<%BANNER%>

Characterizing the Long-Term Lability of Biosolids-Phosphorus

Permanent Link: http://ufdc.ufl.edu/UFE0022710/00001

Material Information

Title: Characterizing the Long-Term Lability of Biosolids-Phosphorus
Physical Description: 1 online resource (85 p.)
Language: english
Creator: Miller, Matthew
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

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

Notes

Abstract: Land-application is a critical disposal route for the biosolids generated by many municipalities and provides farmers with numerous agronomic benefits. However, heightened concern that biosolids land-application can potentially accelerate cultural eutrophication of surface waters threatens to limit land-based biosolids recycling programs. Previous short-term studies successfully distinguished the phosphorus (P) release characteristics of biosolids from highly soluble inorganic fertilizer sources, and illustrated how various biosolids have vastly different P release characteristics. However, the studies did not address the long-term lability of biosolids-phosphorus (biosolids-P). A prolonged (16-month) greenhouse study using large columns was conducted to characterize the long-term phytoavailability and environmental lability of biosolids-P (collectively referred to as overall P lability) and to provide understanding of the ultimate fate of biosolids-P. Seven biosolids and triple super phosphate (TSP) were used as P-sources. The biosolids selected represent a wide-range of P-solubility?percent water extractable-P (PWEP) values ranging from 0.58 to 47%. Sources of P were applied to a P-deficient, low P retention capacity Immokalee fine sand at three application rates: 56 kg P ha-1 (P-based rate), 112 kg P ha-1, and 224 kg P ha-1 (N-based rate). Bahiagrass (Paspalum notatum Flugge) grew continuously for 498 days after planting. Bahiagrass tissue was harvested every 4-8 weeks to characterize P uptake, and columns were leached immediately after each tissue harvest to characterize environmental P lability. Less soluble-P biosolids (PWEP ? 8.4%) were significantly less environmentally labile than biological P removal (BPR) biosolids, BPR-like biosolids, and TSP. The environmental lability of BPR and BPR-like biosolids-P was similar to TSP-P. The relative P phytoavailability (RPP) of less soluble-P biosolids was ~40-80% that of TSP, but BPR and BPR-like biosolids were 98-131% as phytoavailable as TSP. Less soluble-P biosolids pose significantly less environmental P risk than BPR and BPR-like biosolids, and TSP. Biosolids application rates should account for the relative P phytoavailability of less soluble-P biosolids, but no P-based application rate adjustment is warranted for BPR and BPR-like biosolids. Data from the greenhouse study suggest that the ultimate lability of less soluble biosolids-P is less than TSP-P, but BPR and BPR-like biosolids-P is ultimately as labile as TSP-P. A correlation of overall P lability to the labile P load (biosolids P saturation index x total-P load) suggests that biosolids P saturation index (PSI) is a useful a priori indicator of overall biosolids-P lability. Biosolids PSI can also provide a useful indication of the relative phytoavailability (RPP) of biosolids-P.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Matthew Miller.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: O'Connor, George A.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022710:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022710/00001

Material Information

Title: Characterizing the Long-Term Lability of Biosolids-Phosphorus
Physical Description: 1 online resource (85 p.)
Language: english
Creator: Miller, Matthew
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

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

Notes

Abstract: Land-application is a critical disposal route for the biosolids generated by many municipalities and provides farmers with numerous agronomic benefits. However, heightened concern that biosolids land-application can potentially accelerate cultural eutrophication of surface waters threatens to limit land-based biosolids recycling programs. Previous short-term studies successfully distinguished the phosphorus (P) release characteristics of biosolids from highly soluble inorganic fertilizer sources, and illustrated how various biosolids have vastly different P release characteristics. However, the studies did not address the long-term lability of biosolids-phosphorus (biosolids-P). A prolonged (16-month) greenhouse study using large columns was conducted to characterize the long-term phytoavailability and environmental lability of biosolids-P (collectively referred to as overall P lability) and to provide understanding of the ultimate fate of biosolids-P. Seven biosolids and triple super phosphate (TSP) were used as P-sources. The biosolids selected represent a wide-range of P-solubility?percent water extractable-P (PWEP) values ranging from 0.58 to 47%. Sources of P were applied to a P-deficient, low P retention capacity Immokalee fine sand at three application rates: 56 kg P ha-1 (P-based rate), 112 kg P ha-1, and 224 kg P ha-1 (N-based rate). Bahiagrass (Paspalum notatum Flugge) grew continuously for 498 days after planting. Bahiagrass tissue was harvested every 4-8 weeks to characterize P uptake, and columns were leached immediately after each tissue harvest to characterize environmental P lability. Less soluble-P biosolids (PWEP ? 8.4%) were significantly less environmentally labile than biological P removal (BPR) biosolids, BPR-like biosolids, and TSP. The environmental lability of BPR and BPR-like biosolids-P was similar to TSP-P. The relative P phytoavailability (RPP) of less soluble-P biosolids was ~40-80% that of TSP, but BPR and BPR-like biosolids were 98-131% as phytoavailable as TSP. Less soluble-P biosolids pose significantly less environmental P risk than BPR and BPR-like biosolids, and TSP. Biosolids application rates should account for the relative P phytoavailability of less soluble-P biosolids, but no P-based application rate adjustment is warranted for BPR and BPR-like biosolids. Data from the greenhouse study suggest that the ultimate lability of less soluble biosolids-P is less than TSP-P, but BPR and BPR-like biosolids-P is ultimately as labile as TSP-P. A correlation of overall P lability to the labile P load (biosolids P saturation index x total-P load) suggests that biosolids P saturation index (PSI) is a useful a priori indicator of overall biosolids-P lability. Biosolids PSI can also provide a useful indication of the relative phytoavailability (RPP) of biosolids-P.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Matthew Miller.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: O'Connor, George A.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022710:00001


This item has the following downloads:


Full Text

PAGE 1

CHARACTERIZING THE LONG-TERM LABILITY OF BIOSOLIDS-PHOSPHORUS By MATTHEW L. MILLER A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2008 1

PAGE 2

2008 Matthew L. Miller 2

PAGE 3

To my Dad whom GOD gave a st eadfast commitment to family, abundant patience, and most importantly, a sense of humor so that he could endure my time in school. 3

PAGE 4

ACKNOWLEDGEMENTS First, I must thank GOD. He instructs: b e strong and of a good courage; be not afraid, neither be thou dismayed: for the LORD thy God is with thee whithersoev er thou goest (Joshua 1:9). I know that, because I believe these words, I never lack His love, guidance, or help. Next, I must thank my advisor Dr. OConnor. Much is said about how demanding Dr. OConnor is of his students, but li ttle is said about how demanding he is of himself. No one can ever truthfully say Dr. OConnor demands more fr om others than he demands from himself. I thank Dr. OConnor for his contagious work ethic and for never allowing me to settle for less than my best effort. I would also like to thank my committee members (Dr. Harris and Dr. Elliott) for providing meaningful insights into my work, and for furnishing intellectual and physical resources to aid in my research. The environmental soil chemistry group (past and present) provided me with an unbelievable amount of help. Sampson, Scott, Sarah, Wally, Liz, Daniel, Xiaolin, Jaya, Manmeet, Augustine, and Ying are gr eat friends and coworkers. I could not have asked to be associated with better colleagues. I thank Sarah for showing me the ropes and making lab as entertaining as possible during my first year in the group. Thanks also go to Gavin for helping with the tissue nitrogen determinations. I greatly a ppreciate the help I received from Jason at the greenhouse. Special thanks must go to the two lab managers I had the privilege of working under, Dr. Sampson Agyin-Birikorang and Scott Brinton. Sampson and Scott are extremely patient, hard-working individuals who never hesitated to help me. Most of my research success can be attributed to the hands-on educa tion I received from Scott and Sampson. 4

PAGE 5

I would like to thank Milorganite, Inc., fo r generously supporting my research. Thanks also go to the Soil and Water Science Departme nt and the College of Agricultural and Life Sciences for providing my res earch assistantship. Finally, I thank my amazing family and friends. I thank my mother and sister for their unconditional love and support. My uncle, Dan, treated me like his own son, and provided more for my education than I ever deserved. My fath er earned every bit of gr atitude that I can give and more. I will never forget that Dad worked more than 25 hours over little more than 2 days helping me build rainfall simulation boxes dur ing one of my weeke nd trips home. Few young men have the privilege to be blessed with a father who gives so much to his sons education. 5

PAGE 6

TABLE OF CONTENTS Page ACKNOWLEDGEMENTS.4 LIST OF TABLES...7 LIST OF FIGURES.8 LIST OF ABBREVIATIONS ABSTRACT...................12 CHAPTER 1 INTRODUCTION.........14 Biosolids Land Application under P-based Management..16 Hypotheses and Research Objectives Study Approach.............................20 2 MATERIALS AND METHODS.......................22 Greenhouse Study Design..............................22 Leachate and Bahiagrass Tissue Analyses.....................25 Soil Analysis..............................26 Root Mass Determinations.........................27 Quality Assurance/Quality Control....27 Statistical Analysis.................28 3 RESULTS AND DISCUSSION....................32 Tissue Yield...33 Tissue P Concentrations.35 Environmental P Lability...............................38 Biosolids-P Phytoavailability....................40 Overall P Lability.......................48 4 CONCLUSIONS...................74 APPENDIX A R SQUARED VALUES FOR VARIOUS CORRELATIONS.78 B ADDITIONAL BIOSOLIDS PSI CORRELATIONS...79 LIST OF REFERENCES...................81 BIOGRAPHICAL SKETCH.85 6

PAGE 7

LIST OF TABLES Table Page 2-1 Treatment processes used to produce the selected biosolids ....30 2-2 Various measures of P for the P-sources utilized..............30 2-3 Selected chemical properties of the Immokalee A horizon and base sand....................31 3-1 Time series analysis of agronomic and environmental P lability for the N-based application rate...56 3-2 Time series analysis of agronomic a nd environmental P lability for the P-based application rate...57 3-3 Slope-ratio estimates of biosolids relative P phytoavai lability for harvest 4 and harvest 12........64 3-4 Point estimates of biosolids relative P phytoavailability following harvest 4 and harvest 12....65 3-5 Slope-ratio estimates of bi osolids relative P lability fo r harvest 4 and harvest 12....70 3-6 Point estimates of biosolids relative P lability for harvest 4 and harvest 12.....71 A-1 Relationship between various measures of biosolids-P with cumulative P uptake and overall P lability..........78 A-2 Relationship between various measures of biosolids-P with estimates of biosolids relative P phytoavailability (RPP) and biosolids relative P lability (RPL)........................78 7

