FATE AND TRANSPORT OF PHOSPHORUS IN DEVELOPING LANDSCAPES UNDER LONG TERM EFFLUENT IRRIGATION By GRANT B. WEINKAM 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 2015
Â© 2015 Grant B. Weinkam
3 ACKNOWLEDGMENTS Thanks to the people, places, and things that made this possible. T o Jerry , Marcia , Greg , and Todd for th eir efforts to making me more competent, capable, and caring. To the many University of Florida professors who se assistance and discussions helped fuel and incentivize research progression : Mark T. Brown, Mark Clark, Matt Cohen, David Kaplan, Willie Harris , Wendy Graham, Tom Frazer, and Mark Brenner . To Florida water management and department of environmental protection agents Eric h Marzolf, D. Albrey Arrington, Shanin Speas Frost , and experimental design consultant Uncle Edward Sombati. To w astewater t reatment plant personnel at U niversity of F lorida , Tallahassee, Lake Ci ty, Leesburg, Alachua, and Live Oak. Last, but not least , friends and fellow members of the 2011 Water Institute Graduate Fellows cohort : Thomas Elliott Arnold, Charlie Nealis, Wesley Henson, Joelle Laing, and .
4 TABLE OF CONTENTS page ACKNO WLEDGMENTS ................................ ................................ ................................ .. 3 LIST OF TABLES ................................ ................................ ................................ ............ 6 LIST OF FIGURES ................................ ................................ ................................ .......... 7 LIST OF ABBREVIATIONS ................................ ................................ ............................. 8 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 Plan of Study ................................ ................................ ................................ .......... 14 Review of the Literature ................................ ................................ .......................... 16 Objectives ................................ ................................ ................................ ............... 19 2 E FFECTS OF LONG TERM EFFLUENT IRRIGATION ON SOIL CHEMISTRY AND PHOSPHORUS LABILITY IN FLORIDA SPRAYFIELDS ............................... 21 Chapter Abstract ................................ ................................ ................................ ..... 21 Introductory Remarks ................................ ................................ .............................. 22 Methods ................................ ................................ ................................ .................. 27 Description of Study Sites ................................ ................................ ................ 27 Soil Sampling and Analysis ................................ ................................ .............. 28 Data Analysis ................................ ................................ ................................ ... 2 9 Results and Discussion ................................ ................................ ........................... 30 Effluent Irrigated and Control Soil Concentrations ................................ ............ 30 Differences in Effluent Irrigated and Control Soils with Depth .......................... 32 ................................ ............. 35 Concl usions ................................ ................................ ................................ ............ 40 3 ASSSESSMENT OF SOIL PHOSPHORUS LOSS INDICATORS AT PREDICTING DISSOLVED ORTHOPHOSPHATE LEACHING FROM FLORIDA EFFLUENT SP RAYFIELDS ................................ ................................ ... 42 Chapter Abstract ................................ ................................ ................................ ..... 42 Introductory Remarks ................................ ................................ .............................. 43 Methods ................................ ................................ ................................ .................. 47 Description of Study Sites ................................ ................................ ................ 48 Soil Sampling and Analysis ................................ ................................ .............. 48 Leaching Experiment Methodology ................................ ................................ .. 49 Data Analysis ................................ ................................ ................................ ... 50
5 Results and Discussion ................................ ................................ ........................... 51 M3P vs. WEP ................................ ................................ ................................ ... 51 P SR vs. WEP ................................ ................................ ................................ ... 53 M3P, WEP, and PSR vs. Leached Orthophosphate ................................ ......... 55 SPSC Accu racy at Predicting Leached Orthophosphate ................................ .. 58 Conclusions ................................ ................................ ................................ ............ 63 4 ELEVATED PHOSPHORUS LEACHING ASSOCIATED WITH LONG TERM EFFLUENT IRRIGATION ................................ ................................ ....................... 64 Chapter Abstract ................................ ................................ ................................ ..... 64 Introductory Remarks ................................ ................................ .............................. 64 Methods ................................ ................................ ................................ .................. 69 Description of Study Sites ................................ ................................ ................ 69 Soil Sampling and Analysis ................................ ................................ .............. 70 Leaching Experiment Methodology ................................ ................................ .. 71 Data Analysis ................................ ................................ ................................ ... 74 Field Scale Mass Balance Calculations ................................ ............................ 75 Results and Discussion ................................ ................................ ........................... 76 Soil Bound Nutrients ................................ ................................ ......................... 77 Leached Orthophosphate Concentrat ions ................................ ........................ 79 Uptake of Effluent Orthophosphate Percolating Through Soils ........................ 81 Spatiotemporal Phosphorus Leaching ................................ .............................. 82 Field Scale Phosphorus Mass Balance ................................ ............................ 83 Conclusions ................................ ................................ ................................ ............ 87 5 SUMMARY AND CONCLUSIONS ................................ ................................ .......... 89 LIST OF REFERENCES ................................ ................................ ............................... 97 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 113
6 LIST OF TABLES Table page 2 1 Sprayfield reference numbers, area, loading rates, and sampling information. .. 28 2 2 Analytical soil chemistry methods. ................................ ................................ ...... 29 3 1 Sprayfield reference numbers, area, loading rates, and sampling information. .. 48 3 2 Average SPSC with depth for effluent irrigated and control soils. ....................... 59 4 1 Expe rimental steps to generate leachate samples. ................................ ............ 72 4 2 Total differences in M3P and WEP (kg ha 1 ) between effluent irrigated soils and paired control soils. ................................ ................................ ...................... 77 4 3 Percentage of total M3P and WEP accumulated in surface horizons. ................ 79 4 4 Calculated p hosphorus mass leaching from effluent irrigated soils. ................... 83 4 5 Effluent sprayfield P mass balance over entire application timespan. ................ 84
7 LIST OF FIGURES Figure page 2 1 Sprayfield reference numbers and sampling locations ................................ ....... 27 2 2 Average and range of parameter concentrations in all effluent irrigated and non effluent irrigated soils. ................................ ................................ ................. 32 2 3 Differences between mean values of effluent irrigated and control soil parameters on a mass per area basis (kg ha 1 ) with depth. ................................ 35 2 4 Time series showing the effect of 31 years of effluent application on pH. .......... 36 2 5 Time series showing the effect of 31 years of effluent application on extractable P concentrations. ................................ ................................ ............. 38 3 1 Relationship between M3P and WEP for individual samples of effluent irrigated surface (0 0.3 m) and subsurface (0.3 2.0 m) soils. ............................. 53 3 2 Relationship between WEP and PSR for individual effluent irrigated soil samples below and above threshold PSR of 0.1. ................................ ............... 54 3 3 Relationship between WEP and PSR for individual control soil samples below and above threshold PSR of 0.1. ................................ ............................. 55 3 4 Correlation of M3P and leached orthophosphate in effluent irrigated soils. ........ 56 3 5 Correlation of WEP and leached orthophosphate in effluent irrigated soils. ....... 57 3 6 Relationship of PSR and leached orthophosphate in effluent irrigated soils. ...... 57 3 7 SPSC vs. leached orthophosphate for individual effluent and control soil columns. ................................ ................................ ................................ ............. 61 4 1 Orthophosphate leached from effluent irrigated and control soils for experiments A, B, and C. ................................ ................................ .................... 81 4 2 Percentage change of orthophosphate concentration after percolating through effluent irrigated and control soils. ................................ ......................... 82
8 LIST OF ABBREVIATIONS Al Aluminum Ca Calcium DI De Ionized FAC Florida Administrative Co de FDEP Florida Department of Environmental Protection Fe Iron ICP Inductively Coupled Plasma N Nitrogen NNC Numeric Nutrient Criteria NPDES National Pollution Discharge Elimination System NRC National Research Council M3P Mehlich 3 Phosphorus Mg Magnesium MGD Million Gallons per Day O rtho P Orthophosphate OM Organic Matter P Phosphorus PSR Phosphorus Saturation Ratio SPSC Soil Phosphorus Storage Capacity TKN Total Kjeldahl Nitrogen TN Total Nitrogen TP Total Phosphorus US United States USEPA United States Environmental Protection Agency
9 WEP Water Extractable Phosphorus WW Wastewater WWTP Wastewater Treatment Plant
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 FATE AND TRANSPORT OF PHOSPHORUS IN DEVELOPING LANDSCAPES UNDER LON G TERM EFFLUENT IRRIGATION By Grant B. Weinkam December 2015 Chair: Mark T. Brown Major: Environmental Engi neering Sciences In Florida the reuse of treated wastewater (WW) and improved surface water quality are closely tied because of the significant phosphorus (P) loads present in the large volume of land applied effluent . U nder long term WW irrigation soils can exhibit propert y changes that can influence P transport rates and potentially negatively affect receiving water s . This r esearch characterize d and quantified soil changes associated with more than 25 years of WW irrigation and how resultant alterations have, and can in the future, influence P lability over extended time frames . In order to make determinations on the significance of long term loading on soil chemistry and P leaching rates , soil samples from f ive WW sprayfields were collected up to three m eters in depth and analyzed alongside nearby non irrigated, control sites . Analysis of soil s revealed that , when compared to controls, pH and concentrations of P, Ca, N were higher, and Al was lower. The influence of t he effluent loading to chemical proper ties were predicted and confirmed to influence P accumulation and transport in the system. C oncentrations of total, water extractable, and Mehlich 3 P at all depths sampled were higher than controls. Increases in P lability and transportability were valida ted through column experimentation where leached P rates were 8 times higher from effluent irrigated soils
11 than nearby non irrigated soils. Results highlight the significant reduction in the long term WW receiving soils ability to sequester further applied P loads , and may have the potential to influence groundwater P concentrations in shallow systems. These results show a need for careful attention in the long term management of the environmental buffer s that are traditionally considered an inexhaustible l ong term P endpoint . F indings are useful for increasing the efficiency and accuracy of nutrient transport model s and best management practices used to control P movement in municipalities attempting to meet societal and environmental goals in landscapes di sposing of large , and increasing, volumes of wastewater .
12 CHAPTER 1 INTRODUCTION In the mid twentieth century decreasing surface water quality and an increasing demand on limited freshwater supplies led to the encouragement of alternative wastewater (WW) disposal techniques. Federal regulations and incentives began emphasizing this desire in the 1970 s , requir ing that land application as a disposal method be considered before grants were approved for wastewater treatment plant (WWTP) construction (Crites et al., 1977; Wright & Rovey, 1979) . This concept of reusing treated WW to meet demands for non potable water and decreasing direct discharges to surface waters, w as heavily embraced by Florida; currently leading the nation reusing over 700 million gallons per day (MGD) (2.6 million m3 per day) , of which greater than 60% is used for landscape irrigation (FDEP, 2014) . While the benef its of this disposal method, including reduced demand on groundwater reserves and decreased surface water pollution, are valuable there are also concerns related to the potential effects of long term loading of contaminants asso ciated with WW effluents. Re search suggests that under perpetual, long term applications of phosphorus (P), receiving soil systems can reach a state at which they are no longer able to assimilate further loading (Elliott & Jaiswal, 2011; Kim et al. , 2007; Lin et al., 2006) potentially resulting in landscapes that are actively leaching and eroding P rich particulate matter to receiving hydrologic systems . At the current rate of applying over 550 MGD (2.1 m illion m3 per day) of effluent to land surfaces, an estimated 1.5 million kg of P are reintroduced to the landscape yearly (assuming ef fluent concentration of 2 mg P L 1 ), compared to only 22 ,000 kg that would be applied if
13 irrigated with ground water (at 0.03 mg P L 1 ). This load will continue to increase, and reach 2.3 million kg of P of re using 1 billion gallons per day (3.8 million m3 per day) will be achieved within the near future (FDEP, 2012a) . e water nutrient concentration targets (FDEP, 2012b) this research focus ed on P loading associated with reclaimed water irrigation. A n accurate understan ding of how long term WW effluent loading may influence the achievement of water quality goals is warranted as the proposed numeric nutrient criteria (NNC) legislation (FAC 62 302.400) has recently established target concentrations of nitrogen ( N ) and phos phorus ( P ) in many state waters . As alternative watershed management plans and upgrades to municipal treatment systems to meet these standards are estimated to cost between $3.1 $8.4 billion dollars (Florida Association of Counties, 2015) , but make no specific mention of how reclaimed water will be a ffected and/or exempt (Arrington & Melton, 2010) , further research is needed for sustainable and cost efficient WW reapplication that helps achieve surface water quality goals . Due to the limited research and many uncertainties regarding the long term fate and transport of P that result from land applying reclaimed water in different locations a bette r understanding of potential ground and surface water implications is warranted. This research seeks to quantify the long term m agnitude of land applied effluent P and the assimilative capacity of Florida soils leading to a better understanding of risks , b enefits , and alternative management practices that can allow for land applying reclaimed water sustainably over the long run. Findings address research needs for a
14 more complete understanding of systems under effluent irrigation (Jacangelo et al., 2012; N RC , 2012) , and provide the information necessary to improve upon strategies targeted at reducing non point P loss from municipal WW irrigated landscapes. Plan of Study Although there are a number of different constituents present in a typical WW effluent this study focus ed on P for two primary reasons . One, i n much of Florida P is the limiting nutrient in surface waters, and is considered by some to be the most critical element i n controlling undesirable eutrophication events in freshwater systems (Conley et al. , 2009; Schindler, 1977; Sharpley et al., 1994; Smith & Schindler, 2009) , since some algae species are able to sequester nitrogen from the atmosphere . Secondly, while P is a nutrient that may complex with other minerals and undergo sequestration within so ils and waters, it does not have an open cycle like N thus making mass balance and accuracy of long term fate determination easier and potentially less uncertain. I hyp othesize that soil properties, in long term municipal WW receiving landscapes, have been altered in manners that have affected key P cycling processes (e.g. solubility, retention, leaching and lateral flow of P) leading to increased P lability . If effluent receiving soils are identified as increasingly limited in their ability to sequester fu rther P (according to specific saturation indexes ), elevated P leaching is hypothesize d to occur . If soils are actively leaching elevated P , how and why is this occurring, and what are effective management stra tegies that can be applied to reduce P transpo rt to protect sensitive groundwater and surface water systems in the future? W hat suggestions can be made to improve the efficiency and accuracy of watershed loading models and
15 management practices in a state disposing of a continuously increasing volume o f WW through land application ? This investigation, looking at the fate and transport of long term effluent applied P, occurred in three primary steps: 1. Soil sampling and analysis Soil samples collected from Florida locations where WW had been continuously applied for greater than 25 years are used to provide insights into responses and shifts of specific elements that influence P transport . The goal is to determine the extent of P accumulation and changes in P lability that have occurred deep in to receiving soil s (up to three meters) by comparing effluent irrigated soils with simila r , nearby , non irrigated soils . Through the analysis of multiple forms of P (total P (TP), Mehlich 3 P (M3P), and water extractable P (WEP)) , and other chemical constituents that affect P transport ( i.e., pH, Fe, Al, Ca, Mg), conclusions can be drawn on how receiving soils have responded and what are potential long term implications for other areas of the state under effluent irrigation . 2. Analysis of P leaching risk assessment tool s In order to more accurately determine the applicability of commonly applied P leaching risk soil metrics in long term effluent irrigated landscapes soil analysis and leaching experimentation was conducted. Soil analytes (M3P, WEP) and formulations (PSR, SPSC) were correlated with leached orthophosphate values generated through co lumn experimentation to determine what measures (including M3P, WEP, and iron and aluminum based soil saturation indices P saturation ratio (PSR) and soil P storage capacity (SPSC ) ) are most appropriate and predictive. Th is research addresses the
16 following questions : In long term effluent irrigated soils, which commonly used P leaching risk indicators (M3P, WEP) and formulations (PSR) most accurately predict P leaching risk? Does t he SPSC valuation accurately predict the fate of further applied P? R ecommendations and corrective factors are suggested for assessed metrics to help improve accuracy and applicability of the techniques. 3. Leaching e xperimentation Soil s at effluent irrigated , and non irrigated, sites were collected to depths of groundwater in many locations around Florida to determine: Is elevated P leaching occurring in Florida soils under long term effluent irrigation? S oil cores , up to two meters, were reconstructed for co lumn leaching experimentation to compare differences and similarities between sampled sites . Through compari son of expected dissolved orthophosphate leaching outputs ( from the analyzed soil sample concentrations) to experimentally generated value s , determinations can be made on how systems are behaving differently than other heavily P loaded systems ( i.e., m anure, biosolid, and granular fertilizer loaded landscapes ) . R esults from the soil analysis and experimental leaching r uns provide information necessary for develop ing more accurate P loss estimates in landscapes under long term effluent irrigation . Increased system understanding and accuracy of landscape loading models can improve effectiveness of best management practices targeted at reducing P transport in similar environments . Review of the Literature P revious researchers have noted that the long term impact of applying WW effluent to landscapes is still relatively uncertain and of theoretical concern for agronomic and environmenta l systems (O'Connor, Elliott, & Bastian, 2008; Rosabal et al., 200 7; Scott, Faruqui, & Raschid Sally, 2004; Xu et al. , 2010) .
17 Most soil systems not receiving additional P loads for long periods of time are below their P saturation capacity and have the capability to sequester upwards of 90% of applied P within the upper 0.1 5 1.0 meter of the soil layer. U nder continuous, long term effluent applications however soil layers can become saturated and leach undesirable concentrations of P and other contaminants (Andres & Sims, 2013; He et al. , 1999; Lin & Banin, 2005; Moura et al., 2011; Walter, 1996) . Examples display ing the finite ability of a given soil to assimilate additional P loads were seen in Tallahassee, Florida where extractable P in upper soil layer s increased from 7.1 to 72.8 mg kg 1 over an 8 year period of irrigation with reclaimed water (Allhands & Overman, 1995) , and was shown by Lin et al. (2006) who des cribes soils where a P accumulation rate of ~28 mg kg 1 soil per year dropped to ~2.2 mg kg 1 soil per year from early operation to after 20 25 years o f effluent irriga tion. This trend towards P saturation, where the total mass of P held in the soil achieves a maximum finite quantity , can occur in as few as 10 15 years of irrigation ( Elliott & Jaiswal, 2012 ; Moura et al. , 2011) . The actual fate of the applied P however w ill be heavily dependent on a number of factors, including: soil characteristics, effluent quality, loading rates, plant uptake, and management practices (Elliott & Jai swal, 2011; Lin et al., 2006; Maurer, 1994) . greatest potential for leaching P and other contaminants to ground and, eventually, surface waters within the basin (Kim et al., 2007) . The issue of soils leaching contaminants is of increased concern in Florida due to the high amount of sands , often inherently low in iron (Fe) and aluminum (Al) based minerals ( Nair & Graetz, 2004) , and strong groundwater surface water interconnectivity .
18 T his hydrologic system can allow for more P movement from the surface to groundwater flow path (Maurer, 1994; O'Connor et al., 2008; Perry, 1991) and increase losses of dissolved P, via leaching and lateral flow (Church et al. , 2010; Heckrath et al. , 1995; Maguire & Sims, 2002; Nelson, Parsons, & Mikkelsen, 2005) , and ultimately contribute to surface water impairment (Lin & Banin, 2005; Moura, 2009; Nair & Graetz, 2002) . Florida sandy soils in particular provide a substrate for rapid infiltration and comparatively few er surface binding sites than many other soil classes. T he P seques tration rate will be highly influenced by the relation of P relative to Al and Fe, termed the P saturation ratio (PSR ) (Nair et al. , 2004) . This PSR value can then be related to a calculated soil P storage capacity (SPS C ) value which can be used to quantitatively describe historic P loading in the area and estimate how much P mass a soil will be a ble to assimilate in the future (Nair & Harris, 2004) . The PSR and SPSC formulation s have been found to be accurate predictive measures of P loss risk for acidic soils, where P is more strong ly bound to Fe and Al compounds. In long term vegetated effluent receiving sites however, changes in soil chemistry may influence P lability and transporta bility and make the application and accuracy of these metrics uncertain . As the potential of soils to build up P over time has been known , the state has made re commendations for soil testing to limit or eliminate P fertilization in soils that rank n soil P concentration. Under irrigation scenarios using Floridan aquifer ground water (containing substantially lower dissolved P concentrations) t his was a practical approach for managing P enriched soils (Adams ki & German, 2004) . I n the current system , however, where secondarily treated effluent water that contains 1 15
19 mg TP L 1 (Jacangelo et al., 2012) , the ability to reduce the input of additional P to the system i s limited by the quantities applied in the irrigation water . This switch from irrigation with groundwater to reclaimed water can increase P load application by upwards of two orders of magnitude, and often results in landscape application of P in excess of typical v egetative nutrient requirements (Elliott & Jaiswal, 2011) . This research seeks to fill specific gaps in the sciences related to fate and transport of P in landscapes under long term (>25 years) irrigation w ith municipal WW effluent. Resulting outputs can be applied by agricultural, biological, and envi ronmental engineer s and municipalities in need of a better understand ing on the influences that WW disposal decisions can have on the lo ng term attainment of surface water quality goals. Objectives 1. Objective: Quantitatively determine the influence of > 25 years of effluent irrigation on soil constituents (pH, P, Al, Fe, Ca, Mg) that strongly influence P fate and transport. a. H ypothesis: Accum soil horizons above show significant accumulation in M3P a nd a , as calculated by PSR/SPSC valuations . 2. Objective: Determine if M3P, WEP, PSR, and the SPSC index are effective measures of predicting soil P l eaching risk following long term effluent irrigation. a. Hypothesis: M3P, WEP, and PSR have reduced predictive accuracy of P leaching rates in long term effluent irrigated soils. b. Hypothesis: If effluent irrigated soil SPSC is positive, leached P concentratio ns should be equivalent to unsaturated control soils. 3. Objective: Determine influence of long term effluent irrigation on point scale P leaching rates to generate landscape scale loss estimates and wastewater disposal management suggestions. a. Hypothesis: If soil is P saturated it will leach a greater concentration of orthophosphate than non saturated soils.
