and Sims (2000) have shown that Fe and Al can control P sorption capacity by contributing to P retention. This occurs mainly because of the formation of chemical bonds between orthophosphates and Fe -Al at the soil surface, a ligand exchange reaction producing an orthophosphate -surface complex (Pierzynski et al., 2005). While PSR is often used to make environmental risk assessments, it does not give information about how much more P can be added to the s oil without creating an environmental problem. Hence, the concept of soil P storage capacity (SPSC) was introduced, which refers to the mass of P that can be added to a mass of soil before the soil becomes an environmental concern (Nair and Harris, 2004; Chrysostome et al., 2007). This information is important to this study as it will identify soil
73 horizons (surface and subsurface) that are capa ble of storing P, and predict the amount of additional P that can be accommodated by a given soil horizon before t he added P becomes an environmental concern. Under inundated conditions the P release from soil and sediments has only been evaluated with in the top few centimeters or even millimeters, that are considered to be important for most biological and chemical activity (Bostrm et al., 1982). However, wetlands are hydrologically linked to deeper soil horizons by fluctuating water tables. Periodic drying of soils has been shown to chemically and physically alter soil horizons up to 40 cm (De Groot and Van Wijck, 1993). During the dry season, the receding water table can transport dissolved P from surface horizons into deeper zones, while subsequent flooding can potentially mobilize P from shallow horizons (Kleinman et al., 2003) The fact that wetland soils below the surface may be active in the process of P release has implications for the prediction of internal loading to the wetland water column and deserves additional investigation. An evaluation of P release and storage capacity of surface and subsurface soil horizons can be useful in the design of future wetland restoration and management programs with long term goals. Thus, this study was conducted with the following objectives: (i) to determine the maximum P sorption capacity (Smax) of surface and subsurfac e soil horizons (up to 160 cm) based on a batch equilibrium isotherm experiment, (ii) determine the equilibrium P concentration ( EPC0) values from the e xperiment, and evaluate the effect of changing soil:solution ratio on EPC0 values (iii) to determine soi l P storage capacity (SPSC) based on oxalate -extractable Fe, Al, and P con centrations Sampling and Analytical Methods In this study I conducted two separate experiments on soil samples derived from cores 1, 2, 3, and 4. O nly selected sub -sections were use d for analyses and experimental purpose s Soil
74 samples used for the experiments were selected starting at the surface (0 2 cm), and then at 10 cm depth increments (10 12cm, 20 22cm, 30 32cm, and so on) until the bottom of the core. Batch I ncubation and EPC0 C alculations Soil samples were freeze -dried and ground to pass a 0.8 -mm (20 -mesh) sieve for use in the EPC0 experiment. Incubation was performed under aerobic conditions, where 1 g of soil was placed in a 50-ml centrifuge tube and treated with 20 ml of 0 .01 M KCl solution containing several concentration levels of P (0, 0.1, 0.5, 1.0, 5.0, and 10.0 mg P L1) as standard phosphate solution. The tubes were placed on an automatic shaker for a 24 h equilibration at low speed. Following equilibration, soil sam p les were centrifuged for 10 minutes at 6000 rpm, and filtered through a 0.45 m membrane filter. Filtrates were analyzed for soluble reactive P (SRP) using a Technicon Autoanalyzer ( Standard EPA method 365.3). Phosphorus not recovered in the solution was assumed to be retained by the soil. Water soluble P (WSP) was calculated based on the concentrations recovered from samples that contained 0 mg P L1. Analytical precision was calculated based on 5% duplicates of the entire batch experiment and calculatin g relative percent differen ce (RPD) for individual concentration levels (RDP 3%) Phosphate retained by the soils and the EPC0 was calculated as follows: M V C V C Seq/ '0 (4 1) where S = P sorbed by soil (mg kg1); C0 = concentration of P added (mg L1); V = volume of liquid (L); Ceq = concentr ation of P in solution after 24 h (mg L1); M = mass of dry soil (kg). Calcu lations expressed in equation (4 1 ) do not account for the native soil P in the sorbed phase, because it is primarily used t o calculate P a dsorbed or desorbed with regard to the P added. The total amount of P retained by the soil can be calculated as:
75 0' S S S (4 2) where S = total amount of P sorbed (mg kg1); S0 = initial soil P present in adsorbed pha se (mg kg1). The S0 can be estimated using a least squares fit of S measured at low equilibrium concentrations Ceq. At these concentrations the relationship between S and Ceq is typically linear (Rao and Davidson, 1979) and described by the following eq uation: 0' S C K Seq d (4 3) where S0 = the y axis intercept representing the initial soil P present in the adsorbed phase (mg kg1); Kd = linear adsorption coefficient (L kg1). The EPC0 is defined as the concentration of P in solution w here adsorption equals desorption (i.e., S = 0). Thus, d eqK S C /0 (4 4) where Ceq = EPC0 (mg L1). For individual depth intervals the P sorption data were fitted to the Langmuir model to help characterize the isotherms. Quantitative fitting parameters such as maximum P sorption capacity (Smax), and an affinity constant related to the bonding energy ( k ) were derived using the linearized form of the Langmuir equation: max max/ / 1 / S C kS S Ceq eq (4 5) By regressing Ceq/S against Ceq, the slope corresponds to 1/ Smax and the y-intercept equals to 1/( kSmax). Effect of Soil:Solution R atio on EPC0 The rationale for selecting a particular soil:solution ratio (SSR) in studies relating to sorption of inorganic P is rarely, if ever given (Hope and Syers, 1976), e xcept for the choice of a wide ratio to simulate runoff environments (Ryden et al., 1972). The fact that variation in SSR
76 can influence the final concentration of P in solution (Hope and Syers, 1976; Nair et al., 1984) requires addition al investigation if P management strategies are based on EPC values derived using sorption isotherms. The purpose of this exercise was to try to determine the EPC0 valu es based on SSR that are representative of field conditions so that EPC0 values could b e used to predict pore water concentrations if the soil and pore water were under equilibrium conditions. To determine t he effect of SSR at equilibrium which represent natural field condition s without measuring the equili brium isotherms we used the data from the isotherm experim ent performed on a 1:20 SSR and recalculated S0 and Kd based on the new SSR The rationale Hope and Syers (1976) used to evaluate t he effect of SSR on P remaining in solution was, that if the final solution concentration (Ceq) afte r 24 h equilibrium using a 1:40 SSR was x, and the soils adsorbed P, then at a 1:1 SSR the Ceq should be x/40; if however, the soils desorb P, th en the final solution concentration (Ceq) at 1:1 SSR should be 40x. They used four separate SSR (1:1, 1:5, 1:10; 1:40) to test t he effect of SSR on P sorption by soils. We used the same rationale to evaluate the degree of change in EPC0 v alues by adjusting the SSR to represent field conditions. Calculations were done on all four cores at the same depth increments ( 0 2, 1012, 2022, and so on) as those done for the EPC0 e xperiment us ing the 1:20 SSR In order to esti mate the SSR under field conditions, we converted the mass of 1 g of soil to volume based on the bulk density, then using changes in soil porosity with depth we calculated th e new SSR at individual depth increments. Us ing the new SSR we estimated EPC0 values representing field conditions. Oxalate Extraction, Analyses, and SPSC C alculations Oxalate -extractable (ox) P Fe, and Al were determined at the surf ace (0 2 cm) followed by every 10 cm depth interval for all four cores, using 0.1 M oxalic acid + 0.175 M ammonium oxalate solution, equilibrated at a pH of 3.0. Phosphorus, Fe and Al in the oxalate solution were
77 determined using inductively coupled argon spectroscopy (Thermo Jarrell Ash ICAP 61E). The degree of P saturation (DPS) and phosphorus saturation ratio (PSR) were calculated as: 100 Al Fe ox P ox DPS (4 6) Al Fe ox P ox PSR (4 7) where P, Fe, and Al are molar concentrations o f these elements and is a dimensionless P saturation factor whose value is often assumed to be 0.5 (Schoumans and Groenendijk, 2000; Daly et al., 2001). Nair et al. (2004) used a statistical approach in identifying a 0.15 PSR threshold for Florida soils as corresponding to an approximate P solution concentration of 0.10 mg L1. Soil P storage capacity (SPSC; Nair and Harris, 2004) can be calculated based on this threshold PSR, as follows: SPSC = (0.15 PSR) o x(Fe + Al) 31 (4 8) w here Al and Fe c oncentrations are expressed as mol L1 and SPSC is a calculation of how much additional P (mg kg1) can be added before reaching the PSR threshold. Results and Discussion Phosphorus Sorption and Equilibrium C oncentrations The soil p hosphorus sorption chara cteristics for all four cores were well described (r2 0.88) using the Langmuir model for C0 10 mg P L1 (Figure 4 1). At zero initial concentration level (C0 = 0 mg P L1) the soils from every horizon desorbed P after 24 h equilibration. The surface horizons by virtue of having higher P released the high est amount of P (S0) at the zero concentration level, which may also account for the higher EPC0 concentrations at the surface. At the highest initial concentration level (C0 = 10 mg P L1) soils at every horizon adsorbed P after equilibration. The S10 val ues only provide a relative indication of the maximum sorption
78 capacity (Smax) at different depths, because Smax were higher than the S10 for all soil horizons (Smax = 1.9S10); however the correlation was weak (r = 0.61) suggesting that the feasibility of using a s ingle point isotherm for a data set should be done using C0 > 10 mg P L1. The amount of P adsorbed at the high initial concentration level (typically at C0 = 100 mg L1) is usually a good indication of the soils maximum adsorption capacity base d on a single point isotherm (Harris et al., 1996; Reddy et al., 1998a). The highest maximum sorption capacities (Smax) calculated using linearized fits were observed at the surface (0 2 cm), at 588, 250, and 260 mg kg1 for cores 2, 3, and 4 respectively. Periodic drying of the surface soils may have oxidized ferrous iron to amorphous ferric oxides, and increase d the P sorption capacities at the surface. There was a positive correlation between Smax and amorphous metal concentrations of Fe and Al that were best fit using a linear regression (r = 0.84) (Figure 4 2). This type of positive effect is not uncommon (Reddy et al., 1998a; Pant et al., 2002; Kang et al., 2009), implying that the presence of non -crystalline Fe and Al is a good proxy for P sorption ca pacities regardless of soil depth. To evaluate the effect of OM, the relationship between ox(Fe+Al) and Smax was divided into two groups: low OM (< 4%, n = 33) and high OM (>4%, n = 28). There was a 10% diminishing effect of ox(Fe+Al) at the higher OM cont ent based on the change in linear regression from r = 0.90 for low OM, to 0.80 for high OM content. Similar findings were also observed by Kang et al. (2009), suggesting that the contribution of oxalate -extractable Fe and Al tends to be less effective in s oils with relatively higher OM. The OM content at t he surface was 38% (14) from the wetland, and 48% (2) from the upland, and <2% at depths beyond a meter (Chapter 2). Values of Smax were also generally elevated (but not always) in the deepest horizons, i.e., below 120 cm in cores 1, 2, and 3, and below 140 cm in core 4. This increase in Smax may be associated with the higher clay content (up to 38%) present in those zones. Similar significant