PAGE 8

LIST OF FIGURES Figure Page 3-1 Cumulative bahiagrass tissue yields following the 16-month greenhouse study.. 3-2 Bahiagrass tissue N contents for harvests 1-4 3-3 Bahiagrass tissue N concen trations for harvests 5-12 3-4 Yield-weighted tissue N concentrat ions for N-based rate treatments 3-5 Bahiagrass yield-weighted tissue P concentrations (harvests 1-11) for the P-based and control treatments......................58 3-6 Bahiagrass root masses for N-based (224 kg P ha-1) and P-based (56 kg P ha-1) rate treatments ......................59 3-7 Cumulative P leached as a function of P-source for P-based rate (56 kg P ha-1) treatments and N-based rate (224 kg P ha-1) treatments....................................................60 3-8 Environmental P lability (P leached) differences among P-sources at different rates...61 3-9 Cumulative bahiagrass P uptake as a f unction of P-source for P-based rate (56 kg P ha-1) treatments and N-base d rate (224 kg P ha-1) treatments.............................62 3-10 Cumulative P uptake diff erences among P-sources ..........................63 3-11 Relative P phytoavailability curves for P-sources.64 3-12 Cumulative P uptake plotted as a function of the labile P load (biosolids PSI*total-P load)...66 3-13 Biosolids relative P phytoavailabi lity (RPP) values plotted as a function of biosolids phosphorus saturation index (PSI) values.66 3-14 Relationship between cumulative P uptake and the labile P load for two previously conducted s hort-term studies greenhouse studies..........67 3-15 Long-term estimates (Oladeji, 2006; cu rrent greenhouse study) and short-term estimates (OConnor et al., 2004; Chin ault, 2007) of bios olids relative P phytoavailability (RPP) plot ted as a function of bioso lids phosphorus saturation index (PSI) values..67 3-16 Overall P lability (agronomic and en vironmental) of various P-sources for P-based rate (56 kg P ha-1) treatments and N-base d rate (224 kg P ha-1) treatments.....68 8

PAGE 9

3-17 Overall P lability (agronomic and en vironmental) differences among P-sources. 3-18 Relative P lability curves for P-sources 3-19 Overall P lability plotted as a function of the labile P load...........72 3-20 Biosolids relative P lability (RPL) plo tted as a function of biosolids phosphorus saturation index (PSI)....72 3-21 Relationship between overall P lability and the labile P load for two previously conducted short-term studies greenhouse studies......................................................73 3-22 Cumulative P leaching from a 5.5 laboratory incubation study plotted as a function of the environmentally effective P load. B-1 Long-term estimates of biosolid s relative P phytoavailability (RPP) plotted as a function of biosolids PSI values.79 B-2 Short-term estimates of biosolids re lative P phytoavailability (RPP) plotted as a function of biosolids PSI values.80 B-3 Cumulative P uptake from a short-term greenhouse study utilizing a Candler soil with a relative P adsorption capac ity of ~15% (OConnor et al., 2004)............80 9

PAGE 10

LIST OF ABBREVIATIONS Al Aluminum AN Ammonium nitrate Biosolids-P Biosolids-phosphorus BPR Biological phosphorus removal Ca Calcium CRD Completely randomized design EC Electri cal conductivity EPA Environmental Protection Agency FDEP Florida Department of Environmental Protection Fe Iron GLM General linear model GRU Gainesville Regional Utility K Potassium Mg Magnesium MCP Monocalcium phosphate N Nitrogen NRCS National Resource Conservation Service OCUD S Orange County south P Phosphorus PAN Plant available nitrogen PSI Phosphorus saturation index PSR Phosphorus saturation ratio 10

PAGE 11

PWEP Percent water-extractable phosphorus QA/QC Quality assurance/quality control RPA Relative phosphorus adsorption RPP Relative phosphor us phytoavailability RPL Relative phosphorus overall lability SAS Statistical analysis software SRP Soluble reactive phosphorus S Sulfur TC Total carbon TN Total nitrogen TP Total phosphorus TSP Triple super phosphate USEPA United States Envir onmental Protection Agency WEP Water-extractable phosphorus 11

PAGE 12

Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science CHARACTERIZING THE LONG-TERM LABILITY OF BIOSOLIDS-PHOSPHORUS By Matthew L. Miller August 2008 Chair: George A. OConnor Major: Soil and Water Science Land-application is a critic al disposal route for the biosolids generated by many municipalities and provides farmers with numerous agronomic benefits. However, heightened concern that biosolids land-a pplication can potentially acceler ate cultural eutrophication of surface waters threatens to limit land-based biosolids recycling programs. Previous short-term studies successfully distinguish ed the phosphorus (P) release char acteristics of biosolids from highly soluble inorganic fertiliz er sources, and illustrated how various biosolids have vastly different P release characteristics. However, th e studies did not address the long-term lability of biosolids-phosphorus (biosolids-P). A prol onged (16-month) gree nhouse study using large columns was conducted to characterize the l ong-term phytoavailability and environmental lability of biosolids-P (collectively referred to as overall P lability) and to provide understanding of the ultimate fate of biosolids-P. Seven biosolids and triple super phosphate (TSP) were used as P-sources. The biosolids selected repr esent a wide-range of Psolubilitypercent water extractable-P (PWEP) values ranging from 0.58 to 47%. Sources of P were applied to a P-deficient, low P retention capacity Immokalee fine sand at three application rates: 56 kg P ha-1 (P-based rate), 112 kg P ha-1, and 224 kg P ha-1 (N-based rate). Bahiagrass (Paspalum notatum Flugge) grew continuously for 498 days after planting. 12

PAGE 13

Bahiagrass tissue was harvested every 4-8 weeks to characterize P uptake, and columns were leached immediately after each tissue harvest to characterize environmental P lability. Less soluble-P biosolids (PWEP 8.4%) were significantly less e nvironmentally labile than biological P removal (BPR) biosolid s, BPR-like biosolids, and TSP. The environmental lability of BPR and BPR-like biosolids-P was similar to TSP-P. The relative P phytoavailability (RPP) of less soluble-P biosolids was ~40-80% that of TSP, but BPR and BPR-like biosolids were 98131% as phytoavailable as TSP. Less soluble-P biosolids pose significan tly less environmental P risk than BPR and BPR-like biosolids, and TSP. Biosolids application rates should account for the relative P phytoavailability of less solubl e-P biosolids, but no Pbased application rate adjustment is warranted for BPR and BPR-like biosolids. Data from the greenhouse study suggest that the ultimate lability of less soluble biosolids-P is less than TSP-P, but BPR and BPR-like biosolids-P is ultimately as labile as T SP-P. A correlation of overall P lability to the labile P load (biosolids P saturation index x total-P load) suggests that biosolids P saturation index (PSI) is a useful a priori indicator of overall biosolids-P lability. Biosolids PSI can also provide a useful indication of the relative phytoavailability (RPP) of biosolids-P. 13

PAGE 14

CHAPTER 1 INTRODUCTION Land-based biosolids recycling programs are critical to municipalities. Wastewater treatment produces 6.5 x 106 dry Mg of biosolids annually in the US, and production is expected to increase to 8.2 x 106 Mg by 2010 (USEPA, 1999; Epstein, 2003). Traditionally, municipalities relied on three main biosolids disp osal options: land application, landfill disposal, and incineration (Elliott et al., 2007). Curtaili ng of biosolids landfill disposalmostly due to high costs and limited available landfill capacit yis occurring in many states (Epstein, 2003), although there appears to be a resurgence in land-fill ing in other states (Elliott et al., 2007). New Jersey currently allows biosolids landfill disposal only under emergency conditions (Elliott et al., 2007). Incineration is constrained by high operating and capital cost s, and air quality concerns. The limitations of land-fill disposal and incineration are prohibitive to many municipalities, and emphasize the importance of maintaining viable land-based recycling programs. Converting municipal wastes into biofuels is an increasingly attractive di sposal option (Champagne, 2007). Municipalities are often attracted to processes used to convert biosolids into biofuels because of advantages unrelated to energy production. A deep-well bioso lids injection pr oject in Los Angeles, California reduced poli tical opposition from communities that might receive biosolids because municipal wastes are treated and managed onsite (Attai et al., 2008). Land application is the most common biosolids disposal route utilized by municipalities. Fifty to sixty percent of biosolids produced annu ally in the US are land applied (Epstein, 2003). Florida land applies 66% of the more than 3.0 x 105 dry Mg of biosolids produced annually (FDEP, 2007). Land application is federally regulated under the EPA Title 40 CFR Part 503 rule, and extensive risk assessment documents the safety of biosolids land application practices conducted in accordance with existing regul ation (USEPA, 1993a; USEPA, 1995; NRC, 2002; 14

PAGE 15

Moss et al., 2002). Land applying biosolids provides agronomic benefits, making land application programs mutually bene ficial to farmers and municipa lities. The organic matter in biosolids (most biosolids are ~ 50-80% organic matter) increases so il cation exchange capacity, and beneficially affects soil physical properties such water retention, porosity, and bulk density (Moss et al., 2002). Biosolids contain macronutrients, micronutrients, and trace elements, and are useful as low grade fert ilizers (Elliott and OConnor, 2007) Low costs and agronomic benefits make biosolids an economically favorable alternative to inorga nic fertilizers. Nutrients are the major benefit to farmers i nvolved in land-based recycling programs, yet paradoxically, nutrient concerns are the primary th reat to biosolids land application (Shober and Sims, 2003; Brandt et al., 2004; Elliott and OConnor, 2007). Bioso lids are generally applied to meet crop nitrogen (N) requirements in accordan ce with the EPA part 503 rule (USEPA, 1993a), which restricts application ra tes exceeding agronomic N dema nds. Applying biosolids at Nbased rates generally exceeds the rate of crop phosphorus (P) removal, resulting in an accumulation of soil-P (OConnor et al., 2004). Exce ss soil P is not harmful to crops (Peterson et al., 1994); however, accelerated eutrophication of most freshwat ers is limited by P inputs, and high-P soils represent an increase d risk for non-point pollution of surface waters (Sharpley et al., 1994; Sharpley et al., 1996). Increasingly stringent effluent P regul ations are exacerbating soil-P accumulation problems in biosolids amended soils (Elliott et al ., 2002). Decreased effluent P concentrations result in more P partitioning to biosolids, decreas ing the N:P ratio of the materials (Stehouwer et al., 2000). Decreasing N:P ratios in biosolids increases P loadi ng and soil-P accumulation when biosolids are applied at N-based rates. The wastewater treatm ent method used to meet the 15

PAGE 16

increasingly stringent effluent P concentration affects the Prelease characteristics of the biosolids (Maguire et al., 2001; Brandt et al., 2004). Heightened concern that agricultural P losses are adversely affecting surface waters is prompting regulators to utilize a variety of P management tools to reduce non-point source P pollution. The P index is a P management tool de veloped by the U.S. Department of Agriculture Natural Resource Conservation Service (NRCS) to assess the potential for P loss from agricultural fields (Lemunyon and G ilbert, 1993). The P index uses a field-scale scoring matrix to identify sites vulnerable to P-lo ss. Most states, including Flor ida, chose to adopt a phosphorus indexing approach to environmental P management (Elliott and OConnor, 2007). Biosolids Land Application under P-based Management The P index dictates applicat ion rates based on crop P demands (P-based rates) on soils with high P loss risks so that P loading is reduced on soils particularly susceptible to P losses. Accurate determinations of biosolids relative P phytoavailabilitythe plant availability of biosolids-phosphorus (biosolids-P) compared to fertilizer-Pis necessary if the P index forces land-based recycling programs to apply biosolids at P-based rates. Accounting for relative P phytoavailability (RPP) determinatio ns assures that biosolids-P is applied at rates agronomically equivalent to fertilizer-P recommendations (OConnor et al., 2004). Sarkar and OConnor (2004) found that a high native-P Millhopper sand with high P-sorption capacity masked the effects of P-source additions; however, P-source effects were significant in the moderate Psorption capacity P Candler sand, and clearly expressed in a lo w P-sorption capacity Immokalee fine sand. The relative efficacy of P-sources is of particular interest in Florida where many sandy soils possess little P retention capacity to mask P release differences among P-sources. 16