20 b. Hypothesis: A mendment of effluent irrigated surface soils with an a luminum substrate will lower the concentration of orthophosphate leached during efflue nt loading .
21 CHAPTER 2 EFFECTS OF L ONG T ERM E FFLUENT I RRIGATION ON S OIL C HEMISTRY AND P HOSPHORUS LABILITY IN FLORIDA SPRAYFIELDS Chapter Abstract Applications of large volumes of wastewater effluents can result in alterations to soil chemical properties i n land treatment system operations. I evaluated the effects of long term (28 38 years) wastewater application on chemical properties and phosphorus lability in five municipal effluent sprayfields in Florida . Soils receiving effluent were sampled to a dept h of 3 meters to quantify P accumulation and changes in soil properties that could increas e P lability. E leven soil chemi cal attributes related to wastewater loading and P transport were sampled : three phosphorus (P) species [ Mehlich 3 P (M3P) , water extra ctable P (WEP), total P], four m etals [Fe, Al, Ca, Mg] , two N species [nitrates nitrites (NOx), total Kjeldahl N (TKN)], organic matter (OM) content, and pH. E ffluent sprayfield soils differed significantly from nearby non irrigated soils for many of the a nalyzed constituents . Overall , effluent receiving soils to maximum depths had higher pH, WEP, M3P, TP, Ca, and NOx; lower Al; and a negligible difference i n OM content, Mg, Fe, and TKN. WEP differed the most , with concentrations ~70% higher in effluent irr igated soils than control soils, with accumulation identified below the root zone. Time series over 31 years, conducted with historical data in one field , showed clear upward trends in surface and subsurface soils for pH and Mehlich extractable P . I conclu de that the long term loading of soils with municipal effluents ha s altered soil chemistry properties in a manner that has increased soil P to a depth of three meters , with implications for P transport .
22 Introduct ory Remarks Wastewater (WW) can be applied t o landscapes to reduce direct surfac e water discharges of effluents that contain elevated nutrient concentrations (Feigin, Ravina, & Shalhevet, 1991) and provide a reliable source of irrigation water. Under long ter m application however physical, chemical, and biological characteristics of the receiving environment can be altered, influencing phosphorus (P) retention and transport . As reuse of WW is expected to increase nationally and globally (Bixio et al., 2008; Miller, 2006; USEPA, 2012) it is important to understand the long term implications on receiving soil and water systems. A lteration of soil properties depends on effluent constitue nt s , application practices, and characteristic s of the receivin g environment which can have both beneficial and adverse influences on the ability of the system to be maintained as an effective plant matrix and environmental buffer . Resulting soil changes can have positive benefits o f increas ed organic matter (OM) and nutrient concentrations (Mann et al. , 2012; Menzies, Skilton, & Guppy, 1999; N RC , 1996) that can improve soil fertility (Bahri, 1998; Kiziloglu et al., 2008; Westcot & Ayers, 1985) and crop yields (Krauss & Page, 1997) . N egative conseque nces can also result, such as permeability reduction, and accumulation of specific undesirable elements ( i.e., salts, metals , nutrients, emerging micro pollutants) to toxic and/or leach able concentrations (Bouwer & Idelovitch, 1987; Feigin et al., 1991; NRC , 1996; USEPA, 2004) . Many studies on the effects of WW to receiving soils are conducted only on s mall spatial and temporal scales ( i.e., 1 5 years) (Falkiner & Smith, 1997; Greenberg & Thomas, 1954; Rosabal et al., 2007) , with the long term effects on soils and crops receiving less attention (Kalavrouziotis et al., 2009) . Additionally, much of the research
23 at long term effluent receiving sites is at locations that dispose of large volumes of WW through rapid rate ap plication methods, justifying further review at slow rate, irrigation operations. To address research needs that ensure long term high quality performance of WW reuse systems ( NRC , 2012) , m y objective w as to identify significant chemical shifts and accumulation patterns of effluent receiving soils , and conclude how changes may influence P leachability . I hypothesize that water exchangeable P ( WEP) will increase with long term loading , poten tially influencing dissolved P leaching rates . As r eclaimed water is relatively high in inorganic P compared to nitrogen (N) (Thompson & Milbrandt, 2014) , P is typically applied to landscapes in excess of plant requirements (Elliott & Jaiswal, 2011) . Under this application s cenario the ability of the receiving soils to perpetually sequester further added P may compromised . In early estimates (USEPA, 1981) and confirmed field conditions (Elliott & Jaiswal, 2011; Lin & Banin, 2005) receiving surface soil s (0.3 m) can become P saturated in ~10 years, limiting the useful life of a heavily loaded soil profile with coarse textured soils, and shallow underdrains discharging to sensitive water bodies to 20 60 years (USEPA, 1977; USEPA, 2006) . While some landscapes have b een identified with soil properties and loading rates w here long term P removal capabilities from continuously applied WW may be maintained (Gerritse, 1996; Kardos & Hook, 1976) , the ability to sequester further loading without saturati ng other receiving environmental buffer s is uncertain . The shift from rapid physico chemical P sorption to minimal P uptake has been identified in (Hooda et al., 2000; Pautler &
24 Sims, 2000) . Below the change point, a large fraction of P will be strongly bound to soils, with an expected rapid increase in P release above the identified point ( Lin et al., 2006) soil characteristics, effluent constituent makeup, hydraulic loading rates, and management practices (Agyin Birikorang, O'Connor, & Brinton, 2008; Lin & Banin, 2005; Walter, 1996) . V arious processes that will directly and indirectly influence the uptake, sequestration, and transport of P in the receiving sys tem s include physical, chemical, and biological influences on solution surface reactions, sorption desorption, precipitation dissolution, chemical oxidation reduction, particle interactions, and biological activity and uptake (Fox, 2002; Medema & Stuyfzand, 2002; USEPA, 2004) . The most influential factor affecting fate and transport of P in sandy soil is the concentration of charged molecules able to surface sorb or precipitate dissolve d phosphates in the system. This P retention is primarily associated with adsorption onto surface coatings of iron (Fe), aluminum (Al), and manganese (Mn) oxides and hydroxides , interactions with clays, OM, and/or precipitation with calcium (Ca) and magnes ium (Mg) (Allen et al. , 2006; Lin & Banin, 2005; Mann et al., 2012 ; Parfitt, Atkinson, & Smart, 1975; Rajan, 1975; Ryden & Pratt, 1980; Yao & Millero, 1996) . Under acidic soil conditions phosphates will preferentially bind with Al and Fe, with Al typically being the primary factor on P fate and transport (Eckert & Watson, 1996; Kleinman & Sharpley, 2002) . With increasing soil alkalinity Ca and Mg can become increasingly influential on P sorption through the precipitation of Ca and Mg phosphate minerals (Hu et al. , 2005; Moura, 2009) , such as Ca3(PO4)2 and Ca5(PO4)3OH
25 (hydroxyapatite ) (Lin & Banin, 2005) . The strength and stability of P sorbed to soils can also be highly influenced by pH and ion loading, a s P bound to Ca and Mg are generally considered more water soluble, exchangeable , and weaker adsorption mechanism s than Fe Al bound P under natural, acidic conditions ( Cho, 1991; Isensee & Walsh, 1972; Nair et al., 1995) . S orption mechanisms typically result in tightly bound P that accumulate s in the top 15 cm of soils (Hayes, Mancino, & Pepper, 1990; Tesar, Knezek, & Hook, 1982) . In studies by Allhands & Overman (1995) and Menzies et al., (1999) , P accumulation in two different surface soils (0 15 cm) under high effluent P loading ( i.e., 300 kg P ha 1 per year ) increased from 7 to 73 mg P kg 1 and from 29 to 108 k g P ha 1 over 8 and 20 year s of application , respectively . These increases in surface soil P can create an environment where further applied P is no longer sequestered, and thus transported to lower soil horizons. This reduced ability to bind and ho ld further applied effluent P has been identified in the surface sands (15 cm) of rapid rate effluent disposal sites where P sequestration decreased from 28 mg P kg 1 in the first 3 years of operation to less than 2.3 mg P kg 1 after 20 25 years (Lin et al., 2006; Lin & Banin, 2005) . P transport to surface waters is typically predicted solely on sediment associated losses that occur during surface runoff events . G roundwater transport of P is impl icitly considered negligible . However, vertical and lateral subsurface flow of dissolved phosphates to aquifer systems can represent a major mechanism for off site P loss (Agyin Birikorang et al., 2008; Lin & Banin, 2005; Walter, 1996) . The impact of landscape water drainage on P loss depend on factors such as soil texture, OM content, and irrigation practice s (Sims, Simard, & Joern, 1998) . In Florida, the topographic and
26 pedological la ndscape results in an environment where percolation rates are high and contact time between solutes and soil surfaces can be minimal. Sandy soils are common throughout the state an d contain lower surface areas for reaction. While some parts of state contai n subsurface horizons hi gher in silts and clays , soil texture in many areas results in significantly lower P sequestration capabilities prior to encountering a saturation change point (He et al., 1999) . Solutes can be tra nsported during percolation occurrences coinciding with precipitation or irrigation events. In Florida where average precipitation is 100 150 cm yr 1 (40 60 in yr 1 ) (DeWiest & Livingston, 1999) , and intense st orms are common, groundwater loading associated with these events can be substantial. Most agricultural and landscape irrigation rates range between 1.3 7.6 cm per week , with many states specifying a maximum hydraulic loading rate of 5.1 cm per week (USEPA, 1977; USEPA, 1981) . While this effluent is required to receive at least secondary treatment and disinfe ction prior to land application (USEPA, 2004) , there are generally no P effluent standards imposed (Angelakis et al., 1999; Lazarova & Bahri, 2004; USEPA, 2004) , nor is there much current review of the performance of systems in long te rm operation. Through this research conclusions can be drawn on the long term effects of applying WW to Florida soils, and inform land users and municipalities on the limitations and management adjustments necessary to maintain the sustainability of the p ractice as a socially and environmentally responsible method of municipal WW disposal.
27 Methods Description of Study Sites Study sites were five agricultural sprayfields, located across northern and centr al Florida (Fig . 2 1) , where secondarily treated muni cipal WW effluent had been land applied through pivot irri gation for 28 38 years (Table 2 1). Prior to effluent application sites were used as low intensity pasture operations where a pplications of P and other amendments were assumed negligible. Effluent h ydraulic loading rates ranged from 2.5 5.0 cm/week ( 1 2 in/week ) , and P concentrations dropped from system inception in the 1970s and 1980s (~10 mg L 1 ) to present day ( 1 2 mg L 1 ). Each of the sprayfields was under cultivation, predominantly growing and h arvesting grasses. A t times in Field 1 crops of pearl millet, sorghum, sudangrass, kenaf, c orn, and soybean have also been planted (Overman et al., 2003). F igure 2 1. Sprayfield reference numbers and sampling locations (Google Earth, 2015).
28 Table 2 1. Sprayfield reference numbers, area, loading rates, and sampling information . Area Time of operation Sample depth range Avg sample depth No. of soil samples Field Municipality ( ha ) (year) (cm) (cm) E ffluent : C ontrol 1 Tallahassee 810 32 300 300 24 : 24 2 Lake City 140 28 100 165 135 11 : 24 3 Leesburg 130 34 150 300 275 23 : 24 4 Alachua 45 38 180 300 250 24 : 22 5 Live Oak 63 28 70 115 90 17 : 14 Soil S ampling and A nalysis At each sprayfield, six soil cores were taken to a maximum depth of thre e meters. When clay horizons were encountered , sample depth was limited to the top of the clay horizon (Table 2 1). S ix cores from each sprayfield , to be used for comparative analysis, included: three cores from areas receiving effluent irrigation, and thr ee cores from nearby (0.1 not receiving effluent. Control soils were considered representative of soil unaffected by effluent application or other major soil alterations or amendments. Locations for soil cores were randomly selected within the sprayfield areas. In Field 2, one sampling site was initially thought to be an effluent receiving site, but was later identified as a non effluent receiving site. Due to this sampling error , in Field 2 two effluent receiving sites and fou r control sites were collected. In Field 1 all effluent irrigated soils were randomly sampled from within a 56 ha (140 acre) irrigation pivot . Fie ld 1 data were used for comparative analysis between effluent irrigated and non irrigated soils, and also adde d to p revious samples collected at the same pivot (Overman et al., 2003) for generation of a time series analysis highlighting changes in pH and extractable P at the site over 31 years .
29 Soils were collected with a 7.62 cm diameter h and auger and sectioned i nto eight depth increments ((cm): 0 15, 15 30, 30 60, 60 90, 90 120, 120 150, 150 230, 230 300 ) . In total, fourteen cores from effluent irrigated locations and sixteen cores from non irrigated, control locations were collected and analyzed. S ub samples wer e taken from each core depth increment and analyzed for constituents listed in Table 2 2 . From each increment 500 1000 g of mixed soil was sieved ( 2 mm ) and oven dried at 85 Â°C for 2 4 hours. Dried soil was homogenized and 5 10 g were analyzed . Due to difficulties in accurately measuring s oil bulk de nsity for all soils were assumed at 1500 kg m 3 . Table 2 2. Analytical soil chemistry methods. Analyte Soil Extraction Method Analytical Method Analytical Instrument Ca Mehlich 3 EPA 200.7 Inductively Coupled Plasma Spectrophotometer (ICP) (Spectro Arc os) Mg Mehlich 3 EPA 200.7 ICP Fe Mehlich 3 EPA 200.7 ICP Al Mehlich 3 EPA 200.7 ICP P Mehlich 3 EPA 200.7 ICP Total P Digestion EPA 365.1 Continuous Flow Autoanalyzer (Alpkem Flow IV) Ortho P (WEP) Water EPA 365.1 AQ2 Discrete Analyzer ( Seal Analytical) NO x Water EPA 353.2 Continuous Flow Autoanalyzer (Alpkem Flow IV) TKN Digestion EPA 351.2 Continuous Flow Autoanalyzer. A2 Analyzer (Astoria Pacific International) pH Water EPA 150.1 Orion Benchtop pH Mete r OM Oven Dry Loss on ignition N/A Data A nalysis Data were analyzed to highlight field specific and overall differences between soil parameters in effluent irrigated and non irrigated samples. A summary table was generated to show general trends and maj or differences in the systems by taking the
30 average and range of samples , at all depths, at each sprayfield (Fig. 2 2) . In all other figures samples were separated by depth to highlight differences in surface and subsurface soils over time. Differences bet ween effluent irrigated and control soils were calculated by subtracting averaged effluent values from averaged control values at each field and pooled by depth increments . I compared effluent irrigated soils to non irrigated s oils using a two tailed T te st. Field values were calculated by averaging data from the three soil cores extracted at each field, under each condition. Differences in depth were calculated by averaging field and condition values for the given depth increment. If the calculated T valu e was greater than the T critical value calculated at 1% ( = 0.01) and 5% ( = 0.05) differences were considered statistically significant. The n ull hypothesis was that there is no difference between the parameters tested for effluent irrigated soils and non irrigated soils. Results and Discussion I observed clear differences betwe en effluent irrigated and non effluent irrigated soils ( Fig. 2 2 ), with effects vary ing with depth ( Fig. 2 3 ). Two time series are shown , incorporating previously published data from Field 1, that highlight impacts from 31 years of effluent loading on pH a nd P with depth (Fi g. 2 4 , 2 5 ) . Effluent Irrigated and Control Soil Concentrations The mean and range for all soil parameters for each sprayfield ( Fig. 2 2 ) provides information on the general influences of effluent loading. F ields clearly have inherent n atural and manage ment induced variation in soil chemical properties, but effluent loading effects we re clearly identifiable for some constituents . Average pH at all irrigation fields was higher (7.3 Â± 0.4 ) than corresponding control site s (6.2 Â± 0.4 ). In
31 F ield 5, pH was the only parameter showing a significant difference. In Fields 1 through 4, average WEP, or orthophosphate (OrthoP), concentrations were significantly higher in irrigated fields (3.0 Â± 2.0 mg kg 1 ) than control sites (0.9 Â± 0.9 mg kg 1 ). Oth er parameters with significant concentration differences in multiple fields includes: higher M3P (Fields 1, 2); higher NOx (Fields 1, 2, 4); higher Ca (Fields 1, 2); and lower Al (Fields 1, 4, 5). No significant effect s w ere observed on Mg, Fe, TKN, and OM . OM content (not included in the figure ) ranged from 2 5% in all surface soils (0 0.3 m) . I n Field 4 naturally present concentrations of M3P and TP in the soils were m uch higher (10 100 times ) than other fields, especially in deeper horizons and control s . Ef fects of effluent loading in this field are clearer in Fig. 2 3 that separate s surface and subsurface samples. The higher pH and Ca values identified are in general agreement with Hu (2006) who documented elevated values in effluent receiving systems. Excl uding Field 4, all sites exhibited increased Ca concentrations to, minimally, 0.6 m depth. In Fields 1, 2, and 3, where Ca is elevated to maximum sampled depth, 43% of increased Ca occurred in the upper profiles (0 0.6 m), highlighting an accumulation of C a in surface soils.
32 Figure 2 2. Average and range of parameter concentrations in all effluent irrigated and non effluent irrigated soils. Differences in Effluent Irrigated and Control Soils with Dep th M ean values for soil parameters on a mass per area basis (kg ha 1 ) that are separated by depth also show important differences ( Fig. 2 3 ) . Positive and negative values indicate higher or lower analyte mass es in irrigated soils when compared to control soils. Soil concentrations were converted to a mass per area value, with an assumed bulk density (1500 kg m 3 ) , in order to determine the total mass difference between effluent irrigated and control soils . As most research focuses on near surface soils, the figure hi ghlights the influence of long term efflu ent loading at surface soi ls (0 0.15 m, 0 0.3 m, 0 0.6 m) and subsurface soils (0 .6 3.0 m). Impacts to surface soils (0 0.15 m) can influence soil fertility, plant productivity, and potentially increase runoff P.