79 direct effects between clay content, Al, and Fe on the Smax ha ve also been observed in the North Carolina Coastal Plain soils that ultimately resulted in a good regression model (Kang et al., 2009). The EPC0 determined from sequential P sorption isotherm data is often used as an estimate of wetland floodwater P conce ntration, where adsorption equals desorption between soil -water interactions (Reddy et al., 1998). The EPC0 values in all four cores were greatest at the surface (0 2 cm), compared to the deeper horizons (Figure 4 3). The values were between 0.813.28 mg L1 from the three wetland cores, and are well within the range of estimates calculated by Pant and Reddy (2003) for a series of surface agricultural soils from a proposed storm water treatment site located in the Lake Okeechobee basin (1.3 to 3.4 mg L1 SR P). Greatz and Nair (1995) reported EPC0 values for four different soil horizons (A, E, Bh, and Bw) from a pasture located on an active dairy operation. They found decreasing EPC values with depth ranging from A = 5.3 (3.5), E = 2.2 (2.7), Bh = 0.1 (0.3) and Bw = 0.0 ; similar in behavior to our findings. Dunne et al. (2006) reported EPC values of 1.08 ( 0.063) mg L1 for these soils based on a 28 d incubation experiment on 20 cm soil cores that were initially spiked at 1 mg L1. They found that completely flooded cores corresponded to lower P flux than partially flooded cores. This would imply that changes in wet and dry conditions have the potential to affect EPC0 values depending on the time of year and the environment (flooded o r completely drained) f rom which soil samples are collected. The 0 2 cm surface soil layer from upland core 4 is unique from the other cores because it has a very high EPC0 value of 8.99 mg L1 based on non linear (polynomial) fit using the six point isotherm (r2 = 0.94, Figure 4 3). The high EPC0 value of upland soil surrounding the wetland would imply that t he surface horizon (0 2 cm ) of the uplands consists of labile P that could potentially behave as a source of P to the wetland if the concentration of P in
80 the incoming water s (runoff) is lower than 8.99 mg L1. H owever, overland sheet flow runoff is uncommon in this region due to the low relief and sandy nature of the soils, but, it is quite possible for a hurricane or a severe storm to induce runoff over short durations. It is more common for P to move laterally along the A and E horizons before ultimately being intercepted by the wetland or ditches (Allen, 1987). Below 40 cm the EPC0 values were <0.004 mg L1, implying that the deeper horizons do not pose an imminent threat to wetland water quality resulting from vertical (upward) P loading of ground water. However, the downward diffusion, and advection associated with water table fluctuations may be responsible for the redistr ibution of P from the surface to deeper (Bh or Bt ) horizons. Effect of Soil:Solution R atio on EPC0 Values and Pore Water C oncentrations Trying to translate EPC0 values to pore water concentrations is a challenging task because of the transient hydrologic conditions which allow for the constant redistribution of soluble P within the soil profile. Additionally, biotic transformations of P by plants and microbes could significantly alter pore water concentrations regardless of EPC0 values of the soils (Kadlec and Wallace, 2008) The EPC0 values calculated us ing 1:20 SSR from cores 1, 2, and 3 were compared to in situ pore water SRP concentrations and total P from the soils up to 110 cm below the soil -water interface (Figure 4 4). P ore water P concentrations at all times were exponentially elevated compared to the EPC0 values based on a regression fit (r2 =0.64) between EPC0 and pore water. The regression was as: EPC0 = 8 104 exp (1.65 pore water) The effect of adjusting the SSR from 1:20 (used in the e xperiment ) to those of field conditions based on soil p orosity is summarized in Table 4 2 T he EPC0 values were significantly higher (p<0.001) when using SSR that were increased to represent the field conditions. The average SSR for core 1 was 1:0.67 (0.03), core 2 was 1:0.75 (0.1), core 3 was 1:0.74 (0.1) and core 4 was 1:0.63 (0.1). T he relation ship between field condition versus 1:20
81 SSR was explained using a log normal regression shown in Figure 4 5 (r2 = 0.76) Nair et al. (1984) found inconsistent results based on their study using t wo separate SS R s (1:10 versus 1:25) to evaluate its effect on EPC0; their relation ship between the two SSR s was linear. Ho pe and Syers (1976) found that t he concentration of P in solution at equilibrium (Ceq) was not always exactly equal at varying SSR. They concluded t hat at higher SSR the greater mass of soil offers more P -sorbing sites if kinetics involved the fast adsorption of P from s olution into the solid phase. I found that the deviation in EPC0 values using the field condition ratios (versus 1:20 SSR) was most d ominant at low initial P concentrations. Generally, at the higher initial concentration ( typically between 0 22 cm) the increase in EPC0 using field conditions was 5.8 (2.8) compared to 1:20 SSR and below 22 cm the increase in EPC0 values was 19.5 (9.9) compared to 1:20 SSR Inc reasing the SSR to mimic field conditions showed convergence towards the measured pore water concentrations especially at the 10 12 cm depth. The pore water concentrations below 12 cm were still higher than the new field conditio n EPC0 values and this could be because of a combination of the downward diffusion of high concentration pore water at the surface and the transient hydrology. Soil P Storage C apacity Soil P storage capacity (SPSC) was estimated using oxalate -extractable (ox -) P, F e, and Al for all four cores. I found a good correlation (r = 0.91) between ox-P and ox-(Fe+Al) from all four cores (Figure 4 2). My intention was to evaluate the surface and subsurface distribution of ox DPS in the soil profile, and determine wh ether the DPS in subsurface samples (irrespective of horizon type) could be related to water soluble P ( WSP ). The DPS values were highest at the surface horizons (0 2 and 1012 cm) for the wetland cores 16.6% ( 3.2), and the DPS value in the upland core w as 20.2% at the surface (0 2 cm), and as high as 35.2% at 2022 cm depth. Below 12 cm the DPS val ue in the wetland was 4.3% (2.7), compared to slightly higher DPS
82 values in the upland 9.6% (11.3). The higher DPS values in the upland suggest that the upla nd soils have higher oxP content compared to ox (Fe+Al), and possibly higher P storage capacities. Reddy et al. (1998) had estimated that 60 70% of the P imported to the Okeechobee watershed was stored within the Bh horizon in upland soils. Although these soils could still have up to 75% of their maximum P storage capacity remaining, Capece et al. (2007) suggested that they are already a significant source of non -point P, despite their total storage capacity. The relationship between WSP and DPS from surfa ce and subsurface horizons seems to generate a change point at around 12% (Figure 4 6 ), corresponding to a PSR of 0.06. Nair et al. (2004) were able to show a change -point of 20% between the oxalate extractable calculated DPS and WSP from surface and sub surface soil horizons (including upper Bt) from the middle Suwannee River Basin, with relatively good confidence (r2 = 0.88) based on a nonlinear fit. DPS values of >30% in topsoils have been reported as a threat to water quality for Mid-Atlantic Soils, a ssociated with P losses from runoff (P autler and Sims, 2000). D eviations in P release indices have also been closely linked with the presence of OM, and the soil test method used to evaluate the P indices (Mukherjee et al., 2009). The soils within the wetl and (cores 1, 2, and 3), and upland (core 4) all seem to have significant remaining capacity to serve as a sink for P based on SPSC (Figure 4 7 ). Excluding the surface horizon, these results along with TP values that are not appreciably elevated suggest th at the soils have been only minimally impacted by cattle farming activity. Except for the E horizon (20 cm) within the upland the soils seem to behave as sinks for P. Horizons with SPSC < 100 mg kg1 may be susceptible to behaving as sources in the future provided the soils cease to retain additional P, and if continued application of manure persists on these sites. The high SPSC observed at the surface of core 2 and between 70 90 cm in core 4 is a result of the high Fe -Al
83 concentration observed at those de pths. Based on these SPSC values it appears that the high pore water P concentration is not entirely controlled by in situ soil conditions, and that ground water hydrology may have a much greater influence on pore water P concentrations than anticipated. A highly fluctuating w ater table at the wetland site could cause that P and Fe become associated as the water table drops. Specifically, some entrained Fe is oxidized as air enters the profile and some entrained P becomes associated with it. Organic matter in the > 1 m of stratified soil would promote rapid reduction in that zone when the water table rises. The P associated with Fe would be released upon Fe reduction resulting in higher solution P concentrations. This redox -driven P immobilization and releas e would be cyclic. Unusually high P concentrations in the pore water would be the result of the scavenging of P by Fe in the oxidizing phase of the cycle. The crystallinity of Fe would be constantly perturbed and its rate of oxidation or reduction would be rapid. The Fe would be in its scavenging state when the sample is dried and Fe is oxidized, resulting in low WS P, PSR (positive SPSC), and EPC. Summary and Conclusions The P sorption capacity and resulting soil characteristics from this study were well de scribed using the Langm u i r model. The highest sorption capacities (Smax) were restricted to the surface (0 2 cm) horizons in both the wetland and upland soil cores, probably because of the high Fe and Al concentration at those depths. In the deeper horizons the increase in Smax may be a result of the accumulation of finer clay content pres ent below 120 cm. There was a strong positive correlation between the oxalate -extractable Fe+Al and Smax (r = 0.84) and oxalateextractable P (r = 0.91) from all four core s and horizons. However, the presence of > 4% O M in the soils seem to have a 1 0 % diminishing effect between oxalate -extractable (Fe + Al) and Smax, probably due to the number of available adsorption sites being reduced as a result of metal complexing with organic acids.
84 The EPC0 values from all four cores were highest at the surface horizon, between 0.813.28 mg P L1 in the wetland, and 8.99 mg P L1 in the upland. Below the surface horizon the EPC0 value dropped until it reached an asymptotic value below 40 cm. The high EPC0 values in the surface layer may potentially act as a source of labile P that could be redistributed vertically within the soil profile, or even laterally by the movement of groundwater. The EPC0 values of the wetland cores were signifi cantly lower than the P concentration in pore water measured at all depth horizons. This may be because under field conditions factors such as transient local hydrology, biotic uptake and release, and complex reactions such as precipitation are just as dom inant as adsorption -desorption; making it difficult to assess pore water P dynamics based on EPC0 values alone. At most EPC0 can be used as a good first estimate for the mobilization potential of P under natural conditions. Based on our investigation, c ha nging SSR did increase the on the final EPC0 values however they were still lower than those measured in the pore water This investigation has relevance to several practical aspects of the chemistry of soil P. In particular, the effec t of varying SSR in soil testing procedures for P needs to be explained more rationally. We do not believe that there is a single best method to measure P sorption isotherms and estimating EPC0 values. Nonetheless, it is in the appl ication of the values if used as a tool to assess P loading that scientists need to pay added attention in correla ting the results in the future. Except for the 20 22 cm horizon in core 4, the positive SPSC values would suggest that these soils can safely sequester additional P before the added P becomes an environmental concern. However, elevated concentrations of pore water P at the surface horizons indicate that the soils and surrounding p ore water may not be in complete equilibrium, probably because of the transient behavior of groundwater hydr ology. It may be possibl e that prolonged dry -outs cause soil oxidation and subsequent mobilization and export of soil -bound P, in which case the
85 wetland simply alternates between storage and release during the wet and dry periods of the year. Based on our findings we recommend retaining water in the wetland for prolonged periods by reduci ng loss from drainage. This could allow for sufficient time for groundwater to kinetically equilibrate with the soils thereby removing some of the P from solution prior to reaching their maximum sorption capacities.
