MONITORING PHOSPHORUS TRANSPORT AND SOIL TEST PHOSPHORUS FROM TWO DISTINCT DRINKING WATER TREATMENT RESIDUAL APPLICATION METHODS.

Transcription

1 MONITORING PHOSPHORUS TRANSPORT AND SOIL TEST PHOSPHORUS FROM TWO DISTINCT DRINKING WATER TREATMENT RESIDUAL APPLICATION METHODS A Thesis Presented in Partial Fulfillment of the Requirements for the Degree of Master of Science in the Graduate School of the Ohio State University By Jason S. Undercoffer, B.S. * * * * * The Ohio State University 2009 Master s Examination Committee Dr. Nicholas Basta, Advisor Dr. Elizabeth Dayton Dr. Brian Slater Dr. Kevin King Approved by Advisor Graduate Program in Soil Science

2

3 ABSTRACT Applications of manure and soils with elevated amounts of phosphorus (P) can result in surface transport of P leading to eutrophication of surface waters. Drinking water treatment residuals (WTR) have been identified as a potential best management practice to reduce the loss of P from agricultural fields. Two field simulated rainfall studies were used to investigate the efficacy of WTR to reduce P transport, reduce soil test P (STP), and determine if relationships between STP and runoff dissolved P (RDP) are altered by soil applied WTR. In the first field study, WTR was co-blended with poultry litter to achieve a range of phosphorus to aluminum molar ratios, or phosphorus saturations (P sat ), two weeks prior to land application. Blending WTR at rates of 0, 7, 21, 84 g WTR kg-1 manure resulted in phosphorus saturations of 1860% (0WTR), 600% (LWTR), 200% (MWTR) and 50% (HWTR), respectively. Manure soluble P was reduced by 33, 62, and 96% by the LWTR, MWTR, and HWTR, respectively. The treatments were broadcast at 11.3 Mg ha -1 on field plots (2m X 2m) and simulated rainfall was performed prior to, immediately following application and at 1 month intervals for 3 months. Immediately following treatment application, RDP was reduced by 68% by the MWTR treatment and 97% by the HWTR treatment when compared to the RDP of the 0WTR treatment (32.9 mg L -1 ). These large reductions relative to previous research suggest co-blending WTR with manure prior to land application, rather than broadcasting ii

4 each material separately, may be a more effective use of WTRs P-binding capability. Currently, Ohio s P-index uses total manure P as an indicator of P transport risk. Results of this study show that the P-index should be adjusted for WTR treatments. Phosphorus source coefficients determined by soluble manure P is currently used by several states and could be used in Ohio to reflect reduced P transport from WTR co-blending. Coblending WTR with manure to achieve a final blended P sat < 100% may provide the best protection of water quality and be a useful tool for WTR/manure co-blending calibration. In the second field simulated rainfall study, WTR (10 Mg ha -1 ) was incorporated into field plots (2m x 2m) with a wide STP range. Soil incorporated WTR reduced STP for all soil test methods following WEP (74.8%) > P sat (50.2%) > M3P (40.2%) > B1P (39.5%) and RDP (39.4%), one day after WTR application. We observed positive linear relationships between Mehlich-3 P (M3P), Bray-1 P (B1P), water extractable P (WEP), and phosphorus saturation (P sat ) with RDP for all runoff events. Relationships between M3P, B1P, or P sat and RDP were not significantly altered by soil incorporated WTR while the relationship formed by WEP and RDP had a significantly higher slope when soil incorporated WTR was present. Ohio s P-index currently utilizes B1P and M3P as an indicator of environmental risk of P transport. Results from this study support the use of B1P and M3P to categorize P transport risk, regardless of previous WTR applications. iii

5 ACKNOWLEDGMENTS I would like to thank my advisor, Dr. Nicholas Basta, for allowing me the opportunity to study as a member of the Soil Chemistry group, for his encouragement, for his advice, and for his feedback throughout this research. I would also like to thank Dr. Elizabeth Dayton for being instrumental in the design, analysis, and reporting of this research project and for providing her invaluable research experience. I would like to express my appreciation to Dr. Brian Slater and Dr. Kevin King for serving as my committee members. I would also like to thank all of those who provided their thoughts, advice, and labor to help complete this research project. Shane Whitacre for his help with sample analysis, and for always being there to answer my questions. Thanks to David Henry and Nathan Weber for their assistance with the field portion of this research. Last but not least, I would like to thank my wife and family for their support and encouragement throughout this process. iv

6 VITA July 26, Born Barberton, Ohio B.S. in Environmental Science, The Ohio State University 2002-present... Graduate Teaching and Research Associate, The Ohio State University Major Field: Soil Science FIELDS OF STUDY v

7 TABLE OF CONTENTS Page Abstract... ii Acknowledgments... iv Vita... v List of Tables... viii List of Figures... x List of Abbreviations... xi Introduction... 1 Chapters: Drinking Water Treatment Residuals (WTR)... 7 Soil Incorporated WTR as a Best Management Practice... 9 Co-blended WTR and Manure as a Best Management Practice Study Objectives Co-blending WTR with poultry litter as a Best Management Practice to reduce P transport over a growing season Specific Objectives Materials and Methods WTR Collection and Characterization Field Plots Statistical Analysis Results Characterization of Background Soil and Materials Used Field Runoff Study Discussion vi

8 2. Soil Test Phosphorus and Runoff Dissolved Phosphorus in the Presence of a Soil Incorporated WTR Best Management Practice Specific Objectives Materials and Methods Field Runoff Study Statistical Analysis Results WTR Characterization Field Runoff Study Discussion Effects of Soil Incorporated WTR on Runoff Dissolved P Effects of Soil Incorporated WTR on Soil Test P Relationships between Soil Test P and Runoff Dissolved P Conclusions List of References Appendices: Appendix A Appendix B vii

9 LIST OF TABLES Table Page 1.1 Select properties of materials used and background soil at field location Mean values of soil test P and runoff P for each treatment across sampling events Results from soil P extractions at T0, before manure application, and at T1, one year after manure application Results from soil testing and RDP at T2, one day after application of WTR treatment Results from soil testing and RDP at T3, one week after application of WTR treatment A.1 Deionized water extractable phosphorus from the field site in Celina, OH A.2 Bray-1 extractable P (B1P) from field site in Celina, OH A.3 Mehlich-3 extractable P (M3P) from field site in Celina, OH A.4 Phosphorus saturation (P sat ) from field site in Celina, OH A.5 Runoff dissolved phosphorus (RDP) from field site in Celina, OH A.6 Runoff total phosphorus (RTP) from field site in Celina, OH A.7 Oxalate extractable phosphorus (P ox ) from field site in Celina, OH A.8 Oxalate extractable aluminum (Al ox ) from field site in Celina, OH viii

10 Table Page A.9 Total phosphorus from field site in Celina, OH B.1 Soil test phosphorus and runoff dissolved phosphorus for the T0 sampling event at the field site in Columbus, OH B.2 Soil test phosphorus and runoff dissolved phosphorus for the T1 sampling event at the field site in Columbus, OH B.3 Soil test phosphorus and runoff dissolved phosphorus for the T2 sampling event at the field site in Columbus, OH B.4 Soil test phosphorus and runoff dissolved phosphorus for the T3 sampling event at the field site in Columbus, OH ix