PAGE 17

The USEPA (1995) assigns a 50% agronomic r elative effectiveness value to biosolidsP compared to fertilizer-P. Wen et al. (1997) found sewage sludge-P was less available than manure-P to crops, and hypothesized the decreased phytoavailability was due to the greater Fe and Al content in biosolids. OConnor et al (2004) suggested groupi ng biosolids into three categories of phytoavailability rela tive to the inorganic fertili zer triple super phosphate (TSP): low (<25% of TSP), moderate (25-75% of T SP), and high (>75% of TSP). A four-month greenhouse study identified most biosolids as bei ng in the moderate category, BPR biosolids as being in the high category, and so me biosolids with uniquely high total Fe+Al concentrations as being in the low category. A tw o-year field study validated th e short-term RPP differences suggested by the greenhouse study (OConnor and Elliott, 2006). Chinault (2007) further verified short-term RPP differences among a dditional P-sources in a four-month greenhouse study: the relative P phytoavailab ility of BPR and BPR-like materi als was similar to TSP, and conventional biosolids-P was less phytoavailable than TSP-P. Results of the studies suggest increasing application rates of non-BPR biosolids are necessary to supply plant available P in quantities equivalent to TSP. The long-term availa bility of biosolids-P, however, is incompletely characterized. The P index considers both P transport m echanisms and P source characteristics in evaluating environmental risks, and maintains that significant P loss only occurs where a P transport pathway and a labile P source coexist (S harpley et al., 2003). Most states utilizing the P-index have soils with apprec iable P sorbing capacity and P leaching is not considered an important P loss mechanism (Peterson, 1994). Elliott et al. (2002) showed that even soils with only moderate P sorbing capacity can limit P leac hing. Florida is dominated by sandy soils with low native P-sorbing capacity, and P-leaching is a critical P loss mechanism (He et al., 1999). 17

PAGE 18

Phosphorus can migrate through sandy Florida so ils to shallow groundwat er tables that are hydrologically linked to surface wa ter bodies (Harris et al., 1996; Sims et al., 1998; Lu and OConnor, 2001; Elliott et al., 2002 ). The potential for P-leachi ng results in high P transport risks in the soils; thus, the envir onmental lability of P-sources is a critical factor controlling P loss. Studies that account for P-leaching losses and that utilize soils with minimal P-retention capacity (to reduce the soils abi lity to mask P-source effects) are the most relevant for accurately distinguishing environmental and agrono mic P lability differences among P-sources. Column P-leaching from an Immokalee fine sand (sandy, siliceous, hyperthermic Arenic Alaquods) amended with six conventional biosolid s at N-based rates was less than 1% of Papplied and not statistically differe nt from controls. In contrast, 21% of the P-applied leached in columns amended with TSP (Elliott et al., 2002). Chinault (2007) conducted a laboratory incubation study to assess the ultimate lability of biosolids-P in a worsecase scenario; the study soil had minimal P-retention capacity, and no crops were grown to remove labile P. Leaching continued until there was no further change in drainage-P concentration, suggesting ultimate release. Results suggested that conventional biosolids are ultima tely less labile than TSP, and ultimate P-release from BPR and BPR-like biosolids is similar to TSP. The unique P-release characte ristics of biosolids is not explicitly acknowledged by the P indices used by most states, de spite abundant evidence suggesting P lability depends critically on whether P originates from inor ganic fertilizers, manures, or biosolids (Elliott et al., 2002; OConnor and Elliott, 2006; Oladeji, 2006). Two clear challenges remain to the inclusion of factors accounting for the P-release characteristics of biosolids in P indices. The first challenge is that limited long-term data exist to validat e short-term lability differences among P-sources. 18

PAGE 19

Concerns remain that all biosolids-P will ultimat ely become plant available, and that regulatory distinctions permitting some biosolids to be appl ied at rates in excess of crop P demands are not justified. McLaughlin and Champion (1987) found the relative efficacy of sewage sludge P increased over time (8 croppings over 209 days ) from 44 to 90% of monocalcium phosphate (MCP) and from 64 to over 100% of MCP in two Oxisols. A long-term study is necessary to distinguish ultimate P-lability differences am ong P-sources and to address concerns that biosolids-P is ultimately bioavailable. The second challenge is determining a useful a priori measure of ultimate biosolids-P lability. Empirically determining long-term lability differences among all biosolids is impractical; thus, an a priori measure of ultimate biosolids-P lability is a necessary tool to distinguish P-release differences among P-sources. Studies show water extractable P (WEP) is highly correlated to runoff and leachate P in manures and manure-amended soils, and could provide a useful indication of P loss potential in waste-amended soils (Kleinman et al. 2002; Sharpley and Moyer, 2000). Biosolids typi cally have lower percent water extractable phosphorus (PWEP) values than manures and TSP, suggesting biosolids represent smaller P loss risk than highly soluble inorganic fertilizers or typical animal manures (Brandt et al., 2004). Biosolids PWEP values were strongly correlated to runoff and leaching P loss in two rainfall simulation studies (Oladeji, 2006; Agyin-Birikor ang, 2008). Elliott et al. (2002) found biosolids phosphorus saturation index (PSI)th e ratio of oxalate extractable P to oxalate extractable iron and aluminumis a useful gauge of biosolids P-l eaching potential in low P-sorbing soils. A lab incubation study suggested that PSI and PWEP were good qualitative in dicators of ultimate biosolids-P lability, but establ ished no clear quantitative re lationship (Chinault, 2007). Moreover, no a priori measure exists to determine the relative phytoa vailability of 19

PAGE 20

biosolids-P. Citric acid-extractable P, used to ch aracterize plant available P in most states, is a poor indicator of biosolids RPP (OConnor et al., 2004). OConnor and Elliott (2006) found estimates of RPP tracked well with P-source WEP values in short-term greenhouse study. Yet, a two-year field study identified total-P as the most important variab le in accounting for differences in RPP (Oladeji, 2006). An accurate a priori measure of biosolids RPP would provide a useful tool to adjust biosolids application rates so th at the quantity of biosolids-P applied is equivalent to reco mmended fertilizer-P rates. Hypotheses and Research Objectives Previous studies successfully characterized the short-term P-release characteristics of different biosolids. However, characterizing the long-term phytoavailability and environmental lability of biosolids-P is necessary to validat e P-release differences suggested by short-term studies, and to determine the ultimate lability of biosolids-P Hypothesis 1: The long-term phytoavailability and environmental lability of conventional biosolids-P is less than fertilizer-P (TSP-P). The long-term phytoavailability and environmental lability of BPR and BPR-like biosolids-P is similar to TSP-P. Hypothesis 2: Short-term greenhouse and lab incuba tion studies adequately approximate the relative phytoavailability and envi ronmental lability of biosolids-P. Hypothesis 3: Some measure of biosolids-P is a us eful indicator of ultimate biosolids-P lability and biosolids-P phytoavailability. Objective 1: Characterize the long-term phytoavailabi lity and environmen tal lability of biosolids-P. Objective 2: Compare long-term characterizations of biosolids-P release to previous short-term characterizations. Objective 3: Correlate various measures of bioso lids-P to ultimate biosolids-P lability, and select the best a priori measure of ultimate biosolids-P release. 20

PAGE 21

Study Approach The phytoavailability and environmental lability of eight P-sources (7 biosolids and TSP) were evaluated. The selected biosolids repres ent a wide-range of P solubility: low soluble P (high Fe+Al materials) to high soluble P (B PR and BPR-like biosolids). Bahiagrass ( Paspalum notatum Flugge) was grown in columns containing Imm okalee fine sand (low native P-fertility and P-sorbing capacity) for 16 months. Bahiagrass tissue was harvested every 4 to 8 weeks, and P-uptake was determined to characterize P-source phytoavailability. Environmental lability was characterized by P leaching after each harvest. Overall biosolids-P lability is represented by cumulative mass of P removed from the column in bahiagrass tissue and column leachate. Previously determined characterizations of bios olids-P were correlated to P-uptake and overall P lability. 21

PAGE 22

CHAPTER 2 MATERIALS AND METHODS Greenhouse Study Design A greenhouse experiment was utilized to ch aracterize the environm ental lability and phytoavailability of biosolids-P. The greenhous e study was established as an 8 x 3 factorial experiment in a complete randomized block desig n. The study was arranged in 4 blocks; each block contained 1 treatment replicate. A control column, which received no P-source application, was placed in each of the four blocks. Seven biosolids were selected as P-sources The biosolids selected represent a wide range of wastewater treatment processes and P solubility (Tables 2-1, 2-2). The Gainesville Regional Utility (GRU), Boca Raton, Lakela nd NS, and Orange County South (OCUD S) biosolids possess PWEP values gr eater than 15% and are herein considered as more soluble-P biosolids. The Boca Raton biosolids is a BPR biosolids, and the GRU, Lakeland NS, and OCUD S biosolids are BPR-like biosolids (Chinaul t, 2007). Milorganite and Greenedge are commercial, thermally dried biosolids produced from conventional wastewater treatment processes. Milorganite and Gr eenedge biosolids possess PWEP values less than 1.1% and are herein considered as less solubl e-P biosolids. The Disney is a composted biosolids with a PWEP value of 8.4% and is also herein considered as a less soluble-P material Phosphorus saturation index (PSI) values also vary greatly among th e biosolids selected, ranging from 0.64 to 2.9. Triple super phosphate (TSP) was al so selected as P-source to comp are the lability of biosolids-P to highly soluble (PWEP = 85 %) inorganic fertilizer-P. The P-sources were mixed with 4 kg of the A horizon of the Immokalee fine sand. The Immokalee fine sand was select ed to represent a typical sa ndy Florida soil with minimal P content and minimal P-sorption capacity (Chinau lt, 2007). A base sand commonly used to top22

PAGE 23

dress turfgrass greens was selected to represent an minimally P-reactive E horizon underlying the Immokalee fine sand A horizon. The Immokal ee A horizon and the base sand contained very low quantities of native-P, and possessed little P sorption capacity (Table 2-3). Utilizing low P retention capacity, low native-P soils was important for characterizing ultimate P lability differences among P-sources (de Haan, 1980; Sa rkar and OConnor, 2006). The sandy, low P sorption capacity soils were also chosen to maximize environmental P leaching risks. Sources of P were applied to the Immokalee A horizon at three rates, equivalent to 56 kg P ha-1, 112 kg P ha-1, and 224 kg P ha-l. The 56 kg P ha-1 and 224 kg P ha-1 application rates represent P-based and N-based rates, respectively. The 112 kg P ha-1 rate was included to better define the biosolids relative P phytoavailability and lability response curves. The biosolidsamended Immokalee A horizon was wetted to ap proximately field capacity, and allowed to equilibrate in zip-lock bags for 2 week s prior to use in the greenhouse study. Large columns were utilized as soil containers for the greenhouse study. The columns consisted of 45 cm long, 15 cm di ameter polyvinyl chloride (PVC ) pipes with a screen at the bottom to support overlying soil. Columns were fitte d with PVC caps that contained tubes at the bottom to direct leachate into collection bottles. Columns were filled with 30 cm (~ 8 kg) of base sand to simulate a native E horizon (bulk density 1.5 g cm-3) and to allow adequate rooting depth for the pasture grass. The base sa nd was saturated and allowed to drain to remove readily soluble constituents. The 4 kg of incubated, P-source-amended Immokalee A horizon (bulk density 1.5 g cm-3) was placed on top of the 30 cm of base sand in each column. Soil columns were seeded with 5 grams of bahiagrass seed (~ 29 Mg seed ha-1 equivalent). The seeding rate exceeded the 11 Mg seed ha-1 field rate recommended by Chambliss et al. (2001) to ensure thorough bahiagrass coverage of the soil and to maximize P-uptake potential. Seed was 23