33 I ncrements of 0 0.3 m and 0 0.6 m represent typical soil sampling and grass/crop rooting depths , respectively . The 0 3.0 m profile highlights overall differences between the two systems and where alteration is occurring below typical plant uptake depths and available for subsurface leaching and loss. In all surface soils, to a depth of 0.6 m, higher M3P concentrations were identified. In Fields 1, 2, 3, and 5 60% of increases are in the upper 0 0.6 m. In all fields , except Field 4, M3P a ccumulation occurred to maximum depth . Higher concentrations of surface M3P indicates that it is unlikely th at the harvested biomass, at an estimated 35 kg P ha 1 per year (Elliott et al., 2011), is regularl y taking up all e ffluent P, applied from 30 300 kg P ha 1 per year . Higher WEP was identified to maximum depths in four of five fields, excluding Field 5. Only 29% of the higher WEP was associated with the top 0 0.6 m of soil . As higher WEP was predicted to correlate well with M3P increases, the substanti al rise in WEP deep into the profiles is a result that is contradictory to the proposed hypothes is that increase of WEP P saturated . Further verification on this finding are highligh ted in C hapter 3 . In Field 4 naturally occurring phosphatic rich soil deposits (Harris, 2015) in the deeper, control soil horizons, may have resulted in the M3P and TP values observed . These subsurface findings however make the WEP increase in Field 4 even more striking as it is the only P parameter to increase to maximum depths sampled. Concentration , as well as the percentage of ph osphorus in the WEP form , increased with depth in effluent irrigated soils . P ercentage of T P as WEP was 0.2 0.4% across control soils . In effluent irrigated soils , this percentage ranged from 0.7% Â± 0.6%
34 in surface soils, and increased to 1.7 2.8% in deepe r horizons (1.2 3.0 m). In these systems , some of the applied P appears to be accumulating in the more labile, water extractable form prior to substantial increase in M3P, especially in deeper soil profiles. This finding is contradictory to the hypo until the soil horizons above show significant accumulation in M3P. T he influence of effluent loading on Field 5 was lower than other sites, but exhibited some characteristic shifts associated with the effluent l oading in surface soils (Ca) and to maximum depth sampled (pH, M3P, Al). The reduced influence may be associated with lower effluent P concentrations and application rates, and presence of a, relatively, shallow clay lens that may allow for periodic satura tion, and flushing, of loosely bound soil elements during storm events. The reduction in extractable Al is possibly associated with an increase in insoluble Al precipitates associated with applied WW anions. This shift, to a less acidic, Ca enriched enviro nment with lower concentrations of reactive Al, can strongly influence strength, binding preference, and ultimately, lability and transportability of P, as binding shifts from stronger, less soluble Fe Al bound P to a generally weaker Ca Mg associated adso rption mechani sm (Cho, 1991; Isensee & Walsh, 1972 ; Nair et al., 1995) . If only considering the increased concentrations for WEP and NOX it would be expected that these soils woul d leach higher concentrations of P and N, compared to non effluent irrigated soils. Most of the applied N is assumed to be denitrified, taken up through the plant root systems, or leached from the system. Influence on TKN appear s to be site specific and do es not identify clear trends. NOx elevation was found to be higher in surface
35 horizons in Fields 1 through 4, where 55% of mass accumulation occurred. R esults indicate that most, or all, applied N is either held in surface soils or lost to the environm ent, with minimal increase occurring as organic N and ammonium forms deeper in the profiles. Figure 2 3. Differences between mean values of effluent irrigated and control soil parameters on a mass per area basis (kg ha 1 ) with depth. Note: Negative valu es indicate lower concentrations for effluent irrigated soils. Time P with Depth For Field 1, effluent irrigated soil data were added to h istorical data (Overman & Leseman, 1982; Overman et al., 2003) to extend time series of pH (Fig . 2 4 ) and extractable P (Fig . 2 5 ). Four previous sets of soils were sampled to 7 meters depth in the same irrigation pivot (Pivot 5) in 1982, 1985, 1990, and 2000. In 2013 samples were collect ed in the same irrigation pivot to determine current status . The effect that time of
36 WW irrigation has had on pH and P concentration with depth can clearly be seen in these findings . A substantial pH shift has occurred since system inception, when soils wer e highly acidi c ( ~ 5.3) throughout the profile , to present, where the pH has risen to ~7.5 to , in some cases , three meters below the surface (Fig. 2 4 ) . This shift in pH is strongly aligned with the pH of the effluent waters applied. As the pH shift after 18 years (Year 2000) appears to have influenced soil chemistry to a depth of about 5.5 meters, it is reasonable to assume that the current pH alteration would be at least to that depth, and probably deeper. This shif t in pH can influence P binding and decrease overall st ability and strength of P stored in the system. Figure 2 4 . Time series showing the effect of 31 years of effluent application on pH. Figure 2 5 emphasizes the rapid accumulation, and then vertical transport, of extractable P at the site, using historic al (Overman et al., 2003) and sampled data. Initial effluent loads, containing ~10 mg P L 1 , are shown to accumulate P in the surface soils over the first 8 years of system operation. Following this initi al application period, effluent P concentrations dropped to ~5 mg L 1 , where the influence is shown through
37 decreased P concentrations in surface soils by 2000 . The initial high P load (i.e., 300 400 kg P ha 1 per year for 10 years) plausibly saturated sur face soils and, after 18 years of loading, indicates a downward moving layer of elevated P evident at ~ 225 cm. Th is increase in soil element concentration with depth over application time was also observed for other io ns commonly associated with effluent i rrigat ion ( i.e., Ca, Mg, K) ( Overman et al., 2003 ) , highlighting that numerous variables are influenced in long term effluent receiving systems. Phosphorus sample s collected in 2013 were extracted using Mehlich 3 (M3), rather than the previous researchers Mehlich 1 (M1) extraction methodology. As M1 extracted concentrations from the same soils are typically lower than M3 concentrations (Alva, 1993) our data is within a reasonable range of expected values, and suggest a continuing accumulation and transport of P through the receiving soils . Overall, the Mehlich extractable P results shows vertical transport rates of, roughly, one meter every 10 years . Trends of accumulation and dow nward transport of P have also been recognized up to 10 meters (Lin et al., 2006) in systems applying effluent at high rates (Moura et al., 2011) , but in other long term ef fluent irrigated agricultural operations elevation in soil P is often not reported at depths greater than 0.5 m (Jaiswal, 2010) .
38 Figure 2 5 . Time series showing the effect of 31 years of effluent application on extractable P concentrations. Early reports at one sprayfield system found that the soils were effectively removing biochemical oxygen demand, fecal coliform, suspended OM, and phosphate (Overman, 1979) , but the result s of this study indicate that current P uptake mechanics have been altered since system inc eption. Major findings include that sequestration and loss of analyzed constituents is occurring throughout the profiles, and that P transport is occurring is occurr ing below the root zone in many circumstances . M ost of the applied P was expected to accumulate in the tightly bound , acid extractable M3P form, and then upon saturation begin to accumulate in a more loosely bound water extract able form. Contrary to the hypothesis, in these systems accumulation of WEP was occurring in surface and subsurface soils without the prerequisite accumulation and saturation of M3 P. Because of the soil chemistry changes that have occurred in soils receiving long term effluent loadi ng , lower sorptive
39 strength of P with soil particles was likely the cause of the increased lability and transportability of P. F indings can also be partially explained through the primarily, highly labile dissolved ortho P form that is applied through effl uent irrigation (Jacangelo et al., 2012) . It has been proposed that the major removal mechanism for applied effluent P i s through various surface s orption mechanisms and not precipitated into the more recalcitrant and ins oluble complexes (Lin et al., 2006) that can reduce expected P transportability . Assuming that effluent receiving soils are behaving like other soils heavily loaded with organic was tes that increase pH, Ca, and P , it is expected that much of the P in this system has shifted from a Fe Al bound state ( i.e., control soils) to upwards of 70% as Ca Mg bound P (Hu et al., 2006; Nair et al., 1995) . Given the pH and Ca incr eases, identified in Fig. 2 2 and 2 3 , the presence of P sorbed to CaCO3 and Ca P precipitates most likely increased . Currently Ca bound P species may dominate P mineralogy in the system, but are identified as l ess effective long term storage mechanisms wh en considering t he surface and subsurface increase s identified in M3P and WEP. OM content was expected to increase with contributions of dissolved and particulate OM associated with the effluent, and ultimately increase soil P sorptive capacity (Menzies et al., 1999) . As OM content did not increase in the soils, its impact on P retention in these systems is considered negligible. Insignificant accumulation was possibly associated with harvesti ng and removal of crop biomass, increased microbial activity in nutrient enhanced effluent irrigated landscapes, and the presence of coarse sands, which prevent extended periods of saturation, thus promoting oxic microbial degradation (Falkiner & Polglase, 1997; Falkiner & Smith, 1997) . One management
40 option for improved nutrient and water holding capacity i n surface soils would be through targeted OM content increase , through the return and incorporation of pl ant residue to irrigated plots (Mann et al., 2012; NRC, 1996) . While the returning of the organic P pools to the s oil will increase the overall mass of P re tained in the system, OM can act as a long term reservoir for applied P and , additionally, reduce leachable nitrate loads through increase d denitrification. These results are aligned with previous research (USEPA, 1981; Elliott & Jaiswal, 2011; Lin & Banin, 2005) in other effluent receiving surface soils , where over 25 40 years the sequestration capabilities have decreased s o substantially that the environmental buffers are no longer functioning as water quality protection measures (USEPA, 1977; USEPA, 2006) . Previous researchers have determined that the concentration of WEP correlates well to the expected soi l solution P . Assuming that WEP (mg kg 1 ) times 0.1 is equal to the expected soil solution P ( mg L 1 ) (Moura et al., 2011) , WEP in Fields 1 through 4 , ranging f rom 1 7 mg kg 1 , would expect to generate soil solution P fro m 0.1 0.7 mg L 1 . S oil solution P has been well correlated with P leaching (Zhang et al., 2002) , and all these dissolved P concentrations are above, or near, Florida surface water standards (FDEP, 2012b) . If limited P sequestration occurs along the remaining flow path s to surface water systems these additional loads can be influential on meeting water quality goals. Conclusions Long term irriga tion with effluent wastewater has resulted in significant alterations to soil chemical properties and P lability , both in surface soils (0 0.3 m), and , in most sites, to depths up to three meters. I ncrease d WEP with depth was the most important finding , as the results indicate unique system behavior where soil P saturation
41 is not necessarily required prior to significant increases in leachable soil P . P rimary factors that played a role in the accumulation and increased lability of P in the soils incl uded: high concentrations of P in initial system operation, high hydraulic loading rates, and alteration of chemical species that can heavily influence P fate . Soil chemistry changes , such as lower extractable Al and higher pH , can help explain the occurre nce of higher WEP concentrations deep in soil profiles . Increases in effluent associated Ca were expected to limit P transport in the systems, but appeared to have little influence on the transport of WEP deep in soil profiles . Many of soil analytes measured in this study can strongly influence P fate and transport , but are typically not sampled to such a depth in agricultural systems. Testing these parameters to depths identified can provide additional information to improve accuracy of predictive crop yield and nutrient transport models. Findings can help improve long term management strategies attempting to meet groundwater quality and crop productivity goals in effluent irrigated landscapes.
42 CHAPTER 3 ASSSESSMENT OF SOIL PHOSPHORUS LOSS INDICATORS AT PREDICTING DISSOLVED ORTHOPHOSPHATE LEACHING FROM FLORIDA EF FLUENT SPRAYFIELDS Chapter Abstract P hosphorus (P) leaching from wastewater effluent disposal sites is important to quantify for protecting and improving water quality. Leaching potential is commonl y based on soil characteristics such as soil test P (STP), water extractable P (WEP), and the capacity of the soil to sorb added P, often calculated as a P saturation ratio (PSR) or soil P storage capacity (SPSC) utilizing extractable P relative to iron and aluminum. These indicators have all been validated as me trics to determine susceptibility of P leaching in landscapes receiving long term applications of organic and inorganic P , but have not been confirmed in landscapes where long term wastewater irrigation has altered typical P fate and transpo rt mechanisms . This research determined the relative effectiveness of STP (as Mehlich 3 P), WEP, PSR , and SPSC as predictive measures of P leaching in five Florida sprayfields with ~30 years of municipal wastewater applications. Through analytical soil tes ting, of surface and subsurface soils, and soil column leaching experimentation the most accurate P leaching rate predictors were determined to be PSR [for subsurface soils (r 2 = + 0.80) and surface soils (r 2 = + 0.72)], WEP [for surface soils (r 2 = + 0.70) and sub surface soils (r 2 = + 0.62)], and lastly STP (r 2 = + 0.58) . The SPSC calculation accurately predicted P leaching risk in 3 of 5 sprayfield s , compared to 100% accuracy of control (background) soils sampled . R esults highlight that the SPSC may have limited applica bility in long term effluent irrigated soils , but can be improved through the incorporation of corrective fact ors that appropriately consider the influences of effluent loading on soil chemistry and P lability .
43 Introductory Remarks While appropriate in mos t soils , specific factors associated with wastewater (WW) irrigation m ay influence the applicability of phosphorus (P) leaching risk metrics in in long term effluent receiving soils . P leaching risk determinants such as soil test P (STP), water extractable P (WEP), P saturation ratio (PSR) and soil P storage capacity (SPSC) are commonly used for estimating P loss, but can be influenced by pH shifts and additional Ca and base anion loading ( Ch. 2 ) . These factors may alter soil P holding capacity from conditi ons present at the onset of loading . As many land waste disposal typically applies at rates based either on a (USEPA, 1981) (Robinson & Sharpley, 1996) , biosolid (Alleoni et al., 2008) , and effluent disposal operations (Elliott & Jaiswal, 2011) . In such systems P can be applied at 10 100 times plant needs (Alleoni et al. , 2008; Robinson & Sharpley, 1996) ( Barrow, 1983) . The resulting P saturation condition can produce an environment where dissolved P is able to travel through soils unattenuated, enter groundwater systems, and ultimately influence surface water quality (Harris et al., 1996; Moura, 2009; O'Connor et al., 2005) . Moreover, P in effluent is more labi le , mobile, and bioavailable than in gr anular fert ilizer (Holloway et al., 2001; Lombi et al., 2004; Shuman, 2001) , biosolid, and manure forms (Agyin Birikorang et al., 2008; McDowell & Sharpley, 2001; O'Connor et al., 2005) . As such , commonly used P loss estimation metrics may not be as accurate or appropriate in effluent irrigated soils (Elliott & Jaiswal, 2011; Moura et al., 2011) .
44 In this paper I focus on interpreting P fate and transport findings in soils receiving WW effluents to advance predictive accuracy of commonly used P attenuation metrics. Objectives are to determine relative accuracy of STP (as Mehlich 3 extracted P) , WEP, PSR , and SPSC at predicting dissolved P concentrations generated through leachate experimentation. It is hypothesized that these metrics have reduced predictive accuracy of P leaching rates in long term effluent receiving soils . Results can be interpreted to determine if incorporation of addition al soil parameters , alternative sampling depths , or other corrective factors can improve predictability . Phosphorus applied to the land can adsor b onto soil surfaces, be taken up by plant systems, and, under certain circumstances, be transport ed to receiving ground and surface waters . In many environments, leaching of applied P is not a major concern, with greater than 90% of the applied mass rapidly bound to sediments (Dobermann et al., 2004) and made unavailable for transport via percolating waters. Most soils receiving P can be relied upon as long term environmental buffer s against water quality degradation, but under specific conditions, subsurface P loss can occur and contribute to watershed loading (McCobb et al., 2003; Nair & Graetz, 2002; Nair et al., 2004; Sims et al. , 1998) . In areas with sandy soils and shallow water tables, hydrologic connectivity is high and soil P sequestration is likely reduced . Inorganic P can be bound to various ionic species present in the soils , where bond strength is dependent on relative e lemental concentrations and chemistry of the system . In low organic carbon, acidic soil systems retention typically depends primarily on concentrations of iron (Fe) and aluminum (Al) oxides and hydroxides (Eckert & Watson, 1996; Kleinman & Sharpley, 2002) , clays, and
45 organic matter (Nair et a l., 1995; Pautler & Sims, 2000) , with calcium (Ca) and magnesium (Mg) becoming increasingly influential in alkaline and calcareous soils (Eckert & Watson, 1996; Lin & Banin, 2005; Parfitt et al., 1975; Rajan, 197 5; Ryden & Pratt, 1980; Sharpley, 1983) . Several indicators are commonly used in agricultural systems to determine P loss from the soil u pon additional loading. Soil test P (STP) is a determination of P content in the soil, and has been broadly researche d as a P risk assessment tool . It correlate s well with subsurface P losses and is relatively inexpensive ( Maguire & Sims, 2002) . Zhang et al. (2002) observed a strong correlation between P leaching and water extractable P (WEP), a measure of labile and mobile soil P content . Others have improve d predictions by incorporati ng additional soil factors, such as Fe and Al, where P loss is based on soil P concentration s relative to sorptive capacity (Nair & Harris, 2004; Nair et al., 2004) (Chrysostome et al., 2007b) . The P Saturation Ratio (PSR ; Eq. 3 1 ) (Chrysostome et al., 2007a; Maguire & Sims, 2002; Nair & Harris, 2004) suggests that the molar concentration of extractable P in the soil relative to molar concentrations of extractable Fe and Al predict s the fate and leaching risk of further applied P . This has been validated using oxalate, Mehlich 1, and Mehlich 3 extracting solutions (Kleinman & Sharpley, 2002; Nair et al., 2010) . PSR = P / (Fe + Al) * 100 ( 3 1) (Nair, 2004) The empirical alpha ( ) factor in the denominator accounts for the fraction of Al and Fe responsible for P sorption for soils in a given region, but is omitted for most systems (Maguire & Sims, 2002; Nair & Harris, 2004) .
46 While the STP, WEP, and PSR values are valuable for determining current status of P enrichment, the information is most useful to land owners as a numerical loading factor that indicates whe n soil retention has been saturated . The PSR is limited in its applicability, since it does not inform the land user how much more P is expected to be sorbed in the receiving environment . In response, researchers have proposed a soil phosphorus storage cap acity (SPSC) (Eq . 3 2) for terrestrial environments receiving ongoing P loading . SPSC has been referred to as a soil can sorb before exceeding a threshold soil equilibrium concentration (Nair & Harris, 2004; Nair et al., 2010) . Previous studies have validated SPSC as an effective measure for determining leaching risk at biosolids and f ertilizer amended soils (Chrysostome et al., 2007a; Chrysostome et al., 2007b; Maguire & Sims, 2002 ; Oladeji, 2006) . However, it has not been validated for predicting leaching in long term effluent application sites (Moura, 2009) . SPSC (mg kg 1 ) = (thres hold PSR calculated PSR) x (Al + Fe ) x (P molar mass) ( 3 2) Note: Al and Fe as molar concentrations of extractable aluminum and iron . 3 2, is the hypothesized value where the potential mobilit y of P in a soil is expected to vary greatly . T he based on analyzed soil values . If the calculated PSR is b elow the threshold PSR , P is believ ed to be strongly bound to high energy sites and not readily released. Soils with calculated P SRs above this threshold are characterized by low energy , more easily released, P sorption (Heckrath et al., 1995 ; Maguire & Sims, 2002; Nair & Harris, 2004) . Nair a nd Harris (2004) originally proposed a PSR threshold value of 0.15 for Florida sandy soils that corresponds to the critical P solution concentration of
47 0.10 mg L 1 (USEPA, 1986) , with the value typically ranging from 0.05 for soils with lower P binding capacity ( i.e., coarse uncoated sands) to 0.15 for soils with greater P sequestration capabilities (i.e. , coated fine sands). SPSC concentration s can be converted into a depth dependent measure , which can be used to e stimate fate of additional ly applied P . SPSC (kg ha 1 ) = SPSC (mg kg 1 ) * [Depth of the soil horizon (m) * bulk density (kg m 3 ) / 100] (3 3) While SPSC may be an effective tool to determine P leaching risk in acidic soils receiving solid P forms, its ap plicability is untested in some environments. F actors that may influence transport and lead to an overestimation of soil sink capacity include hydraulic conductivity and cal careous and phosphatic rich clay soils (Menz ies et al., 1999) . Assimilation and cycling of P is continuously occurring within soil vegetative systems (Evanylo et al., 2010) , but with continuous loading, the soil can eventually cease to provide additional P storage, allowing downward transport (Heckrath et al., 1995; Kleinman et al., 2000) . Due to such factors and additional uncertainties in vegetated landscapes receiving high WW loads et al. , 2008) I hypothesize that the SPSC value is not be an appropriate measure for estimating P leaching risk. Methods T o determine the effectiveness and applicability of commonly used P loss metrics at determining P leaching values, I collec ted soil cores from five WW irr igation fields (Table 3 1) , analyzed samples for constituents listed in Table 2 2 , and conducted a leach ing experiment .
48 Description of S tudy S ites The study was conducted at five WW effluent sprayfields, located across northe rn and centr al Florida (Fig . 2 1), where secondarily treated municipal WW effluent had been land applied through pivot irrigation for 28 38 years (Table 3 1). Prior to effluent application , fields were used as low intensity pasture operations where applica tions of P and other amendments were minimal. Each of the sprayfields was under cultivation, predominantly growing and harvesting grasses. Effluent hydraulic loading rates ranged from 2.5 5.0 cm per week ( 1 2 in per week ) . Historic P concentration in efflu ents were unknown, but assumed to exhibit water quality trends similar to other systems in the United States, dropping substantially from system inception (10 15 mg L 1 ) to present day (1 2 mg L 1 ). Table 3 1. Sprayfield reference numbers, area, loading r ates , and sampling information. Field Area Time of operation Sample depth range Avg sample depth (ha) (years) (cm) (cm) 1 809 32 200 200 2 142 28 100 165 135 3 129 34 150 200 190 4 47 38 180 200 195 5 63 28 70 115 90 Soil S ampling and A nalysis At each sprayfiel d, three soil cores were taken in duplicate to a maximum depth of two meters. In addition three soil cores were taken in duplicate at nearby (0.1 1.0 km) , but not receiving effluent irrigatio n or soil unaffected by effluent application or other major soil alterations or amendments. Soil core locations were randomly selected within t he sprayfield and control si tes.