11 LIST OF FIGURES Figure Page 1.1 Dissolved and total phosphorus in surface runoff from WTR/poultry litter co-blended treatments Bray-1, Mehlich-3, and deionized water extractable P from plots treated with co-blended poultry litter Oxalate extractable phosphorus, oxalate extractable aluminum, and phosphorus saturation from plots treated with co-blended across sampling events Total phosphorus from T0, before treatment application, and T4, three months after treatment application Mean % alfalfa tissue P from plots treated with co-blended WTR/poultry litter Relationships between Bray-1 extractable P (A), Mehlich-3 extractable P (B) and runoff dissolved phosphorus at T3, one week after WTR application Relationships between water extractable P (A), phosphorus saturation (B) and runoff dissolved phosphorus at T3, one week after WTR application Relationships between Bray-1 extractable P (A), Mehlich-3 extractable P (B) on the untreated side of the split plots and the amount of soil test reduction on the WTR treated side of the split plot, one day and one week after WTR treatment application Relationships between water extractable P (A), phosphorus saturation (B) on the untreated side of the split plots and the amount of soil test reduction on the WTR treated side of the split plot, one day and one week after WTR treatment application Oxalate extractable aluminum from the untreated and WTR treated plots, one day and one week after WTR treatment application x

12 LIST OF ABBREVIATIONS DP... Dissolved Phosphorus STP... Soil Test Phosphorus RDP... Runoff Dissolved Phosphorus RTP... Runoff Total Phosphorus B1P... Bray-1 extractable Phosphorus M3P... Mehlich-3 extractable Phosphorus WEP... deionized Water Extractable Phosphorus P sat... Phosphorus Saturation 0WTR co-blended WTR application amount LWTR... Low - co-blended WTR application amount MWTR... Medium - co-blended WTR application amount HWTR... High - co-blended WTR application amount xi

13 INTRODUCTION Phosphorus is an important plant macronutrient essential for attaining profitable yields in agricultural systems. Deficient crops exhibit stunted growth and poor grain or fruit development. For this reason it is important to maintain sufficient phosphorus levels in the soil to ensure that phosphorus is not a limiting nutrient, maintaining optimum growing conditions. However, when phosphorus is transported from soil to surface waters it can cause eutrophication, which has been identified by the USEPA as the most imposing threat to surface waters of the United States (USEPA, 2000). This process impairs water quality both aesthetically and limits suitability for water use; it has negative impacts on fishing industries, limits recreational use, and increases costs of drinking water treatment. The USGS (1999) reported intensively managed agriculture, mainly livestock production, as the leading cause of phosphorus pollution. Phosphorus can be transported from a field by surface runoff, lateral flow and leaching to subsurface drainage; however, surface runoff is thought to be the main pathway for phosphorus transport in most landscapes (Vadas et al., 2005). Phosphorus is transported during a runoff event as dissolved phosphorus (DP) or as particulate 1

14 phosphorus (PP). The amounts of both DP and PP in runoff are controlled by the quantity of phosphorus available for transport and by site-specific hydrologic conditions (Sharpley, 1995). Transport of particulate phosphorus occurs when sediment and organic matter is moved in runoff water. This process is governed by the same processes that control soil erosion. Land management practices including tillage, fertility management, vegetative cover, soil type, and slope are all important in understanding and predicting the amount of PP in runoff water. Since phosphorus is a reactive molecule that prefers to be attached to soil particles or as solid minerals, the majority of phosphorus transported to surface waters in most agricultural situations is attached to sediments (Sharpley and Smith, 1989). However, in some situations the amount of dissolved phosphorus entering surface water is of environmental concern (Sharpley et al., 1994). The amount and solubility of P in the upper 5 cm of the soil, which interacts with rainfall and overland flow, is a primary factor in the risk of DP transport (Torbert et al., 1998). Elevated P levels in the soil surface lead to situations that are at high risk for DP transport (Pautler and Sims, 2000). This is frequently the case in livestock production when relatively small tracts of land receive continual manure additions or when manure additions are targeted at achieving crop nitrogen needs, which leads to an over application of P. This over application of P, with time, leads to high soil P levels that increase the risk of DP transport to surface waters. 2

15 Regardless of soil P status, recent applications of material containing significant amounts of soluble P can increase the risk of DP transport (Kleinman et al., 2002; DeLaune et al., 2004). Termed event-specific losses, heavy rains after application of manure or inorganic fertilizers have the potential to transport large amounts of RDP to surface waters (Hart et al., 2004). These events can significantly contribute to water quality problems. The Natural Resource Conservation Service (NRCS) has developed the P-risk Index to assess the risk of P transport from agricultural sites (USDA-NRCS, 2002). The goal of the index is to identify areas that have high risk of P transport aiding in selection of management practices that will lead to protection of water quality. Based on the design of Lemunyon and Gilbert (1993) the index is a scoring matrix of well established source and transport factors that influence P transport. The Ohio P-risk index currently assesses risk based on 9 site characteristics: soil erosion potential, runoff class, connectivity to water, soil test P, inorganic and organic P application rate, inorganic and organic P application method, and the presence of a filter strip. Each of these factors is given a weighted value based on relative contribution to P transport. The sum of these values place a site in a level of perceived risk. One of the factors used in the P-risk index, soil test P, is based on the amount of Bray-1 or Mehlich-3 extractable P in the soil. Mehlich-3 (M3P) and Bray-1 (B1P) were developed to determine plant available phosphorus in soils and to predict crop response to fertilizer additions (Bray and Kurtz, 1945; Mehlich, 1984). These extractions use acidic ammonium fluoride solutions to dissolve available P bound to Ca, Al, and Fe and 3

16 promote release of Al-bound P by decreasing Al activity in solution through formation of Al-F complexes. Since these agronomic soil tests were developed without measuring P transport there is no reason to assume a correlation between the test results and the amount of P in runoff (Sims et al., 2000). However, agronomic soil tests have been shown to have a strong positive linear relationship with RDP in small plot simulated rainfall experiments (Sharpley et al., 1996; Pote et. al., 1996; Sims et al., 2002; Andraski and Bundy, 2003). In a review paper, Vadas et al. (2005) looked at the relationships between M3P, B1P and RDP in 17 studies using similar methods to the National Phosphorus Research Protocol for simulated rainfall research. It was determined that the slopes of linear regression lines formed by M3P (mg kg -1 ) or B1P (mg kg -1 ) and RDP (mg L -1 ) did not significantly differ for 26 of the 31 soils analyzed, ranging from to However, the correlation between agronomic soil extractions and RDP has been shown to improve by the inclusion of site specific P retention properties of the soil (Schroeder et al., 2004). Developing and analyzing environmental soil tests with a more robust relationship to P transport by extracting an amount of P proportional to the amount available for transport has been proposed for assessing the risk of P transport. One potential alternative is to extract soil phosphorus with water or weak salt solutions. This method is designed to mimic rainfalls ability to desorb phosphorus from the soil. Therefore with a relatively simple soil extraction, analysis of the potential for P transport may be accomplished. Pote et al. (1996) determined that the deionized water (WEP) extraction has a significant linear relationship with RDP from small scale simulated rainfall plots. This research showed that the WEP had a higher correlation (r 2 = 4

17 0.82) with runoff DP than Mehlich-3, although Mehlich-3 was still significantly correlated (r 2 = 0.72). While analyzing the ability of soil tests to predict P desorption from soil with contrasting properties and management, Hooda et al. (2000) determined that a WEP can account for 96 percent (p-value<0.001) of the differences in P release from soil. Another soil test, P saturation (P sat ) accounts for differences in soils P retention properties within the test itself giving an indication of the amount of P that is desorbable and available to be transported. Phosphorus saturation measurements take on multiple forms, but in general it is the ratio between the amount of phosphorus in a soil and the soil s P sorption capacity (P max ). One method to calculate P max is to calculate a Langmuir adsorption maximum from adsorption isotherms. However, this method is a time consuming and expensive process which many soil testing facilities do not currently offer (Sims, 2002). To increase the utility of the test it was necessary to find an alternative to measuring P max for every soil. The solution was to find the specific chemical characteristics of soil which contribute to P max and measure them. The acidified ammonium oxalate extractable Al and Fe can be used to estimate P max because (1) amorphous Al and Fe oxides are the primary source of soil phosphorus retention in many agricultural soils, and (2) the acidified ammonium oxalate extraction is used to estimate the amorphous oxide of Al and Fe content in soils (McKeague and Day, 1993; Zhang et al., 2005). The measure of P saturation then becomes the molar ratio of oxalate extractable P over the oxalate extractable Al and Fe in the soil. P sat (%) = [P ox /(Fe ox + Al ox )] * 100 5