86 Table 4 1. A summary of the Langmuir sorption parameters based on batch incubation experiment, and oxalate extractable data for P, Fe and Al. BDL=below detection limit Core 1 (wetland) Depth (cm) WSP (mg kg 1 ) ox P (mmol kg 1 ) ox Fe (mmol kg 1 ) ox Al (mmol kg 1 ) S max (mg kg 1 ) k (L mg 1 ) K d (L kg 1 ) EPC 0 (mg L 1 ) DPS (%) PSR 0 2 8.19 1.2 2.0 7.7 57 12.6 2.6 3.280 25.3 0.127 10 12 10.98 3.2 3.9 26.9 172 1.7 7.4 0.970 20.8 0.104 20 22 0.70 1.8 2 .5 40.5 172 8.2 19.4 0.040 8.3 0.041 30 32 0.32 1.5 2.3 40.0 172 11.6 19.0 0.013 7.1 0.036 40 42 0.13 0.9 1.9 30.1 145 9.8 18.6 0.004 5.5 0.028 50 52 0.08 0.3 2.3 21.8 114 5.8 19.3 0.004 2.8 0.014 60 62 0.11 0.1 2.8 17.4 99 4.6 18.3 0.002 1.4 0.007 70 72 0.12 0.02 2.7 12.1 81 2.3 19.2 0.004 0.3 0.002 80 82 0.05 BDL 4.0 13.0 91 3.9 18.5 0.001 90 92 0.12 BDL 6.6 9.8 93 5.1 19.2 0.051 100 102 0.05 BDL 10.7 12.6 132 7.6 20.1 0.005 110 112 0.10 BDL 17.3 16.9 167 15.0 19.0 0.002 120 122 0.05 BDL 17.4 13.9 170 8.4 19.7 0.002 130 132 0.08 BDL 18.5 15.4 172 7.3 19.9 0.004 138 140 0.02 BDL 12.4 13.5 143 8.6 19.4 0.000
87 Core 2 (wetland) Depth (cm) WSP (mg kg 1 ) ox P (mmol kg 1 ) ox Fe (mmol kg 1 ) ox Al (mmol kg 1 ) S max (mg kg 1 ) k (L mg 1 ) K d (L kg 1 ) EPC 0 (mg L 1 ) DPS (%) PSR 0 2 13.56 13.6 42.5 128.0 588 3.4 14.3 0.810 15.9 0.080 10 12 6.04 1.0 2.5 13.9 74 1.5 4.3 0.640 12.8 0.064 20 22 0.99 2.0 3.6 38.5 172 5.3 16.9 0.054 9.5 0.047 30 32 0.17 1.1 2.3 36.5 159 9.0 18.8 0.0 06 5.8 0.029 40 42 0.11 1.4 2.7 47.2 185 10.8 18.5 0.003 5.5 0.028 50 52 0.04 0.6 1.6 26.8 115 6.7 19.2 0.002 4.5 0.022 60 62 0.06 0.6 2.1 30.3 137 7.3 19.7 0.003 3.8 0.019 70 72 0.06 0.4 3.3 27.8 127 3.9 18.5 0.000 2.4 0.012 80 82 0.06 0.2 4.9 30.3 1 27 6.1 19.1 0.002 1.0 0.005 90 92 0.03 BDL 11.6 19.7 132 8.4 20.0 0.002 100 102 0.08 BDL 1.1 3.9 44 2.5 17.0 0.003 110 112 0.02 BDL 0.8 4.1 46 2.3 20.0 0.001 120 122 0.02 BDL 10.8 11.4 125 5.3 19.7 0.001 130 132 0.01 BDL 16.9 12.7 149 16.7 20. 1 0.001 142 144 0.05 BDL 15.7 10.3 154 13.0 19.9 0.002
88 Core 3 (wetland) Depth (cm) WSP (mg kg 1 ) ox P (mmol kg 1 ) ox Fe (mmol kg 1 ) ox Al (mmol kg 1 ) S max (mg kg 1 ) k (L mg 1 ) K d (L kg 1 ) EPC 0 (mg L 1 ) DPS (%) PSR 0 2 11.48 5.4 19.4 53.7 250 2.2 8.3 1.200 14.7 0.074 10 12 2.80 0.9 2.4 14.3 63 2.0 9.4 0.310 10.2 0.051 20 22 0.36 1.4 2.8 32.2 147 7.5 18.3 0.015 8.2 0.041 30 32 0.07 1.5 2.7 45.2 189 13.3 19.4 0.003 6.2 0.031 40 42 0.08 1.1 2.0 37.7 159 12.6 19.3 0.003 5.8 0.029 50 52 0.06 1.0 3.8 46.9 175 14.3 19.2 0.002 4.0 0.020 60 62 0.05 0.5 4.0 30.6 137 8.1 19.0 0.001 3.0 0.015 70 72 0.04 0.1 3.1 17.0 90 5.8 18.6 0.001 1.1 0.005 80 82 0.05 0.1 5.7 10.5 70 4.8 18.8 0.000 0.9 0.004 90 92 0.03 BDL 7.6 12.3 92 21.8 20.0 0.002 100 102 0.03 BDL 7.9 10.9 89 5.9 18.7 0.001 110 112 0.04 BDL 8.1 11.3 81 3.6 19.4 0.000 120 122 0.02 BDL 16.1 15.6 105 3.5 19.6 0.001 128 130 0.02 BDL 13.6 11.7 98 5.6 18.7 0.001
89 Core 4 (upland) Depth (cm) WSP (mg kg 1 ) ox P (mm ol kg 1 ) ox Fe (mmol kg 1 ) ox Al (mmol kg 1 ) S max (mg kg 1 ) k (L mg 1 ) K d (L kg 1 ) EPC 0 (mg L 1 ) DPS (%) PSR 0 2 74.40 6.8 40.1 27.3 263 0.5 6.8 8.990 20.2 0.101 10 12 1.82 0.5 6.8 4.5 40 2.4 8.1 0.250 9.4 0.047 20 22 3.02 1.8 4.1 6.2 71 10.1 10.7 0 .420 35.2 0.176 30 32 0.88 0.9 3.7 3.4 42 10.4 7.1 0.070 26.4 0.132 40 42 0.16 3.3 44.4 14.3 196 25.5 18.4 0.004 11.3 0.056 50 52 0.03 BDL 30.8 17.1 135 6.7 20.4 0.001 60 62 0.03 BDL 35.9 20.0 159 10.5 19.6 0.001 70 72 0.03 BDL 76.1 16.6 145 9.8 1 9.7 0.001 80 82 0.04 0.1 101.9 12.0 116 4.7 20.1 0.002 0.2 0.001 90 92 0.03 BDL 12.0 11.0 78 6.1 20.5 0.002 100 102 0.03 0.04 5.0 11.7 82 10.1 20.6 0.001 0.4 0.002 110 112 0.02 0.1 4.8 11.9 86 7.8 19.5 0.001 1.7 0.009 120 122 0.02 0.2 4.8 11.1 86 3.6 20.0 0.001 2.2 0.011 130 132 0.03 0.3 4.7 11.2 93 15.4 20.4 0.001 4.1 0.021 140 142 0.02 0.5 5.2 12.9 109 9.2 20.4 0.002 5.7 0.028 150 152 0.04 1.0 6.1 17.1 143 23.3 20.1 0.002 9.0 0.045 158 160 0.05 1.2 6.3 18.9 152 22.0 19.2 0.002 9.5 0.048
90 Ta ble 4 2. A comparison between (i) effect of soil:solution ratio on EPC0 for soil samples from all four cores. (ii) field condition EPC0 values with pore water concentrations for cores 1, 2, and 3 Core 1 EPC 0 (mg L 1 ) Increase in EPC 0 Pore water Dept h (cm) soil/soln ratio (1:20) soil/soln ratio (field condition) (mg L1) 0 2 3.28 6.4 (1:0.71) 2 10 12 0.97 5.1 (1:0.66) 5 3.92 (2.1) 20 22 0.040 0.22 (1:0.73) 6 2.65 (1.8) 30 32 0.013 0.19 (1:0.72) 15 2.47 (1.3) 40 42 0.004 0.088 ( 1:0.71) 22 0.99 (0.5) 50 52 0.004 0.058 (1:0.67) 17 0.34 (0.2) 60 62 0.002 0.082 (1:0.66) 37 0.82 (0.4) 70 72 0.004 0.078 (1:0.70) 21 0.64 (0.4) 80 82 0.001 0.037 (1:0.67) 34 0.72 (0.4) 90 92 0.051 0.085 (1:0.66) 2 1.24 (0.7) 100 102 0.005 0.033 (1:0.64) 7 110 112 0.002 0.071 (1:0.65) 39 2.71 (1.3) 120 122 0.002 0.034 (1:0.64) 21 130 132 0.004 0.052 (1:0.63) 14 138 140 0.001 0.015 (1:0.62) 15 Core 2 EPC 0 (mg L 1 ) Increase in EPC 0 Pore water Depth (cm) soil/soln ratio (1:20) soil/soln ratio (field condition) (mg L 1 ) 0 2 0.81 4.9 (1:0.91) 6 10 12 0.64 4.0 (1:1.05) 6 3.36 (1.4) 20 22 0.054 0.33 (1:0.84) 6 1.22 (0.6) 30 32 0.006 0.11 (1:0.80) 18 0.67 (0.3) 40 42 0.003 0.072 (1 :0.83) 21 0.40 (0.2) 50 52 0.002 0.021 (1:0.78) 12 0.65 (0.4) 60 62 0.003 0.040 (1:0.76) 14 0.19 (0.1) 70 72 0.001 0.036 (1:0.73) 36 0.17 (0.1) 80 82 0.002 0.040 (1:0.71) 21 0.08 (0.1) 90 92 0.002 0.021 (1:0.69) 12 0.06 (0.1) 100 102 0.003 0.053 (1:0.66) 19 110 112 0.001 0.022 (1:0.65) 18 0.04 (0.03) 120 122 0.001 0.007 (1:0.64) 7 130 132 0.001 0.013 (1:0.64) 19 142 144 0.002 0.03 (1:0.63) 15
91 Core 3 EPC 0 (mg L 1 ) Increase in EPC 0 Pore water Dept h (cm) soil/soln ratio (1:20) soil/soln ratio (field condition) (mg L 1 ) 0 2 1.2 5.2 (1:0.92) 4 10 12 0.31 1.9 (1:0.68) 6 2.23 (0.4) 20 22 0.015 0.24 (1:0.83) 13 1.53 (0.3) 30 32 0.003 0.043 (1:0.84) 14 0.90 (0.4) 40 42 0.003 0.053 (1:0.80) 21 0.60 (0.2) 50 52 0.002 0.035 (1:0.77) 18 0.59 (0.2) 60 62 0.001 0.036 (1:0.74) 53 0.74 (0.1) 70 72 0.001 0.032 (1:0.72) 36 0.37 (0.04) 80 82 0.001 0.032 (1:0.71) 32 0.30 (0.1) 90 92 0.002 0.025 (1:0.72) 11 0.44 (0.0 4) 100 102 0.001 0.021 (1:0.66) 36 110 112 0.001 0.026 (1:0.65) 26 0.37 (0.01) 120 122 0.001 0.016 (1:0.65) 15 128 130 0.001 0.022 (1:0.63) 32 Core 4 EPC 0 (mg L 1 ) Increase in EPC 0 Depth (cm) soil/soln ratio (1:20) soil/soln rat io (field condition) 0 2 8.9 9.96 (1:0.87) 1 10 12 0.25 1.73 (1:0.74) 7 20 22 0.42 2.92 (1:0.68) 7 30 32 0.070 0.38 (1:0.71) 5 40 42 0.004 0.11 (1:0.65) 25 50 52 0.001 0.02 (1:0.58) 14 60 62 0.001 0.023 (1:0.60) 16 70 72 0.00 1 0.024 (1:0.59) 18 80 82 0.002 0.026 (1:0.58) 14 90 92 0.002 0.024 (1:0.58) 16 100 102 0.001 0.011 (1:0.60) 9 110 112 0.001 0.012 (1:0.59) 15 120 122 0.001 0.017 (1:0.60) 17 130 132 0.001 0.018 (1:0.60) 15 140 142 0.002 0.018 ( 1:0.59) 8 150 152 0.002 0.026 (1:0.59) 14 158 160 0.002 0.042 (1:0.58) 18
92 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 0-2 cm 20-60 cm 70-130 cm Ceq(mg L1)Ceq/S (g L1)Core 3 Smax= 161 21 k = 11 3 Smax= 250 k = 2 Smax= 89 11 k = 7 6 0 20 40 60 80 100 120 140 160 180 0 2 4 6 8 10 0-2 cm 10-40 cm; 100-140 cm 50-90 cm Ceq(mg L1)Ceq/S (g L1)Smax= 57 k = 13 Smax= 96 12 k = 4 1 Smax= 161 16 k = 9 4Core 1 0 20 40 60 80 100 120 140 160 180 200 0 2 4 6 8 10 0-2 cm 10-90 cm; 120-144 cm 100-110 cm Ceq(mg L1)Ceq/S (g L1)Core 2 Smax= 558 k = 3 Smax= 138 29 k = 8 4 Smax= 45 1 k = 2 0.1 0 40 80 120 160 200 240 0 3 6 9 12 0-2 cm (Ap) 10-30 cm (E) 40-160 cm (Bt and Btg) Ceq(mg L1)Ceq/S (g L1)Core 4 Smax= 51 18 k = 8 5 Smax= 121 37 k = 12 7 Smax= 263 k = 0.5 Figure 4 1. Lan gm u i r plots for P sorption of all four cores. The P sorption maxima, Smax (mg kg1) and bonding energy constant, k (L mg1) are shown next to the regres sion derived for different soil horizons. Overall R2
93 0 2 4 6 8 10 12 14 16 18 20 0 40 80 120 160 200 ox(P) Smax Smax= 1.69 + 0.077 ox(Fe+Al ) R = 0.84 ox(P ) = 1.18 + 0.08 ox(Fe+Al ) R = 0.91ox[Fe+Al ] ( mmol kg1)( mmol kg1) Figure 4 2. Relationship between (i) oxalate -extractable P (n=39) versus oxalate -extractable (Fe+Al), and (ii) P sorption maxima (Smax) versus oxalate -extractable (Fe+Al) for all four cores an d depth horizons (n=61). Adsorption Desorption Sorption (mg kg1) -80 -60 -40 -20 0 20 40 60 80 100 120 0 2 4 6 8 10 12 core 1 core 2 core 3 core 4 R2 = 0.9959 R2 = 0.9985R2 = 0.9382 R2 = 0.9771Initial concentration (mg L1) Figure 4 3. Surface horizon (0 2 cm) EPC0 values for cores 1, 2, 3, and 4 calculated using a six point isotherm as a function of SRP adsorption and desorption. X axis corresponds to initial P concentration (mg L1). Regression based on linear fit using all six points for cores 1, 2, 3, and a non-linear fit for core 4.
94 Core 1 Core 2Core 3 Figure 4 4. Depth distribution of EPC0 values (mg SRP L1) compared with seasonally variable pore water concentrations (mg SRP L1), and soil total P from wetla nd cores 1, 2 and 3. EPC0 = 8 104 e(x) (x = 1.65 pore water) (r2 = 0.64) based on all four soil cores. 0 2 4 6 8 10 12 0 2 4 6 8 10 EPC0 at soil:soln ratio = field conditionEPC0at soil:soln ratio = 1:20 y = 0.68Ln(x) + 4.26 R2= 0.76 Figure 4 5. Relationship between EPC0 (mg L1) at field condition versus those measured using 1:20 soil:solution ratio for all four soil cores ( n=61).
95 0 10 20 30 40 50 60 70 80 0 5 10 15 20 25 30 35 40 0-12 cm below 12 cm change point (12%)DPS (%)WSP (mg kg1) Figure 4 6 Relationship between the water soluble P (WSP) and the degree of P saturation (DPS) calculated using an oxalate extraction from surface and subsurface horizons of all four cores.
96 Figure 4 7 Soil P storage capacity for all four cores, calculated using the PSR values and oxalate extractable Fe, Al concentrations in the soil.