49 Samples were collected in duplicate for soil analysis and reconstructed soil core leaching experiment s . Soils were primarily coarse and fine sands with low OM content, and minimal stratification and structure. Soils were extracted with a 7.62 cm diameter hand auger and fractionated into, maximally, seven depth increments of (cm): 0 15, 15 30, 30 60, 60 90, 90 120, 120 150, and 150 200. When clay horizons were encountered maximum sample depth was limited to the top of the clay horizon (Table 3 1). In total, four teen locations from effluent sprayfields and sixteen locations from control sites were collected and analyzed. Due to sampling error in Field 2, two effluent receiving sites and four control sites were collected. From each depth increment 500 1000 g of soi l was sieved ( 2 mm ) and oven dried at 85 Â°C for 2 4 hours. Dried soil was homogenized and 5 10 grams from each depth increment were sub sampled for analy t es listed in Table 2 2. STP was determined using a Mehlich 3 extraction method (M3P). Soil bulk density for all samples was assumed 1500 kg m 3 . Leaching E xperiment M ethodology Column experiments were designed to determine effects of long term effluent loading on leachable ortho p hosphate (ortho P) . Correlations between l eachate generated dissolved ortho P and soil values for STP (as M3P) , WEP, and PSR were compared to determine the most accurate P leaching risk metric for soils under long term effluent loading . Six, 2.0 meter (height) by 7.62 cm (diameter), PVC cylinders were outfitted with fiberglass screen bottoms and attached upright for a leach ing experiment. Each depth fraction was added to the column, with light repacking, in the order that it was extracted
50 in the field. Soil reconstruction is meant to be representat ive of soils above clay horizons, when present. Soil columns received ~8 liters (~ 178 cm ) of liquids in total. W ater was applied over one hour as 350 mL ( 7.62 cm ) applications , and repeated every 24 48 hours for a total of three weeks. Assuming a soil porosity of 0.4, 3 7 po re volumes were applied over the course of the experiments. For each column , six leachate samples in total were collected from percolated water , filtered through PTFE 0.45 m filters, and refrigerated prior to analysis . In some soils , high concent rations of fine, suspended colloidal solids were generated and sample r eplicates were limited to a minimum of three samples in initial sampling events . Data A nalysis S oil concent rations (Fig . 3 1, 3 2, 3 3) were averaged within individual sampling points . As all soil in the columns influenced the concentration of P leached, data points for figures that incorporated leaching results (Fig . 3 4, 3 5, 3 6, 3 7) are generated from the averaged soil concentrations values within the entire soil column. All soil concentration, PSR, and S PSC calculations were determined from averages o f three effluent irrigated soils and three control soils per field . S oil samples were divided into surface ( 0 0.3 m) and subsurface (0.3 2.0 m) fractions to determine whether common surface sampling methods are an adequate measure at predicting P leaching in effluent irrigated soils; and if not, could deeper sampling be used as an alternative technique to improv e accuracy of the predictive soil metrics. Soil findings are correlated with column leachate experimentation to determine which measures are most applicable in effluent irrigated soils. Leached P
51 concentrations are plotted against SPSC (kg ha 1 ) values to determine if the value accurately predicts when soils have surpassed the point of further P sequestration. For all PSR and SPSC graphs and calculations a change point PSR of 0.10 was assumed accurate , as was conducted in previous experimentation (N air, 200 4). To calculate the SPSC on a mass per area basis (kg ha 1 ), concentrations were multiplied by soil bulk density and depth of the soil profile. Differences between effluent irrigated soil and non irrigated soil concentrations were evaluated using a one ta iled T test. If the calculated T value w as greater than the T critical value calculated at the 1% ( =0.01) level of significance, differences w ere considered statistically significant. The null hypothesis is that there is no significant difference between effluent irrigated soils and non irrigated soils. Results and Discussion Soil samples highlight differences betwee n surface and subsurface horizons in effluent irrigated and control soils. As WEP is often considered a reliable indicator of soil solution P (Moura et al, 2011) and potential leachability of P from soils (Zhang e t al., 2002) , Fig . 3 1, 3 2, and 3 3 show the correlation betwe en M3P vs. WEP and PSR vs . WEP. M3P vs. WEP In control soils, average values for WEP were significantly ( <0.01) lower than for effluent irrigated soils. As an average across all fields, pooled across all depths , WEP in control soils was 0.8 Â± 0.8 mg kg 1 compared to effluent irrigated soils at 2. 8 Â± 2.0 mg kg 1 . Due to these low concentrations in the control soils, c orrelations between WEP vs. M3P were very low for surface ( r 2 = + 0.04) and subsurface horizons ( r 2 = + 0.24).
52 Effluent soils showed distinctive behavior between surface and subsurface correlations of WEP vs . M3P (Fig . 3 1). At e ffluent sites , WEP is hi ghly correlated with M3P in surface horizons ( r 2 = + 0.87), but not in subsurface soils ( r 2 = + 0.44). These results highligh t that surface and subsurface soils appear to be reacting differently to the P loading. Comparing effluent irrigated and control soil s id entified increases in M3P and WEP . Average concentrations in effluent irrigated surface soils (0 0.3 m) had higher M3P ( 1 7 8 vs . 1 2 8 mg kg 1 ) and WEP (2 . 8 vs . 0 .8 mg kg 1 ) than control soils ( Ch. 2 ) . In subsurface soils (0.3 2.0 m) there was no clear differ ence in M3P concentrations between effluent loaded and control soils ( 77 vs. 76 mg kg 1 ) , but WEP was significant ly higher ( 2.9 vs . 0.5 mg kg 1 ) ( Ch. 2 ) . This substantial increase of WEP was observed in 4 of 5 fields (Fields 1 through 4), indicating that l ong term effluent irrigation has increase d WEP in these soils to at least two meters in depth. If WEP is assumed a reliable predictor of P leaching in effluent receiving soils , these results suggest that M3P may be a poor predic at or of leaching risk.
53 Fi gure 3 1. Relationship between M3P and WEP for individual samples of effluent irrigated surface (0 0.3 m) and subsurface (0.3 2.0 m) soil s . PSR vs. WEP D ifference s between effluent irrigated and control soils become clearer when comparing WEP vs . PSR. If effluent receiving soils are behaving similarly to other P loaded systems cited in the literature , a clear PSR change point value should be encountered where above and below the PSR value is expected a significant increase or decrease in WEP per unit PSR. E ffluent irrigated soils show no clear change point (Fig. 3 2) . If these soils were behaving like those previously studied , the s lope below the threshold PSR (<0.1) would be near zero. I n contrast, the fitted slope ( m= 28.3 ) below the PSR threshold was high er than above it (m= 9.0), when i.e., PSR>0.1) . The low correlation between WEP and PSR in the effluent irrigated soils suggests that PSR may be of limited utility at these sites, or applied with a PSR change point of 0.05 or
54 below . The red, vertical line represents a commonly applied threshold PSR of 0.1 (Nair, 2004). Figure 3 2 . Relationship between WEP and PSR for individual effluent irrigated soil samples below and above threshold PSR of 0.1. In contrast to efflu ent soil s , non irrigated control soils behaved similar to other P loaded landscapes in the literature and aligned well with the change point of 0.1 (Fig. 3 3). B elow the P saturation point (<0.1) , WEP concentrations are very low , and uncorrelated with PSR .
55 Figur e 3 3 . Relationship between WEP and PSR for individual control soil samples below and above threshold PSR of 0.1. M3P , WEP, and PSR vs. Leached Orthophosphate Out of the 16 control soil columns , only two generated average leachate ortho P concentrations g reater than a surface water quality target of 0.1 mg P L 1 (Breeuwsma et al., 1995; USEPA, 1986) ; one from Field 3 ( 0.86 mg L 1 ) and one from Field 4 ( 0.27 mg L 1 ) . Du e to the low concentrations of dissolved P generated fro m control soils , M3P , WEP, and PSR were all weakly correlated to leachate concentrations . H ighest correlation for all tested soil metrics against ortho P in control soils was r 2 = + 0.57 for WEP of surface soils , and r 2 = + 0.53 for PSR of surface soils. All oth er r 2 val ues fell between 0.05 and 0.29. I n effluent irrigated soils correlations between leached ortho P concentrations and soil P metrics ( STP (as M3P ) , WEP, PSR ) were stronger (Fig 3 4 th r ough 3 6) . T he metrics with the highest r 2 are identified as the most reliable P leaching rate predictors for soi ls with similar properties and loadi ng history .
56 Of the three indicators, PSR of subsurface soils had the highest correlation at r 2 = + 0.81 , and PSR of surface soil had the second highest correlation ( r 2 = + 0.72 ) (Fig . 3 6) . Similarly correlated, WEP of surface soils was at r 2 = + 0.70 (Fig . 3 5) . STP , extracted as M3P, was found to be the least predictive at r 2 = + 0.57 for all depths (Fig . 3 4) . Differences in slopes were substantial for M3P surface (m=0.007) and sub surface (m=0.12) and PSR surface (m=8.30) and subsurface (m=12.97), when compared to WEP surface (m=0.42) and subsurface (m=0.48) slope values . Figure 3 4 . Correlation of M3P and lea ched orthophosphate in effluent irrigated soils.
57 Figure 3 5 . Correl ation of WEP and leached orthophosphate in effluent irrigated soils. Figure 3 6 . Relatio nship of PSR and leached orthophosphate in effluent irrigated soils. The i ndicators, and associated formulas, that most closely c orrelate with leached ortho P concentration s are shown in Eq . 3 6 , 3 7 , and 3 8 . T hese formulations can be used to predict P leaching risk in effluent receiving soils where PSR or WEP values are known .
58 Ortho P (mg L 1 ) = 12.97 *PSR 0.37 , for PSR of subsurface soil s (3 6) Ortho P ( mg L 1 ) = 8.30* PSR 0.83 , for PSR of surface soil s (3 7) Ortho P ( mg L 1 ) = 0.48*WEP 0.03 , for WEP (mg kg 1 ) of surface soils (3 8) The findings in the PSR graphs highlight the importance that the depth of sampling can have on expected leaching concentrations. For example if calculated PSR values were determined at PSR=0.2 for a surface soils, the expected leached P would be ~0.75 mg L 1 , as opposed to subsurface soils at PSR=0.2 that would predict ortho P = ~2.25 mg L 1 . Given these findings, if applying formulas t o determine deep P leaching rates, and only surface soil data is available , the WEP may be a more reliable tool, as the relationship between WEP and leached ortho P is more consistent between surface and subsurface soils . In other research a pplying these meas ures to determine P surface runoff loss, relationship s indicat e d WEP would also be the most accurate predictive measure ( r 2 = + 0.82, r 2 = + 0.91) (Agyin Birikorang et al., 2008) , followed by PSR ( r 2 = + 0.77), and STP ( r 2 = + 0.72). SPSC Accuracy at Predicting Leached Orthophosphate According to the SPSC values ( Table 3 2 ) the soils vary in their ability to sequester additional P. N egative SPSC values calculated for all effluent soils, down to 0. 3 m depth , along with positive values at the same depths in control soils supports the hypothesis that effluent irrigation has decreased P sequestration potential . N egative values indicate that additional P applications will not be retaine d in surface soils, and thus could leach to lower horizons. In Fields 1, 2, and 5, positive SPSC values in deeper horizons indicate that dissolved P leached from surface soils would be sequestered during movement through the soil profile. In contrast , diss olved P loads would be expected to move through the entire soil profil e in Fields 3 and 4 , where negative SPSC
59 values throughout the profile indicate P saturation . Field 4 was an anomaly exhibiting negative SPSC values throughout the soil profile, in both effluent irrigated and control sites, suggesting a strong influence of soil horizons that were geologically high in P . According to the SPSC values calculated for control soils some surface horizons exhibited saturation, but given the positive SPSC values deeper in the horizons at sites 1, 2, 3, and 5 additionally loaded P would be expected to be sequestered and not found in the leachate. Table 3 2 . Average SPSC with depth for effluent irrigated and control soils. SPSC (kg ha 1 ) per sprayfield Depth (m) 1 2 3 4 5 Effluent Irrigated Soils 0 0.15 157 157 339 684 138 0.15 0.3 75 57 228 600 2 0.3 0.6 14 88 229 683 83 0.6 2.0 480 488 292 1164 243 Control Soils 0 0.15 99 71 247 443 6 0.15 0.3 141 121 166 375 104 0.3 0.6 301 277 48 277 167 0.6 2.0 925 608 281 2351 515 N ote: 2.0 meter is a maximum value, limited by sampled soil depth . SPSC , calculated by summing the mass retention (kg ha 1 ) values for all depths in each column (Table 3 2), against average leached ortho P concentration (Fig. 3 7) was hypothesized to provide values where soil columns with a positive SPSC would leach the lowest concentrations of ortho P . The dotted line , representing the best fit line between leachate concentration and SPSC for effluent irrigated soils , has a y intercept of 573 kg ha 1 ability to sequester further added P . If the threshold PSR at 0.1 and SPSC were assumed accurate for soil s at effluent receiving landscapes, the researcher or land manage r would conclude th at th e soils can be continually loaded until the SPSC value reaches 0. These findings show
60 that the hypothesis that SPSC is an accurate measure of predicting elevated P leaching at long term at PSR=0.1 . Given these results, i f the SPSC is applied in long term effluent loaded landscapes adjustments should be made that incorporate soil chemistry shifts that have increased P transportability . Taken directly from Fig . 3 7, Eq . 3 9 gives the fitted line with adjustments necessary to implement SPSC in soils sampled to 1 2 meters in depth. As most sampling is typically only conduct ed in surface soil s , Eq . 3 10 was generated for soils collected to 0.3 meter depth . This value was calculated under the assumption that the SPSC adjustment could be applied linearly, without weighting for specific horizons. SPSC (kg ha 1 ) = 1087*Ortho P ( mg L 1 ) + 573 to 1.6 m (3 9) SPSC (kg ha 1 ) = 1087*Ortho P ( mg L 1 ) + 108 at 0.3 m (3 10) C ontrol soils, in contrast, react ed similarly to othe r P loaded landscapes where SPSC is accurate at predicting P leaching . When excluding Field 4 values , c ontaining naturally high soil P concentrations , SPSC valuations accurately identified expected P leaching trends in 7 of 11 effluent columns and 10 of 11 control columns .
61 Figure 3 7 . SPSC vs. leached ortho phosphate for individual effluent and control soil columns . The SPSC calculations accurately predicted leaching P in 3 of 5 spray fields , but for 5 of 5 matched control s . Fields 3 and 4 were expected to be saturated in their a bility to sequester further applied P and leached higher P concentrations. In Field 4, SPSC values may not be appropriate for estimating P leaching risk, as phosphate rich calcareous soils produced S PSC values that indicate d P saturation in both effluent a nd control soils (ranging from 2500 to 3600 kg ha 1 ) . These Field 4 samples were the only control values where SPSC did not accurately predict P leaching . Field 5 was expected to still have additional sequestration capabilities and displayed this w ith P concentrations comparable to control s . In Fields 1 and 2 , SPSC was not accurate in predict ing P leaching, where values suggested seques tration potential of an additional 150 250 kg P ha 1 prior to expected increase s in leached P concentrations . In o nly one of the fourteen effluent irrigated soil columns are leached dissolved ortho P concentrations below critical sol ution concentrations identified for many streams (USEPA, 1986) and shallow ground waters (Breeuwsma et al., 1995) . Florida location
62 specific surface water standards range from 0.01 0.5 mg L 1 (FDEP, 2012b) , making the . As the P leached from the experiments are within the ranges found at other P loaded landscapes ( 0.3 to 13 mg L 1 ) (Elrashidi et al., 2001 ; Gerritse , 1996; Preedy et al., 2001 ; Shuman, 2001; Wong et al., 1998) it is clear th at long term effluent irrigation can contribute higher P loads to receiving waters than nearby non irrigated soils. The PSR indicator may have had the best alignment with leached ortho P b ecause of its integration of higher M3 P and lower Al identified in effluent irrigated soils ( Ch. 2 ) , but numerous other influential factors ar e not considered. In order to improve predictive accuracy incorporation of shifts , associate d with pH, Ca, and WEP , may be necessary . S hift s in these variables can be substantial under effluent irrigation ( i.e., pH ( 7.3 vs. 6.2) and Ca (394 vs. 203 mg kg 1 ) in effluent irrigated soils vs. control soils ) ( Ch. 2 ) and can strongly influence P fate and transport. As systems with significant pH shifts and Ca accumulation are suspected to contain a higher proportion of P in the more soluble and rapidly exchangeab le Ca/Mg fraction s (Cho, 1991; Isensee & Walsh, 1972) , than Fe/Al bound P increased t ranspo rtability is expected . In efflue nt loaded soils , even the P associated with clays and Fe / Al oxides ha ve been found to be in a higher proportion of less crystalized , more labile forms (Lin & Banin, 2005) than in other non effluent irrigated soils . Due to the soil characteristics altered by long term effluent irrigation , most soils were found to have lower P saturation change points, limited SPSC applicability, and elevated P leaching rates. In order to improve accuracy of these P loss metric s i t may be appropriate to integrate quantitative variables , such as , (Chrysostome et al.,
63 2007b) tive solubility of P in applied materials, respectively (Elliott et al., 2006; Leytem et al., 2004) . As effluent P is applied , predominantly, in a dissolved, highly labile ortho P form, capable of behaving di fferently than P applied in solid phases such as granular fertilizer, manure, a nd biosolid forms (McDowell & Sharpley, 2001; O'Connor et al., 2005) , increased loss potential (Ag yin Birikorang et al., 2008 ; Lombi et al., 2004) should be incorporated for watersheds where large areas are devoted to effluent irrigation . Conclusions These findings help provide a better understanding of contaminant attenuation and system function of terrestrial environmental buffers under long term effluent loading, and exposes potential weaknesses of commonly utilized P leaching metrics . Correlations of M3P , WEP, and PSR a gainst lea ched P concentrations suggest the most accurate indicators were PSR of sub surface soils (0.3 2.0 m) , PSR of surface soils (0 0.3 m) , WEP of surface soils, WEP of subsurface soils, followed by M3 P as the worst predictive indicator of leaching risk in the sampled sprayfield s . The SPSC valuation was found to a poor predictive indicator in long term effluent receiving landscapes, compared to other heavily loaded P systems, and may require incorporation of corrective factors to account for the infl uences of pH, calcium loading , and increased P lability. Given the increasing commonality of the effluent surface application method in Florida, identifying the most accurate P leaching metrics will be important for improved model accuracy and design of efficient best management practices .
64 CHAPTER 4 ELEVATED PHOSPHORUS LEACHING ASSOCIATED WITH LONG TERM EFFLUENT IRRIGATION Chapter Abstract Soils that have been heavily irrigated with treated municipal wastewater effluent can have altered characteristics leading to increased transportability of phosphorus (P) to ground and surface waters. In order to determine this potential in Florida landscapes, soil profiles were collected from effluent irrigated and paired control fields to be used for soil analysis and leaching experimentation. Concentrations of dissolved orthophosphate resulting from water and effluent flushes, under background and amended soil conditions, were determined and extrapolated to estimate landscape loading effects associate d with the practice of long term effluent irrigation. Four, out of the five, fields with a history of efflue nt loading showed higher rate s of dissolved P in leachate, regardless of applied water type or amendment conditions , indicating that the r eceiving environmental buffer s h ave a reduced ability for stripping dissolved P during effluent irrigation events . E ffluent irrigated soils were found to leach approximately 8 kg P ha 1 per year, from a 1 2 meter soil depth, compared to non effluent irriga ted sites that leached only 1 kg P ha 1 per year. Results can help nutrient transport modelers and watershed managers more accurately determine what wastewater treatment and reuse methods are most beneficial for achieving long term social and environmental goals. Introduct ory Remarks Diffuse transport of nutrients, as experienced with land application of reused wastewater (WW), can be a major (Elliott et al., 2002; O'Connor et al., 2005) , and not fully understood (Jacangelo et al., 2012) , contributor of loads to surface water systems. Many reuse systems are applying P in excess of plant needs (Elliott & Jaiswal, 2011)
65 and have been identified as probable contributing factors to increased surface water P concentrations in some developed landscapes (Thompson & Milbrandt, 2014; Thompson, 2015) . As research at effluent receiving landscapes is limited (Menzies et al., 1999) or conducted during the early stages of system operation , which present only high soil P sequestration rates (Menzies et al., 1999) , determining potential of increasing leaching P loads can be important given natural and managed surface water and groundwater interconnectivity influences on nutrient transport in the state (Jacangelo et al., 2012) . P applied to landscapes can be chemically adsorbed onto soil particles, taken up by plants, or leached from the soil profile (Barton et al., 2005) . While the l eached fraction of applied P is usually considered a negligible amount for water shed loading calculations, in areas with coarse textured soils of low P sorbing capacity and shallow groundwater, downward movement of P from organic wastes is potentially sign ificant (Agyin Birikorang et al., 2008; Eghball et al., 1996; Lu & O'Connor, 2001; O'Connor et al., 2005) . S urface and groundwater systems in Florida are often hydrologically linked due to various natural ( i.e., s and y soils, preferential flow pathways, near surface ground water table s, extreme seasonal storm conditions ) and management induced ( i.e., landscape change, development, stormwater ponds, canals, subsurface drainage systems) influences. Because of these condit ions P mass transport to surface waters via lateral subsurface flow (He et al., 1999) can be substantial contributor s to watershed P loading, and have been identif ied in other agricultural (Schoumans & Groenendijk, 2000; Sims et al., 1998; Turner & Haygarth, 2000) and effluent infiltration systems (Lin e t al., 2006; Walter, 1996) .