18 P sat has been shown to make accurate predictions of DP in runoff across differing soil types. While comparing 17 studies which analyzed the relationships between soil test and runoff DP, it was suggested that P sat may have the most potential to work across a wide range of soils (Vadas et al., 2005). For the 10 soils investigated in this analysis, the slope of the regression line between P sat and RDP did not significantly differ across soil types. The combined data indicated a split line relationship with an inflection at 12.5% saturation. Above 12.5% the amount of RDP increased rapidly indicating that this point may be used as an environmental threshold. Additionally, Hooda et al. (2000) determined that P sat can make accurate predictions of the amount of P that is extractable by water from soils with contrasting properties showing that P sat is a robust measure of P solubility. The relationships between B1P, M3P, WEP, P sat and RDP from simulated rainfall plots has been extensively studied and well documented. The agronomic soil extractants, M3P and B1P, are well correlated to RDP; although, research suggests that measures of P saturation and WEP may be more robust and work across a wider range of soil with differing P retention properties. This may have implications when assessing P transport risk or trying to predict RDP from fields amended with WTR, which alter the P retention properties of the soil (Dayton and Basta, 2005b). Recent applications of material containing significant amounts of soluble P can increase the risk of DP transport, regardless of soil P status. DeLaune et al. (2004) showed RDP concentrations of 8.8 to 33.0 mg L -1 one day after poultry litter surface applications ranging of 2.24 to 8.96 Mg ha -1. Dayton and Basta (2005b) surface applied 6

19 8.8 Mg ha -1 poultry litter and reported RDP concentrations of 31.1 mg L -1 from simulated rainfall plots. These RDP concentrations from manure applied plots far exceed RDP values reported from high STP soil in the absence of recently applied manure (Vadas et al. 2005). STP values have little effect on RDP in the presence of recently applied manure (DeLaune et al., 2004). Because of the large potential for DP transport from surface applied manure, Ohio s P-risk index uses organic P application amounts and application methods as a P transport contributing factor. Drinking Water Treatment Residuals To reduce P transport from agricultural land, Best Management Practices (BMP) have been developed, many targeted at reducing soil erosion. These BMPs are effective in reducing the transport of PP; however, they are not as effective in reducing the loss of DP (Daniel et al., 1998). New best management strategies need to be developed to control these specific situations when the risk of DP transport is elevated. One promising development in recent years is the beneficial use of drinking water treatment residuals (WTR). Drinking water treatment residuals (WTR) are the by-product of the coagulation process in drinking water treatment plants using surface water as a source. The coagulation process uses salts of Al and Fe (typically alum and ferric chloride, respectively) to flocculate particles, clarifying the raw water. The resulting sludge that settles in the tanks is a mixture of sediments and organic matter from the raw water and reaction products of the Al and Fe salts, amorphous Al or Fe hydroxides. Currently, 7

20 many facilities dispose of these materials in landfills. Beneficial use of these materials could save landfill space and financial resources that municipalities spend for landfilling. In early WTR research it was apparent that WTR has the ability to bind and reduce the solubility of phosphorus (Bugbee and Frink, 1985; Elliott and Dempsey, 1991; Peters and Basta, 1996; Dayton and Basta, 2001). Amorphous hydroxides of Al or Fe are the component of WTR that is responsible for P adsorption (Dayton and Basta, 2003). Aluminum and iron hydroxides form an inner-sphere complex with orthophosphate in soil. The Al ox -P bond that is formed is strong, not readily desorbable, and stable over time and during changes in environmental conditions (Sims and Pierzynski, 2005; Agyin- Birikorang and O Connor, 2007). Basta and Dayton (2005a) determined that the amount of oxalate extractable Al (1:100 WTR to solution), used to extract amorphous Al-oxides, is predictive of the differences in WTR s P sorption maxima in an Al-WTR (McKeague and Day, 1993). The reaction between P and the Al ox surfaces are fast, on the scale of a few minutes to a few hours (Bolan et al., 1985). The P sorption rates of WTR are influenced by the size of WTR particles as smaller particles have increased reactive surface areas which lead to faster P sorption kinetics and higher P sorption maxima (Novak and Watts, 2005a). A second, slow solid state diffusion reaction of adsorbed P into WTR micropores may influence long-term P sorption capacities and immobilization (Bolan et al., 1985; Makris, 2005). These materials have been shown to decrease the solubility of phosphorus in soil reducing STP and the risk of DP tranport from high STP soils (Codling et al., 2000; Novak and Watts, 2005b; Peters and Basta, 1996; Dayton and Basta, 2005b; Haustien et 8

21 al., 2000; Agyin-Birikorang et al., 2007). Drinking water treatment residuals have also been shown to reduce the solubility of P in manure and biosolids reducing the risk of DP tranport from surface applications of manure or biosolids (Ippolito et al., 1999; Gallimore et al., 1999; Elliott et al., 2002; Dayton and Basta, 2005b; Makris et al., 2005; Codling et al., 2000; Dao et al., 2001; Oledeji et al., 2008). To ensure the proper implementation and management of this potential BMP it is necessary to evaluate P transport, and changes to the factors that influence P transport, in the presence of WTR. Soil Incorporated WTR as a Best Management Practice Additions of WTR to soil will not reduce the total amount of P; rather it changes the solubility of P by specifically adsorbing phosphate ions to Al or Fe oxide surfaces. This decreases the amount of P in soil solution and the ability of soil extractants to extract P, altering soil test results. Peters and Basta (1996) showed reductions of M3P from 553 to 250 mg kg -1 when soil is amended with 100 g kg -1 Al-WTR in a laboratory study. In another study, incorporating multiple WTRs into a phosphorus soil with 315 mg kg -1 M3P at rates of 25, 50, and 100 g kg -1 was shown to reduce M3P extractable by 6.93 to 86.7 percent depending on the application amount and Al ox of the material used (Dayton and Basta, 2005). In this study, M3P was decreased linearly with increasing additions of Al ox. Haustein et al. (2000) reported M3P reductions of more than 100 mg kg -1 by surface applying 18 Mg ha -1 of Al-WTR in a field study. Codling et al. (2000) incubated three high P soils for 7 weeks with 10, 25 and 50 g kg -1 WTR and showed B1P reductions as high as 83% with the 50 g kg -1 addition. Collectively, this research demonstrated that 9

22 agronomic soil tests can be significantly reduced under increasing WTR additions. These reductions could potentially be used to lower soil test values where phosphorus levels exceed environmental thresholds (Novak and Watts, 2005b). Under the same application rates that were shown to reduce M3P extractable P by 6.93 to 28.4 percent (25 g kg -1 ), Dayton and Basta (2005b) reduced weak salt extractable P (0.01 M CaCl 2 ) by 60.9 to 96 percent. This reduction is far greater than that of M3P. Similar results were demonstrated by Novak and Watts (2005b) when incorporating between 1 and 6 percent WTR with a high P soil and reported WEP reductions between 45 and 91 percent and M3P reductions from 26 to 40 percent. Peters and Basta (1996) similarly reduce WEP by a greater magnitude relative to M3P, and Codling et al. (2000) reduced WEP by a greater magnitude relative to B1P with additions of WTR to soil. This difference has been attributed to the dissolution of P bound to the amorphous Al surfaces by the acid ammonium fluoride solutions of B1P and M3P (Basta et al., 2000). Additions of WTR to soil with elevated amounts of phosphorus, which would normally have a high risk of DP transport, not only reduced STP values it has been shown to be effective at reducing RDP. Haustein et al. (2000) demonstrated this by performing simulated rainfall on field plots containing 150 to 300 mg kg -1 M3P amended with Al-WTR (46.7 g Al kg -1 ) at rates of 1.1, 2.2, 9.0 and 18 Mg ha -1. Rates of 9.0 and 18 Mg ha -1 significantly reduced RDP in runoff one day, one month and four months after application. The RDP reductions were approximately 57 and 43% for the 18.0 and 9 Mg ha -1 applications, respectively, one day after the treatment application. The RDP was further reduced one month after treatment application. The RDP from the control 10