97 CHAPTER 5 PRACTICAL IMPLICATIONS AND RECOMMENDATIONS Soil Erosion and Stratification The soils encountered within the Okeechobee Basin wetland and upland were generally sandy consisting of high P and OM con tent at the surface that gradually decrease d with depth. What was very striking about the wetland soils was the presence of un disturbed stratification observed in all three wetland soil profiles, which seeme d unusual at first; h owever I believe that continued soil erosion and subsequent deposition of the upland soils by wind and cattle activity may have resulted in the formation of these features. Typically during the dry season surface soils are susceptibl e to being exposed to wind erosion thereby being displaced and redistributed. D uring numerous field visits I have observed large areas of both wetland and upland soils that have been exposed as a result of erosion. Overgrazing by cattle in a particular are a has also been known to cause ero sion, by reducing the vegetation cover and exposing the topsoil. Organic matter along with b oth inorganic and particularly organic P is displaced as a result of soil erosion. Most farm management plans for P assume that i f erosion is controlled, so are P losses. One way to address this issue is through soil conservation measures such as crop rotations or by making sure that there is a good vegetation cover year round K nowing the crop nutrient limit of the soil s and prop er manure management plans can help to retain vegetation through out the year thereby reducing soil erosion Stratifica tion within soil profiles may not allow for a B horizon to form as new material is being constantly deposited over the existing soil pr ofile. All three wetland cores (1, 2 and 3) were lacking the p resence of a well -formed Bh or Bt horizon s seen in t he upland core This could have implications for P sequestration, as Bh horizons have a high capacity to retain P One strategy for wetland re storation would be to increase the wetland
98 hydroperiod. This could potentially help reduce the erosion of soil by wind, and would also allow for denser vege tation cover if soil moisture were higher than present Scaling -up to Basin wide P Loads Based on Using Local Flux Measurements The total P fluxes (diffusive and advective) calculated in Chapter 3 were use d to scale up to estimate basin -wide P loads associated with isolated wetlands within the Lake Okeechobee drainage basin. These estimates could be usef ul to predict the amount of P load reduction to Lake Okeechobee that would occur if wetland restoration work can increase P retention in soils and reduce P fluxes exiting the wetland via overland modes. To estimate the amount of P associated with isolated wetlands that could potentially be stopped (controlled) from reaching Lake Okeechobee I used P fluxes numbers from this study and extrapolated them to the four priority basins of the Lake Okeechobee watershed. These priority basins (S 65E, S 65D, S 154, and S 191) have consistently produced the highest levels of P concentrations of all the tributary basins to Lake Okeechobee (Flaig and Reddy, 1995). These four basins are dominated by beef and dairy operations and account for 12% of the land area of the Okee chobee basin, yet contribute 35% of the P to the Lake (Dunne et al., 2006) In 1999, the Lake Okeechobee Action Plan was put in place to address the issue of reducing high P loads to the L ake. Between 1997 and 2001, estimated P loads to the Lake from 25 contributing basins was ~303 tons yr1 (Hiscock et al., 2003). In 2004, SFWMD reported the total P load entering the Lake to be ~543 tons yr1. It is estimated that if the four priority basins met their target loads, then P loading to Lake Okeechobee could be reduced by over 100 tons yr1. Basin w ide P Loads f rom Isolated W etlands The four priority basins cover an estimated area of 3340 km2 within the Okeechobee drainage basin, of which isolated wetlands represent about 7%, totaling 234 km2 (McKee, 2005). Ass uming the average size of a wetland to be about 0.011 km2 (size of Wetland 1 on Larse n
99 Ranch), I estimate there are some 21, 273 such isolated wetlands scattered within the four priority basin s Bevc (2007) had also estimated a comparable number (22, 400) of isolated wetlands within the priority basin s By using the average of the total mass load of P entering and leaving the wetland (from Chapter 3) multiplied by the number of wetlands in the basins I estimated a basin -wide effect of loading associated with diffusion and advection into the wetland, and ground water out of the wetland (Table 5 1). T o estimate P loads associated with ditch flows, I assumed that only about 45% of the total isolated wetlands in the priority basin are ditched (McKee, 2005; Dunne e t al., 2007b). On a basin -wide extent the total P load out of wetlands via advective and diffusive fluxes accounts for 131 ton yr1. This does not mean that all 131 tons reach the Lake. By restoring the wetlands it should be possible to reduce this load, d epending on the efficiency of the restored wetlands to retain P. The total amount of P load exiting the wetlands is a lot higher, 250 ton yr1. The estimated amount of P exiting the se wetlands via ditch flow is 24 ton yr1. D amming the ditches to provide f or greater hydroperiods could (i) reduce the P load associated with fluctuating water tables, and (ii) substantially reduce the export via ditches This would however increase P loading into the ground, assuming that P inputs to the wetland remain the same. However, b ased on the soil P storage capacities the wetland soils still have substantial capacity to retain P. Perkins (2007) had estimated between 0.2 and 9.6 metric tonnes per year of P load entering Lake Okeechobee from isolated wetlands with the wate rshed. His estimates were significantly smaller compared to other basin -scale estimates of 415 tonnes yr1 (Boggess et al., 1995) and 582 tonnes yr1 (SFWMD, 2001). In his defense, he acknowledges that not all historically drained isolated wetlands functio n hydrologically similarly within the landscape or have had similar historic and current land use practices (fertiliz er, cattle stocking, etc.). His estimate s of basin -scale load are only pertinent to the fraction of runoff passing through
100 depressional we tlands and out ditches. It does not incorporate P load related to regional drainage from ditch networks, land use contribution of dissimilar nature (dairies and other relatively high intensity operations), and P load from tributary and in-stream internal loads. Literature estimates of P reduction due to water detention in wetlands range between 4 80% (Zhang et al., 2006). With such a wide range in reduction, it is obv ious that more research is necessary to confirm the effectiveness of water detention in iso lated wetlands to reduce P loads. However, the estimated 100 ton yr1 load reduction from the four priority basins discussed in the Lake Okeechobee Action Plan of 1999 may not be achievable by water detention alone, and other possible strategies need to be considered Perkins (2007) showed that P treatment in these wetlands was comparable to that using the landscape -scale ps eudo2d modeling approach He concluded that u nder current managed wetland hydrology, the utility of these wetlands to retain P does no t fully result in P load reduction to Lake Okeechobee but with measures to increase retention time in the wetlands; higher treatment potential might be expected. The duration of P treatment under a modified hydraulic regime, such as blocking the export of P loading via ditches, is not known and would require a more in-depth look at the mechanisms and limits of the P cycle. Basin w ide P Loads f rom D itches It has been proposed that drainage ditches behave similar ly to isolated wetlands and can serve as either a source or sink for P (Dunne et al., 2007b) Ditches can serve as temporary P sinks; as overland flow transports with it both organic and inorganic forms of P. If the water in these ditches can be retained then the ditches could function as sinks, provi ded the soils underlying these ditches have not been impacted with high concentrations of P. Bevc (2007) estimated the range of P load associated with drainage ditches and canals within the four prio rity basins to be between 31 and 362 ton yr1. The high d egree of variability
101 was associated with the range in P fluxes associated with ditch soils. There was al so some discrepancy with regard to the total areal cover of ditches occupied within the basins. Com paring the P loads associated with isolated wetlands with drainage ditches it is evident that best management practices sho uld be applied to the ditches in the same manner in which they target isolated wetlands. Perkins (2007) had estimated that groundwater outflow comprised approximately 35% of the total h ydraulic outflow from the study wetlands, and that the principal component responsible for the majority of P export was ditch flow, which yielded between 7.6 and 43.7 % (wetlandaveraged) of the total TP outflow. This finding wa s critical in the development of a conceptual model of the P dynamics in these wetlands becau se the treatment effectiveness wa s a function of not only what comes out the ditch as surface water export, but also the load to groundw ater that will ultimately reach the lake. Kinetic Contr ol of Dissolved P in Aquatic Systems An understanding of the chemical parameters req uired to regulate P uptake and release in natural aquatic systems requires a detailed assessment of the controls on the bioavailability of P. These efforts are hampered by the difficulty of distinguishing between inert and bio reactive P. The ultimate source of dissolved phosphate is the weathering of continental rocks, where it may exist in a variety of insoluble minerals at a concentration of about 0.1% (Froelich, 1988). I n lentic aquatic systems such as lakes, wetlands, and ponds, it is generally assumed that some of these minerals are in equilibrium with water, thus controlling the concentrations of dissolved P. Seldom, however, is equilibrium proved by laboratory manipul ations demonstrating the solubility control of the reactive solid phase phosphates in natural systems, except in a few cases (Emerson and Widmer, 1978; Golterman, 1995), due to lack of sufficient P concentrations in the sol id phase. Presently, a large pot ential reservoir of available reactive P to natural aquatic systems is from anthropogenic sources lik e fertilizers and manure, due to the growing demands