66 U nder Florida standard s for impaired waters about 27% of estuaries and coastal waters, 39% of lakes, and 33% of streams are nutrient impaired (Jacangelo et al., 2012) , w ith numbers expected to increase wi th the passage of new numeric nutrient criteria (NNC) legislation (FDEP, 2012b) . W ith this in mind, the objectives of this investigation are to better understand how effluent irrigation may influence P release and transport in receiving soils over long time frames , and to determine potential impacts on receiving groundwater and surface water systems . Agricultural sprayfield soils receiving effluent applications for >25 years were sampled for chemical analysis and column leaching experimentation to determine the validity of the hypothesis that , if soil is P saturated due to effluent loading it will l each a greater concentration of orthophosphate than non saturated soils. Results provide i nformation on how the re ceiving systems have responded in order to more accurately simulate landscape P cont ributions from effluent irrigated soils in sensitive watersheds. Column leaching experimentation was conducted under t hree conditions: (1) potable/groundwater irrigation, (2) effluent irrigation, and (3) irrigation with effluent under amended soil conditio ns. D eterminations are made on: current leaching rates, current sequestration rates of applied effluent P, and the potential to reduce leaching P through the incorporation of an aluminum rich surface soil amendment . E xperiments will allow for test ing of the hypotheses that , if soil is P saturated, amendment of surface soils with an aluminum substrate will lower the concentration of P leached during effluent loading. Calculated values for P a ccumulated i n the soils and lo ss of P through leaching is used to fulfill field scale mass balance s that attempt to identify P fate in long term effluent irrigated sprayfields.
67 The primary P removal mechanism in these systems of surface sorption onto soil particles, rather than c hemical precipitation (Lin & Banin, 2005) , has been shown to be an effective method of stripping dissolved P from infiltrating waters (Arias et al ., 2001) in short term, low P loading scenarios. In systems receiving high rates of treated effluent, steady state saturation can occur in surface soils (0 0.3 m) within 10 15 years (Elliott & Jaiswal, 2011; Lin et al., 20 06; USEPA, 1981) , limiting the useful life of sites, with coarse textured soils and shallow underdrains discharging to sensitive water bodies, to ~20 60 years (USEPA, 1977) . As soils are continually loaded, significant accumulation and reduced P sorbing capacity can occur (Hayes et al., 1990; Lin et al., 2006; Menzies et al., 1999; Pescod & Arar, 1988) and lead to incr eased dissolved P losses via leaching and lateral flow transport pathways (Church et al., 2010; Heckrath et al., 1995; Maguire & Sims, 2002; McDowell & Sharpley, 2001; Nelson et al., 2005) . In some effluent receiving landscapes, applied through rapi d rate infiltration and irrigation methods, groundwater P concentrations can become elevated. Concentrations below rapid rate systems have been found to increase to 0.2 4.0 mg P L 1 (Andres & Sims, 2013; McCobb et al., 2003; Meinikmann et al., 2015; Moura et al., 2011; Walter, 1996) , with concentrations in subsurface flow from effluent irrigated turf environments ranging from 0.3 3.0 mg P L 1 (Elrashidi et al., 2001; Preedy et al., 2001; Wong et al., 1998) . In column experiments using effluent irrigated soils , Gerritse (1996) and Shuman (2001) iden tified concentrations as high as 13 to 20 mg P L 1 . Elevated groundwater P, mostly found in the form of highly bioavailable orthophosphate (ortho P) (Nash & Halliwell, 2000; Tur ner & Haygarth, 2000) , is then available for transport to surface waters via lateral subsurface flow (Burgoa et al., 1991; Harris et al., 1996; Mansell et al.,
68 1991) where algal species proliferation, and subsequent water quality reduction, can be significant (Sharpley & Menzel, 1987) . Overall, the reclaimed water nutrient load to landscapes will be dependent on irrigation rates and concentration of constituents in the water supply. Most states that utilize WW reuse on turf plots specify a maximum hydraulic loading rate of 5.1 cm per week (2 inch per wee k), and require a minimum of secondary treatment before land application (USEPA, 2004) . I rri gation rates can often be much higher than recommended , up to 5 10 cm per week (2 4 inch per week) in some instances . Under year round, high hydraulic loading dissolved consti tuents can be applied in excess of the receiving system s kinetic adsorption limit s, and result in decreased P adsorption capacity and increased nutrient transport in sandy soils (Menzies et al., 1999) . In addition to the effects of excessive irrigation, the concentration and form of P in WW can influence long term transport phenomena. G enerally, there are no P effluent standards imposed on irrigation waters (Angelakis et al., 1999; Lazarova & Bahri, 2004; USEPA, 2004) . While over the past 30 40 years P concentrations ha ve decreased substantially in municipal effluents ( from 10 15 mg L 1 to 1 5 mg L 1 ), t he form of P in ef fluent (~90% ortho P) (Jacangelo et al., 2012) can be expected to have lower fixation strength , increased lability, and thus increased transport rates when c ompared to other solid and semi solid inorganic and organ ic phase P sources applied to landscapes ( i.e., biosolids, m anures, and granular fertilizer) (Agyin Birikorang et al., 2008; Holloway et al., 2001; Lombi et al., 2004; Pierzynski et al., 2005; Sharpley, 1996) . Under reclaimed water irrigation, annual P applications can range from quantities insufficient to meet plant needs (<10 kg P ha 1 ) to rates of significant over application
69 (>300 kg P ha 1 ) . As typical crop needs per year can range from 5 100 kg P ha 1 , depending on soil conditions and crop type, P is often applied in excess of crop needs (Elliott & Jaiswal, 2011) most common effluent receiving turfgrass landscapes which receive g reater than 65% of all reused WW in the state (FDEP, 2014) . In these systems, which typically require no more than 5 40 kg P ha 1 per year (Elliott & Jaiswal, 2011; Evanylo et al., 201 0) , P is applied annua lly at rates of 10 100 kg P ha 1 (under light irrigation) up to 50 150 kg P ha 1 (under heavy irrigation) . While many uncertainties exist in the natural and managed landscape s , experimental results can help quantify unknown or unclear P loading values associated long term effluent irrigation, and provide suggestions for improvement in nutrient transport modeling accuracy and best management practice efficiency. Methods Description of S tudy S ites The stu dy was conducted at five WW effluent sprayfields, located across northern and centr al Florida (Fig . 2 1), where secondarily treated municipal WW effluent had been land applied through pivot irrigation for 28 38 years (Table 3 1). Prior to effluent applicat ion fields were used as low intensity pasture operations where applications of P and other amendments, prior to initiation of WW disposal, were assumed minimal but were uncertain. Each of the sprayfields was under cultivation, predominantly growing and har vesting grasses. Effluent hydraulic loading rates ranged from 2.5 5.0 cm per week. Historic P concentration in effluents were unknown, but assumed to exhibit water quality trends similar to other systems in the United States,
70 dropping substantially from sy stem inception (maybe as high as 10 15 mg L 1 ) to present day (1 2 mg L 1 ). Soil S ampling and A nalysis At each sprayfield, three soil cores were taken in duplicate to a maximum depth of two meters. In addition three soil cores were taken in duplicate at ne arby (0.1 1.0 km) not receiving effluent, but with the same soil classification and not under unaffected by effluent application or other major soil alterat ions or amendments. Soil core locations were randomly selected within the sprayfield and control sites. Samples were collected in duplicate for soil analysis and reconstructed soil core experiments. Soils were primarily coarse and fine sands with low OM co ntent, and minimal stratification and structure. When clay horizons were encountered maximum sample depth was limited to the top of the clay horizon (Table 3 1). Soils were extracted with a 7.62 cm diameter hand auger and frac tionated into, maximally, seve n depth increments of (cm): 0 15, 15 30, 30 60, 60 90, 90 120, 120 150, and 150 200 . Soil bulk density for a ll soils was assumed at 1500 kg m 3 . In total, fourteen locations from effluent sprayfields and sixteen locations from control sites were collected and analyzed. Due to sampling error in Field 2, two effluent receiving sites and four control sites were collected. Cores were sub sampled from each depth increment and analyzed for constituents listed in Table 2 2. From each depth increment 500 1000 g of soil was sieved ( 2 mm ) and oven dried at 85 Â°C for 2 4 hours. Dried soil was homogenized and 5 10 grams were sub sampled for analysis .
71 Leaching E xperiment M ethodology Six, 2.0 meter (height) by 7.62 cm (diameter), PVC cylinders were outfitted with fibergla ss screen bottoms and attached upright for comparative leachate experimentation. Each depth fraction was added to the column, with light repacking, in the order that it was extracted to recreate soil horizon layering in the field. Soil reconstruction is me ant to be representative of soils above clay horizons, when present. Soil columns received ~8 liters (~ 178 cm ) of liquids, applied as 350 mL ( 7.62 cm ) applications over one hour, and repeated every 24 48 hours for a total of three weeks. Volume of liquid a pplied was equivalent to six months of loading under circumstances of 5 cm per week ( 2 inch per week) of irrigation water and 127 cm per year ( 50 in ch per year ) of precipitation. Assuming a soil porosity of 0.4, 3 7 pore volumes were applied over the cours e of the leaching experiments, as the pore volume of the 0.6 2.0 met er soil columns ranged from 1.1 3.6 liters. Table 4 1 provides information on the water sources, and volumes, applied over the experiments conducted (A, B, C).
72 Table 4 1 . Experimental ste ps to generate leachate samples. Part Action Volume applied (mL) Water Applied (cm) No. of leachate samples per column A1 Initial DI water flush 1050 23 2 48 hour dry period A2 DI water flush 700 15 2 48 hour dry period B1 Effluent for 7 days 2450 5 5 2 B2 48 hour dry period DI water flush 700 15 1 C1 Soil amendment Effluent for 7 days 2450 5 5 2 48 hour dry period C2 DI water flush 700 15 1 TOTAL 8050 17 8 10 From each soil column four samples were collected from Part A , to determine dissolved P concentrations from s oils under DI water flushing conditions. (A1) Initial water applications saturated soils with 700 1050 mL of DI water over 24 hours to generate percolating waters. Some initial flushes produced high concentrations of fine suspended sediments and were disc arded as the samples were unfilterable and not expected to be representative of in situ conditions . Due to this, total number of samples for Part A were limited to on ly three replicates . (A2) F ollowing 48 hours of percolation and air drying a subsequent DI water application was conducted. From the percolated water s a 20 mL subsample was filtered through a PTFE 0.45 m syringe filter and refrigerated prior to analysis. For Part B , soil columns were flushed with effluent water for seven days and then flushed with DI water. In total two effluent samples and one DI water sample was collected. (B1) Following the DI water flushing and air drying associated with Part A , soil
73 columns were loaded with 350 mL (7.62 cm ) of WW effluent per day, for seven consecutive days. During two of these days percolated effluent w ater w as collected from each column . E ffluents were collected from the University of Florida campus WW treatment facility and averaged 1 2 mg P L 1 . (B2) After 48 hours of air drying , the soi ls received 700 mL of DI water , and one sample was collected . For Part C, the soil columns were amended with an aluminum rich substrate, r eceived effluent water for seven days, and then flushed with DI water . In total two effluent samples and one DI water sample was collected. (C1) Following 48 hours of air drying (after Part B) the top 15 cm (6 in) of column soils w as augered out and homogenized with a relatively high rate of coarsely graded, aluminum rich drinking water treatment plant residual. The al uminum rich soil amendment s elected has been shown effective at reducing dissolved P concentrations in landscapes receiving manure, biosolid, and/or granular phosphate fertilizers (Agyin Birikorang et al, 2007; O'Connor, 2002; Silveira et al., 2006) . R esidual w as applied at 56 Mg ha 1 (equivalent to 2.5% of extracted surface soil mass) with elemental concentra tions assumed e quivalent to previously analyzed samples containing Al = 11 89 g kg 1 , Fe = 2 4 g kg 1 , and TP = 2.8 g kg 1 (Agyin Birikorang et al., 2009; Miyittah Kporgbe, 2004) . P contributions from the amendment to leached P w as expected to be of minimal significance as desorption from the residual has been identified a t <1 % (Makris, 2004) . Following the repacking of the soils, columns were loaded with 350 mL ( 7.62 cm ) of WW efflue nt per day for seven consecutive days. During two of these days percolated effluent w ater w as collected as samples. (C2) After 48 hours of air drying the soils were flushed with 700 mL of DI water, and one sample was collected.
74 In total, each field loca tion and condition (effluent irrigated or control) had twelve DI flush samples for part A, six effluent flush and three DI flush values for part B, and six effluent flush and three DI flush values for part C amended soils. Due to issues encountered with hi gh concentrations of unfilterable suspended colloidal solids and failure of one column for Field 2 , number of replicates were reduced for some fields. F or part A , a minimum of eight samples were collected, and for each part B and C a minimum of four efflue nt samples and two DI flush samples were collected. Data A nalysis Data was analyzed to highlight field specific and overall differences between soil parameters and P leaching rates in effluent irrigated and non irrigated soils. All soil concentration and SPSC calculations were determined from averages of three effluent irrigated soils and three control soils per field. To calculate the SPSC on a mass per area basis (kg ha 1 ), all average concentration values were multiplied by soil bulk density and depth o f the soil profile. T o determine the influence of long term effluent load i ng on the soils ability to sequester further added P , values were calculated by subtracting concentration of applied effluent ortho P from leachate outflow ortho P, divided by applied ortho P. Dif ferences of significance were determined by comparing concentrations of o rtho P l eaching from effluent irrigated soils to t hose from non irrigated soils using a one tailed T test. If the calculated T value is greater than the T c ritical value calculated at the 1% ( = 0.01) level of significance, differences w ere considered statistically significant. The null hypothesis states that there is no significant difference between concentrations generated from effluent irrigated soils and non effluent irrigated soils.
75 Dissolv ed P concentrations generated during DI leachate experiment s were extrapolated to la ndscape scale according to Eq . 4 1 to estimate total mass l eached per area per year. [ Kg of P leached per ha per year ] = [avg. concentration o f P leached ( mg L 1 ) ] * [height of water to percolate through soil depth per year] * Conversions ( 4 1) Conversions: (mg to kg), (L to m3), (m2 to ha) Field Scale Mass Balance Calculations Using Eq . 4 2 a field scale mass balance for P was conducted using soil analysis, experimental leaching results, and a number of assumptions related to P loading and removal over time. Assumptions were based on current and historical documentation, personal discussions with WW treatment plant personnel, and general WW quality trends. INPUTS = OUTPUTS (4 2) (A) = (B) + (C) + (D) Effluent applied = Soil accumulation + Plant Harvest + Leached (A) P inputs to the system were assumed to be completely associa ted with effluent loading (Eq . 4 A). Effluents were assumed to be homogenously applied at 2.54 cm per week ( 1 inch per week ) over the entire area and operational timespan. P concentrations were assumed consistent throughout the year b ut to decrease ove r time from 10 mg L 1 (1976 1986), 8 mg L 1 (1986 1996), 5 mg L 1 (1996 2006), to 2.5 mg L 1 (2006 2014); generating application rates ranging from 32 to 127 kg P ha 1 per year. Effluent P applied (kg P ha 1 per year) = Irrigation rate (m yr 1 ) * Effluent P conc. (mg P L 1 ) * Conversions (4 A)
76 Conversions for: (L to m3), (m2 to ha), (mg to kg) Fate of applied P was assumed associated with three pathways: (B) accumulation in M3P and WEP fractions of receiving soils, (C) uptake and removal in harvested plant biomass, or (D) leached from the soil profile. (B) Accumulation of P was considered the differences between concentrations of M3P and WEP in effluent irrigated and control soils, and was calculated by subtracting averaged effluent values from averaged control v alues at each field. The kg P ha 1 values were calculated by weighting analyte concentration s by depth increment s . (C) Plant harvest was assumed to consistently remove 35 kg P ha 1 per year , similar to other grassland P harvesting estimates (Elliott and Jais wal, 2011) . (D) Leaching P mass per area was calculated according to Eq. 4 1 . Leaching P rates were expected to increase over time and therefore three generalized time phases were constructed where 20% (Year 0 10), 50% (Year 10 20), and 100% (Year 20 38) of DI leaching values were a ssumed legitimate . Total P leached w as calculated assuming that 0.635 m (25 inches) of water per year percolate through soil depths ; e quivalent to 25% of total hydraulic loading in landscapes receiving 2. 54 cm per week of irrigat ed water and 127 cm per year of precipitation . Results and Discussion Results highlight soil findings and leaching output from the sampled l ocations . Following discussion on how the soil findings can influence P transport in the effluent i rrigated lands , the leachate results are presented to test hypotheses on differences between soil conditions. Leaching results are extrapolated to landscape scale, and
77 incorporated into field mass balances to determine potential influenc e that P leaching can have on w atershed loading m odel accuracy . Soil Bound Nutrients E ffluent receiving soils were substantially influenced by the long term loading as shown by the d ifference s in total mass of M3 P and WEP per area (Table 4 2) . Differences, calculated as effluent irrigated soils subtracted by control soils as higher or lower (negative) values, are separated by field and depths fractions of 0 0.15 m, 0 0.6 m, and 0 2.0 m (or max imum soil depth). Relative increases are separated by depth to highlight response s in: surface soils (0 0.15 m), maximum g rass rooting depth (0 0.6 m ), and maximum depth sampled (0 2.0 m). Fields 1, 2, and 3 reacted similarly to the long term effluen t loading. Field 4 also clearly showed an effluent loading influence, identifiable in higher M3P to 0.6 m and substantial ly higher WEP , to maximum depths sample d . Total influence on Field 4 soils is somewhat uncertain as control soil samples were (unknowin gly) located above naturally present P rich soils. Field 5 also exhibited higher M3P, indicative of over application of P , a nd was the only field not to display elevated WEP. Table 4 2 . Total difference s in M3P and WEP (kg ha 1 ) between effluent irrigated soi ls and paired control soils . Field 1 2 3 4 5 Mean Â± Stdev Analyte Dept h (m) kg ha 1 increase or ( ) decrease from effluent irrigation M3P 0 0.15 135 225 75 100 93 126 Â± 60 0 0.6 370 540 320 140 170 308 Â± 162 0 2.0 725 700 805 3457 200 608 Â± 275* WEP 0 0.15 3 4 3 9 1 4 Â± 3 0 0.6 15 10 14 44 0 17 Â± 16 0 2.0 75 15 39 134 1 52 Â± 54 *=excluding Field 4
78 Table 4 3 shows the percentage of the total elevated kg P per ha (identified in Table 4 2) present in upper horizons of 0 0.15 m and 0 0.6 m . Surface soil accum ulated P is associated with increased plant available resources and increased nutrien t runoff a nd leaching loss rate potential . Increases from 0 0.6 m indicates the vast majority of nutrient accumulation available to plant rooting systems . A ccumulation below 0. 6 m is considered inaccessible to p lant matter , but can accumulate as semi stable ( i.e., M3P under acidic conditions) or l eachable ( i.e., WEP) forms . On average, only 8% of the total elevated WEP was located in the upper 0 0.15 m, with only 31% found in the upper 0.6 m. Therefore 69% of the elevated WEP identified is occurring below typica l surface soil sampling depths. F rom t hese results it appears that the systems are reacting more similarly to effluent receiving rapid infiltration basins (Andres & Sims, 2013; Moura et al., 2011) than those receiving manures, biosolids, and other granular fertilizer applications. The trend in elevated M3P is more similar to other P loaded systems . A n average 26% of the elevation occurs in the upper 0.15 m, followed by substantial accumulation in soils directly below the su rface (0 0.6 m), h arboring 63% of total elevated M3P . As many landscape loading models assume that ~100% of WW appl ied P is accumulated and freely available for plant uptake in receiving surface soils , these findings indicate that actual accumulation and plant available a mounts c an be substantially lower than total load applied .