PAGE 24

covered with a thin layer of soil, and misted about every 3-4 hours until germination occurred. Following germination, the grass was watered using daily to semi-daily applications of pH 5.0 tap water. The tap water was analyzed prior to use in the greenhouse study, and the soluble P concentration in the tap water was be low the detection limit of 0.01 mg P L-1 (Chinault, 2007). Soil moisture content was maintained at ~80% of column pot-holding capacity mass throughout the study by bi-weekly weighing of the columns and adding water as necessary. Excessive watering was carefully avoided to prevent accidental leaching even ts. Columns were rotated one position within their respective blocks every week to minimize potential greenhouse positioning effects. Biosolids-N concentrations varied among bios olids; thus, applying biosolids at uniform P-based rates resulted in varying applicati ons of plant available nitrogen (PAN). A 40% biosolids-N mineralization rate was assumed fo r all biosolids (OConnor and Sarkar, 1999). Ammonium nitrate (AN) was used to supplem ent biosolids-N supply so that a uniform N application rate of 300 kg N ha-1 was supplied to all columns per cropping (4 harvests). Total N additions were substantially greater than the 179 kg N ha-1 rate recommended by Kidder et al. (1998) for bahiagrass to ensure that PAN supply was uniform among all treatments and that N supply would not limit P uptake. Ammonium nitr ate was applied in split-application; 75 kg N ha-1 was applied each harvest (300 kg N ha-1 was applied per cropping). A fertilizer known as sul-po-mag (22% S, 18% K, and 11% Mg) was also added to columns at the equivalent rate of 444 kg ha-1 after each harvest to supply bahiagrass with ample and uniform amounts of sulfur, potassium, and magnesium. 24

PAGE 25

Leachate and Bahaigrass Tissue Analyses Bahiagrass was planted June 12, 2006 and the grass grew continuously for 498 days until the study was terminated on October 18, 2007. Bahi agrass tissue harvests occurred every 4-8 weeks. Bahiagrass tissue was cut to a height of 3.8 cm above the soil surface for each harvest. Harvested tissue was placed in preweighed paper bags and dried at 68o C to constant weight as a measure of grass yield Immediately following each tissue harvest, columns were leached with sufficient pH 5.0 water to yield approximately 500 mL ( pore volume) of drainage (leachate), collected in 1 L collection bottles beneath the columns. The bo ttles were weighed to determine the exact volume of leachate. A ~250 mL aliquot of leachate was used in laboratory analyses. Soluble reactive P (SRP), el ectrical conductivity (EC), and pH were determined for all leachate samples within 48 hours of leachate colle ction. The mass of P leached was determined as the product of SRP in the leachate and l eachate volume. The concentration of SRP was determined using the ascorbic acid method (Murphy and Riley, 1962). Total P (USEPA, 1993b) was also determined on darkly colored leachate samples to confirm that SRP constituted the vast majority of total leachate-P. Plant uptake of P was determined as the pr oduct of plant tissue P concentration and dry tissue yield. Dried plant tissue was ground in a Wiley mill to pass through a #20 (0.85 mm) sieve, ashed, and the ash digested following A ndersen (1976). Total P was determined in the digest using the ascorbic acid method (Murphy an d Riley, 1962). Total-N concentrations were determined on the harvested bahiagrass tissue fro m the N-based treatments and controls. The ground bahiagrass tissue was ball-milled to a fine powder, and N concentrations determined by combustion at 1010oC, according to Nelson and Sommer s (1996), using a Carlo Erba TC/TN analyzer (NA-1500 CNS, Carlo Erba Milan, Italy). 25

PAGE 26

Soil Analysis Three 2.5 cm diameter soil cores were taken from each column following the last harvest. The first soil core was taken from 0-45 cm and represented the entire soil profile. The 0-45 cm soil core was utilized for a mass balance analysis The second soil core was taken from ~0-15 cm to characterize the Immokalee A horizon. The third soil core was taken from the base sand layer at a depth interval of ~15-45 cm. A st ark contrast existed between the Immokalee A horizon and the base sand, and great care was taken not to mix the Immokalee A horizon and base sand during sampling. The Immokalee A horizon and base sand were analyzed separately to determine the quantity of P retained in the base sand layer, and to characterize the quantity of P that moved out of the Immokalee A hor izon initially amended with P-sources. Wet soil samples were stirred regularly and allowed to air-dry for 3 weeks. Following air-drying, the 0-45 cm soil sample s were sieved (0.85 mm) to remove plant roots. Plant roots were ground using a household coffee bean grinder, and then thoroughly mixed back into the soil samples. Roots were included in the 0-45 cm sa mples so that the mass balance analysis would represent the total-P remaining in the column so il profile. Roots were generally present in clumps, and obtaining a representative sample of the soil/root mixture was impossible if roots clumps remained intact. Roots were ground so th at an even mixing of so il and roots would occur and a representative sample could be obtained. Plants roots present in the 0-15 cm and 15-45 cm samples were removed mostly by sieving so that essentially, only soil total-P was determined. Soil samples were ashed and digested accordi ng to Andersen (1976), and P was determined using the ascorbic acid method (Murphy and Riley, 1962). 26

PAGE 27

Root Mass Determinations Column root masses were quantified following soil sampling to determine if Papplication rate and/or P-source lability affected root prolifer ation within the columns. The bahiagrass roots mats were sufficiently thick to a llow the root mats to be carefully pulled from the columns. The columns were wetted thorough ly to promote separating the roots mats from the columns. The roots mats were placed on a #10 (2.00 mm) sieve and soil was washed from the roots using a garden hose. The roots mats were washed over a large plastic wash tub to collect material passing through the sieve. The roots were repeatedly washed until a visual assessment indicated the soil was removed. Fine roots that washed through the sieve floated on the surface of the water in the large plastic wash tub. The wash tub water was passed through a #20 sieve to collect fine roots. The coarse and fine roots were placed into pre-weighed bags, and dried at 68oC to constant weight. Quality Assurance/Quality Control All data were obtained in strict accordance to standard quality assurance/quality control (QA/QC) protocol (Kennedy et al ., 1994). Total-P recoveries fr om soil and tissue digestions were determined using standard reference ma terials (Standard reference materials 1547 and 2709, National Institute of Standards and Technology). Total-P recoveries from the standard reference materials were 85% to 101% of the certified va lue. Blanks and replicates were also incorporated into soil and tissue digestions according to standard QA/QC protocol. All calibration curves for colorimetric P determinations achieved an r2 .9999. Method reagent blanks, spikes, replications, and certified standards were included in colorimetric P determinations. Recoveries from certified standards and spikes were within 5% of the expected 27

PAGE 28

value. The relative standard deviations of rep licated samples were less than 5%. Analyses that did not satisfy standard QA/QC protoc ol were repeated until all QA/QC crit eria were satisfied. Statistical Analysis Differences among treatments were statistically analyzed as a factorial experiment with a completely randomized design (CRD), using th e general linear model (GLM) of the SAS software (SAS Institute, 2007). The means of the various treatments were separated using Tukeys mean separation procedur e (Tukey, 1949) at a significance ( ) level of 0.05 The data were tested for the normal distribution and consta nt variance assumptions of analysis of variance using the normal probability plots and the residual plots, respectively (SAS Institute, 2007). Data that did not conform to the normality and hom ogeneity assumptions were appropriately transformed based on the results of the BoxCox transformation procedure (Box and Cox, 1964) before statistical analysis. Regression analyses were performed using the PROC REG pr ocedure of the SAS software (SAS Institute, 2007). Prior to the re gression analyses, stepwise regressions were utilized to determine best predictor(s) of bi osolids-P uptake and lability using the procedure described by Hocking (1976). Time series analysis was conducted usi ng the PROC TSCSREG procedure of the SAS software (SAS Institute, 2007).Time series analysis contrasts were performed at harvests 4, 8, 11 for the P-based rate treatments and 4, 8, and 12 fo r the N-based rate treatments. Contrasts among all harvests yielded li ttle useful information. Conseque ntly, sub-division s of the greenhouse study were necessary to practica lly assess changes over time. Tr aditional greenhouse studies are often a single growing season in length and composed of four tissue harvests; thus, every fourth 28

PAGE 29

29 harvest (except for harvest 11 for the P-based treatments) was selected to perform time series analysis contrasts for the greenhouse study.

PAGE 30

Table 2-1. Treatment processes used to produce th e biosolids selected for use in the greenhouse study. Material Classa Digestion BPRb (Yes/No) Final treatment Milorganite AA Aerobic No Thermally dried Greenedge AA Anaerobic No Thermally dried Disney AA Composted Yes Composted material GRU B Aerobic No Thickened Boca Raton B Anaerobic Yes Thickened, dewatered Lakeland Northside AA Aerobic No Thickened, dewatered, ATAD system OCUD S Cake B Anaerobic No Thic kened, dewatered, bio-N removal a Determined by the biosolids quality sta ndards established by the 40 CFR Part 503 rule (USEPA, 1993a). Class AA biosolids are considered exceptional quality biosolids, while Class B biosolids meet less stringent standards for pathogens. b Biological phosphorus removal Table 2-2. Various measures of P for the P-sources utilized in the greenhouse study (Chinault, 2007). -----------Total P--------Fe-strip ----Oxalate extractable---P Source Determined Producer P WEP PWEP P Al Fe PSI -------------------g kg-1----------------------%-------------g kg-1-----------Milorganite 21 23 0.08 0.12 0.58 16 1.2 25 1.0 GreenEdge 17 19 0.12 0.19 1.1 13 5.0 13 1.0 Disney Compost 11 27 0.95 1.2 8.4 11 6.6 20 0.6 GRU 31 48 1.7 7.9 26 21 6.4 3.7 2.1 Boca Raton 26 39 0.02 3.9 15 33 14 7.3 2.1 Lakeland NS 29 29 0.35 14 47 22 3.2 8.3 2.0 Orange County South 23 30 0.14 4.8 21 23 5.0 4.4 2.9 Triple Super Phosphate (TSP) 190 210 NDa170 85 186 11 6.8 NAb a Not determined b Not applicable 30