102 on the f ood-crop and live -stock industries The environmental and economic consequences of P in soils resulting from anthropogenic sources have prompted a significant long-term effort to understand soil -solution phosphate reactions. In solution, phosphate reacts quickly with a wide variety of surfaces, being taken up by and released from particles through a complex series of sorption reactions. In the field of soil sciences and agricultural research, phosphate sorption has been known to occur both fast and slow, with reactions being reversible or irreversible depending on the sorption (or buffering) capacit y of the solid phase, and the residence time in contact with the solution (Barrow, 1983; Froelich, 1988). The simplest plausible model capable of explaining the wide variety and complexity of observed behavior of phosphate soil -solution reactions involves the uptake and release of dissolved phosphate via a two-step process (Barrow, 1983; Mansell, et al., 1995; Chen, et al., 1996): rapid surface adsorption or desorption of phosphate onto or off of reactive surfaces, followed by a slow penetration or release by solid state diffusion of this surface phosphate into or out of the interior of the particles. The initial fast reaction has time scales of minutes to hours, and the second slow reaction has time scales of days to months, or even years (Barrow, 1983). Th e fast reaction can account for 50 90% of the total P response, and the loglog plots of P sorbed vs. time can often be fit to one or more equilibriu m models such as Langmuir or Freundlich. Studies related to impacted soils involving short -term effect of P (Graetz and Nair, 1995; Reddy et al., 1998; Pant et al., 2002, 2003) typically ignore the consequences of two-step kinetics, and plot P sorption aft er 24 h to be near equilibrium These plots, often referred to as adsorption isotherms, are taken to repr esent the loci of points of a series of equilibrium conditions between solid (soil sample) and solution. Each poin t on these diagrams is termed an equilibrium phosphate (phosphorus) concentration (EPC) (as shown in Figure 4 3 in Chapter 4).
103 True equilibriu m is seldom attained on laboratory time scales except for conditions chosen to entirely mimic those of natural systems (Sharpley et al., 1981; Froelich, 1988). The phosphate retention characteristics of the soils are determined based on the adsorption isot herms. The literature is full of incomplete interpretations of data on P sorption because of the synergistic effect of kinetics on the other variables associated with P sorption experiments. Rao et al. (1980) may have been the first group to point out that the procedures used to determine the P adsorption over the years had varied widely, making the comparison of isotherm coefficients from different research difficult, if not impossible. Phosphate adsorption has been shown to vary with soil:solution ratios (SSR), as shown in this study and by others (Hope and Syers, 1976; Barrow and Shaw 1979; Nair et al., 1984), ionic strength (Ryden et al., 1977); and cation species of the supporting electrolyte (Helyar et al., 1976; Barrow et al., 1980., Nair et al., 1984). Even though adsorption is rapid, varying the equilibration period between 0.5 to 7 d has also resulted in differences in adsorption capacities (Rao et al., 1980). In addition, Nair et al., (1984) offered other possible variables that coul d affect equilibration, such as condition and duration of sample storage prior to experiment, range of initial P concentrations used in the experiment, geometry of mixing (volume of soil suspension to headspace volume in equilibration vessel), rate (low, medium, high) and type of shaking device (reciprocal vs. end -over end). In order to eliminate some of the inconsistencies associated with P adsorption procedures, Nair et al. (1984) proposed a standard P adsorption protocol for soils and sediment based on the ability of multiple laboratories to produce consistent results over a wide range of soil samples. Their procedure included: a Weight of soil/sediment 0.5 to 1.0 g b Soil:Solution ratio 1:25 c Extraction time 24 h
104 d Electrolyte 0.01M CaCl2 e Initial dissolved inorganic P c oncentrations 0, 6.45, 16.13, 32.26, 161.3, 323 mol P L1 as KH2PO4 or NaH2PO4 f Temperature 24 to 26 C g Microbial inhibition 20 g L1 chloroform h Equilibration vessel 50 ml to provide at least 50% head space i Shaking end -over -end shaker j Separation filt er through 0.45 m (0.2 m for clays) k Analysis any procedure for determination of orthophosphate, manual or automatic, capable of detecting >750 mol P l Replication duplicate equilibrations, and single analysis of orthophosphate in solution Nair et al. (1 984) concluded that P sorption was generally lower (up to 40%) with K+ as the supporting electrolyte cation compared with Ca2 +, consistent with the finding of Barrow et al. (1980) and Heylar et al. (1976). The effect was most pronounced at the higher initi al P concentration. They also found tha t changes in SSR had an effect on EPC0 values, and even though this effect was generally smaller than the effect of the supporting electrolyte, they concluded that it was large enough to warrant standardizing both fac tors in the procedure. Role of pH, S oil and Solution C o mposition in Evaluating P M obilization pH: The dominant P species in solution of natural environments in the pH range of 4.86.0 is the H2PO4 ion (White and Taylor, 1977). Koski -Vahala and Hartikainen (2001) showed that the effect of pH on the mobilization of P was dependent on the concentration of suspended solids in the system. Their objective was to evaluate the role of pH at different concentrations of suspended solids in solution, so as to assess the risk of P loading due to resuspended sediments. They found that at solid concentrations between 17 to 170 mg L1, P release increased sharply below pH 5 and above pH 7. At a high concentration level of 1700 mg L1 no effect was
105 observed at pH ranging f rom 3 to 8.5, and no significantly enhanced P mobilization was found until above pH 9. As the concentration of suspended solids diminis hed, the acid induced increase in the P release (pH < 5) was mo re distinct than the base induced release (pH > 7). While the effect of pH on P mobility were not evaluated in this study, its effect on suspended solids can have practical implications for this study because wetland soils located on dairy and beef farming facilities are constantly being re -suspen ded due to cattl e activity. Already there have been best management practice programs in place which requires the implementation of fencing ponds, troughs, or tanks used for cattle watering as well as the setting of stocking rates for individual pastures based on P loadi ng characteristics of the site. It may be feasible in the future to limit (if not completely eliminate) the access of cows to open water wetlands. Soil composition: The initial P adsorption process is dependent on charge balance and surface area, and can o ccur on a variety of natural particle surfaces. The second slow diffusion step is very dependent on the composition of the solid phase (Froelich, 1988 ). Soils containing Fe and Al hydroxides, or surfaces coated with these phases, display a much higher slo w sorption capacity due to the reaction of phosphate with the oxides. The effect of time and concentration on P sorption by Fe and Al hydroxide surfaces has been well documented (Barrow 1983; Bolan et al., 1985). Soils in which P adsorption is mainly associated with Fe -oxides are especially likely to show slow uptake into particles, particularly with the poorly crystallized oxides. Further, the capacity of these oxides to sorb P decreases with their age and apparent crystallinity (McLaughlin et al. 1981) There is a recognizable relationship between extractable Fe and Al hydroxides and the P sorption capacity, as shown in this study and in many pre vious studies ( Richardson, 1985; Reddy et al., 1998a; Pant et al., 2002). These metal oxides have show high P sorption capacity over short experimental time scales, however over longer time periods
106 they may lose some of the P if the redox condition within the system is altered (Baldwin et al., 2000; Watts, 2000; Klotz and Linn, 2001). This could potentially af fect the resulting EPC values depending on the redox condition (potential) of the soil when sampled. I therefore highly recommend measuring the redox potential of the soil at the time of sampling. Solution composition: The effect of ionic strength in the s olution phase on P mobilization has received only sporadic attention (Nair et al., 1984; Koski -Vahala and Hartikainen, 2001). In general, sorption increases as the ionic strength of the solution increases (Nair et al., 1984). However, this may not be the c ase in the presence of anions such as SO4 2 -, F-, OHthat compete with phosphate (or H2PO4 -) ions for surface sorption sites, reducing the amount of P sorbed during the fast reaction, and consequently affect ing the rate and mass of phosphate taken up or re leased during the slow reaction. In order to minimize artifacts associated with solution composition, experiments are generally performed using a buffering electrolyte such as 0.01M CaCl2 or KCl. Some experiments have also relied on using natural water fro m the site itself to overcome the effect of solution composition. The E PC values calculated using site -specific waters are represented as EPCw, and do not necessarily equal to EPC0. For example, Dunne et al. (2006) calculated EPCw for the Larson wetland soils using water from the Kissimmee River as the buffering solution. Their EPCw values were lower (nearly 64%) than those estimated using 0.01M KCl solution from this study. I therefore recommend using either CaCl2 or KCl solution as long as there is consis tency while comparing data. Strategy to Reduce P Loads Plans to restore isolate d wetlands within the watershed have been proposed as a viable option to store water, increase wetland hydroperiod, and sequester P. One way to do this is to build a control str ucture at the head of the ditch or simply damming it. T his dissertation and other related studies (Dunne et a l., 2006; and Perkins, 2007), indicate believe that P loads exiting the
107 wetlands via the ditches could be reduced by damming ditch es that drain wat er out of wetland s This plan was put in place in April 2009, where the ditch connected to wetland 1 was dammed to reduce (if not completely prevent) loss of water from the wetland, thereby increasing the wetland hydroperiod (Figure 5 1 ). Sampling work and experiments in the future will be conducted so as to evaluate the effectiveness of damming the wetland to mitigate P loss. Results will be compared with wetland 2, whose undammed ditch will serve as a control for comparison. Table 5 1. Basin -wide estima te of P load from isolated wetlands associated with diffusion, advection, ditch flow, and ground water Wetland wide load (kg yr 1 ) Basin wide load (ton yr 1 ) Into the wetland Diffusion 5.2 111 Advection 0.93 20 Out of the wetland Ditch flow 2.5 24 Ground water 10.6 226