23 plots had a mean RDP of approximately 1.4 mg L -1 during both simulated rainfall events. Agyin-Birikorang et al. (2007) simulated rainfall on indoor soil boxes containing soils that were amended with 114 Mg ha -1 Al-WTR (29.7 g Al ox kg -1 ) 7.5 yrs prior. Reductions in flow weighted RDP on the WTR treated sites were approximately 60% less than sites with no WTR application even after 7.5 years of time elapsed since the application, demonstrating the stability of WTR bound P and the potential for long term water quality benefits. The results of these experiments demonstrate the ability of surface applied WTR to reduce the transport of DP. From this research it can be concluded that WTR additions will reduce soil test and RDP values proportional to the amount of WTR added and the P max of the WTR used, or the amount of Al or Fe hydroxides added (Dayton and Basta, 2005b). However, the magnitude of soil test reduction will differ depending on extraction method. The dissimilar response of soil tests to the presence of WTR may have implications when trying to assess the environmental risk by monitoring previously set environmental thresholds or when calculating the contribution of soil test values to P-risk assessments. For this reason it is important to address the relationships between soil testing and potential for DP transport in the presence of soil applied WTR. Co-Blended WTR and Manure as a Best Management Practice Blending manure with Drinking Water Treatment Residuals (WTR) has been suggested as a way to mediate the water quality risk associated with surface applications of manure. Multiple laboratory studies have analyzed the reductions of soluble P in 11

24 manure or biosolids from additions of WTR (Ippolito et al., 1999; Codling et al., 2000; Dayton and Basta, 2005b; Makris et al., 2005; Dao et al., 2001). Ippolito et al. (1999) coblended WTR with biosolids at dry weight ratios (WTR:biosolids) from 0 to 40:1 to determine the ability of WTR to adsorb soluble P. Co-blending ratios greater than 8:1 (WTR:biosolids) reduced soluble P by 99.7%. Codling et al., (2000) reported reductions in soluble P ranging from 25 to 87% when poultry litter was incubated with WTR at 25, 50, and 100 g kg -1. Dao et al. (2001) composted poultry litter with additions of 590 g WTR kg -1 manure, reduced soluble P in manure from 2034 mg kg -1 in the untreated manure to 651 mg kg -1 in the WTR co-blended manure. Dayton and Basta (2005b) showed large reductions ranging from 56.9 to 98.4 % in soluble P by incubating poultry litter and biosolids with 25, 50 and 100 g kg -1 WTR. This research showed a strong relationship between the reduction of soluble P and amount of Al ox added. Makris et al. (2005) reported reduced P solubility in poultry litter co-blended with WTR and that reductions were exceptionally high when molar Al:P ratio of co-blended material was 1 suggesting that Al:P ratios or P sat may be a useful method for calibrating WTR additions. Reductions in manure soluble P from co-blended additions of WTR translate into reduced DP transport from surface applications of manure (Gallimore et al., 1999; Elliott et al., 2002; Oledeji et al., 2007) Gallimore et al. (1999), using simulated rainfall on 1.8 m x 9.8 m plots amended with 6.72 Mg ha -1 poultry litter and two amounts of WTR (44.8 and 11.2 Mg ha -1 ), analyzed RDP concentrations replicated with two WTRs differing in Al oxide content (50.5 and 11.7 g kg -1 Al ox ). The 44.8 Mg ha -1 application of the 50.5 g Al ox kg -1 WTR reduced RDP in runoff by 42.7% compared to control plots. The same 12

25 application of the lower Al ox content WTR reduced the RDP; however, this reduction was not significant. This research reinforces the importance of amorphous Al in reducing the transport of P in surface runoff. Oledeji et al. (2007) performed simulated rainfall on indoor soil boxes with surface applications of poultry litter at rates of 56 and 224 kg P ha - 1, co-blended with an Al-WTR. The mean mass of DP transport was reduced from 47 mg in the absence of WTR to 24 mg by 10 g kg -1 WTR additions. Co-blending WTR has also been shown to reduce P leaching from soil amended with 224 kg P ha -1 biosolids (Elliott et al., 2002). Co-blending WTR (25 g kg -1 ) with biosolids reduced P leaching so that it was not statistically different than the leaching columns with no applied biosolids. Collectively, this research has demonstrated substantial reductions in manure P solubility and P transport when co-blended with WTR. Ohio s P-risk index does not consider relative differences in manure P solubility when calculating the contribution of applied manure P. Manure P solubility has been shown to influence P transport from simulated rainfall plots (DeLaune et al., 2004; Vadas et al., 2007). By considering only total manure P and omitting differences in manure P solubility, Ohio s P-risk index does not allow WTR co-blending as a best management practice to influence the P-risk score and may not adequately reflect P transport risk (Dayton and Basta, 2005). Elliott et al. (2002) suggested using index score modifying factors based on manure WEP, P source coefficients, to weight index score contributions from manure applications. This method was shown to improve relationships between manure applications and RDP and should improve the ability of P-risk index tools in identifying areas that have a high risk for P transport (Elliott et al., 13

26 2002). This approach, or a similar manure P contribution weighting scale, would allow for the inclusion of WTR co-blending in the P-risk index and give credit to land managers for utilizing this potential best management practice. Study Objectives Assessing the risk of DP transport to surface water must take into account many factors which alter DP transport. Analyzing the potential for soil and manure to enrich runoff waters with P is a very important. Many studies have shown the utility of soil testing in predicting RDP and the P transport contributions from surface applied manure. Strong relationships built on simulated rainfall research exist. Research has also explored the use of WTR as an effective tool to reduce soil test results, manure P solubility and P transport. However, relatively little research has analyzed relationships between P sources and DP transport in the presence of a WTR application. Without an understanding of how these relationships are altered by WTR applications, the utility of WTR additions as a BMP for water quality protection is hindered. An improved understanding will allow for better selection of WTR treatment placement, calibration of WTR application rates, and successful monitoring of previous applications. The objectives of this study are to 1) evaluate the effectiveness of WTR coblended with poultry litter, as a best management practice, to reduce STP, P solubility and P transport from a field with surface applied manure, 2) determine the efficacy of the WTR co-blended BMP over a growing season, 3) evaluate the effectiveness of WTR incorporated into a soil with a wide STP range, as a best management practice, to reduce 14

27 STP and P transport, and 4) assess the relationships between soil test methods and P transport and determine if the relationship changes with a WTR BMP. The objectives will be accomplished with two field simulated rainfall projects. The first, located in Celina, OH, will analyze the efficacy of WTR co-blended with poultry litter as a BMP over a growing season. The second, located at Ohio State University s Waterman Farm in Columbus, OH, will analyze changes in STP, RDP and in the relationships between STP and RDP from a soil incorporated WTR BMP. Accomplishment of these objectives will provide a basis for monitoring previous WTR applications. Additionally this research will provide important data and concepts for the use of WTR as a best management practice and for the inclusion of WTR BMPs in the Ohio s P-risk Index. 15