PAGE 31

Table 2-3. Selected chemical properties of the Immokalee A horizon and base sand utilized in the greenhouse study (Chinault, 2007). Immokalee Base Parameter Units A horizon sand Sand % 95 100a pH 4.8 5.1 Organic matter g kg-1 7.0 NDbTotal P mg kg-1 15.5 12.0 Mehlich-1 extractable P mg kg-1 1.5 NDbWater extractable P mg kg-1 1.1 0.2 Oxalate extractable P mg kg-1 13.1 3.7 Oxalate extractable Al mg kg-1 40.1 17.6 Oxalate extractable Fe mg kg-1 85.6 10.0 Phosphorus Saturation Ratioc 0.1 0.1 Soil Phosphorus Storage Capacityd mg kg-1 0.8 0.2 Relative Phosphorus Adsorption (RPA)e % 2.0 8.6a Estimated from field characterization b Not determined c The molar ratio of oxalate extractable-P to oxalate extractable Al+Fe in soils (Maguire and Sims 2002; Nair and Harris, 2004) d A measure of the amount of P that can be added to the soil before the soil becomes an environmental concern (Nair and Harris, 2004) e The relative amount of P sorbed from a 400 mg P kg-1 load (Harris et al., 1996) 31

PAGE 32

CHAPTER 3 RESULTS AND DISCUSSION The greenhouse study characterized the fate of biosolids-P over more than 16 months. The residual value of biosolids-P was characterize d for more than one year following the initial 4 month short-term characterization conducted by Chinault (2007). Twelve tissue harvest and column leaching events occurred for the 112 kg P ha-1 and 224 kg P ha-1 amended columns, and 11 events occurred for the 56 kg P ha-1 amended columns. An implicit study objective was to determine the ultimate fate of biosolids-P, a nd no predetermined study en d-point existed at the beginning of the study. Bahiagrass is commonly f ound in grazed pasture systems of Florida, and consideration was given to mainta ining tissue nutrient contents that are sufficient for grazing cattle. The study continued until bahiagrass ti ssue N and P concentrations declined below suggested minimum concentrations (described more thoroughly in later discussions). We regarded the characterization of biosolids-P lability from th e 16-month greenhouse study as a long-term biosolids-P lability characterization be cause the supply of labile biosolids-P sufficient for most agronomic conditions wa s exhausted and additional P fert ilization would be necessary. A mass balance analysis was conducted to confirm that tissue harvests and column leaching characterized all P losses from the study columns, and that no egregious P losses or additions occurred during the st udy. The average P recovery for all columns was 92%. Triple super phosphate (TSP) is a highly soluble inorganic fert ilizer, and near complete dissolution of TSP pellets is expected after 14-16 months of water additions. Mass balance for TSP treatments was expected to be the least affected by column sampling error due to high degree of expected TSP dissolution, and a mass balance analysis for TSP-amended columns was expected approach 100% recovery. The average recovery for TSPamended columns was 97%, indicating the study successfully quantified column-P losses. 32

PAGE 33

Tissue Yield Significant differences in cumulative tissu e yield existed among P-sources across the rates of P application (F ig. 3-1). Apparent yield differen ces among P-sources early in the study raised concerns that the nutrient supply wa s not uniform among P-sources (Chinault, 2007). Biosolids-N mineralization was assumed to be 40 % of total biosolids-N for all biosolids, and ammonium nitrate supplemented biosolid s-N so that PAN supply was uniform among treatments. If biosolids-N mi neralization was incorrectly esti mated, PAN supply differences are likely most apparent at the N-ba sed application rate where bioso lids-N represented the greatest proportion of calculated PAN supply. Thus, tissueN concentrations were measured for columns amended at the N-based application rate to gu age if PAN supply was sufficient and uniform across P-sources. Tissue-N concentrations for harvests 1-4 suggested some differences in PAN supply among P-sources (Fig. 3-2). Further, the data suggested the biosolids PAN supply was underestimated for harvest 1, and overestimated for harvests 2-4 (China ult, 2007). Tissue N concentrations following harvest 4 were near the 11 g N kg-1 regarded as minimally sufficient for grazing beef cattle (NRC, 1996). Th is initial tissue N concentrati on analysis increased concern that PAN supply was not adequate and uniform ac ross P-sources, and suggested a change to the N fertilization regime was necessary. Following harvest 5, 300 kg PAN ha-1 of ammonium nitrate, applied in 4 split applications was supplied to columns to ensure PAN supply was adequate, and to compensate for PAN supply differences that existed among P-s ources. Supplying the additional N increased tissue N concentrations, and tissue N concentrations exceeded the suggested minimally sufficient tissue concentration of 11 g N kg-1 for harvests 6-8 (Fig. 3-3). Tissue N concentrations became 33

PAGE 34

increasingly difficult to maintain above 11 g N kg-1 after harvest 8 (despite additional N-fertilizer additions), and only harvest 11 exceeded the suggested minimum concentration. Tissue N concentrations in bahiagrass are affected by photoperiod (Sinclair et al ., 2003), and decreased tissue N concentrations during l onger day length monthsthe m onths corresponding to harvests 8, 9, 10, and 11was expected. The seasonal patt ern in tissue N concentrations reported by Sinclair et al. (2003) suggested that tissue N concentrations sh ould increase from harvest 11 to harvest 12 because the photoperiod decreased; however, tissue N concentrations declined appreciably from harvest 11 to harvest 12. The pe rsistent tissue N con centrations below 11 g N kg-1in spite of abundant PAN supplysuggested that the bahiagrass could no longer maintain sufficient N uptake. Concern that insufficien t N uptake could confound P uptake interpretations resulted in the decision to terminate the study following harvest 12. Yield-weighted tissue N concentrations were calculated using equation 3-1 to determine if overall nitrogen uptake was uniform among P-sources, and bahaigrass tissue N concentrations were sufficient for grazing beef cattle. where for any given P-source: Yi = yield at ith harvest Ni = tissue N concentration at ith harvest (Y 1 *N 1 ) + (Y 2 *N 2 )... (Y 12 *N 12 ) = Yield-weighted tissue N concentration Y1 + Y2Y12 (3-1) No significant differences in yield-weighted N concentrations existed among P-sources (Fig. 34), and yield-weighted tissue N concentrations exceeded the suggested minimum of 11 g N kg-1. The data suggest the greenhouse study fertiliza tion regime adequately controlled PAN supply, and that potentially confounding effects of N insu fficiency on P uptake were minimized, overall. Nevertheless, incorrect estimations of biosolidsN mineralization in early harvests and decreased 34

PAGE 35

N uptake in later harvests illustrate that in teractions among plant nutrients in prolonged greenhouse studies are difficult to control. The relatively uniform yield-weighted tissu e N concentrations also suggest that the overall differences in cumulative tissue yield am ong P-sources were not due to differences in Nuptake. Poor tissue growth occurred in 3 of 4 columns amended with Lakeland NS biosolids at the N-based rate, suggesting some property of the material negatively affected the bahiagrass at high application rates. However, the differences in cumulative tissue yields for other sources are likely due to differences in long-term P-labi lity. McLaughlin and Champion (1987) noted ryegrass ( Lolium multiflorum ) yields were greater for sewage sludge amended sesquioxic soils than for soils amended with monocalcium phosphate and attributed the increased yields to a prolonged supply of phytoavailable P from sewage sludge. Tissue P Concentrations Bahiagrass tissue P concentrati ons are useful indicators of wh en the residual value of Psources is no longer sufficient to meet agronomic demands, and when additional P fertilization is necessary. Oladeji (2006) proposed an agr onomic threshold for bahiagrass of 2.0 g P kg-1. Adjei et al. (2000) suggested ba hiagrass tissue P concentr ations of 1.6 to 1.7 g P kg-1 are agronomically limiting. Critical P tissue concentr ations are also determined by considering the P nutritional requirements of grazing beef cattle. If the current live weight of a grazing beef steer is just maintained and bahiagrass tissue is the only Psource, a minimum tissue P concentration of 0.9 to 1.1 g P kg-1 is suggested by Ternouth et al. (1996) The suggested tissue concentration can increase considerably to values beyond 2.0 g P kg-1 if live weight gains or milk production are desired. OConnor and Elliott (2006) suggested a cr itical bahiagrass tissue P concentration of 35

PAGE 36

1.0 g P kg-1. The 1.0 g P kg-1 value represents a conservative estimation of the critical P tissue concentration of bahiagrass and is used as the indicator value for this study. The residual-P value of all P-sources app lied at N-based rates maintained tissue P concentrations above 1.0 g P kg-1 for more than 16 months harvestsof continuous bahiagrass growth (Table 3-1). Although the tissue P concentrations exceed the critical value, the tissue P concentrations are below most agronomic standards (Adjei et al., 2000; Oladeji, 2006) and are likely agronomically limiting. Plants translocate stored P from older tissues to younger leaves under P-deficient conditions (S chachtman et al., 1998), suggesting tissue P concentrations could remain elevated some time after the plant is subjected to P limited conditions. The low tissue P concentrations at even the N-based application rates following 12 harvests suggest the ultimate agronomic value of biosolids-P was suffici ently characterized. The time series analysis suggests the P-releas e characteristics of M ilorganite is unique from other biosolids and TSP at N-based applica tion rates (Table 3-1). Tissue P concentrations are not significantly different at harvest 4, 8, and 12. The relatively constant tissue P concentrations for the N-based rate of Milorg anite is consistent with the slow P release characteristics previously shown for heat-dri ed, high Fe+Al biosolids such as Milorganite (OConnor et al., 2004; Chinault, 2007). Tissue P concentrations decline significantly from harvest 4 to 8 for other P-sources (Table 3-1). Tissue P concentrations for the P-based ra te treatments did not change significantly following harvest 8 (Table 3-2) Tissue concentrations are 1.0 g P kg-1 for harvests 8 through 11, suggesting that P-sources applied at P-base d rates have little residual agronomic value beyond 8 harvests. The persistently low tissue P concentrations followi ng harvest 8 suggested 36

PAGE 37

the ultimate phytoavailability of P-sources ame nded at P-based rates was characterized, and columns amended at the 56 kg P ha-1 rate were terminated after the 11th harvest. Yield-weighted tissue P concentrations were determined for the P-based rate treatments and compared to the critical P concentration of 1.0 g P kg-1 to assess if the overall P-supply was sufficient to exceed the minimum agronomic P de mand. Yield-weighted tissue P concentrations for Milorganite and Greenedge biosolidsheat-d ried, high Fe+Al biosolidsare approximately 1.0 g P kg-1 (Fig. 3-5), suggesting minimally sufficient P supply to meet the 14 month agronomic P demand. An additional P-based application of P within 14 months of the initial application is likely necessary to assure sufficien t P supply to grazing beef cattle. Abundant P supply is associated with increase d root growth (Havlin et al., 1999), and the differences in P-supply among treatments were expected to affect below-ground biomass. Bahiagrass root masses were quantified for the P-based and N-based treatments following study termination to gauge the hypothesi zed effect of P-supply differences on below-ground biomass. No significant differences in bahiagrass root ma ss existed among P-sources at either rate, and only the root mass for the N-based Boca Raton tr eatment was significantly different from the controls (Fig. 3-6). Average N-based treatment r oot masses tended to be greater than the average root mass of P-based treatments; however, only th e controls were signifi cantly different from Nbased rate treatments. Quantifying root masses following the studys termination makes interpreting P-supply effects on root masses difficult. The P-based treatment tissue P concentrations were deficient for some time (Tab le 3-2) before root masses were quantified, and plants can increase root length and root hair numbers in response to P-deficiency (Vance et al., 2003). Plant physiological responses in P de ficient conditions could confound root mass interpretations. 37