108 Figure 5 1 Dam structure constructed to reduce the amount of water exiting wetland 1 via ditch flow.
109 LIST OF REFERENCES Ahn, H., and R. T. James. 2001. Variability, uncertainty, and sensitivity of phosphorus deposition load e stimates in South Florida. Water, Air, and Soil Pollution. 126: 3751. Allen, L. H. 1987. Dairy-siting criteria and other options for waste water management on high water table soils. Soil and Crop Science Society. Fla. Proc. 47: 108 127. Alloush, G. A., G D. Boyer, D. P. Belesky, and J. J. Halvorson. 2003. Phosphorus mobility in a karst landscape under pasture grazing system. Agronomie. 23: 593600. American Society of Limnology and Oceanography, 1998. Denitrification in estuarine sediments determined by membrane inlet mass spectrometry. Limnology and Oceanography. 43: 334339. Baldwin, D. S., A. M. Mitchell, and G. N. Rees. 2000. The effects of in situ drying on sediment phosphate interactions in sediments from an old wetland. Hydrobiologia. 431:3 12. Barrow, N. J., J. W. Bowden, A. M. Posner, and J. P. Quirk. 1980. Describing the effects of electrolyte on adsorption of phosphate by a variable charge surface. Australian Journal of Soil Research. 18:395404. Barrow, N. J. 1983. A mechanistic model for descr ibing the sorption and desorption of phosphate by soil. Journal of Soil Science. 34: 733750. Berner, R. A. 1980. Early Diagenesis. A Theoretical Approach Princeton University Press, New Jersey. Bevc, E. 2007. Nutrient transport in groundwater near isolat ed wetlands and drainage ditches: Implications to best management practices. .M.E. Thesis: University of Florida, Gainesville. Bezkorowajnyj, P. G., A. M. Gordon, and R. A. McBride. 1993. The effect of cattle foot traffic on soil compaction in a silvo -past oral system. Agroforestry Systems. 21: 110. Boggess, C. F., E. G. Flaig, and R. C. Fluck. 1995. Phosphorus budget basin relationshi p s for Lake Okeechobee tributary basins. Ecolog ical Engineering. 5: 143162. Bolan, N. S., N. J. Barrow, and A. M. Posner. 1 985. Describing the effect of time on sorption of phosphate by iron and aluminum hydroxide. Journal of Soil Science. 36: 187197. Bostic, E. M., and J. R. White. 2006. Soil phosphorus and vegetation influence on wetland phosphorus release after simulated d rought. Soil Science Society of America Journal. 71: 238244. Bostrm, B., M. Jansson, and C. Forsberg. 1982. Phosphorus release from lake sediments. Arch. Hydrobiol. Beih. 18: 5 9.
110 Bottcher, A. B., T. K. Tremwel, and K. Campbell. 1999. Phosphorus manageme nt in flatwood (spodosols) soils: p 405424. In: Reddy, OConnor, and Schelske (ed.). Phosphorus Biogeochemistry in Subtropical Ecosystems. Lewis Publishers, Boca Raton, Florida. Braskerud, B. 2002. Factors affecting phosphorus retention in small construct ed wetlands treating agricultural non -point source pollution. Ecological Engineering. 19: 4161. Breeu w sma, A., J. G. A. Reijerink, and O. F., Schoumans. 1995. Impact of manure on accumulation and leaching of phosphate in areas of intensive livestock farmi ng. p 239251. In K. Steele (ed.) Animal waste and land water interface. Lewis Publishing -CRC, New York. Campbell, K. L., J. C. Capece, and T. K. Tremwel. 1995. Surface/subsurface hydrology and phosphorus transport in the Kissimmee River Basin, Florida. Ec ological Engineering. 5: 301330. Capece, J. C. 1994. Hydrology and contaminant transport on flatwoods watershed. Ph.D. dissertation. University of Florida, Gainesville, Florida. Capece, J. C., K. L. Campbell, P. J. Bohlen, D. A. Graetz, and K. M. Portier. 2007. Soil phosphorus, cattle stocking rates, and water quality in subtropical pastures in Florida, USA. Rangeland Ecology and Management. 60: 19 30. Carignan, R., F. Rapin, and A. Tessier. 1985. Sediment porewater sampling for metal analysis: A comparisi on of techniques. Geochim. Cosmochim. Acta. 49: 24932497. Carleton, J. N. 2002. Damkhler number distributions and constituent removal in treatment wetlands. Ecological Engineering. 19: 233248. Chen, J. S., R. S. Mansell, P. Nkedi -Kizza, and B. A. Burgao 1996. Phosphorus transport during transient, unsaturated water flow in an acid sandy soil. Soil Science Society of America Journal. 60: 4248. Chrysostome, M., Nair, V. D., Harris, W. G., and D. R. Rhue 2007. Laboratory validation of soil phosphorus storage capacity predictions for use in risk assessment. Soil Science Society of America Journal. 71: 15641569. Corstanje, R., and K. R. Reddy. 2004. Response to biogeochemical indicators to a drawdown and subsequent reflood. Journal of Environmental Quality. 33: 23572366. Daly, K., D. Jeffrey, and H. Tunney. 2001. The effect of soil type on phosphorus sorption capacity and desorption dynamics in Irish grassland soils. Soil Use and Mangement. 17: 1220. Daniels, R. B., and E. E. Gamble. 1967. The edge effect in some Ultisols in the North Carolina Coastal Plain. Geoderma. 1: 117 124.
111 Davis, S. M. 1991. Growth, decomposition, and nutrient retention of Cladium jamaicense Crantz and typha domingensis Pers. i n the Florida Everglades. Aquatic Botany. 40: 203224. D ean, W. E. 1974. Determination of carbonate and organic matter in calcareous sediments and sedimentary rocks by loss on ignition: comparison with other methods. Journal of Sedimentary Petrology. 44: 242248. DeBusk, W. F., and K. R. Reddy. 2003 Nutrient and hydrology effects on soil respiration in a northern Everglades Marsh. Journal of Environmental Quality. 32: 702710. De Groot, J. C. and C. Van Wijck. 1993. The impact of desiccation of a freshwater marsh (Graines Nord, Camargue, France ) on sedi ment -wat er -vegetation interactions, Part 1 Sediment Chemistry. Hydrobiologia. 252: 8394. Dreccer, M. F., and R. S Lavado. 1993. Influence of cattle trampling on preferential flow paths in alkaline soils. Soil Use and Management. 9: 143 148. Dolan, T. J., S. E. Ba yley, J. Zoltek, and A. J. Hermann. 1981. Phosphorus dynamics of a Florida freshwater marsh receiving treated wastewater. Journal of Applied Ecology. 18: 205 219. Dougherty, W. J., N. K. Flemming, J. W. Cox, and D. J. Chittleborough. 2004. Phosphorus trans fer in surface runoff from intensive pasture systems at various scales: A review. Journal of Environmental Quality. 33: 19731988. Dunne, E. J., K. R. Reddy, and M. W. Clark. 2006. Phosphorus release and retention by soils of natural wetlands. Internationa l Journal of Environment and Pollution. 28: 496516. Dunne, E. J., J. Smith, D. B. Perkins, M. W. Clark, J. W. Jawitz, and K. R. Reddy. 2007 a Phosphorus storages in historically isolated wetland ecosystems and surrounding pasture uplands. Ecological Engineering. 31: 1628. Dunne, E. J., K. A. McKee, M. W. Clark, and K. R. Reddy. 2007b. Phosphorus in agricultural ditch soil and potential implications for water quality. Journal of Soil and Water Conservation. 62:244252. El Dewiny, C. Y., K. S. Moursy, and H I. El -Aila. 2006. Effect of organic matter on the release and availability of phosphorus and their effects on spinach and radish plants. Research Journal of Agriculture and Biological Sciences. 2: 103108. Emerson, S., and G. Widmer. 1978. Early diagenes is in anaerobic lake sediments 2. Thermodynamic and kinetic factors controlling the formation of iron phosphate. Geochim Cosmochim Acta. 42: 13071316. Fetter, C. W. 1994. Applied Hydrogeology. Chapter 7: Ground water flow to wells. Prentice Hall Inc. New Jersey, N.J. Fisher, M. M., and K. R. Reddy. 2001. Phosphorus flux from wetland soils affected by longterm nutrient loading. Journal of Environmental Quality. 30: 261271.