79 Tabl e 4 3 . Percentage of total M3P and WEP accumulated in surface horizons . Field 1 2 3 4 5 Mean Â± Stdev Analyte Depth (m) % of total increase M3P 0 0.15 18 30 9 45 26 Â± 16* 0 0.6 51 77 39 85 63 Â± 22* WEP 0 0.15 3 23 7 7 0 8 Â± 9 0 0.6 19 66 36 33 0 31 Â± 24 *=excluding Field 4 Leached Ort hophosphate Concentrations D issolved ortho P concentrations from DI water flushes , in all e xperiments , show ed significantly higher values fro m effluent irrigated soils when compared to control soils (Fig. 4 1) . In Fields 1 through 4, effluent irrigated soil ortho P concentrations were st atistically ( <0. 01) higher than paired control s , with Field 5 being the only site showing no significant difference s . Averag ing all values, the mg ortho P L 1 concentrations for Fields 1 to 5 were 1.0 Â± 0.3, 0.6 Â± 0.2, 2. 1 Â± 0.5, 2.5 Â± 0.7, and 0.1 Â± 0.1 , respectively. Leached c oncentrations matched well with the number of years under irrigation, with Fields 4, 3, 1 (irrigated for 38, 34, and 32 years, respectively) having the highest values , that decreased to Field 2 and 5 values, irrigated for 28 years (Table 3 1). Control soil o rtho P concentrations g enerated from individual e xperiments are not shown, but averaged values can be identified by the horizontal line s ranging from 0.01 0. 4 mg ortho P L 1 . Result s emphasize t h e validity of th e hypothesis , if soil is P saturated it will leach a greater concentration of orthophosphate than non saturated soils . Higher concentrations identified in some initial (Part A) flushes is possibly associated with the exposure of soil surfaces and loosely bound P generated from soil disturbance and transfer . As Styles (2006) showed that dry soils can extract two to three
80 times more dissolved P than wet soils, the higher a mounts of w ater saturation in Parts B and C can also partially explain the higher release rates identified in Part A . Leached concentrations from Part B (effluent applied to non amended soil ) and Part C (effluent appli ed to amended soil ) were found to not be statistically ( <0.01) different. Given these findings it is concluded that the soil a mendment did not reduce ortho P concentration in leachate. Some of the applied effluent P may have been sequestered to the a mendment particle surfaces , but as WEP concentrations were found to be substantially ele vated at depths much greater than 0.15 m , P transport is likely occurring well below amended depths. Incorporation of such soil amendments may help sequester future P loads applied and extend the time till P saturation is reached , but were contradictory to the stated hypothesis that if soil is P saturated, amendment of surface soils with an aluminum substrate will lower the concentration of orthophosphate leached during effluent loading . T he method was found to not be a satisfactory technique to decrease ongoing P leaching where effluent loaded subsurface horizons ha v e substantial increases in labile WEP . If the soil amendment approach is applied in developing landscapes utilizing reclaimed water irrigation, amendment of surface soils, to minimally 30 cm (1 2 in), prior to turf installation may help increase overall P sequestration. Incorporation of an Al, Fe, or OM rich soil amendment placed along receiving surface waters as subsurface permeable barrier systems have also been found effective at intercepting lateral subsurface flow (McCobb et al., 2003; Sibrell & Tucker, 2012) .
81 Figure 4 1. Orthophosphate leached from effluent irrigated and control soils for experiments A, B, and C. Uptake of Effluent Orthoph osphate Percolating Through Soils T he percentage change in effluent ortho P concentration after percolating through the soil column s for each field emphasizes that a significant ( <0.01) change has occurred in the dissolved P sequestration capabili ties of Fields 1 through 4 (Fig. 4 2) . Field 5 was the only field where no significant difference in up take efficiency was identified. Effluent was applied to columns with an average ortho P concentration of 1.2 Â± 0.58 mg L 1 (ran ge 0.1 2.1), and exited F ield 1 through 4 columns a t 0.6 2.3 mg L 1 , for an average increase of 16% Â± 75%. All control soils decreased effluent ortho P by an average 88% Â± 8%, equivalent to a leached concentration of 0.13 mg L 1 . Control soil uptake rates are similar to th ose identified in much of the literature for effluent receiving soils , generally on ly studied on limited time scales ( i.e., 1 5 years) .
82 Figure 4 2. Perce ntage change o f orthophosphate concentration after percolating through efflu ent irrigated and control soils . Spatiotemporal P hosphorus L eaching Assuming the leached concentrations can be scaled linearly, the P mass leaching from the bottom of a soil layer (ranging from 0.6 2.0 m) per year can be estimated by the volume of percolating water (Table 4 4) . Values were calculated according to Eq. 4 1 using average P concentrations ( Fig . 4 1 ) and under assumptions of a homogeneous landscape with consistent application prac tices. Total mass of P leached from each effluent irrigated soils ranged from 0.2 mg (Field 5) to 26.1 mg, with an average loss of 11.5 Â± 8.3 mg. Control soils, in contrast, released an average 1 .2 Â± 1.3 mg, and ranged from 0.1 7.3 mg. Values calculated for Field 5 and control s are similar to those identified by previous researchers in agricultural landscapes t hat
83 identified annual leaching rates of 0.5 1.9 kg P ha 1 (Culley et al., 1983) and 0.25 kg P ha 1 (Kirchmann, 1998) . G iven these substantial differences, it is concluded that long term effluent irrigated landscapes can potentially leach higher rates than non effluent irrigated and traditionally fertilized agricultural lands. Table 4 4 . Calculated p hosphorus mass leachi ng from effluent irrigated soils. 1 2 3 4 5 Mean Â± Stdev Percolating w ater (m / yr) P Leached (kg / ha * yr) 0.635 6.4 3.8 13.3 15.9 0.6 8.0 Â± 6.4 1.27 12.7 7.6 26.7 31.8 1.3 16.0 Â± 12.8 1.905 19.1 11.4 40.0 4 7.6 1.9 24.0 Â± 19.3 Note: Expected leaching from soil column depth of 0.6 2.0 m, depending on field. Many natural and managed spatiotemporal variables, that can significantly influence water flow and nutrient transport rates under in situ conditions, a re not considered in the extrapolation of these small scale experiment al results . Applying these findings to field scale should consider important influential site specific variables, such as: infiltration rate, soil porosity, antecedent moisture content, hydraulic conductiv ity, and preferential flow affected by clay layers, sinkholes, and fractured groundwater flow paths. Field Scale Phosphorus Mass Balance The field scale P mass balance (Table 4 5 ) was calculated according to Eq. 4 2, using soil P accumulation (Table 4 2 ) and expected leaching rate values (Table 4 4 ). Soil values, taken as M3P and WEP accumulat ed to soil depths sampled, accounted for an average 27% Â± 11% of the total effluent P applied. Leached P values were calculated according to the l owest cal culated percolat ion rate s (0.635 m/yr) , and accounted for an average 6.3% Â± 5.1% of total effluent P applied to the system . If rates of percolation under field conditions are higher than 0.635 m (25 in) per year leaching
8 4 rates w ould be e xpected to be higher . Unaccounted for P averaged 19% Â± 11% of a ll P applied . P in the unaccounted for fraction can possibly be accumulat ing bel ow sampled depths, lost during aeolian and surface runoff events, or present in soil layers sampled but in forms not extracted by the methods selected . The highest unaccounted for P was in Field 5, with 38%, and may b e attributed to higher rates of surface runoff at the field , given common soil saturation events that occur in the landscape with the shallow est restrict ive clay layer (0.7 1.15 m). Table 4 5 . Effluent s prayfield P mass balance over entire application timespan. Effluent Applied Soil Accumulation Plant Harvest Leached Unaccounted Field kg P ha 1 over operational timespan kg P ha 1 % 1 2413 800 1120 121 372 15 2 1905 715 980 57 153 8 3 2667 844 1190 280 353 13 4 3175 742 1330 397 706 22 5 1905 200 980 10 715 38 Physico chemical responses of the receiving system to long term effluent irrigation were clear in both the soil analysis and column leaching results. If receiving soils were reacting similarly to o ther systems major P accumulation would be expected to occur primarily in surface soils with substantial decrease in accumulation with depth. In addition, WEP trends would closely align with M3P, only increasing when soil P sequestration capabilities were saturated. Differences in P fate and transport can partially be explained through soil analysis findings ( Ch. 2 ) that highlighted shifts in characteristics associated with P transport, such a s increases in pH, from 6.2 Â± 0.4 to 7.3 Â± 0.4, and decreases in Al, from 564 to 465 mg kg 1 . The alterations are expected to
85 have decreased overall binding site capacity (reduced Al) and sorpti v e strength (rise in pH), ultimately i ncreasing P lability and transport, as evidenced by the leachate findings. Fields 1 through 4 leach ed P that was above concentrations of receiving surficial aquifers ( 0.025 0.275 mg L 1 ) (FDEP, 2010) , and critical val ues identified for shallow ground waters (Breeuwsma et al., 1995) and Florida streams (0.06 0.18 mg L 1 in panhandle and 0.3 0.5 mg L 1 in north central a nd west central FL) (FDEP, 2012b; USEPA, 1986) . The elevated P leaching rates can be most d etrimental in heavily loaded landscapes and/or watersheds with nutrient sensitive waters. In these landscapes regular monitoring and preventative, or remedial, actions should be taken to reduce accumulation and active transport of P. To ensure the long term sustainability of the practice numerous passive and active recommendations are suggested. Control and harvest of plant vegetation is a phytoremediation method that has been identified as the most inexpensive, yet highly effective, technique available (Marti nez et al., 2011) . Depending on the depth, and significance, of P accumulation, planting of deep rooted vegetation, such as pine and other tree species (Falkiner & Smith, 1997; Martinez et al., 2011) , can be effective at providing a long term, near surface, storage for applied N and P (Minogue et al., 2012; Stewart et al., 1990) and help increase uptake from lower soil layers (Chakraborty et al., 2011; Ibrikci et al., 1994; Nair & Graetz, 2004; Rechcigl et al., 1992) . As over 140,000 ha (350,000 acres ) o f effluent receiving soils in Florida are turf landscapes (FDEP, 2014) that only provide a temporary and rarely removed P storage , the incorporation of stable vegetative sinks c an assist in lon g term nutrient control.
86 Natural and human induced subsurface alterations are also potential methods of nutrient transport control. Fluctuating groundwater tables , associated with groundwater pumping and climatological conditions, have been shown to influence r eactivity and solubility of dissolved ortho P through change s in soil redox potential and/or exposure of previously submerged organic ( i.e., spodic) (Graetz & Nair, 1995; Harris et al., 1996) or Fe rich horizon s (Zurawsky et al., 2004) . In systems where aquife r levels are trending downwards horizons of high sequestration potential may be consistently held above groundwater tables and act as long term environmental bu ffers against percolating constituents (Graetz & Nair, 1995 ; Harris et al., 1996) . In these environments, the long term transport of P may be controlled naturally, without human interventi on, but require s specific knowledge on aquife r levels and subsurface horizon pedology . In effluent receiving landscapes at risk of nutrient over enrichment critical factors to incorporate in watershed management plans c an include the amount and timing of the nutrient supply in the reclaimed water and the uptake patterns of the receiving vegetation (Martinez et al., 2011) . As applications of effluent, even at low P concentrations (~1 mg L 1 ), have been found sufficient to meet nutritional needs for citrus (Parsons et al., 2001) , alfalfa, radish, tomato (Shahalam, Zahra, & Jaradat, 1998) , and turf (Fan et al., 2014; Marzolf, 2011; Morgan et al., 2008) , education of the public on t he irrigati on and fertilization n eeds when using reclaimed water can help avoid undesirable vegetative growth and groundwater contamination (Jacangelo et al., 2012; NRC, 1996) meter reclaimed water service or employ volume based rates (Jacangelo et al., 2012) , increased metering, installation of moisture sensors, and attention to poorly functioning irrigation systems ( that ove rspray
87 onto driveways, sidewalks, and roads) can help reduce excessive irrigation and associated P overloading (Arrington & Melton, 2010) . When applied at fertilization and irrigation rates constrained by agronomic limitations, receiving soils can act as effective long term environmental buffers. Studies in short term (8 24 month) experiments irrigating with high WW nutrient concentrations, but maintained at irrigation rates based on the ev aporation replenishment needs of bermudagrass, saw no accumulation to a 0.6 m soil depth while still meeting turf P needs (Geber, 2000; Heidarpour et al., 2007) . In a long term investigation, of golf courses applying hig hly treated WW (0.5 mg TP L 1 ) for 4 33 years, minimal P accumulation was also found (Qian et al., 2007) . In some long term effluent receiving landscap es groundwater concentrations have been identified as similar to background concentrations (Arrington & Dent, 2008; Overman & Leseman, 1982; Overman et al., 2003) , but may not be predictive of future trends under c ontinued application . While nearly all computational models, justifiably, assume that P loss is completely attributed to dissolved and particulate P in surface runoff (Tomer, James, & Isenhart, 2003) , these results highlight that leached P can be significant in systems with heavy effluent loading, high hydraulic conductivity , and minimal topographic relief. Incorporation of these findings into simulations can help i ncrease precision o f nutrient transport models, and ultimately improv e economic and environmental e ffectiveness of selected remedial and legislative decisions targeted at water quality improvement. Conclusions This research provides a more complete understanding of contami nant attenuation in environmental buffers receiving treated effluents, and highlights potential limitations in long term system operation. Results show that while the method of
88 reapplying WW to landscapes can directly improve surface water quality, it may affect groundwater quality over extended time frames. Since P concentrations in municipal effluents have decreased substantially over the irrigation time frame over application of P may be of lower concern in newly developed l andscapes , a nd validates further inv estigation into appropriate effluent concentration goals given the end use of the effluent. Conformational sampling may be justified in turf landscapes historically applying effluents for longer than 20 years, especially in impaired wa tersheds undergoing r estoration.
89 CHAPTER 5 SUMMARY AND CONCLUSIONS In their Guidelines for Water Reuse, the USEPA (2012) suggests remove nutrients during treatment for reuse depends on the intended use of the product water . Based on the findings in this study I suggest addition al language as follows, W hen land applied, effluent nutrient standards should be based on the sequestration capabilities of the receiving landscape with an acknowledgement that the receiving system characteristics can be spatially and temporally variable . If treated and reused according to location specific constraints, reclaimed WW effluent can act as a reliable , low cost source of water to help achieve legislative, social, and environmental water quality and resource goals (Cuthbert & Hajnosz, 1999; Rygaard, Binning, & Albrechtsen, 2011) . Objectives and research questions were addressed through field data collection and experimentation t hat highlighted the significant influence that long ter m loading has had on effluent irrigated systems . A number of soil characteristics were fo und to be significantly altered and showed to influence P leaching rates to a point that dissolved ortho P concentrations were well above many surface water quality limits . C hapter 2, term effluent irrigation on soil chemistry and phosphorus lability in Florida spra yfields , term irrigation with WW effluent has altered receiving soils and increased P labilit y up to three meters belo w surface soils . In addition to lower Al, c oncentrations of M3P, TP, Ca, NOx, and pH were all found higher in effluent irrigated soils . These variables are thought to control P storage and retention, and their changes (direction and magnitude) are clearly
90 consistent with accelerated P losses through the soil to the groundwater. T he soil chemistry shifts , associated with Al, M3P, TP, and pH, are suggested to strongly influence the increase in P lability, or WEP concentrations, identified to maximum sampled d epths in four of five fields. These factors are also thought to explain why WEP increased more substantial ly than M 3P , which was cont rary to the stated hypotheses . Findings emphasize that long term effluent irrigated soils react differently than m any other heavily P loaded systems when discussing P chemistry and transportability . M ajor findings from Chapter 3, predicting dissolved orthophosphate leaching fro m Florida effluent sprayfields , t hat the soil chemistry alterations associated with the long term effluent loading were suggested to influence the reliability of values commonly utilized to determine P leaching risk . Through analysis of surface and subsurface soils, and leachi ng experimentation, the most accurate P leaching predictors were determined to be PSR [for subsurface soils (r 2 =+0.80) and surface soils (r 2 =+0.72)], WEP [for surface soils (r 2 =+0.70) and subsurface soils (r 2 =+0.62)], and lastly STP (r 2 =+0.58). While all effluent irrigated surface soils were clearly influenced by loading, identified through the lower SPSC values, m any of the soils leach ed elevated P prior to reaching the expected change point . Due to the M3P, WEP, Al, and pH soil chemistry ch anges in the system, t he PSR value of 0.1 was found to overestimate the ability of the effluent receiving soil s to sequester additional P. I f PSR and SPSC methods are to be utilized in these landscapes to estimate leaching risk t he threshold PSR, or change point, value should be applied e xtremely conservative ly, at maximally, 0.05 . If SPSC is calculated using a
91 PSR threshold value of 0.1 , P leaching risk accuracy can be improved through an adjustment of SPSC values by + 575 kg ha 1 to 1.6 m depth, or + 100 kg h a 1 in surface soils (0 0.3 m). Chapter 4, term effluent irrigated sprayfields , show e d the impact that the extended loading has had on soil P storage and retention . Orthophosphate values measured in column experimentat ion i dentified that effluent irrigated soils were p ercolating significantly higher concentrations of ortho P than control soils when flushed with potable water . The ability of effluent loaded soils to further sequester applied P was also significantly reduced in four of five fields . In three of the four fields , leached P concentrations were even higher than the effluent applied . This value was especially drastic when compared to the removal efficiency of control soils that sequestered ~90% of applied P . Aluminum s oil enrichment to en hance P sequestration was not effective for decreasing P leaching. In these systems the saturation of surface soils and accumulation of highly mobile WEP throughout the soil profiles is thought to be the main factor in the limited success of the remedial technique. As soils inherently have a limite d capacity to sequester continuously applied P the incorporation of Al amendment into soils during development may a cost effective management option to temporarily extend the effective lifespan of receiving environmental buffers (Bixio et al., 2008; NRC, 2012) . All of the se findings can be applied by watershed managers to more accurately predict the long term fate and transport of P in Florida soils under effluent irrigation . To confirm results additional sampling and exp erimentation is s uggested to clarify the significance and extent of current, and future, impacts on soils and waters around the
92 state. In Florida there are currently over 25 sprayfields, numerous golf courses, and other turf landscapes that have been receiving efflu ent for over 25 years where the influence of long term loading on P leaching should be more thoroughly investigated. Investigations can provide answers to questions associated with the impact on receiving groundwater and surface water systems, as well as information on the forms, phases, and lability of P s pecies present . As systems are increasingly well understood these effluent P legacy loads , and associated increased P transport rate s , can be identified as potentially influential nutrient s ource contributors to watershed loading calculations. In this experiment many soils were extracted above groundwater tables so further information on the potential P sequestration beneath sampled depths can advance additionally advance understanding on long term fate and transport phenomena . Using these findings, t he accuracy of P transport models can be improved through alteration of variables and rates of transfer be tween soil, soil pore water, vegetation, and microbial biomass storages . The field sampling results can be used to further calibrate model parameters and provide information on level s of uncertainty in v arious effluent receiving environments. P arameter adjustments can build upon accepted values from the literature and P simulation models ( i.e., WAM, CREAMS, GLEAMS, HSPF ) to better pre dict and visualize long term P tran sport with effluent irrigation. The infl uence of effluent irrigated landscape P load contribution s can be especially significant in small storm water basins (i.e., golf courses , residential, and commercial properties) where effluent is typically applied. The shorter travel time, and infl uence of lateral flow into nearby surface water s may prove influential on system
93 performance over the long term , if effluent quality and irrigation rates are not appropriate for the receiving landscape. Accurate monitoring and incorporation of effluent ass ociated P loads into model s where large areas of the watershed are dedicated to effluent a pplication can be a valuable and cost effective manner t owards maintaining water quality, when compared to dealing with the complexities involved in widespread pollution sources (Sharpley, 1995) . One study of currently implemented water quality improvement plans identified that a majority are not delivering results in the time scales predicted (GAO, 2013) , r equiring signi ficant adjustment to achieve objectives . Results from this study are i n agreement , stating that , to better achieve long term water quality improvement, P watershed loading plans need to be based on multi decadal scale s . Benefits and Impediments to Increased Wastewater Reuse : quarter century, a recurring thesis in environmental and water resources engineering has been that improved WW treatment provides a treated effluent of such quality that it (Tchobanoglous, Burton, & Stensel, 2003) . With an American population that is expected to grow by over 50 percent between 2010 and 2060 (U.S. Census., 2010) , urban systems will increasingly i ncorporate WW reuse into water portfolio and water self sufficiency planning. As rates of treated municipal WW reuse are growing at an estimated rate of 15% per year (Miller, 2006), it can be important that municipalities and land owners are aware of the l ong term fate of effluent applied nutrients to maximize the value of the resource. Agricultural and landscape irrigation systems can benefit from th is (Kraus s & Page, 1997) and
94 reduced fertilization costs (USEPA, 2012) . As agricultural lands around the world are becoming increasingly reliant on reserves of limited inorganic fertilizer s to meet plant needs (Ayoub, 1999; Cordell, 2009) , recycling of nutrients will become increasingly important to meeting the goal of sustainable, closed loop horticultural systems. Further potential for efficient reuse in the United States is also p resent as many water treatment facilities will soon reach the end of their design life . As hundreds of billion s will need to be i nvested in WW infrastructure in the coming decades (USEPA, 2002) , the question becomes how best to allocate funds to meet long term societal and environmental water resource goals ? Historically upgrades primarily involve d updates to ce ntralized treatment facilities , but as issues related to increasing energy expense, growing and expanding urban populations, and increasing knowledge of aquatic ecosystem sensitivity and greenhouse gas emissions impacts (Daigger, 2009; West, 2001) , a strong justification can be made to minimize piping and pumpi ng costs through multiple smaller, decentralized treatment operations (van Roon, 2007) . While centralized treatment systems produce only a single quality effluent, decentralized systems can meet a varie ty of constituent standards (NRC, 2012) and meet specific, end use based , effluent qualities . With advances in membrane bioreactor and sequencing batch reactor systems (Asano et al., 2007; Vuono et al., 2013) , nanotechnology, and microbial genetic modification, small scale operations have been shown capable of providing effluent water of varying quality under flexible, changing conditions, and thus tailorable to the standards needed of the vegetated landscape (Sevostianova & Leinauer, 2014) or client.