PAGE 38

Environmental P Lability Eleven leaching events occurred for P-ba sed rate treatments, resulting in 2.75 pore volumes of leachate collected. Phosphorus leachi ng from any P-sources a pplied at the P-based rate was minimalless than 1% of the P applied. Nevertheless, environmental P lability differences existed among P-source s at the P-based application rate (Fig. 3-7A). Milorganite, a high Fe+Al biosolids, leached si gnificantly less P than BPR and BPR-like biosolids, and TSP. Notably, the mass of P leached from the less soluble-P biosolidsMilorganite, Greenedge, and Disneytreatments was not significantly different than the mass of P-leached from the controls. More soluble-P biosolids (BPR and BPR-like biosolids) and TSP leached significantly more P than the controls. Although significant differe nces in environmental P lability exist among Psources, the cumulative mass of P leached ( 0.27 mg P) suggests P-sources applied at P-based application rates pose minima l environmental risk. Twelve leaching events resulted in 3 por e volumes of leachate from N-based rate amended columns, and much greater P leaching occurred (Fig. 3-7B). Significantly more P leached from BPR biosolids, BPR-like biosolids, and TSP than from the less soluble-P biosolids at the N-based rate. The mass of P leached from the more soluble-P biosolids and TSP treatments represented more than 12% of the total P applied. Remarkably, the mass of P leached from Milorganite N-based rate treatments was not significantly different than the controls, and represented < 0.1% of the total P applied. The leaching data s uggest that Milorganite poses minimal P leaching risk, even at application rates approximately four times plant P requirements. The time series analysis suggests P-release kineti cs also differed among P-sources (Table 3-1). An appreciable portion of the P leaching occurred in the first 4 harvests for the more soluble-P biosolids and TSP, whereas P-leaching is more gradual for less sol uble-P biosolids. 38

PAGE 39

Overall differences in environmental P labi lity exist among P-sources across application rates (Fig. 3-8). The environmental lability of heat-dried, high Fe+Al Milorganite and Greenedge biosolids, and the composted Disney material is significantly less than BPR biosolids, BPR-like biosolids, and TSP. Fertilizer-P (T SP) leached the greatest mass of P; however, the environmental lability of BPR and BPR-like biosolid s-P is not statistically different from TSP-P, across application rates. Relative environmental P lability differe nces among P-sources are well documented (Elliott et al., 2002; Oladeji, 2006; Chinault 2007). Elliott et al. (2002) determined environmental P lability differences among bios olids by quantifying P leach ing in a short-term greenhouse study. Yet, concern rema ined that short-term environmental lability differences are not necessarily indicative of long-term environm ental P lability differences. The relative differences in environmental P lability among bi osolids determined in this 16-month greenhouse study are similar to the short-term (4-month) Ellio tt et al. (2002) studyless soluble biosolids-P is significantly less environmentally labile than BPR biosolids-P and TSP-P. Despite the similar determinations of relative biosolids-P leaching risks, the fraction of Psource total P that is environm entally labile is markedly di fferent between the two greenhouse studies. Columns amended with TSP at the 224 kg P ha-1 and 56 kg P ha-1 rate leached approximately 21% and 14% of the P-applied, re spectively, in the 4 month Elliott et al. (2002) study. Approximately 13% and 0.2% of P-applied leached at the 224 kg P ha-1 and 56 kg P ha-1 rate, respectively, in the current greenhouse st udy. The substantially greater mass of TSP-P leached in Elliott et al. (2002) studydespite leaching two times less leachate than the current greenhouse studyraised concern that the base sand in the current greenhouse study retained more P than Elliotts base sand. Columns in bot h studies were amended at the same N-based 39

PAGE 40

rate, and utilized a similar Immokalee fine sand as surface (0-15 cm) soils. However, the base sand underlying the Immokalee A horizon material di ffers between the studies. The Elliott et al. (2002) study utilized a Myakka (Sandy, siliceous, hyperthermic Aeric Alaquods) E horizon with a low P-sorbing capacityRPA 3.2%. The current greenhouse study utilized a base sand capable of sorbing almost three times as much PRPA 8.6% (Table 2-3). Although the 8.6% RPA value of the base sand represents a low P sorption capacity, the relatively large mass of base sand in the columns (~8 kg) resulted in an appreciable sink (~275 mg) for labile P, representing 267%, 133%, and 67% of the to tal-P applied at the 56, 112, and 224 kg P ha-1 treatments, respectively. Thus, the base sand us ed in the current study would be expected to retard P-leaching to a greater extent than the ba se sand used by Elliott et al. (2002), and explains the differences in P leachate losses. Biosolids-P Phytoavailability Eleven harvests, over 14 months of continuous bahiagrass growth, characterized P phytoavailability in the P-base d rate treatments. Uptake of P from all P-sources was significantly greater than contro l P uptake (Fig. 3-9A). Uptake of P from only two biosolids treatments was significantly different than TSP at the P-based application rate; Milorganite-P uptake was significantly less than TSP-P uptake, and OCUD S-P uptake was significantly greater than TSP-P uptake. Twelve harvests, over 16 months of continuous bahiagrass gr owth, characterized P phytoavailability in the N-based treatments. Up take of P from almost all BPR and BPR-like biosolids was not statistically different than TSP-P uptake (Fig. 3-9B); P uptake from OCUD S biosolids treatments was significan tly greater than TSP treatments. Uptake of P from the less soluble-P biosolids was significan tly less than P uptake from the more soluble-P biosolids and 40

PAGE 41

TSP treatmentsnotably, P uptake from Milorganite treatments was significantly less than P uptake from all other P-sour ce treatments. Differences in P uptake across rates of appli cation were analyzed to determine overall Psource phytoavailability differences (Fig. 3-10). Phytoavailability trends present in individual Pbased and N-based rate analyses were also distinguished in th e overall P-source phytoavailability analysis. Uptake of P from BPR and BPR-like bi osolids was similar to, or greater than, P uptake from TSP across application rates. Uptake of P from OCUD S biosolids treatments was significantly greater than from TSP treatments acro ss application rates. Uptake from Milorganite and Greenedge treatments was significantly less than P uptake from more soluble-P biosolids and TSPbahaigrass P uptake from Milorganite tr eatments was significantly less than uptake for all other P-sources. The envi ronmental lability of Disney bios olids-P was distinguished from more soluble biosolids-P and TSP-P (Fig. 3-8), but P uptake from Disney treatments was not significantly different from TSP and GRU biosolids treatments. OConnor et al. (2004) and Chinault (2007) utilized a slope -ratio approach to estimate the relative P phytoavailability (RPP) of biosolid s compared to TSP by regressing P uptake as a function of application rate. The linear regression lines were forced through a y-intercept equal to the average P uptake of the control treatment. The ratio of the slop e of biosolids-P uptake regression lines to the TSP-P uptake regression line reflected the relative P phytoavailability of the biosolids. Chinault (2007) al so utilized point es timates of RPP to determine the RPP of biosolids where a linear regressi on poorly described the relations hip between biosolids-P uptake and application rate for some treatments. A similar slope-ratio approach was utilized here to determine the long-term RPP of biosolids; however, the OConnor et al. ( 2004) and Chinault (2007) slope-ratio method was 41

PAGE 42

modified to distinguish treatment effects on P uptake and to a ccount for the difference in the number of harvests (11 for the 56 kg P ha-1 rate and 12 the 112 and 224 kg P ha-1 rates). Cumulative P uptake from the control treatment wa s subtracted from cumulative P uptake for the various P-source treatments to isolate the contribu tion of P-sources (Fig. 3-11). The origin in a plot of P uptake versus P-source application ra te represents zero P-source uptake when no Psource is applied, and linear regressions of P upt ake from P-sources versus P application rate were forced through the origin. One study objective was to determine the long-term phytoavailability of biosolids-P, and characteri zing the ultimate agronomic value of biosolids-P is implicit to achieving the objec tive. Tissue P concentrations suggested 11 harvests were sufficient to characterize the ultimate agronomic value in the P-based treatments, but an additional harvest was deemed necessary to more fully characterize the residual value of biosolids-P applied at N-ba sed rates. Data for P uptake from the additional 12th harvest could not be utilized if the regression was forced through a y-intercept equal to the average cumulative control P uptake because cumulative control P uptake differs between harvest 11 and 12. Utilizing only harvest 11 data for the 112 and 224 kg P ha-1 rates would ignore almost two months of additional P uptake data. Accounting for the cumulative control P uptake at each rate, and forcing the regression through the origin, perm its the utilization of harvest 12 P uptake data without compromising the determination of relative differences among P-sources. The slope-ratio estimates of long-term RPP suggest less soluble biosolids-P is ultimately less phytoavailable than TSP-P (Table 3-3). M ilorganite-P is ultimatel y less than half as phytoavailable as TSP-P. The RPP values of BPR and BPR-like biosolids-P are similar to, or greater than, values for TSP-P. Notably, the slope-ratio RPP estimates suggest OCUD S biosolids-P is approximately 130% as phytoavailable as TSP-P. The PWEP value for TSP is ~4x 42

PAGE 43

the PWEP value for OCUD S biosolids; yet the percent of total P that is oxalate extractable is 98% and 100% for TSP and OCUD S, respectiv ely (Table 2-2). The water and oxalate extractions suggest that while onl y about 20% of OCUD S total P is water soluble, almost all of the total-P is potentially accessible to plant roots, which can exude organic acids to extract labile P (Vance et al., 2003). Recall that P leaching was greatest in TSP tr eatments and that most of the P leaching occurred in the first leaching event (T ables 3-1, 3-2). Thus, an appreciable portion of the labile pool of TSP likely le ached out of the root zone befo re plant uptake could occur. Conversely, less P leached in OCUD S biosolids amended columns, and the leaching occurred less rapidly. Much (~80%) of the OCUD S biosolidsP labile pool is not water soluble and likely remained in the root zone for plant uptake. Mo reover, OCUD S biosolids treatment tissue yields were significantly greater than all other P-sour ces across application ra tes (Fig. 3-1)despite similar yield-weighted tissue N concentrations among P-sources (F ig. 3-4). Apparently, some additional nutritive quality of the OCUD S bios olids improved bahiagrass yields, and thus, increased P uptake. Point RPP estimates were also utilized to estimate biosolids RPP values (Table 3-4). Point estimates were determined by averaging P uptake at each rate of P-source applied, and dividing by the average P uptake of TSP treatment s at each rate. Similar to slope-ratio RPP estimates, control P uptake values were subtracted from P-source cumulative P uptake values to isolate the contribution of P-sources to P uptake. The point RPP estimates were averaged across rates and compared to the slope -ratio RPP estimates. Strong ag reement exists between sloperatio and point estimate methods of RPP cal culation; RPP point and slope-ratio estimates differed less than 10% for all biosolids, and th e average difference was less than 4%. Empirical P uptake data also supports the sl ope-ratio and point es timates of biosolids RPP. The RPP value 43