112 Flaig, E. G., and K. E. Havens. 1995. Historical trends in the Lake Okeechobee ecos ystem. I. Land use and nutrient loading. Arch. Hydrobiol. Suppl. Monogr. 107: 124. Flaig, E. G., and K. R. Reddy. 1995. Fate of phosphorus in the Lake Okeechobee basin, Florida, USA: overview and recommendations. Ecological Engineering. 5: 127143. Froeli ch, P. N. 1988. Kinetic control of dissolved phosphate in natural rivers and estuaries: A primer on the phosphate buffer mechanism. Limnology and Oceanography. 33:649668. Gathumbi, S. M., P. J. Bohlen, and D. A. Graetz. 2005. Nutrient enrichment of wetland vegetation and sediments in subtropical pastures. Soil Science Society of America Journal. 69: 539548. Gee, G. W., and J. W. Bauder. 1986. Particle -size analysis. p. 383 411. In: A. Klute et al. (ed.) Methods of Soil Analysis Part 1, Physical and Miner alogical Methods (second ed.), ASA, Inc., Madison, WI. Goh, T. B., and P. M. Huang. 1984. Formation of hydroxyl -Al -montmorillonite complexes as influenced by citric acid. Canadian Journal of Soil Science. 64:411421. Golterman, H. L. 1995. The role of the ironhydroxide phosphate -sulphide system in the phosphate exchange between sediments and overlying water. Hydrobilogia. 297: 43 54. Graetz, D. A., and V. D. Nair. 1995. Fate of phosphorus in Florida Spodosols contaminated with cattle manure. Ecological Engi neering. 5: 163181. Graham, S. A., C. B. Craft, P. V. McCormick, and A. Aldous. 2005. Forms and accumulation of soil P in natural and recently restored peatlands upper Klamath Lake, Oregon, USA. Wetlands. 25: 594606. Haan, C. T. 1995. Fate and transport of phosphorus in the Lake Okeechobee Basin, Florida. Ecological Engineering. 5: 331339. Haggard, B. E., P. A. Moore, and P. B. DeLaune. 2005. Phosphorus flux from bottom sediments in lake Eucha, Oklohoma. Journal of Environmental Quality. 34: 724728. Hk anson, L. 2004. Internal loading: A new solution to an old problem in aquatic sciences. Lakes and Reservoirs: Research and Management. 9 : 3 23. Harper, M. P., W, Davison, and W. Tych. 1997. Temporal, spatial, and resolution constraints of in situ sampling devices using diffusional equilibration: Dialysis and DET. Environmental Science and Technology. 31: 31103119. Harris, W. G., R. D. Rhue, G. Kidder, R. C. Littell, and R. B. Brown. 1996. P retention as related to morphology of sandy coastal plain soils. S oil Science Society of America Journal. 60: 15131521.
113 Harris, W. G., and G. N. White. 2008. X ray diffraction techniques for soil mineral identification. p. 81116. In: Barbarick, Roberts, and Logsdon (ed.) Methods of Soil Analysis, Part 5 Mineralogical Methods. Soil Science Society of America, Inc., Madison, WI. Havens, K. E., E. G. Flaig, R. T. James, S. Lostal, and D. Muszick. 1996. Results of a program to control phosphorus discharges from dairy operations in south -central Florida, USA. Environmental Management. 21: 585593. Hayes, W. A., and M. J. Vepraskas 2000. Morphological changes in soils produced when hydrology is altered by ditching. Soil Science Society of America Journal. 64: 1893 1904. Hesslein, R. M. 1976. An in-situ sampler for close inte rval pore water studies. Limnology and Oceanography. 21: 912914. Heylar, K. R., D. N. Munns, and R. G. Barau. 1976. Adsorption of phosphate by gibbsite. I. Effects of neutral chloride salts of calcium, magnesium, sodium, and potassium. Journal of Soil Sci ence. 27: 307314. Hiscock, J. G., C. S. Thourot, and J. Zhang. 2003. Phosphorus budget land use relationships for the northern Lake Okeechobee watershed, Florida. Ecological Engineering. 21: 6374. Hope, G. D., and J. K. Syers. 1976. Effects of solution :soil ratio on phosphate sorption by soils. Journal of Soil Science. 27: 301306. Johns, G. M., and K. ODell. 2004. Benefit -cost analysis to develop the Lake Okeechobee Protection Plan. Florida Water Resources Journal. June: 34 38. Justin, S.W., and W. Ar mstrong. 1987. The Anatomical Characteristics of Roots and Plant Response to Soil Flooding. The New Phytologist. 106: 465495. Kadlec, R. H. 2000. The inadequacy of first -order treatment wetland models. Ecological Engineering. 15: 105119. Kadlec R. H., a nd S. D. Wallace. 2008. Treatment Wetlands. (2nd Ed.). CRC Press, Boca Raton, FL. Kang, J., D. Hesterberg, and D. L. Osmond. 2009. Soil organic matter effects on phosphorus sorption: A path analysis. Soil Science Society of America Journal. 73: 360-366. Kl einman, P. J. A., B. A. Needleman, A. N. Sharpley, and R. W. McDowell. 2003. Using soil phosphorus profile data to assess phosphorus leaching potential in manured soils. Soil Science Society of America Journal. 67: 215224. Klemmedson, J. O., and A. R. Tie demann. 1995. Effects of nutrient stress. p. 414439. In: Bedunah and Sosebee (ed.) Wildland Plant: Physiological Ecology and Developmental Morphology. Society of Range Management, Denver, CO.
114 Klotz, R. L., and S. A. Linn. 2001. Influence of factors associ ated with water level drawdown on phosphorus release from sediments. Lake Reservoir Management. 17: 48 54. Koski -Vahala, J., and H. Hartikainen. 2001. Assessment of risk of phosphorus loading due to resuspended sediment. Journal of Environmental Quality. 30: 960966. Lewis, D. L., K. J. Liudahl, C. V. Noble, and L. J. Carter. 2001. Soil Survey of Okeechobee County, Florida, USDA/NRCS in cooperation with the UF IFAS, Agricultural and Experimental Stations and Soil Science Department, and Florida Department o f Agriculture and Consumer Services. Li, Y. H., and S. Gregory. 1974. Diffusion of ions in seawater and deep-sea sediments: Geochim. Cosmochim Acta. 38: 703714. Line, D. E., W. A. Harman, and G. D. Jennings. 1998. Comparing sampling schemes for monitorin g pollutant export from a dairy pasture. Journal of the American Water Resources Association. 34: 12651273. Lovley, D. R., and E. P. Phillips. 1986. Availability of ferric iron in microbial reduction in bottom sediments of the freshwater tidal Potomac Riv er. Applied Environmental Microbiology. 52: 751757. Maguire, R. O., and J. T., Sims. 2002. Soil testing to predict phosphorus leaching. Journal of Environmental Quality. 31: 16011609. Malecki, L. M., J. R. White, and K. R. Reddy. 2004. Nitrogen and phosphorus flux rates from sediment in the Lower St. Johns River Estuary. Journal of Environmental Quality. 33: 15451555. Mansell, R. S., S. A. Bloom, and P. Nkedi -Kizza. 1995. Phosphorus transport in spodosols impacted by dairy waste. Ecological Engineering. 5: 281299. Martin, B. J., M. K. Hartl, R. D. Corbett, P. W. Swarzenski, and E. J. Cable. 2003. A multi -level pore -water sampler for permeable sediments. Journal of Sedimentary Research. 73(1): 128132. McClymont, E. L., A. M. Garcia, and A. R. Mele. 2007. Benefits of freeze -drying sediments for the analysis of total chlorins and alkenone concentrations in marine sediments. Organic Geochemistry. 38: 10021007. McKee, K. A. 2005. Predicting soil phosphorus storages in historically isolated wetlands within t he Lake Okeechobee priority basins. M.S. Thesis: University of Florida, Gainesville. McLaughlin, J. R., J. C. Ryden, and J. K. Syers. 1981. Sorption of inorganic phosphate by iron and aluminum containing components. Journal of Soil Science. 32: 365377.
115 Mi chael, H. A., J. S. Lubetsky, and C. F. Harvey. 2003. Characterizing submarine groundwater discharge: A seepage meter study in Waquoit Bay, Massachusetts. AGU, Geophysical Research Letters. 30: 12971298. Moore, P. A., and K. R. Reddy. 1994. Role of Eh and pH on phosphorus geochemistry in sediments of Lake Okeechobee, Florida. Journal of Environmental Quality. 23: 955964. Moore, P. A., K. R. Reddy, and M. M. Fisher. 1998. Phosphorus flux between sediment and overlying water in Lake Okeechobee, Florida: Spa tial and temporal variations. Journal of Environmental Quality. 27: 14281439. Mukherjee, A., V. D. Nair, M. W. Clark, and K. R. Reddy. 2009. Development of indices to predict phosphorus release from wetland soils. Journal of Environmental Quality. 38:878886. Mulholland, B., and M. A. Fullen. 1991. Cattle trampling and soil compaction on loamy sands. Soil Use and Management. 7: 189193. Nair, P. S., T. J. Logan, A. N. Sharpley, L. E. Sommers, M A. Tabatabai, and T. L. Yuan. 1984. Interlaboratory comparison of a standard phosphorus adsorption procedure. Journal of Environmental Quality. 13: 591595. Nair, V. D., D. A. Graetz, and K. R. Reddy. 1998. Dairy manure influences on phosphorus retention capacity of Spodosols. Journal of Environmental Quality. 27: 522 -527. Nair, V. D., K. M. Portier, D. A., Graetz, and M. L. Walker. 2004. An environmental threshold for degree of phosphorus saturation in sandy soils. Journal of Environmental Quality. 33: 107113. Nair, V. D., and W. G. Harris. 2004. A capacity factor as an alternative to soil test phosphorus in phosphorus risk assessment. N.Z. Journal of Agricultural Resources. 47: 491497. Nelson, N. S. 1987. An acid-persulfate digestion procedure for determination of phosphorus in sediments. Commun. in Soil Sci. Plant Anal. 18: 35969. Olila, O. G., K. R. Reddy, and D. L. Stites. Influence of draining on soil phosphorus forms and distribution in a constructed wetland. Ecological Engineering. 9: 157169. Pandit, A., and C. C. El -Khazen. 1990. Ground water seepage into the Indian River lagoon at Port St. Lucie. Florida Scientist. 53: 169179. Pant, H. K., K. R., Reddy, and E. Lemon. 2001. Phosphorus retention capacity of root bed media of subsurface flow constructed wetlands. Ecological Engineering. 17: 345355. Pant, H. K., V. D. Nair, K. R. Reddy, D. A. Graetz, and R. R. Villapando. 2002. Influence of flooding on phosphorus mobility in manure -impacted soils. Journal of Environmental Quality. 31: 13991405.