28 CHAPTER 1 CO-BLENDING WTR WITH POULTRY LITTER AS A BEST MANAGEMENT PRACTICE TO REDUCE P TRANSPORT OVER A GROWING SEASON SPECIFIC OBJECTIVES 1. Evaluate the effectiveness of WTR co-blended with poultry litter, as a best management practice, to reduce STP, P solubility and P transport from a field with surface applied manure. 2. Evaluate the efficacy of the WTR BMP over a growing season. MATERIALS AND METHODS WTR Collection and Characterization Drinking water treatment residuals were collected from the Celina Drinking Water Treatment Plant in Celina, Ohio. The Celina Drinking Water Treatment plant treats raw water from Grand Lake/St. Mary s with an aluminum salt for flocculation of particulates. Material from the primary clarifying tank was collected in a smaller secondary settling tank and augmented with <1% aluminum. The WTR was then homogenized in a cement mixer, air dried, crushed and sieved (2 mm) before analysis and co-blending with poultry litter. The Al, Fe and P content of the material was analyzed using an acid ammonium oxalate extraction (100:1 solution to WTR) (McKeague and Day, 1993; Dayton and Basta, 2005a). 16

29 Field Plots Small field plots and simulated rainfall were used to evaluate the ability of WTR co-blended with poultry litter to reduce transport of phosphorus in surface runoff and soil test P. The field site was located in Celina, Ohio in the Grand Lakes/St. Mary s watershed. Plots measuring 2 m x 2 m were established in an alfalfa field on a Glynwood silt loam soil (fine, illitic, mesic Aquic Hapludalfs) with a 4 percent slope. Plot establishment and rainfall simulation followed protocol established by the National Phosphorus Research Project (National Phosphorus Research Protocol, 2007). Each plot received a 4.5 kg (11.3 Mg ha -1 ) application of air dried poultry litter. Total P content of poultry litter was determined by digesting manure according to EPA Method 3050B (U.S. EPA, 1996). Soluble P in manure was determined by a 1g: 200 ml deionized water extraction, shaken for 24 hours (Wolf et al., 2005). The poultry litter (4.5 kg) was co-blended with one of four amounts of WTR: High WTR (HWTR) 84 g kg -1 or 950 kg ha -1, Medium WTR (MWTR) 21 g WTR kg -1 or 250 kg ha -1, Low WTR (LWTR) 7 g WTR kg -1 or 82.5 kg ha -1, and a control treatment of 0 g WTR kg -1 manure (0WTR). The poultry litter/wtr co-blending rates were selected to achieve a wide range of P saturations (0WTR of 1860% LWTR of 600% MWTR of 200% and HWTR of 50%). Manure and WTR treatments were blended in a cement mixer in batches 2 wk prior to application to ensure homogeneity of treatments. The four treatment levels were replicated 5 times, for a total of 20 plots. Simulator rainfall was supplied using a single Teejet ½ HHSS50WSQ nozzle mounted 3 m above the soil surface on a rainfall simulator following the design of Miller 17

30 (1987). The rainfall simulator was calibrated using methods outlined by Humprey et al. (2002) and met recommended criteria for rainfall distribution with a 93% coefficient of uniformity, and rainfall intensity of 60 mm hr -1. This rainfall intensity, maintained for a duration of 1 hr, occurs on average once every 10 years in Ohio (NOAA, 2007). During rainfall simulation, temporary metal borders were installed around the perimeter of the plot to isolate runoff and channel it to a PVC pipe which led downslope to a large collection vessel. All runoff was collected for 30 minutes after it commenced. A 250 ml composite sample of homogenized runoff was collected and a 20 ml subsample was 0.45 µm filtered and acidified (one drop concentrated HCl per 10 ml sample). Runoff dissolved P (RDP) was determined from the filtered sample using inductively coupled plasma atomic emission spectroscopy (ICP). Runoff total P (RTP) was determined by digesting 50 ml of homogenized runoff water with 0.5 g potassium persulfate and 1 ml of concentrated sulfuric acid in a Mars Xpress microwave at 170 C o for 30 min (Pote and Daniels, 2000)). Sodium pyrophosphate check standards (1 mg P L -1 ) were also digested with satisfactory P recovery (99%). Simulated rainfall was performed on each plot 5 times: Time 0 (T0) before any treatment application to evaluate background runoff and field variability in November 2007, Time 1 (T1) immediately after treatment application in June 2008, and three more times at intervals of approximately one month, mid-july (T2), mid-august (T3), and mid- September 2008 (T4). Simulated rainfall took place within a week after crop harvest to ensure consistent crop height (approximately 15 cm) with the exception of the initial rainfall which took place in November, 2007 when no crop harvest was anticipated. 18

31 Composite samples of the surface soil (5 cm depth), taken from ten locations within each plot, were collected before each rainfall simulation. Samples were oven dried (60 C ) and sieved (< 2mm). Soil P status was evaluated using multiple soil extractions. Mehlich 3-extractable P (M3P) and Bray 1-extractable P (B1P) were analyzed with 1 g of soil: 10 ml of corresponding extraction solution shaken for 5 minutes on a rotating shaker (150 evolutions min -1 ) and filtered (<0.45 µm) (Bray and Kurtz, 1945; Mehlich, 1984). Soil % P saturation (P sat ) was measured with a 0.25 g: 25 ml acid ammonium oxalate extraction shaken for 4 hours on an oscillating shaker and filtered (<0.45 µm) (McKeague and Day, 1993; Dayton and Basta, 2005a). Soluble P was evaluated using a 2g soil: 20 ml deionized water extraction (WEP) shaken for 1 hour, filtered (<0.45 µm) and acidified (one drop concentrated HCl per 10 ml of sample; Olsen and Sommers, 1982). Alfalfa samples (top 6 inches) of 10 plants were collected within each plot, following guidelines in the Ohio Agronomy Guide (2005), to determine plant tissue P. Tissue P content was determined by digesting tissue according to EPA Method 3050B (U.S. EPA, 1996). All extracts were analyzed by ICP according to USEPA methods 6010C on a Varian Vista-MPX ICP-OES (Varian Inc., Walnut Creek, CA). Data QA/QC included analysis of an intralaboratory established control sample, initial calibration verification, initial calibration blank, continuing calibration verification every ten samples, continuing calibration blank every ten samples, and a low limit of quantitation verification. All checks were within the quality control limits set in USEPA ILM04.0b. 19

32 Statistical Analysis All data were Log 10 transformed before analysis to control for unequal variances. After the transformation all tests complied with Levene s test for unequal variances (α > 0.10). Outliers within groups were determined using Dixon s test (α 0.10) for outliers which determines if the minimum and maximum values fall outside of the calculated range (Sheskin, 1997). Differences in soil test and runoff results within treatments across sampling events and across treatments within a sampling event were assessed using a one-way ANOVA, with Fisher s least significant difference (LSD) pair-wise comparisons for means separation. Analysis was conducted using SAS, version 9.1 (SAS Institute Inc., 2002). Significant differences are assigned at α Percent change is calculated as the difference between a measured value and of the control (0WTR) and is only shown if the difference is significant. RESULTS Characterization of Background Soil and Materials Used Properties of the WTR, poultry litter, and Glynwood soil at the field site are summarized in Table 1.1. The WTR had an oxalate extractable Al (Al ox ) content of 115 g kg -1, which is within the range ( g kg -1 ) for 18 Al based WTRs reported by Dayton and Basta (2005a), representing an approximately 100 fold more reactive Al than the Glynwood soil at the field site and more than 500 times more than the poultry litter. Oxalate extractable P (P ox ) content of the poultry litter (8.62 g kg -1 ) was 97% of the total P (8.89 g kg -1 ) which was approximately 5 and 14 times greater than the WTR and 20