PAGE 44

of Milorganite is about 50%, suggesting Milorgan ite-P uptake will be sim ilar to TSP-P uptake if twice as much Milorganite-P is applied as T SP-P. Indeed, TSP-P uptake at the 56 kg P ha-1 application rate is equivalent to Milorganite-P uptake at the 112 kg P ha-1 application (P uptake is ~79 mg for both treatments). Biosolids RPP estimates increased appreciably from harvest 4 to harvest 12 (Tables 3-3, 3-4); however, relative differences among biosolids remained similar. The greatest increase in RPP values occurred for less soluble-P biosol ids. Milorganite and Greenedge biosolids maintained an overall moderate RPP classificat ion, but the RPP estimate increased about 20% for both biosolids from harvest 4 to harvest 12. The RPP estimate for the Disney biosolids increased > 30%, and the long-term RPP estimat e suggests the relative phytoavailability of Disney-P changes from moderate to high. Modest increases in RPP estimates occurred for BPR and BPR-like biosolids following the additional year of plant uptake. Co mparison of short-term versus long-term RPP estimates suggests that sh ort-term RPP values successfully distinguish relative differences among biosolids, but longterm RPP estimates are necessary to fully characterize the agronomic value of biosolids-P relative to TSP. Nevertheless, th e categorization (low, moderate, high) of biosolids RPP based on a short-term (4 month) characterization was appropriate in all but one case. McLaughlin and Champion (1987) noted that relative efficiency of sewage sludge-P compared to monocalcium phosphate (MCP) increased over time, and that the relative efficiency of sewage sludge P was ultimately similar to or greater than MCP. The current greenhouse study confirms that biosolids RPP estimates tend to increase with time ; however, less soluble biosolids-P ultimately remained less phytoavail able than TSP-P. McLaughlin and Champion (1987) regarded sewage sludge as a slow-release P fertilizer. Th e appreciable increase in RPP 44

PAGE 45

with time for Milorganite, Greenedge, and Di sney biosolids indicat es the slow-release description is appropriate for th e less soluble-P biosolids, and that a 20-50% increase in less soluble-P biosolids application rate is necessary to supply P at rates agronomically equivalent to TSP. The RPP estimates of BPR and BPR-like biosolids-P were similar to TSP throughout the study, and the slow-release description is not appropriate for such biosolids. No agronomic P application rate adjustment is justified for BPR and BPR-like biosolids. No a priori measure of biosolids-P phytoavailability exists to guide agronomic biosolids application rates, and empirical evidence is necessary to justify application rate adjustment. Chinault (2007) characterized the total-P, PWEP, oxalate extractable P, iron strip-P, and PSI of the biosolids utilized in the greenhouse study. Bi osolids-P uptake was correlated to the various measures of biosolids-P to identify a potentially useful measure of biosolids-P phytoavailability. The load of total-P, oxalate extractable-P, and ir on strip-P that was applied in each treatment was utilized in the regressi on analysis to account fo r rate effects. Biosolids PWEP and PSI values were multiplied by the total-P load to account for rate effects. The product of PWEP and total-P load was termed the environmentally effective P load (Agyin-Birikora ng, 2008), and biosolids PSI multiplied by total-P load is henceforth designated the labile P load. The labile P load was the best predictor of biosolids-P uptake, and explained 85% of the variability in biosolids-P uptak e (Fig. 3-12). The strong, linear relationship between biosolids-P uptake and the labile P load suggests biosolids PSI is a potentially useful a priori measure of biosolids-P phytoavailability. I ndeed, a good linear relationship (r2 = .76) exists between biosolids RPP and biosolids PSI (Fig. 3-13), which suggests that a priori determinations of biosolids PSI could be used to make applicati on rate adjustments for agronomic consideration. Oladeji (2006) determined long-term RPP values for two biosolids in a two-year field study, and 45

PAGE 46

in a greenhouse pot study util izing two croppings of bahiagra ss and one cropping of ryegrass ( Lolium perenne L.). The RPP estimates and PSI values of the two biosolids utilized in the Oladeji (2006) studies were a dded to the dataset from the cu rrent greenhouse study to confirm that biosolids PSI is a good a priori indicator of long-term bioso lids RPP values. A strong linear relationship (r2 = .82) exists between the RPP estimates and PSI values determined in the current greenhouse study and in the Oladeji ( 2006) studies (Fig. B-1). Additional datasets were available to validate that biosolids PSI is a useful a priori indicator of biosolids-P phytoava ilability. OConnor et al. (2004) and Chinault (2007) conducted short-term (4-month) greenhouse studies that utilized methods identical to the current greenhouse study to characterize biosolids-P phyt oavailability: bahiagrass was grown in columns that contained ~30 cm of base sa nd which underlaid ~15 cm of P-source amended Immokalee A horizon. A strong relationship (r2 = .82) existed between cumulative biosolids-P uptake and the labile P load in the two short-term greenhouse studies (Fig. 3-14). The estimates of biosolids RPP determined in the two short-term greenhouse studies also correlated well (r2 = .81) with biosolids PSI (Fig. B-2) OConnor et al. (2004) also quantified P uptake utilizing a Psource amended Candler A horizon (RPA 15%). The relationship between cumulative biosolids-P uptake and the labile P load was less strong on the amended Candler A horizon (r2 = .62, Fig. B-3) than the ame nded Immokalee A horizon (r2 = .82, Fig. 3-14). The RPA value of Candler A horizon is nearly three times the RPA value of the Immokalee A horizon, which suggests that even modest increases in P rete ntion capacity can decrea se the ability of an a priori measure of biosolids-P to predict biosolids-P phyt oavailability once the biosolids are applied to soils. 46

PAGE 47

A single, large dataset that described th e relationship between cumulative biosolids-P uptake and the labile P load was desired; howev er, combining cumulative P uptake data from the short-term greenhouse studies with the prol onged (16-month) greenhouse cumulative P uptake data was not appropriate because cumulative P up take values differed appreciably between the studies. Cumulative P uptake was determined ove r 4 harvests for the OConnor et al. (2004) and Chinault (2007) studies, whereas cumulative P up take was determined over 12 harvests in the current greenhouse study. However, relative esti mates of biosolids-P phytoavailability in the short-term and long-term studies could be combined into a single dataset. Short-term estimates of biosolids RPP (OConnor et al., 2004; Chinault, 2007) were combined with long-term estimates of biosolids RPP (Oladeji, 2006; the current greenhouse study) were combined into a single dataset to determine an overall relations hip between short and long-term estimates of biosolids RPP and biosolids PSI values. A good relationship (r2 = .70) exists between the shortterm and long-term estimates of biosolids RPP a nd biosolids PSI values (F ig. 3-15). These data suggest that, although long-term es timates of biosolids RPP are greater than short-term RPP estimates (Table 3-3), biosolids PSI is a useful a priori indicator of bi osolids-P relative phytoavailability regardless of whether the RPP va lues are long-term or short-term estimates. OConnor and Elliott (2006) suggested PWEP is a potentially useful indicator of biosolids RPP. The PWEP value is likely a good indicator of the immediately available P pool, but does not sufficiently quantify the entire labile P pool ultimately availabl e to plants. Biosolids PSI better describes the ultimately phytoavailable P pool by utilizing an oxalate extraction to quantify labile P and the reactive Fe and Al capable of retaining potentially phytoavailable P. Thus, biosolids with the greates t ratio of oxalate extractable-P to oxalate extractable-Fe+Al (greatest PSI values) represent the most phytoava ilable biosolids-P sources. Notably, the PSI 47

PAGE 48

concept is not applicable to all biosolids. Alka line stabilized biosolids (and the chemistry of P contained therein) are dominated by Ca, and the PSI concept is not applicable to these materials (Elliott et al., 2002). Overall P Lability Previous greenhouse studies often consid ered the environmental lability and phytoavailability of biosolids-P separately and did not address differen ces in overall (agronomic and environmental) lability. Combining cumulative P leached with cumulative P uptake characterizes the overall lability of P-sources. Leaching of P contributed little to the overall P lability at the P-based application rates, and P-source lability differences resembled P-uptake differences. The overall P labi lity values of more solubleP biosolids (BPR and BPR-like biosolids) were similar to values for TSP, and M ilorganite-P was significantly less labile than the more soluble-P biosolids and TSP-P (Fig. 3-16A). Leaching of P represented an appreciable por tion of the overall P lability for the more soluble-P biosolids and TSP applied at the N-based rate. The overall P lability of more soluble-P biosolids was statistically similar to TSP. At the N-based application rate, Milorganite-P was significantly less labile than all other P-s ources, and all less sol uble-P biosolids were significantly less labile than more soluble-P bi osolids, and TSP (Fig. 3-16B). Moreover, the clear difference in overall P lability between less soluble-P biosolids and BPR biosolids, BPRlike biosolids, and TSP was distinguished by overa ll P lability comparisons across rates (Fig. 317). The RPP value describes the relative agronomic effectiveness of biosolids-P and only considers P uptake. Incorporating P leaching with P uptake to quantify overall P lability characterizes the overall relative P lability (RPL ) of the biosolids. Biosolids RPL estimates 48

PAGE 49

provided a better indication of ultimate P releas e differences between biosolids and TSP than biosolids RPP estimates because the RPL consider s P lost to leaching. The slope-ratio method used to determine biosolids RPP values was simila rly utilized to estimate the RPL of biosolids (Fig. 3-18). Point estimates of biosolids RPL we re also utilized to validate slope-ratio RPL estimates. Slope-ratio RPL estimates suggest less solubl e-P biosolids are 37-64% as overall labile as TSP, and that BPR and BPR-like biosolids-P ove rall lability is approximately equal to TSP-P (Table 3-5). The greatest mass of P leached in TSP treatments; thus, incorporating P leaching decreased biosolids RPL estimates compared to RPP estimates. Estimates of RPL were appreciably less than RPP estimates for less solu ble-P biosolids because these P-sources leached relatively little P. Point RPL estimates suggest that BPR a nd BPR-like biosolids are about 90-120% as labile as TSP and that less sol uble-P biosolids are roughly 50-75% as labile as TSP-P (Table 36). Point estimates of biosolids RPL values are similar to slope-ratio estimates of biosolids RPL, although point RPL estimates are approximately 10% greater than slope-ratio estimates of biosolids RPL for all biosolids except GRU. Point RPL estimates at each rate suggest appreciable changes in RPL values occur from P-ba sed to N-based applicati on rates. Relatively little P leaching occurred at the 56 and 112 kg P ha-1 application rates, and RPL estimates are similar to RPP estimates. The greatest relative P leaching difference between biosolids and TSP occurred at the 224 kg P ha-1 application rate, resu lting in smaller RPL estimates at the N-based application rate than the 56 and 112 kg P ha-1 application rates. Slope-ratio and point RPL estimates in creased approximately 20% from the 4th harvest to the 12th harvest. Short-term RPL estimations tended to underestimate the long-term overall 49