116 Pant, H. K., and K. R. Reddy. 2003. Potential internal loading o f phosphorus in a wetland constructed in agricultural land. Water Research. 37: 965972. Pardue, J. H, and W. A. Jackson. 2001. Dialysis sampler evaluation of rhizosphere phytoremediation processes in the Poplar Grove at J -Field, Aberdeen Proving Ground, M aryland. Final Report. p 4 6. Pautler, M. C., and J. T. Sims. 2000. Relationships between soil test phosphorus, soluble phosphorus, and phosphorus saturation in Delaware soils. Soil Science Society of America Journal. 64: 765773. Perkins, D. 2007. Water a nd chemical budgets of depressional wetlands in the Lake Okeechobee basin, FL. Ph.D. dissertation. University of Florida. Gainesville, Florida. Pierzynski, G. M., R. W. McDowell, and J. T. Sims. 2005. Chemistry, cycling, and potential movement of inorganic phosphorus in soils. p. 53 86. In J. T. Sims and A. N. Sharpley (ed.). Phosphorus: Agriculture and the environment. Agron. Monogr. 46. ASA, CSSA, and SSSA, Madison, WI. Rama, and W. S. Moore. 1996. Using the radium quartet for evaluating groundwater input and water exchange in salt marshes. Geochim. Cosmochim. Acta. 60: 46454652. Rao, P. S. C., and J. M. Davidson. 1979. Adsorption and movement of selected pesticides at high concentrations in soils. Water Research 13: 375380. Rao, P. S. C., J. M. Davidso n, V. E. Berkheiser, L. T. Au, J. J. Street, W. B. Wheeler, and T.L. Yuan. 1980. Retention and transformation of selected pesticides and phosphorus in soil water systems: a cri tical review. USEPA, Washington, D.C., EPA/600/3 82/060. Reddy, K. R., Y. P. She ng, and B. L. Jones. 1995. Lake Okeechobee Phosphorus Dynamics Study, Volume 1, Summary, Report, SFWMD, West Palm Beach, Fl. Reddy, K. R. G. A. O Connor, and P. M. Gale. 1998a Phosphorus sorption capacities of wetl and soils and stream sediments i mpacted b y dairy e ffluent. Journal of Environmental Quality. 27: 438447. Reddy, K. R., Y. Wang, W. F. DeBusk, M. M. Fisher, and S. Newman. 1998 b Forms of soil phosphorus in selected hydrologic units of Florida Everglades. Soil Science Society of America Journal. 62: 11341147. Reddy, K. R., G. A. OConnor, and C. L. Schelske. 1999. Phosphorus Biogeochemistry in Subtropical Ecosystems. Lewis Publishers. Boca Raton, Florida. Rengasamy, P., R. S. Greene, and G. W. Ford. 1984. The role of clay fraction in the particle arrangement and stability of aggregates a review. Clay Resources. 3: 53 67.
117 Rhoades, E. D., L. F. Locke, H. M. Taylor, and McIlvain, E. H. 1964. Water intake on sandy range as affected by 20 years of differential cattle stocking rates. Journal of Range Management. 17: 185190. Rice, R. W., F. T. Izuno, and R. M. Garcia 2002. Phosphorus load reductions under best management practices for sugarcane cropping systems in the Everglades Agricultural Area. Agricultural Water Management. 56: 17 39. Richardson, C. J. 1985. Mechanisms controlling phosphorus retention capacity in freshwater wetlands. Science. 228: 14241427. Richardson, C. J., and P. Vaithiyanathan. 1995. Phosphorus sorption characteristics of Everglades soils along a eutrophication gradient. Soil Science Society of America Journal. 59: 17821788. Ryden, J. C., J. K. Syers, and J. R. McLaughlin. 1977. Effects of ionic strength on chemisorption and potential -determining sorption of phosphate by soils. Journal of Soil Science. 28: 6271. Sah, R. N., D S. Mikkelsen, and A. A. Hafez. 1989. Phosphorus behavior in flooded-drained soils. II Iron transformations and phosphorus sorption. Soil Science Society of America Journal. 53: 17231729. Sample, E. C., R. J., Soper, and G. J. Racz. 1980. Reactions of ph osphate fertilizers in soils. p. 263310. In F. E. Khasawneh et al. (ed.). The role of phosphorus in agriculture. ASA Madison, WI. Scholefield, D., P. M. Patto, and D. M. Hall. 1985. Laboratory research on the compressibility of four to p -soils from grassla nds. Soil and Tillage Research. 6: 1 16. Schoumans, O. F., and P. Groenendijk. 2000. Modeling soil phosphorus levels and phosphorus leaching from agriculture land in the Netherlands. Journal of Environmental Quality. 29: 111116. Sen, S., B. E. Haggard, I. Chaubey, K. R. Brye T. A. Costello, and M. D. Matlock 2007. Sediment phosphorus release at Beaver Reservoir, Northwest Arkansas, USA, 20022003: A preliminary investigation. Water Air Soil Pollution. 179: 67 77. Sharpley, A. N., L. H. Ahuja, M. Yamamoto and R.G. Menzel. 1981. The kinetics of phosphorus desorption from soil. Soil Science Society of America Journal. 45: 493 496. Sharpley, A. N., R. W. McDowell, J. L. Weld, and J. A. Kleinman. 2001. Assessing site vulnerability to phosphorus loss in an agr icultural basin. Journal of Environmental Quality. 30: 20262036. Sigua, G. C. 2003. Soils and environment of grazed pasture ecosystem in subtropical United States. Trends and Research Results. Journal of Soils and Sediments. 3: 258260.
118 Sigua, G. C., and S.W. Coleman 2006. Sustainable management of nutrients in forage based pasture soils: Effect of animal congregation sites. Journal of Soils and Sediments. 6: 249253. Sims, J. T., L. Bergstrom, B. T. Bowman, and O. Oenema. 2005. Nutrient management for in tensive animal agriculture: policies and practices for sustainability. Soil Use and Management. 21: 141151. Smith D. G. 1984. Vibracoring fluvial and deltaic sediments: Tips on improving penetration and recovery. Journal of Sedimentary Petroleum 54: 6606 63. Soil Survey Staff, 2007. NRCS -USDA. Web Soil Survey. Available online at http://websoilsurvey.nrcs.usda.gov/ accessed [03/15/2007]. South Florida water Management District (SFWMD), Florida Department of Environmental Protection (FDEP) and Florida Depar tment of Agriculture and Consumer Services (FDACS). 2004. Lake Okeechobee Protection Program: Lake Okeechobee Protection Plan, West Palm Beach, FL. Steinman, A. D., and B. H. Rosen. 2000. Lotic lentic linkages associated with Lake Okeechobee, Florida. Jou rnal of North American Benthological Society. 19: 733741. Sweerts, J. R., C. Kelly, J. Rudd, R. Hasslein, and T. Cappenberg. 1991. Similarity of whole sediment diffusion coefficients in freshwater sediments of low and high porosity. Limnology and Oceanogr aphy. 36: 335342. Taboada, M. A., and R. S. Lavado. 1993. Influence of cattle trampling on soil porosity under alternate dry and ponded conditions. Soil Use and Management. 9: 139143. Tollne r, E. W., G. V. Calvert, and G. Langdale. 1990. Animal trampling effects on soil physical properties of two Southeastern U.S. ultisols. Agriculture, Ecosystems and Environment. 33: 7587. Traina, S. J., G, Sposito, Hesterberg, D., and U. Kafkafi. 1986. Effect of pH and organic acids on orthophosphate solubility in an a cidic, montmorillonitic soil. Soil Science Society of America Journal. 50: 4552. Tweel, A. W., and Bohlen, P.J. 2008. Influence of soft rush (Juncus effucus) on phosphorus flux in grazed seasonal wetlands. Ecological Engineering. 33: 242 251. Villapando, R. R., and D. A. Graetz. 2001 a. Water table effects on phosphorus reactivity and mobility in a dairy manure -impacted spodosol. Ecological Engineering. 18: 7789. Villapando, R. R., and D. A. Graetz. 2001 b. Phosphorus sorption and desorption properties of the spodic horizon from selected Florida spodosols. Soil Science Society of America Journal. 65: 331339.
119 Watts, C. J. 2000. Seasonal phosphorus release from exposed, re inundated littoral sediments of two Australian reservoirs. Hydrobiologia. 431: 2739. White R. E., and A. W. Taylor. 1977. Effect of pH on phosphate adsorption and isotopic exchange in acid soils at low and high additions of soluble phosphate. Journal of Soil Science. 28: 48 61. Wu, M. Y., E. H. Franz, and S. Chen. 2001. Oxygen fluxes and ammonia removal efficiencies in constructed treatment wetlands. Water Environment Research. 73: 661666. Zhang, J. J. G. Hiscock, A. B. Bottcher, B. M. Jacobson, and P. J. Bohlen. 2006. Modelling phosphorus load reductions of agricultural water management practices on a beef cattle ranch. American Society of Agricultural and Biological Engineers. Meeting Presentation, Portland, OR. Zhou, M., R. D. Rhue, and W. G. Harris. 1997. Phosphorus sorption characteristics of Bh and Bt horizons from sandy coastal plai n soils. Soil Science Society of America Journal. 61: 13641369.
120 BIOGRAPHICAL SKETCH Dr. Jehangir H. Bhadha started his education in 1980 when he joined St. Peters High School, Bombay, India He studied there for twel ve years, from kindergarten to tent h grade. In 1992 he joined St. Xaviers College, a part of the University of Bombay and completed the Junior College (high s chool equivalent) level in 1994. He then went on to the Degree College in St. Xaviers where he graduated in May 1997 having earned a Bachelor of Science (Honors) degree majoring in geology and a minor in gemology. In 1997 he began his masters degree program from St. Xaviers College and graduated in June 1999 with a Master of Science (Distinction) degree, maj oring in geology. In Fa ll 2000 he joined the Department of Geological Sciences at the University of Florida (G o Gators!). He graduated in Spring 2003 with a Master of Science degree, major in geology and a minor in environmental enginee ring sciences. Apart from this he also rece ive d a certificate in hydrologic sciences as part of the Hydrologic S ciences Academic Cluster from the University of Florida. In Spring 2005 he joined the Soil and Wat er Science Department at U.F in pursuit of a doctorate degree; and successfully graduate d in Summer 2009.