33 Glynwood soil, respectively (Table 1.1). Oxalate extractable Fe (Fe ox ) content was highest in the Glynwood soil (Table 1.1). All materials have near, to slightly above, neutral ph (Table 1.1). Field Runoff Study The T0 sampling, before any treatment application, indicates minimal variability of soil test P and runoff dissolved P (RDP) at the field site. All soil extractions and RDP were not significantly different between treatments at T0. At T0 mean RDP for all plots was mg L -1 (Table 1.2). Mean Bray-1 Extractable P (B1P), Mehlich-3 Extractable P (M3P), and Water Extractable P (WEP) values were 120, 191, and 15.9 mg kg -1, respectively (Table 1.2). Initial soil P ox and Al ox had mean values of 622 and 1112 mg kg -1, respectively (Table 1.2). Mean phosphorus saturation (P sat ) was 18.8% (Table 1.2). Co-blending WTR with the poultry litter resulted in reductions of manure WEP. The untreated litter WEP was 1.5 g kg -1 and the WEP of the treated manure was 1.05, 0.422, and g kg -1, for the LWTR, MWTR, and HWTR treatments, respectively. The HWTR treatment reduced manure WEP by 86%, MWTR treatment reduced manure WEP by 62% and the LWTR treatment reduced manure WEP by 33%. Results of the co-blended poultry litter/wtr treatments on RDP are summarized in Table 1.2 and Figure 1.1A. At T1 the mean RDP was 32.9, 33.7, 10.5, and 1.00 mg L -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The RDP of HWTR was reduced 97% compared to 0WTR and was significantly lower than all other treatments (Figure 1.1A). The MWTR RDP was reduced by 68% from the 21

34 0WTR treatment and was significantly lower than the 0WTR and LWTR treatments (Figure 1.1A). The LWTR RDP was not significantly different than the 0WTR at T1. At T2 the RDP was 0.886, 0.803, 0.716, and mg L -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The RDP from the HWTR treatments was significantly lower than all other treatments and was 56% lower than the 0WTR treatments (Figure 1.1A). The RDP from the MWTR and LWTR treatments were not significantly different than the 0WTR treatment or each other at T2. At T3 RDP ranged from to mg L -1, and RDP ranged from to mg L -1 at T4 (Table 1.2). There were no significant effects from the WTR treatment at T3 or T4. Results of the co-blended poultry litter/wtr treatments on Runoff Total P (RTP) are summarized in Table 1.2 and Figure 1.1B. At T1 the mean RTP was 35.4, 41.0, 19.9, and 12.7 mg L -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The RTP from the HWTR and MWTR treatments were significantly lower than the 0WTR treatment with 64% and 44% reductions, respectively, though these reductions were not significantly different from each other (Figure 1.1B). The RTP from the LWTR treatment was not significantly different than 0WTR at T1. At T2 the RTP was 2.13, 1.93, 1.70, and 1.30 mg L -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The RTP from the HWTR and MWTR treatments were significantly lower than the 0WTR treatments with 39% and 20% reductions, respectively, though these reductions were not significantly different from each other (Figure 1.1B). The RTP of the LWTR treatment was not significantly different than any 22

35 other treatment at T2. At T3 RTP ranged from 1.63 to 3.50 mg L -1, and RTP ranged from 2.10 to 2.92 mg L -1 at T4 (Table 1.2). There were no significant effects from the WTR treatment at T3 or T4. Results of the co-blended poultry litter/wtr treatments on B1P are summarized in Table 1.2 and Figure 2A. At T1 the mean B1P was 242, 192, 214, and 163 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The B1P of the HWTR and LWTR treatments were significantly lower than the 0WTR treatments with 33% and 21% reductions, respectively, though these treatments were not significantly different from each other (Figure 1.2A). The B1P of the MWTR treatment was not significantly different than the 0WTR or the LWTR treatments at T1. At T2 the B1P was 227, 236, 221, and 167 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The B1P of the HWTR treatment was significantly lower than all other treatments with a 27% reduction from the 0WTR treatment (Figure 1.2A). There was no significant difference in B1P among the other treatments at T2. At T3 the B1P was 156, 185, 145, and 128 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The B1P of the HWTR treatment was significantly lower than all other treatments with an 18% reduction from the 0WTR treatment (Figure 1.2A). The B1P of the LWTR treatment was significantly higher than all other treatments with a 19% increase from the 0WTR treatment (Figure 1.2A). The B1P of the MWTR treatment was not significantly different than the control at T3. At T4 B1P ranged from 131 to 176 mg kg -1 (Table 1.2). There was no significant reduction of B1P due to the WTR treatment at T4. 23

36 Results of the co-blended poultry litter/wtr treatments on M3P are summarized in Table 1.2 and Figure 2B. At T1 the mean M3P was 593, 380, 490, and 306 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The M3P of the HWTR treatment was significantly lower than the 0WTR and MWTR treatments with a 48% reduction from the 0WTR treatment (Figure 1.2B). The M3P of the MWTR and LWTR treatments were not significantly different than the 0WTR treatment or each other at T1. At T2 the M3P was 342, 465, 419, and 314 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). There were no significant reductions in M3P due to the treatment at T2. The M3P of the MWTR and LWTR treatments were significantly higher than the 0WTR treatments with 36% and 22% increases, respectively, though they were not significantly different from each other (Figure 1.2B). The M3P of the HWTR treatment was not significantly different than the 0WTR treatment at T2. At T3 the M3P was 358, 441, 348, and 280 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The M3P of the HWTR treatment was significantly lower than all other treatments with a 22% reduction from the 0WTR treatment (Figure 1.2B). The M3P of the LWTR treatment was also significantly different than all other treatments with a 33% increase from the 0WTR treatment (Figure 1.2B). The M3P of the MWTR treatment was not significantly different than the 0WTR treatment at T3. At T4 the M3P was 339, 398, 295, and 212 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The M3P of the HWTR treatment was significantly different than all other treatments 24

37 with a 38% reduction from the 0WTR treatment (Figure 1.2B). The M3P of the MWTR and LWTR treatments were not significantly different than the 0WTR treatment at T4, though the MWTR treatment was significantly lower than LWTR. Results of the co-blended poultry litter/wtr treatments on WEP are summarized in Table 1.2 and Figure 1.2C. At T1 the mean WEP was 51.7, 26.6, 29.6, and 21.1 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The WEP of the HWTR, MWTR, and LWTR plots were all significantly lower than the control plots with 59%, 43%, and 49% reductions, respectively, though the LWTR, MWTR and HWTR plots were not significantly different from each other. (Figure 1.2C). At T2 the WEP was 33.6, 29.3, 24.9, and 18.4 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The WEP of the HWTR treatment was significantly different than all other plots with a 45% reduction from the 0WTR treatment (Figure 1.2C). The WEP of the MWTR and LWTR treatments were not significantly different than 0WTR or each other at T2. At T3 the WEP was 23.6, 25.0, 21.4, and 14.5 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The WEP of the HWTR treatment was significantly lower than all other treatments with a 39% reduction from the 0WTR treatment (Figure 1.2C). The WEP of the MWTR and LWTR treatments were not significantly different than the 0WTR or each other at T3. At T4 the WEP was 27.7, 27.8, 19.9, and 15.5 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The WEP of the HWTR and MWTR plots were significantly lower than the 0WTR and LWTR treatments with 44% and 28% 25

38 reductions, respectively, though they were not significantly different from each other (Figure 1.2C). The WEP of the LWTR plots were not significantly different than the control at T4. Results of the co-blended poultry litter/wtr treatments on P ox are summarized in Table 1.2 and Figure 1.3A. At T1 the mean P ox was 1302, 996, 1271, and 1565 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The P ox of the LWTR treatment was significantly lower than all other treatments with a 24% reduction from the 0WTR treatment (Figure 1.3A). The MWTR and HWTR treatments did not significantly differ from the 0WTR treatment or each other at T1. At T2 the P ox ranged from 950 to 1170 mg kg -1 (Table 1.2). There was no significant change of P ox due to the WTR treatment at T2. At T3 the P ox was 893, 1045, 1003, and 1196 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The P ox of the HWTR treatment was significantly higher than the 0WTR and MWTR treatments, but not significantly different than the LWTR treatment, with an increase of 34% from the 0WTR treatment (Figure 1.3A). The P ox of the MWTR and LWTR treatments were not significantly different than the 0WTR treatments or each other at T3. At T4 the P ox was 846, 976, 758, and 1303 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The P ox of the HWTR treatment was significantly higher than 0WTR and MWTR treatments with an increase of 54% from the 0WTR treatment (Figure 1.3A). The P ox of the MWTR and LWTR treatments were not significantly different than the 0WTR treatment or each other at T4. 26