95 To maximize the value of the nutrient resource, while minimizing environmental impact, appropriate treatment of the effluent is necessary. Since secondary treatment effluent standards were set to protect surface waters directly receiving municipal discharges, their appropriateness when applied to landscapes is questiona ble . Due to the alternative reuse options now available for WW, site specific acceptable effluent concentrations should be based on the final purpose, receiving landscape, and receiving water body sensitivity, as a method to save time, money, energy, and r educe unnecessary overtreatment. This o of effluent can strongly influence WWTP capital and operating costs, energy usage, greenhouse gas emissions, and the necessity for expensive advanced treatment upgrades, while not significantly improving water quality o r protecting human health (Lundie, Peters, & Beavis, 2004; Schimmoller & Kealy, 2014; USEPA, 2012) provides for effective disease contro l and environmental quality (NRC, 2012) , where appropriate, exemptions from stringent ( i.e., tertiary) nutrient standards can help reduce spending and minimize supplemental landscape fertilization needs (Fan et al., 2014) . Another factor of consideration in appropriate treatment are the many biological, chemical, and physical nutrient removal mechanisms that occur naturally through reclaimed water distribution systems. In one analysis, at Loxahatchee River District, concentrat ions dropped from 10 to 1 mg TN L 1 and 1 to 0.1 mg TP L 1 traveling from the point of effluent discharge to the sprinkler head (Arringto n, 2012) . R aising allowable effluent nutrient concentrations can also increase plant growth (Ben Gal & Dudley, 2003) a nd reduce, or completely eliminate, the need for additional fertilization (Fan et
96 al., 2014; Pedrero et al., 2010; Sevostianova & Leinauer, 2014) . I n lieu of these considerations appropriate effluent standards can be critical to increasing su stainability of WW reuse operations. Incorporating future global concerns of P resource limitations (Cordell , 20 09 ) , and an increasing need to dispose of nutrient rich discharges in ways that d o not negatively impact surface waters, w hen managed appropriately , the reuse of w astewater for irrigation can b e seen as a long term practice to increase global food security and reduce undesirable consequences to human and environmental systems.
97 LIST OF REFERENCES Adamski, J. C., & German, E. R. (2004). Hydrogeology and quality of ground water in orange county, F lorida . United States Geological Survey (USGS). Water Resources Inves tigations Report No. 03 4257. Agyin Birikorang, S., O'Connor, G., & Brinton, S. (2008). Evaluating phosphorus loss from a F lorida spodosol as affected by phosphorus source application methods. Journal of Environmental Quality, 37 (3), 1180 1189. Agyin Bir residuals to control phosphorus in soils. University of Florida Extension Letter SL, 300 . Agyin Birikorang, S., O'Connor, G. A., Jacobs, L. W., Makris, K. C., & Brinton, S. R. (2007). Long term phosphorus immobilization by a drinking water treatment residual. Journal of Environmental Quality, 36 (1), 316 323. Agyin Birikorang, S., O'Connor, G. A., & Obreza, T. A. (2009). Are alum based drinking water treatment residuals safe for land application? University of Florida Extension Letter SL, 299. Allen, S. C., Nair, V. D., Graetz, D. A., Jose, S., & Ramachandran Nair, P. (2006). Phosphorus loss from organic versus inorganic fertilizers used in alleycropping on a florida ultisol. Ag riculture, Ecosystems & Environment, 117 (4), 290 298. Alleoni, L. R., Brinton, S. R., & O'Connor, G. A. (2008). Runoff and leachate losses of phosphorus in a sandy spodosol amended with biosolids. Journal of Environmental Quality, 37 (1), 259 265. Allhand s, M., & Overman, A. (1995). Forage grass response to harvest interval and reclaimed water. Journal of Plant Nutrition, 18 (12), 2723 2737. Alva, A. (1993). Comparison of Mehlich 3, Mehlich 1, ammonium bicarbonate DTPA, 1.0 M ammonium acetate, and 0.2 M am monium chloride for extraction of calcium, magnesium, phosphorus, and potassium for a wide range of soils 1. Communications in Soil Science & Plant Analysis, 24 (7 8), 603 612. Andres, A., & Sims, J. T. (2013). Assessing potential impacts of a wastewater r apid infiltration basin system on groundwater quality: A Delaware case study. Journal of Environmental Quality, 42 (2), 391 404. Angelakis, A., Do Monte, M. M., Bontoux, L., & Asano, T. (1999). The status of wastewater reuse practice in the Mediterranean b asin: Need for guidelines. Water Research, 33 (10), 2201 2217.
98 Arias, C., Del Bubba, M., & Brix, H. (2001). Phosphorus removal by sands for use as media in subsurface flow constructed reed beds. Water Research, 35 (5), 1159 1168. Arrington, D. A., & Dent, R. C. (2008). Maintenance of groundwater quality throughout twenty years of reuse. Water Practice, 2 (3), 1 10. Arrington, D. (2012). Nutrient cycling in a reuse distribution system significantly lowers landscape irrigation nutrient loading estimates. Loxa hatchee River District, Jupiter, FL, USA, Arrington, D., & Melton, K. (September, 2010). Unintended consequences: Numeric nutrient criteria will constrain reuse opportunities. 25th Annual WateReuse Symposium, Washington, D.C. Asano, T., Burton, F., Lever enz, H., Tsuchihashi, R., & Tchobanoglous, G. (2007). Water reuse: Issues, technologies, and applications. McGraw Hill New York. Ayoub, A. T. (1999). Fertilizers and the environment. Nutrient Cycling in Agroecosystems, 55 (2), 117 121. Bahri, A. (1998). Fe rtilizing value and polluting load of reclaimed water in tunisia. Water Research, 32 (11), 3484 3489. Barrow, N. (1983). A mechanistic model for describing the sorption and desorption of phosphate by soil. Journal of Soil Science, 34 (4), 733 750. Barton, L., Schipper, L., Barkle, G., McLeod, M., Speir, T., Taylor, M., et al. (2005). Land application of domestic effluent onto four soil types. Journal of Environmental Quality, 34 (2), 635 643. Ben Gal, A., & Dudley, L. M. (2003). Phosphorus availability unde r continuous point source irrigation. Soil Science Society of America Journal, 67 (5), 1449 1456. Bixio, D., Thoeye, C., Wintgens, T., Ravazzini, A., Miska, V., Muston, M., et al. (2008). Water reclamation and reuse: Implementation and management issues. D esalination, 218 (1), 13 23. Bouwer, H., & Idelovitch, E. (1987). Quality requirements for irrigation with sewage water. Journal of Irrigation and Drainage Engineering, 113 (4), 516 535. Breeuwsma, A., Reijerink, J., & Schoumans, O. (1995). Impact of manur e on accumulation and leaching of phosphate in areas of intensive livestock farming. Animal Waste and the Land Water Interface, S, 239 249.
99 Burgoa, B., Bottcher, A., Mansell, R., & Allen Jr, L. (1991). Distributions of residual soil phosphorus along trans ects for three dairies in Okeechobee County , Florida . Proceedings Soil and Crop S cience Society of Florida (USA). Chakraborty, D., Nair, V., Harris, W., & Rhue, R. (2011). The potential for plants to remove phosphorus from the spodic horizon. University o f Florida IFAS, SL 359 . Cho, C. (1991). Phosphate transport in calcium saturated systems: I. theory. Soil Science Society of America Journal, 55 (5), 1275 1281. Chrysostome, M., Nair, V., Harris, W., & Rhue, R. (2007a). Laboratory validation of soil phosp horus storage capacity predictions for use in risk assessment. Soil Science Society of America Journal, 71 (5), 1564 1569. Chrysostome, M., Nair, V. D., Harris, W. G., & Rhue, R. D. (2007b). Minimizing confounding factors in phosphorus leaching assessment for Dairy and Poultry Manure Amended soils. Communications in Soil Science and Plant Analysis, 38 (7 8), 975 987. Church, C. D., Kleinman, P. J., Bryant, R. B., Saporito, L. S., & Allen, A. L. (2010). Occurrence of arsenic and phosphorus in ditch flow from litter amended soils and barn areas. Journal of Environmental Quality, 39 (6), 2080 2088. Conley, D. J., Paerl, H. W., Howarth, R. W., Boesch, D. F., Seitzinger, S. P., Havens, K. E., et al. (2009). Controlling eutrophication: Nitrogen and phosphorus. Sci ence, 323 (5917), 1014 1015. Cordell, D. (2009). The story of phosphorus: global food security and food for thought. Global Environmental Change , 19(2), 292 305. Crites, R. W., Esmay, J., Griffes, D., Olson, J., Powers, J., Shedden, R., et al. (1977). Proc ess design manual for land treatment of municipal wastewater . EPA 625/1 77 008, 1 3. U.S. Environmental Protection Agency. Cuthbert, R. W., & Hajnosz, A. M. (1999). Setting reclaimed water rates. Journal American Water Works Association, 91 (8), 50 57. Da igger, G. T. (2009). Evolving urban water and residuals management paradigms: Water reclamation and reuse, decentralization, and resource recovery. Water Environment Research, 81 (8), 809 823. DeWiest, D. R., & Livingston, E. H. (1999). The Florida stormwa ter, erosion, and sedimentation control inspector's manual . Florida Department of Environmental Protection, Stormwater/Nonpoint Source Management Section.
100 Dobermann, A., Witt, C., & Dawe, D. (2004). Increasing productivity of intensive rice systems throug h site specific nutrient management . Int. Rice Res. Inst. Eckert, D., & Watson, M. (1996). Integrating the mehlich 3 extractant into existing soil test interpretation schemes 1. Communications in Soil Science & Plant Analysis, 27 (5 8), 1237 1249. Eghball, B., Binford, G., & Baltensperger, D. D. (1996). Phosphorus movement and adsorption in a soil receiving long term manure and fertilizer application. Journal of Environmental Quality, 25 (6), 1339 1343. Elliott, H., Brandt, R., Kleinman, P., Sharpley, A., & Beegle, D. (2006). Estimating source coefficients for phosphorus site indices. Journal of Environmental Quality, 35 (6), 2195 2201. Elliott, H., & Jaiswal, D. (2011). Phosphorus management for sustainable agricultural irrigation of reclaimed water. Journal of Environmental Engineering, 138 (3), 367 374. Elliott, H., O'Connor, G., & Brinton, S. (2002). Phosphorus leac hing from biosolids amended sandy soils. Journal of Environmental Quality, 31 (2), 681 689. Elrashidi, M., Alva, A., Huang, Y., Calvert, D., Obreza, T., & He, Z. (2001). Accumulation and downward transport of phosphorus in Florida soils and relationship to water quality. Communications in Soil Science and Plant Analysis, 32 (19 20), 3099 3120. Evanylo, G., Ervin, E., & Zhang, X. (2010). Reclaimed water for turfgrass irrigation. Water, 2 (3), 685 701. Falkiner, R., & Polglase, P. (1997). Transport of phospho rus through soil in an effluent irrigated tree plantation. Australian Journal of Soil Research, 35 (2), 385 397. Falkiner, R., & Smith, C. (1997). Changes in soil chemistry in effluent irrigated pinus radiata and eucalyptus grandis plantations. Australian Journal of Soil Research, 35 (1), 131 147. Fan, J., Hochmuth, G., Kruse, J., & Sartain, J. (2014). Effects of reclaimed water irrigation on growth and nitrogen uptake of turfgrass. Hort Technology, 24 (5), 565 574. FDEP. (2012a). 2011 reuse inventory. Flor ida Department of Environmental Protection, Tallahassee FL . FDEP. (2012b). . Florida Department of Environme ntal Protection, Tallahassee FL.
101 FDEP. (2014). 2013 reuse inventory . Florida Departme nt of Environmental Protection , Tallahassee, FL . FDEP, Harrington, D., Maddox, G., & Hicks, R. (2010). Florida springs initiative monitoring network report and recognized sources of nitrate. Florida Department of Environmental Protection , Division of Envi ronmental Assessment and Restoration . Feigin, A., Ravina, I., & Shalhevet, J. (1991). Irrigation with treated sewage effluent. Springer . Florida Association of Counties ( 2015 ). 2015 legislative agenda. Fox, P. (2002). Soil aquifer treatment: An assessme nt of sustaina bility. Rotterdam, The Netherlands: AA Balkema Publishers. GAO. (2013). Clean water act: Changes needed if key EPA progra m is to help fulfill the nation's water quality goals. GAO 14 80. Geber, U. (2000). Nutrient removal by grasses irrigat ed with wastewater and nitrogen balance for reed canarygrass. Journal of Environmental Quality, 29 (2), 398 406. Gerritse, R. (1996). Transport times of dissolved inorganic phosphate in soils. Journal of Environmental Quality, 25 (1), 107 110. Google Earth . (2015). Florida map . www.maps.google.com Graetz, D., & Nair, V. (1995). Fate of phosphorus in Florida spodosols contaminated with cattle manure. Ecological Engineering, 5 (2), 163 181. Greenberg, A. E., & Thomas, J. F. (1954). Sewage effluent reclamation for industrial and agricultural use. Sewage and Industrial Wastes, 761 770. Harris, W., Rhue, R., Kidder, G., Brown, R., & Littell, R. (1996). Phosphorus retention as related to morphology of sandy coastal plain soil materials. Soil Science Society of Am erica Journal, 60 (5), 1513 1521. Harris, W. (2015). Soil characteristics in Florida landscapes. Personal communication. Hayes, A., Mancino, C., & Pepper, I. (1990). Irrigation of turfgrass with secondary sewage effluent: I. soil and leachate water quality . Agronomy Journal, 82 (5), 939 943. He, Z., Alva, A., Li, Y., Calvert, D., & Banks, D. (1999). Sorption desorption and solution concentration of phosphorus in a fertilized sandy soil. Journal of Environmental Quality, 28 (6), 1804 1810.
102 Heckrath, G., Broo kes, P., Poulton, P., & Goulding, K. (1995). Phosphorus leaching from soils containing different phosphorus concentrations in the broadbalk experiment. Journal of Environmental Quality, 24 (5), 904 910. Heidarpour, M., Mostafazadeh Fard, B., Koupai, J. A., & Malekian, R. (2007). The effects of treated wastewater on soil chemical properties using subsurface and surface irrigation methods. Agricultural Water Management, 90 (1), 87 94. Hesketh, N., & Brookes, P. (2000). Development of an indicator for risk of phosphorus leaching. Journal of Environmental Quality, 29 (1), 105 110. Holloway, R. E., Bertrand, I., Frischke, A., Brace, D., McLaughlin, M. J., & Shepperd, W. (2001). Improving fertiliser efficiency on calcareous and alkaline soils with fluid sources of P, N and Z n. Plant and Soil, 236 (2), 209 219. Hooda, P., Rendell, A., Edwards, A., Withers, P., Aitken, M., & Truesdale, V. (2000). Relating soil phosphorus indices to potential phosphorus release to water. Journal of Environmental Quality, 29 (4), 1166 1 171. Hu, C., Zhang, T., Kendrick, D., Huang, Y., Dahab, M., & Surampalli, R. (2006). Muskegon wastewater land treatment system: Fate and transport of phosphorus in soils and life expectancy of the system. Engineering in Life Sciences, 6 (1), 17 25. Hu, C. , Zhang, T. C., Huang, Y. H., Dahab, M. F., & Surampalli, R. (2005). Effects of long term wastewater application on chemical properties and phosphorus adsorption capacity in soils of a wastewater land treatment system. Environmental Science & Technology, 3 9 (18), 7240 7245. Ibrikci, H., Comerford, N., Hanlon, E., & Rechcigl, J. (1994). Phosphorus uptake by bahiagrass from spodosols: Modeling of uptake from different horizons. Soil Science Society of America Journal, 58 (1), 139 143. Isensee, A., & Walsh, L. (1972). Influence of banded fertiliser on the chemical environment surrounding the band. II. effect on soil solution cations, cation anion balance and solution phosphorus. Journal of the Science of Food and Agriculture, 23 (4), 509 516. Jacangelo, J. G., Oppenheimer, J., Badruzzaman, M., Pinzon, J., Contreras, A., & Waller, P. (2012). Development of markers to identify nutrient sources impacting . WateReuse Foundation, Project Number WRF 09 08 . Jaiswal, D. (2010). Soil phosph orus dynamics in a sprinkler irrigation system for land application of municipal wastewater effluent. PhD Dissertation, The Pennsylvania S tate University, Univ. Park, PA .
103 Kalavrouziotis, I., Koukoulakis, P., Sakellariou Makrantonaki, M., & Papanikolaou, C . (2009). Effects of treated municipal wastewater on the essential nutrient interactions in the plant of brassica oleracea var. italica. Desalination, 242 (1), 297 312. Kardos, L., & Hook, J. (1976). Phosphorus balance in sewage effluent treated soils. Jou rnal of Environmental Quality, 5 (1), 87 90. Kim, S. M., Park, S. W., Lee, J. J., Benham, B. L., & Kim, H. K. (2007). Modeling and assessing the impact of reclaimed wastewater irrigation on the nutrient loads from an agricultural watershed containing rice paddy fields. Journal of Environmental Science and Health Part A, 42 (3), 305 315. Kirchmann, H. (1998). Phosphorus flows in Swedish society related to agriculture. Kungl.Skogs Och Lantbruksakademiess Tidskrift, 137 (7), 145 156. Kiziloglu, F., Turan, M., Sahin, U., Kuslu, Y., & Dursun, A. (2008). Effects of untreated and treated wastewater irrigation on some chemical properties of cauliflower (brassica olerecea L. var. botrytis) and red cabbage (brassica olerecea L. var. rubra) grown on calcareous soil in turkey. Agricultural Water Management, 95 (6), 716 724. Kleinman, P. J., Bryant, R. B., Reid, W. S., Sharpley, A. N., & Pimentel, D. (2000). Using soil phosphorus behavior to identify environmental thresholds. Soil Science, 165 (12), 943 950. Kleinman, P. J., & Sharpley, A. N. (2002). Estimating soil phosphorus sorption saturation from mehlich 3 data. Communications in Soil Science and Plant Analysis, 33 (11 12), 1825 1839. Krauss, G. D., & Page, A. L. (1997). Wastewater, sludge and food crops. BioCycle (US A). Lazarova, V., & Bahri, A. (2004). Water reuse for irrigation: Agriculture, landscapes, and turf grass . Boca Raton, FL: CRC Press. Leytem, A., Sims, J., & Coale, F. (2004). Determination of phosphorus source coefficients for organic phosphorus sources . Journal of Environmental Quality, 33 (1), 380 388. Lin, C., Eshel, G., Roehl, K., Negev, I., Greenwald, D., Shachar, Y., et al. (2006). Studies of P accumulation in soil/sediment profiles used for large scale wastewater reclamation. Soil use and Management, 22 (2), 143 150.