39 Results of the co-blended poultry litter/wtr treatments on Al ox are summarized in Table 1.2 and Figure 1.3B. At T1 the mean Al ox was 976, 1221, 1618, and 4505 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The Al ox of the HWTR, MWTR, and LWTR treatments were all significantly higher than the 0WTR treatment and significantly different from each other with 362%, 65%, and 25% increases from the 0WTR treatment, respectively (Figure 1.3B). At T2 the Al ox was 1180, 1287, 1576, and 2853 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The Al ox of the HWTR treatment was significantly higher than all other treatments with a 142% increase from the 0WTR treatment (Figure 1.3B). The MWTR treatment was significantly higher than the 0WTR and LWTR plots with a 34% increase from the 0WTR treatment (Figure 1.3B). The Al ox of the LWTR treatment was not significantly different than the 0WTR treatment at T2. At T3 the Al ox was 1041, 1184, 1365, and 2789 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The Al ox of the HWTR treatment was significantly higher than all other treatments with an increase from the 0WTR of 168% (Figure 1.3B). The MWTR treatment was significantly higher than the 0WTR treatment with a 31% increase, though not significantly different from the LWTR treatment. The Al ox of the LWTR treatment was not significantly different than the 0WTR treatment at T3. At T4 the Al ox was 1006, 1098, 1254, and 3271 mg kg -1 for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The Al ox of the HWTR treatment was 27

40 significantly higher than all other treatments with an increase of 225% from the 0WTR treatment (Figure 1.3A). The Al ox of the MWTR and LWTR treatments were not significantly different than the 0WTR treatment or each other at T4. Results of the co-blended poultry litter/wtr treatments on P sat are summarized in Table 1.2 and Figure 1.3C. At T1 the mean P sat was 44.7%, 29.7%, 32.8%, and 22.3% for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The P sat of the HWTR, MWTR and LWTR treatments were all significantly lower than the 0WTR treatment with 50%, 27%, and 34% reductions, respectively, though the LWTR and MWTR treatments were not significantly different from each other (Figure 1.3C). At T2 the P sat ranged from 21.8% to 30.0% (Table 1.2). There was no significant difference in P sat due to the WTR treatment at T2. At T3 the P sat was 29.9%, 34.4%, 29.8%, and 23.5% for the control, low, medium, and high WTR treatments, respectively (Table 1.2). The P sat of the HWTR treatment was significantly lower than all other treatments with a 22% reduction from the 0WTR treatment (Figure 1.3C). The P sat of the MWTR and LWTR treatments were not significantly different from the 0WTR treatment at T3 though the MWTR treatment was significantly lower than the LWTR treatment (Figure 1.3C). At T4 the P sat ranged from 23.5% to 31.1% (Table 1.2). There was no significant difference in P sat due to the WTR treatment at T4. DISCUSSION Co-blending WTR with manure greatly reduced the transport of runoff dissolved phosphorus (RDP) and runoff total phosphorus (RTP) in surface runoff immediately after 28

41 manure surface application when the risk of P transport is greatest. Co-blending WTR as a best management strategy provided significant protection of water quality at T1 where the HWTR treatment reduced RDP and RTP by 97% and 64%, respectively (Figure 1.1). These are substantial reductions, and at WTR application amounts of less than 10%, coblending WTR with manure could be a useful tool for protection of water quality from P transport. Runoff total phosphorus was substantially reduced by the WTR co-blended treatment although not as much as RDP. This is most likely due to the physical transport of WTR bound-p in the runoff water. Previous work has suggested that WTR bound P is not bioavailable and should not adversely affect water quality (Agyin-Birikorang et al., 2007). The RDP and RTP reductions in this study are larger than previous coblending/simulated rainfall research (Gallimore et al. 1999). Gallimore et al. (1999) reported 42.7% reduction of RTP and similar percent reduction of RDP from a 44.8 Mg ha -1 application of WTR compared to the 64% reduction of RTP and 97% reduction of RDP from a 0.95 Mg ha -1 application amount of WTR in this study. Amounts of poultry litter P applied between these to studies were similar, 104 kg P ha -1 in Gallimore et al. (1999) and 101 kg P ha -1 in this study. Two potential reasons for the higher RDP reductions in this study are differences in WTR Al ox content and co-blending methods between the two studies. First, the WTR used in this experiment contained more than two times the Al ox than the one used in the Gallimore et al. (1999) study. The differences in WTR application amounts are not as drastic when put on an Al ox basis, the component of WTR responsible for P sorption (Dayton and Basta, 2003). In this study, HWTR was 29

42 blended to achieve a 2:1 Al ox to P molar ratio while Gallimore et al. (1999) was blended at an approximate molar ratio of 24:1. The Al ox to P ratio was selected for comparison because this ratio has been shown to be strongly related to manure P solubility and DP tranport in previous WTR co-blending studies (Ippolito et al., 1999; Elliott et al., 2002; Makris et al., 2005). Secondly, application methods between the two studies are not the same. In Gallimore et al. (1999) poultry litter and WTR were not co-blended prior to land application. Poultry litter was applied to plots and WTR was broadcast over the litter application. In this study, the WTR and poultry litter was mixed two weeks prior to land application. These results suggest that the increased contact of WTR and manure P prior to field application may increase the effectiveness of WTR co-blending. This is beneficial because less WTR may be necessary to achieve water quality benefits immediately following a surface application of manure when WTR is co-blended with the manure prior to land application. Also, timing the availability and spreading of WTR with the application of manure would not be necessary as the WTR could be directly loaded into manure storage structures where loading, unloading, and agitation equipment could provide thorough WTR mixing. This could make WTR co-blending as a BMP more convenient for both the municipality and the producer. Correct calibration of WTR co-blending is necessary to achieve the desired water quality benefits. If too little WTR is co-blended with manure, significant amounts of P transport can still occur as evidenced by LWTR at T1 in this study (Figure 1.1). Coblending more than enough WTR necessary to protect water quality can potentially waste WTR without additional water quality benefit and potential induce P deficiencies in crops 30

43 (Elliott and Dempsey, 1991; Ippolito et al., 1999). WTR treatment reduce alfalfa tissue P below levels of crop sufficiency (Figure 1.5) WTR co-blending amounts in this study were selected based on projected final P sat of the blended material. The HWTR, corresponding to 50% P sat, provided the greatest water quality benefit (Figure 1.1). This supports previous research which has shown relatively low manure P solubility and DP transport when WTR co-blended materials achieve a molar excess of Al ox relative to P ox (P sat < 100%) when compared to WTR co-blended materials that have a molar excess of P ox relative to Al ox (P sat > 100%) (Elliott et al., 2002; Makris et al., 2005). Using the Al ox :P ox molar ratio, or P sat, could provide a simple method of WTR co-blending calibration based on a calculation, rather than relying on empirical relationships between manure P solubility and WTR co-blending which change depending on WTR and manure properties (Dayton and Basta, 2005b). Runoff dissolved phosphorus from the 0WTR treatment at T1 was similar to values reported in simulated rainfall research which measured P transport from recently applied manure (DeLaune et al. 2004; Dayton and Basta, 2005). Over time the effect of the treatment and the manure application on RDP and RTP was diminished (Figure 1.1). Phosphorus transport from all plots was significantly less for sampling events T2, T3, and T4 (Table 1.2). The reduction in RDP and RTP with time is most likely due to loss of manure P. This process was likely accelerated by 6 inches of natural rainfall that took place between T1 and T2 at the field site, which is approximately 200% above normal rainfall for this region (NOAA, 2009). Significant loss of manure P from the 0WTR treatment is evidenced by the significant reductions in P ox on the 0WTR treatments 31

44 between T1 and T2 (Figure 1.3A; Table 1.2). This demonstrates importance of timing manure applications when runoff producing rainfall events are least likely. However, even immediately after manure application, when the risk of P transport is the greatest, WTR co-blending reduced P transport. After the impact of the manure application on P transport was diminished, the effect of WTR on P transport was no longer evident (Figure 1.1). However, examining the amounts and solubility of P between the 0WTR and HWTR treatments demonstrate a continued treatment effect throughout the duration of the experiment. The 0WTR treatment shows a large increase in P solubility immediately after manure application (Figure 1.2C). P solubility on the 0WTR treatment decreases with time supporting the natural attenuation of manure P. The WTR treatments greatly reduced P solubility along a general trend of decreasing P solubility with increase amounts of blended WTR, following previously established trends (Codling et al., 2000; Elliott et al, 2002). Soluble P remains fairly constant over time for plots treated with WTR. Water extractable P from the HWTR treatment did not significantly change after T2 (Table 1.2) This demonstrates the stability of WTR bound P over a 3 month period, although the stability of WTR bound P has been demonstrated to be stable for as long as 7.5 years (Agyin-Birikorang et al, 2007). This is consistent with the long-term stability of Al-P sorbed complex in the Al ox fraction of the WTR. At T4, soluble P on the HWTR treatment was 44% less than the 0WTR despite the HWTR treatment containing 33% more total P (Figure 1.4). 32

45 P sat has been shown to be a robust measure of the risk for P transport working across a wide range of soil with different P retention properties (Vadas et al., 2005; Hooda et al., 2000). Differences in P sat measurements are the result of interactions between oxalate extractable P and Al between treatments over time. At T1 there are significant reductions in P sat with increasing amounts of WTR, which was expected due to the Al ox content of WTR (Figure 1.3B). At T1 the P sat of the HWTR treatment is 50% of the 0WTR treatment (Figure 1.3B). By T2 this difference is not significant despite the HWTR treatment P sat remaining constant over time (Table 1.2). The 0WTR treatment P sat was dramatically reduced between T1 and T2 (Table 1.2). Since Al ox on the 0WTR plots did not significantly change, the changes in P sat reflect the loss of P due to transport, which can be seen by the significant reductions in P ox between T1 and T2 (Table 1.2). The P ox of the HWTR treatment was not reduced as dramatically and by T4 the HWTR treatment had 54% more P ox than the 0WTR plots (Figure 1.3A). This, along with 33% more total P on the HWTR treatment, is strong evidence to suggest that less P transport occurs over time when WTR is co-blended with manure. The increase in Al ox, and the stability of P ox over time when WTR was applied and the reductions of P ox when WTR was not applied, caused P sat measurements to remain the same across treatments and time after T1 (Table 1.2; Figure 1.3). In Ohio s current P-risk index B1P or M3P are one component used to characterize the risk of P loss. For this reason it is important to understand how WTR affects the results of these soil extractants. Previous research has shown that WTR can reduce M3P and B1P, although these reductions are less than reductions measured by 33

46 WEP because the acid ammonium fluoride solutions used in M3P and B1P extractions dissolve amorphous aluminum surfaces releasing previously absorbed P (Basta et al., 2000; Dayton and Basta, 2005b; Novak and Watts, 2005b). For this reason, M3P and B1P may underestimate the effect of WTR on P transport (Dayton and Basta, 2005b). B1P and M3P were both reduced by the HWTR treatment but not as much as WEP, supporting the results of these previous studies (Figure 1.2). One component of Ohio s P-index used to characterize the risk of P loss is manure application. The manure application rate and application method contribute to the overall P-risk index score. WTR co-blending as a best management practice is not included as a modifying factor for the potential contribution of manure P transport (Dayton and Basta, 2005b). The only potential impact WTR can have on the risk score is by reduction of B1P or M3P. In the case of recent manure application, soil testing does not adequately reflect P transport risk, but rather the amount and solubility of manure P applied (DeLaune et al., 2004). It is apparent with these results that Ohio s P-index should include a modifying factor to the manure contribution component of the overall risk score so the risk score reflects the reduced P transport when WTR is co-blended with manure. One potential solution is to modify the manure contribution based on relative differences between manure s P transport potential. Elliott et al. (2006) suggested the use of source coefficients, or manure contribution modifiers, derived from manure WEP to determine the potential risk to water quality. Inclusion of source coefficients allows 34

47 adjustments of risk index scores to better reflect potential for P transport from recent manure applications. Source coefficients could potentially be derived using the P sat of the manure or co-blended material, as evidenced by the trends in this study. 35

48 Table 1.1. Select properties of materials used and background soil at field location Property units Glynwood Soil WTR Poultry Litter Oxalate Extractable P g kg Oxalate Extractable Al g kg Oxalate Extractable Fe g kg P sat % WEP mg kg NA 1556 ph Phosphorus Saturation = ((oxalate extractable P)/(oxalate extractable Al + oxalate extractable Fe))*100 1g: 200 ml Deionized Water Extractable P 36

49 Table 1.2. Mean values of soil test P and runoff P for each treatment across sampling events. Different letters among the same treatment across sampling events, indicate significant differences at α < RDP RTP BP1 M3P WEP # P ox Al ox P sat mg L -1 mg kg -1 % T 0 Initial nd T 1 0WTR 32.9 A 35.4 A 242 A 593 A 51.7 A 1302 A 976 A 44.7 A LWTR 33.7 A 41.0 A 192 B 380 A 26.6 A 996 A 1221 A 29.7 A MWTR 10.5 A 19.9 A 214 A 490 A 29.6 A 1271 A 1618 A 32.8 A HWTR 1.00 A 12.7 A 163 A 306 A 21.1 A 1565 A 4505 A 22.3 A T 2 0WTR B 2.13 B 227 A 342 B 33.6 B 950 B 1180 A 26.7 B LWTR B 1.93 B 236 A 465 A 29.3 A 1065 A 1287 A 30.0 A MWTR B 1.70 B 221 A 419 AB 24.9 AB 1123 AB 1576 A 28.3 A HWTR C 1.30 B 167 A 314 A 18.4 AB 1170 A 2853 B 21.8 A T 3 0WTR B 2.45 B 156 B 358 B 23.6 B 893 B 1041 A 29.9 B LWTR B 3.50 B 185 B 441 A 25.0 A 1045 A 1184 AB 34.4 A MWTR B 1.91 B 145 B 348 B 21.4 B 1003 B 1365 AB 29.8 A HWTR C 1.63 B 128 B 280 B 14.5 B 1196 A 2789 B 23.5 A T 4 0WTR B 2.92 B 158 B 339 B 27.7 B 846 B 1006 A 27.8 B LWTR B 2.52 B 176 B 398 A 27.8 A 976 A 1098 B 31.1 A MWTR B 2.14 B 135 B 295 B 19.9 B 758 C 1254 B 27.6 A HWTR B 2.10 B 131 B 212 C 15.5 B 1303 A 3271 B 23.5 A Runoff Dissolved Phosphorus Runoff Total Phosphorus Bray-1 Extractable P Mehlich-3 Extractable P # 1g: 10mL deionized water extractable P Oxalate Extractable P Oxalate Extractable Al Phosphorus Saturation = ((oxalate extractable P)/(oxalate extractable Al + Fe))*100 37

50 38

51 39

52 40

53 41

54 42