104 Lin, C., & Banin, A. (2005). Effect of long term effluent recharge on phosphate sorption by soils in a wastewater reclamation plant. Water, Air, and Soil Pollution, 164 (1 4), 257 273. Lombi, E., McLaughlin, M. J., Johnston, C., Armstrong, R. D., & Holloway, R. E. (2004). Mobility and lability of phosphorus from granular and fluid monoammonium phosphate differs in a calcareous soil. Soil Science Society of America Journal, 68 (2), 682 6 89. Lu, P., & O'Connor, G. A. (2001). Biosolids effects on phosphorus retention and release in some sandy florida soils. Journal of Environmental Quality, 30 (3), 1059 1063. Lundie, S., Peters, G. M., & Beavis, P. C. (2004). Life cycle assessment for sust ainable metropolitan water systems planning. Environmental Science & Technology, 38 (13), 3465 3473. Maguire, R., & Sims, J. (2002). Measuring agronomic and environmental soil phosphorus saturation and predicting phosphorus leaching with Mehlich 3. Soil Sc ience Society of America Journal, 66 (6), 2033 2039. Makris, K. C. (2004). Long term stability of sorbed phosphorus by drinking water treatment residuals: Mechanisms and implications. PhD Dissertation, Univ. Fl ., Gainesville, Fl. Mann, K. K., Schumann, A. W., Obreza, T. A., Harris, W. G., & Sartain, J. B. (2012). How to characterize soil variability in Florida citrus groves as it relates to tree growth and yield. University of Florida IFAS, SL556. Mansell, R., Bloom, S., & Burgoa, B. (1991). Phosphorous tr ansport with water flow in acid, sandy soils. Transport processes in porous media (pp. 271 314). Kluwer Academic Publ., Dordrecht, the Netherlands: Springer. Martinez, C. J., Clark, M. W., Toor, G. S., Hochmuth, G. J., & Parsons, L. R. (2011). Accounting for the nutrients in reclaimed water for landscape irrigation. University of Florida IFAS, AE479 . Marzolf, E. (2011). Is reclaimed water a free drink? or are there hidden costs? St. J ohns Water Management District. Maurer, M. A. (1994). Microsprinkler ir rigation of mature, reset, and young grapefruit (citrus paradisi MACF.) tr ees with reclaimed wastewater. PhD Diss ertation, University of Florida . McCobb, T. D., LeBLANC, D. R., Walter, D. A., Hess, K. M., Kent, D. B., & Smith, R. L. (2003). Phosphorus in a ground water contaminant plume discharging to ashumet
105 pond, Cape Cod , Massachusetts , US Department of the Interior, US Geological Survey , Reston, VA. McDowell, R. W., & Sharpley, A. N. (2001). Soil phosphorus fractions in solution: Influence of fertilis er and manure, filtration and method of determination. Chemosphere, 45 (6), 737 748. McDowell, R., & Sharpley, A. (2001). Approximating phosphorus release from soils to surface runoff and subsurface drainage. Journal of Environmental Quality, 30 (2), 508 52 0. Medema, G., & Stuyfzand, P. (2002). Removal of micro organisms upon basin recharge, deep well injection and river bank filtration in the netherlands. Management of aquifer recharge for sustainability proceedings of the fourth international symposium on artificial recharge (pp. 125 131). Lisse: A. A. Belkema Publishers. Meinikmann, K., Lewandowski, J., & Hupfer, M. (2015). Phosphorus in groundwater discharge A potential source for lake eutrophication. Journal of Hydrology, 524 , 214 226. Menzies, N., Sk ilton, J., & Guppy, C. (1999). Phosphorus storage on effluent irrigated land. Journal of Environmental Quality, 28 (3), 750 754. Miller, G. W. (2006). Integrated concepts in water reuse: Managing global water needs. Desalination, 187 (1), 65 75. Minogue, P . J., Miwa, M., Rockwood, D. L., & Mackowiak, C. L. (2012). Removal of nitrogen and phosphorus by eucalyptus and populus at a tertiary treated municipal wastewater sprayfield. International Journal of Phytoremediation, 14 (10), 1010 1023. Miyittah Kporgbe, M. (2004). Phosphorus immobolization in manure impacted soil with alum inum based drinking water treat ment residual. University of Florida Thesis. Morgan, K. T., Wheaton, T. A., Parsons, L. R., & Castle, W. S. (2008). Effects of reclaimed municipal waste w ater on horticultural characteristics, fruit quality, and soil and leaf mineral concentration of citrus. HortScience, 43 (2), 459 464. Moura, D. R. (2009). Phosphorus fate and transport in wastewater applied to rapid infiltratio n basins. Master of Science, University of Florida Thesis . Moura, D. R., Silveira, M. L., O'Connor, G. A., & Wise, W. R. (2011). Long term reclaimed water application effects on phosphorus leaching potential in rapid infiltration basins. Journal of Environmental Monitoring, 13 (9), 2 457 2462.
106 Nair, V., & Harris, W. (2011). Phosphorus sources and risk potential: Organic and inorganic fertilizers. University of Florida IFAS, SL358 . Nair, V. D., Chrysostome, M., & Harris, W. G. (2011). The long term contribution of phosphorus from agri cultural lands to Lake Okeechobee . University of Florida IFAS, SL357 . Nair, V., & Graetz, D. (2002). Phosphorus saturation in spodosols impacted by manure. Journal of Environmental Quality, 31 (4), 1279 1285. Nair, V., & Graetz, D. (2004). Agroforestry as an approach to minimizing nutrient loss from heavily fertilized soils: The Florida experience. New vistas in agroforestry (pp. 269 279) Springer. Nair, V., Graetz, D., & Portier, K. (1995). Forms of phosphorus in soil profiles from dairies of south Flori da . Soil Science Society of America Journal, 59 (5), 1244 1249. Nair, V., & Harris, W. (2004). A capacity factor as an alternative to soil test phosphorus in phosphorus risk assessment. New Zealand Journal of Agricultural Research, 47 (4), 491 497. Nair, V ., Portier, K., Graetz, D., & Walker, M. (2004). An environmental threshold for degree of phosphorus saturation in sandy soils. Journal of Environmental Quality, 33 (1), 107 113. Nair, V. D., Harris, W. G., & Chakraborty, D. (2010). An indicator for risk o f phosphorus loss from sandy soils. University of Florida IFAS, SL333. Nair, V. D., Harris, W. G., Chakraborty, D., & Chrysostome, M. (2010). Understanding soil phosphorus storage capacity. University of Florida IFAS, SL336. Nash, D. M., & Halliwell, D. J. (2000). Tracing phosphorous transferred from grazing land to water. Water Research, 34 (7), 1975 1985. National Research Council ( NRC ). Committee on the Use of Treated Municipal Wastewater Effluents, & Sludge in the Production of Crops for Human Consumpti on. (1996). Use of reclaimed water and sludge in food crop production Natl Academy Pr. National Research Council & Committee on the Assessment of Water Reuse as an Approach for Meeting Future Water Supply Needs. (2012). Water reuse: Potential for expandin g the nation's water supply through reuse of municipal wastewater National Academies Press.
107 Nelson, N. O., Parsons, J. E., & Mikkelsen, R. L. (2005). Field scale evaluation of phosphorus leaching in acid sandy soils receiving swine waste. Journal of Envir onmental Quality, 34 (6), 2024 2035. O'Connor, G., Brinton, S., & Silveira, M. (2005). Evaluation and selection of soil amendments for field testing to reduce P losses. Paper presented at the Soil and Crop Science Society of Florida, Proceedings, (64) . O' Connor, G., Elliott, H., Basta, N., Bastian, R., Pierzynski, G., Sims, R., et al. (2005). Sustainable land application. Journal of Environmental Quality, 34 (1), 7 17. O'Connor, G., Elliott, H., & Bastian, R. (2008). Degraded water reuse: An overview. Jour nal of Environmental Quality, 37 (5_Supplement), S 157 S 168. O'Connor, G. A. (2002). Characterizing forms, solubilities, bioavailabilities, and mineralization rates of phosphorus in biosolids, commercial fertilizers, and manures (phase I) . Water Environme nt Research Foundation. Oladeji, O. O. (2006). Management of phosphorus sources and water treatment residuals (WTR) for environ mental and agronomic benefits. PhD Thesis, University of F lorida, Gainesville, Florida. p.271. Overman, A. R. (1979). Wastewate r irrigation at Tallahassee, Florida. EPA Report No. 600/2 79 151, 319p. U.S. Environmental Protection Agency. Overman, A. R., & Leseman, W. G. (1982). Soil and groundwater changes under land treatment of wastewater. Transactions of the American Society of Agricultural Engineers, 381 387. Overman, A. R., Scholtz II I, R. V., & Brock, K. H. ( 2003). Long term performance of Tallahassee southeast farm . Project No. 88185 C. Gainesville, FL 32611 0570: Agricultural and Biological Engineering Department, Unive rsity of Florida. Parfitt, R. L., Atkinson, R. J., & Smart, R. S. C. (1975). The mechanism of phosphate fixation by iron oxides. Soil Science Society of America Journal, 39 (5), 837 841. Parsons, L., Morgan, K., Wheaton, T., & Castle, W. (2001). Wastewate r and reclaimed water disposal problem of potential resource? Eurasian Soil Sci, 27 (12), 58 67. Pautler, M. C., & Sims, J. T. (2000). Relationships between soil test phosphorus, soluble phosphorus, and phosphorus saturation in Delaware soils. Soil Scien ce Society of America Journal, 64 (2), 765 773. Pedrero, F., Kalavrouziotis, I., AlarcÃ³n, J. J., Koukoulakis, P., & Asano, T. (2010). Use of treated municipal wastewater in irrigated agriculture Review of some practices in Spain and Greece . Agricultural Wa ter Management, 97 (9), 1233 1241.
108 Perry, T. C. (1991). Irrigation of citrus with reclaimed water. Florida Departme nt of Environmental Regulation. Pescod, M. B., & Arar, A. (1988). Treatment and use of sewage effluent for irrigation: Proceedings of the FAO regional seminar on the treatment and use of sewage effluent for irrigation held in Nicosia , Cyprus , 7 9 October , 1985 Elsevier. Preedy, N., McTiernan, K., Matthews, R., Heathwaite, L., & Haygarth, P. (2001). Rapid incidental phosphorus transfers from gr assland. Journal of Environmental Quality, 30 (6), 2105 2112. Qian, Y., Migliaccio, K., Wan, Y., Li, Y., & Chin, D. (2007). Seasonality of selected surface water constituents in the Indian River lagoon, Florida . Journal of Environmental Quality, 36 (2), 416 425. Rajan, S. (1975). Adsorption of divalent phosphate on hydrous aluminium oxide. Nature, (253), 434 436. Rechcigl, J., Payne, G., Bottcher, A., & Porter, P. (1992). Reduced phosphorus application on bahiagrass and water quality. Agronomy Journal, 84 ( 3), 463 468. Robinson, J. S., & Sharpley, A. N. (1996). Reaction in soil of phosphorus released from poultry litter. Soil Science Society of America Journal, 60 (5), 1583 1588. Rosabal, A., Morillo, E., Undabeytia, T., Maqueda, C., Justo, A., & Herencia, J. F. (2007). Long term impacts of wastewater irrigation on Cuban soils. Soil Science Society of America Journal, 71 (4), 1292 1298. Ryden, J. C., & Pratt, P. F. (1980). Phosphorus removal from wastewater applied to land. Hilgardia, (48), 1 36. Rygaard, M ., Binning, P. J., & Albrechtsen, H. (2011). Increasing urban water self sufficiency: New era, new challenges. Journal of Environmental Management, 92 (1), 185 194. Schimmoller, L., & Kealy, M. J. (2014). Fit for purpose water: The cost of overtreating rec laimed water . No. WRRF 10 01 , WateReuse Foundation. Schindler, D. W. (1977). Evolution of phosphorus limitation in lakes. Science, 195 (4275), 260 262. Schoumans, O., & Groenendijk, P. (2000). Modeling soil phosphorus levels and phosphorus leaching from a gricultural land in the Netherlands . Journal of Environmental Quality, 29 (1), 111 116.
109 Scott, C. A., Faruqui, N. I., & Raschid Sally, L. (2004). Wastewater use in irrigated agriculture: Confronting the livelihood and environmental realities . Wallingford, England: CABI. Sevostianova, E., & Leinauer, B. (2014). Subsurface applied tailored water: Combining nutrient benefits with efficient turfgrass irrigation. Crop Science, 54 (5), 1926 1938. Shahalam, A., Zahra, B. M. A., & Jaradat, A. (1998). Wastewater ir rigation effect on soil, crop and environment: A pilot scale study at Irbid , Jordan . Water, Air, and Soil Pollution, 106 (3 4), 425 445. Sharpley, A. (1983). Effect of soil properties on the kinetics of phosphorus desorption. Soil Science Society of Americ a Journal, 47 (3), 462 467. Sharpley, A., & Menzel, R. (1987). The impact of soil and fertilizer phosphorus on the environment. Adv.Agron, 41 , 297 324. Sharpley, A. (1995). Fate and transport of nutrients: Phosphorus. United States Department of Agricultu re, Working Paper No. 8. Sharpley, A. N., Chapra, S., Wedepohl, R., Sims, J., Daniel, T. C., & Reddy, K. (1994). Managing agricultural phosphorus for protection of surface waters: Issues and options. Journal of Environmental Quality, 23 (3), 437 451. Shum an, L. M. (2001). Phosphate and nitrate movement through simulated golf greens. Water, Air, and Soil Pollution, 129 (1 4), 305 318. Silveira, M. L., Miyittah, M., & O'Connor, G. (2006). Phosphorus release from a manure impacted spodosol: Effects of a water treatment residual. Journal of Environmental Quality, 35 (2), 529 541. Sims, J., Simard, R., & Joern, B. (1998). Phosphorus loss in agricultural drainage: Historical perspective and current research. Journal of Environmental Quality, 27 (2), 277 293. Smit h, V. H., & Schindler, D. W. (2009). Eutrophication science: Where do we go from here? Trends in Ecology & Evolution, 24 (4), 201 207. Stewart, H., Hopmans, P., Flinn, D., & Hillman, T. (1990). Nutrient accumulation in trees and soil following irrigation w ith municipal effluent in Australia . Environmental Pollution, 63 (2), 155 177. Tchobanoglous, G., Burton, F., & Stensel, H. (2003). Wastewater engineering. M etcalf & E ddy I nc. New York .
110 Thompson, M., E., & Milbrandt, J. E. (2014). Development of stormwate r runoff coefficients, nutrient concentrations and loading estimates for Sanibel Island , Florida . No. Final Report, 21 pp. SCCF Marine Laboratory, City of Sanibel, FL. Thompson, M. (2015). Development and use of site specific nutrient runoff coefficients and concentrations for Sanibel Island . 24th Annual Southwest Florida Water Resources Conference, Feb. 15, 2015, Ft Myers, FL , Sanibel Captiva Conservation Foundation. Tomer, M. D., James, D. E., & Isenhart, T. M. (2003). Optimizing the placement of riparian practices in a watershed using terrain analysis. Journal of Soil and Water Conservation, 58 (4), 198 206. Toor, G., Condron, L., Cade Menun, B., Di, H., & Cameron, K. (2005). Preferential phosphorus leaching from an irrigated grassland soil. European Journal of Soil Science, 56 (2), 155 168. Turner, B. L., & Haygarth, P. M. (2000). Phosphorus forms and concentrations in leachate under four grassland soil types. Soil Science Soc iety of America Journal, 64 (3), 1090 1099. USEPA. (1977). Process design manual for land treatment of municipal wastewater . EPA 625/1 77 008. Cincinnati, OH: US EPA National Risk Management Research Laboratory. USEPA. (1981). Process design manual for la nd treatment of municipal wastewater . EPA 625/1 81 013. Cincinnati, OH: US EPA National Risk Management Research Laboratory. USEPA. (1986). Quality criteria for water. EPA 440/5 86 001 . Office of water regulations and standards, Washington, D.C. USEPA. ( 2002). Clean water and drinking water infrastructure gap analysis report. EPA 816 R 02 020. USEPA. (2004). Guidelines for water reuse . EPA 625/R 04/108 . Cincinnati. OH: US EPA National Risk Management Research Laboratory. USEPA. (2006). Process design ma nual land treatment of municipal wastewater effluents. EPA/625/R 06 016 . Cincinnati, OH. USEPA. (2012). Guidelines for water reuse. EPA/600/R 12/618 . Office of wastewater management, Washington , D.C.
111 van Roon, M. (2007). Water localisation and reclamati on: Steps towards low impact urban design and development. Journal of Environmental Management, 83 (4), 437 447. Vuono, D., Henkel, J., Benecke, J., Cath, T., Reid, T., Johnson, L., et al. (2013). Flexible hybrid membrane treatment systems for tailored nut rient management: A new paradigm in urban wastewater treatment. Journal of Membrane Science, 446 , 34 41. Walter, D. A. (1996). Geochemical and hydrologic controls on phosphorus transport in a sewage contaminated sand and gravel aquifer near ashumet pond, Cape Cod , Massachusetts . Water Supply Paper No. 2463, USGS. West, S. (2001). Centralised management: The key to successful on site sewerage service. Paper presented at the Conference Proceedings of on Advancing on Site Wastewater Systems: Design an d Maintenance. West, P. C., Gerber, J. S., Engstrom, P. M., Mueller, N. D., Brauman, K. A., Carlson, K. M., et al. (2014). Leverage points for improving global food security and the environment. Science , 345 (6194), 325 328. Westcot, D. W., & Ayers, R. S. (1985). Irrigation water quality criteria in: Pettygrove, G.S., asano, T. (eds.), irrigation with reclaimed municipal Wastewater A guidance manual . Chelsea, MI.: Lewis Publishers. Wong, J., Chan, C., & Cheung, K. (1998). Nitrogen and phosphorus leaching from fertilizer applied on golf course: Lysimeter study. Water, Air, and Soil Pollution, 107 (1 4), 335 345. Wright, K., & Rovey, C. K. (1979). Land application of waste, state of the art. Ground Water, 17 , 47 61. Xu, J., Wu, L., Chang, A. C., & Zhang, Y. (2010). Impact of long term reclaimed wastewater irrigation on agricultural soils: A preliminary assessment. Journal of Hazardous Materials, 183 (1), 780 786. Yao, W., & Millero, F. J. (1996). Adsorption of phosphate on manganese dioxide in seawater. Envi ronmental Science & Technology, 30 (2), 536 541. Zhang, M., He, Z., Calvert, D., Stoffella, P., Li, Y., & Lamb, E. (2002). Release potential of phosphorus in florida sandy soils in relation to phosphorus fractions and adsorption capacity. Journal of Enviro nmental Science and Health, Part A, 37 (5), 793 809.
112 Zurawsky, M., Robertson, W., Ptacek, C., & Schiff, S. (2004). Geochemical stability of phosphorus solids below septic system infiltration beds. Journal of Contaminant Hydrology, 73 (1), 129 143.
113 BIOG RAPHICAL SKETCH Grant B. Weinkam, from Cincinnati, Ohio , acquire d a Bachelor of Science in e nvironmental b iology at Ohio University. Following internships at Ohio EPA Division of Water he attended and graduated from Env ironmental Engineering D epartment with a Master of Science in e nvironmental s cience . Graduate r esearch focused on chemical and biological methods for remediation of contaminated sediments. Post master s he was a n environmental scientist at a USEPA research facility where he was a manager and participant of numerous l aboratory and field projects involving: fate and transport phenomena of regulated and emerging contaminants of concern in wastewater, drinking water, and surface water systems. His r esearch at t he University of Florida was conducted as part of the interdisciplinary Water Institute Graduate Fellows program with subject matter focused on sustainable reuse of wastewater in Florida l andscapes trying to attain surface water nutrient goals . In the futu re he looks to design and actively participate in projects where water quality improvement and sustainable water resource management can be accomplished .