PAGE 50

relative lability of biosolids-P; however, four harvests were sufficient to distinguish the relative differences among P-sources and determine the gene ral category of biosolid s RPL. Despite the increase in RPL estimates from harvest 4 to 12, less soluble biosolids-P remained only about one-half to three-fourths as labile as TSP-P. The long-term over all lability of BPR and BPR-like biosolids-P is similar to TSP-P, and applicati on rates of these more soluble-P biosolids should not exceed agronomic or environmental fertilizer-P recommendations. A stepwise regression was utilized to determine a useful a priori measure of ultimate biosolids-P lability. The same method utilized to correlate cu mulative biosolids-P uptake to the measures of biosolids-P determin ed by Chinault (2007) was used to correlate overall biosolids-P lability to the total-P, PWEP, oxalate extractable P, iron strip-P, and PSI values of the biosolids. Labile P load (biosolids PSI*total-P load) was th e best predictor of overa ll biosolids-P lability, and explained 90% of the variabil ity in overall biosolids-P labili ty (Fig. 3-19). Biosolids RPL values were also correlated to biosolids PSI values to valida te biosolids PSI as a useful a priori indicator of biosolids-P labil ity. A good linear relationship (r2=.79) exists between biosolids RPL values and biosolids PSI values (Fig. 3-20). Recall the labile P load explained 85% of the variability in cumulative biosolids-P uptake (F ig. 3-12). Biosolids PSI was also a useful indicator of biosolids-P uptake because P uptake was the predominant mechanism of P removal. The labile P load correlated with overall P lability better than cumulative P uptake, which suggests that the biosolids PSI most accurately pr edicts ultimately labile biosolids-Psoluble P lost to leaching and P ultimately available to plant roots. Further validation of biosolids PSI value as a useful a priori indicator of ultimate biosolids-P lability was sought. OConnor et al. (2002) and Chinault (2007) conducted shortterm (4-month) greenhouse studi es that utilized methods iden tical to the current greenhouse 50

PAGE 51

study to characterize P uptake and P leaching fr om a P-source amended Immokalee A horizon. A strong, linear relationship (r2 = .80) existed between overall P la bility and the labile P load in the two short-term studies (Fig. 3-21), suggesti ng that the labile P load is also a useful a priori indicator of short-term biosolids-P lability. Chinault (2007) conducted a 5.5 month la boratory incubation study using a bare Immokalee soil to characterize ultimate P e nvironmental lability (release and leaching) differences among biosolids. A clear, quantitati ve relationship existed between cumulative P leaching and the labile P load; how ever, the labile P load only expl ained 73% of the variability in the cumulative biosolids-P leaching (Fig. 3-22). Chinault (2007) utilized a bare soil, whereas biosolids were subjected to plan t root interactions for more than 16 months in the current greenhouse study. Plant roots can ex ude organic acids to increase P availability (Vance et al., 2003), and P leaching from a bare soil likely does not sufficiently characterize biosolids-P that is ultimately labile in a grassed pasture system. Thus, cumulative biosolids-P leaching correlated more poorly with the labile P load than overall biosolids-P lability (P uptake and leaching). Agyin-Birikorang (2008) showed that the en vironmentally effective P load (PWEP*totalP load) an excellent predictor of P loss in rainfall simulation studies utilizing bare soils. However, the environmentally effective P load un derestimated biosolids-P lability in the current greenhouse study (Table A-1). Nevertheless, both the labile P load and the environmentally effective P load are useful indicators of biosolid s-P release, provided the indicators are correctly applied to the appropriate P loss mechanisms. The environmentally effective P load best indicates the readily soluble P that is potential ly lost to runoff and leaching during rainfall events, but the labile P load is a better predictor of ultimately labile biosolids-P, especially when plants are grown. 51

PAGE 52

0 10 20 30 40 50 60 70 80 90 100 MilorganiteDisneyGreenedgeTSPGRULakeland NS Boca Raton OCUD S P-SourcesCumulative Tissue Yield (g) 112 kg P ha-1224 kg P ha-156 kg P ha-1aa aab bc c d a Figure 3-1. Cumulative bahiag rass tissue yields following the 16-month greenhouse study. Cumulative tissue yields were determined after 12 harvests for the 112 and 224 kg P ha-1 rates, and 11 harvests for the 56 kg P ha-1 rate. Letter designations indicate statistical differences among P-sources across application rates. 52

PAGE 53

0 5 10 15 20 25 30 35TSPMilorganiteGreenEdgeDisneyGRUBoca Raton Lakeland NS OCUD S P SourceN Content per Harvest g N kg-1 Harvest 1 Harvest 2 Harvest 3 Harvest 4 Figure 3-2. Bahiagrass tissue N cont ents for harvests 1-4. The solid horizontal line represents a tissue N concentration of 11 g N kg-1, regarded as minimally sufficient for grazing beef cattle (NRC, 1996). 53

PAGE 54

0 5 10 15 20 25 ControlTSPMilorganiteGreenedgeDisneyGRUBoca RatonLakeland NSOCUD S P sourceTissue N Content (g kg-1) Harvest 5 Harvest 6 Harvest 7 Harvest 8 Harvest 9 Harvest 10 Harvest 11 Harvest 12 Figure 3-3. Bahiagrass tissue N concentrati ons for harvests 5-12. The solid horizontal line represents a tissue N concentrati on of 11 g N kg-1, regarded as minimally sufficient for grazing beef cattle (NRC, 1996). The nitrogen ferti lization regime was changed following harvest 5 to supply more PAN as ammonium nitrate. 54

PAGE 55

0 2 4 6 8 10 12 14 16 TSPMilorganiteGreenedgeDisneyGRUBoca Raton Lakeland NS OCUD S P-SourcesYield Weighted Tissue N Conc. (g N kg-1) a a a a a a a a Figure 3-4. Yield-weighted tissue N concentrations for N-based rate treatments determined over 16 months of bahiagrass growth. Th e solid horizontal line represents a tissue N concentration of 11 g N kg-1, regarded as minimally sufficient for grazing beef cattle (NRC, 1996). The letter designations indicate that no significant differences among P-sources existed. 55

PAGE 56

Table 3-1. Time series analysis of agronomic and environmental P labili ty for the N-based application rate. Tissue P concentration Cumulative P uptake Cumulative P leached Cumulative P lability ----Harvest---Time ---Harvest--Time ---Harvest---Time ----Harvest---Time P-Source 4 8 11 series* 4 8 11 series* 4 8 11 series* 4 8 11 series* -----g P kg-1 -----------mg------------mg-------------mg-----TSP 2.37 1.40 1.05 a bb 106 142 177 a bc 33.6 45.0 51.7 a a a 139 187 228 abc Milorganite 1.22 1.21 1.42 a a a 31 53 89 abc 0.1 0.1 0.2 a ab b149 b426 3 5 8a bc Greenedge 2.10 1.65 1.50 a b 5 812a bc 0.1 0.2 0.4 a a b 54 82 126 a bc Disney 2.46 1.97 1.59 a bc 58 93 140 abc 0.1 0.7 1.2 a bb 59 94 141 a bc GRU 2.60 1.71 1.36 a bc 99 137 175 a bc 13.6 25.3 32.5 a a a 113 162 208 abc Boca Raton 2.79 2.05 1.71 abb 95 144 200 a bc 5.1 14.5 20.8 a ab b 100 159 220 a bcLakeland NS 3.44 1.74 1.58 a bb 58 114 168 a bc 6.4 23.6 34.1 a a a 65 138 202 a bc OCUD S 2.74 2.10 1.73 a bc 115 172 226 abc 5.2 14.6 21.3 a bb 120 186 248 a bc Letters indicate statistical differenc es among values determined at the 4th, 8th, and 12th harvests. 56

PAGE 57

57 Table 3-2. Time series analysis of agronomic and environmental P lability for the P-based application rate. Tissue P concentration Cumulative P uptake Cumulative P leached Cumulative P lability ----Harvest---Time --Harvest-Time ---Harvest--Time --Harvest-Time P-Source 4 8 11 series* 4 8 11 series* 4 8 11 series* 4 8 11 series* -----g P kg-1 ----------m----------m g----------mg----TSP 1.16 0.72 0.76 a bb 49 67 79 a bb 0.0 0.1 0.2 a bc 49 67 79 abb Milorganite 1.07 0.79 0.87 a bb 24 42 55 a bc 0.1 0.1 0.1 aa a 24 43 55 abc Greenedge 1.13 0.88 0.93 a ab b30516 6 b1 b7 a bb 0.0 0.1 0.2 a ab 3 5166a bb Disney 1.16 0.80 0.85 a bb 37 57 70 a bb 0.0 0.1 0.2 a ab 3 5770a bb GRU 1.34 0.74 0.98 a bb 45 63 75 a bb 0.1 0.2 0.2 a bb 45 63 75 a bb Boca Raton 1.28 0.96 1.04 abb 45 67 85 a bb 0.1 0.1 0.2 a bc 45 67 85 abb Lakeland NS 1.35 0.80 0.95 a bb 53 76 89 a bc 0.1 0.2 0.3 abc 53 76 89 abc OCUD S 1.24 0.91 0.88 a bb 53 78 95 a bc 0.1 0.2 0.2 abb 53 78 95 a bc Letters indicate statistical differenc es among values determined at the 4th, 8th, and 11th harvests.

PAGE 58

0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 ControlMilorganiteGreenedgeDisneyGRUBoca Raton TSPOCUD SLakeland NS P-SourceYield Weighted Tissue P Conc. (g P kg-1) a b bc cd dede de ee Figure 3-5. Bahiagrass yiel d-weighted tissue P concentrations (harvests 1-11) for the P-based and control treatments. The ho rizontal line represents a critical tissue P concentration of 1 g P kg-1 (OConnor and Elliott, 2006). 58

PAGE 59

0 10 20 30 40 50 60 70 80 TSPMilorganiteGreenedgeDisneyGRUBoca Raton Lakeland NS OCUD SControl P SourceRoot Mass (g) 56 224kg P ha-1kg P ha-1a ab ab ab ab ab ab ab ab b b b b b b Figure 3-6. Bahiagrass root masses for N-based (224 kg P ha-1) and P-based (56 kg P ha-1) rate treatments. Average N-based rate root masses were significantly different from the control treatment root masses across P-sources. 59

PAGE 60

0.00 0.05 0.10 0.15 0.20 0.25 0.30 ControlMilorganiteGreenedgeDisneyTSPBoca Raton GRUOCUD SLakeland NS P-SourcesCumulative P Leached (mg P) ab abc abc bc bc b c c a bcA 0 10 20 30 40 50 60 ControlMilorganiteGreenedgeDisneyBoca Raton OCUD SGRULakeland NS TSP P-SourcesCumulative P Leached (mg P) ab bc c d d d d a dB Figure 3-7. Cumulative P leached as a function of P-source for A) P-based rate (56 kg P ha-1) treatments and B) N-based rate (224 kg P ha-1) treatments. Letters indicate significant differences among P-sources. Note the scale differences in the figures. 60

PAGE 61

0 10 20 30 40 50 60 MilorganiteGreenedgeDisneyBoca Raton GRUOCUD STSPLakeland NS P-SourcesCumulative P Leached (mg P) 112 kg P ha-1224 kg P ha-156 kg P ha-1a a b b b b b a Figure 3-8. Environmental P labil ity (P leached) differences among Psources at different rates. Letters indicate statistical differences among P-sources across application rates. 61

PAGE 62

0 10 20 30 40 50 60 70 80 90 100 ControlMilorganiteGreenedgeDisneyGRUTSPBoca Raton Lakeland NS OCUD S P-SourcesCumulative P Uptake (mg P) b bc c cd cde e f f a de f A 0 50 100 150 200 250 ControlMilorganiteGreenedgeDisneyLakeland NS GRUTSPBoca Raton OCUD S P-SourcesCumulative P Uptake (mg P) b c cd ef e f fg de a gBFigure 3-9. Cumulative bahiagrass P uptake as a function of P-sour ce for A) P-based rate (56 kg P ha-1) treatments and B) N-based rate (224 kg P ha-1) treatments. Letters indicate significant differences among P-sources. 62

PAGE 63

0 50 100 150 200 250 MilorganiteGreenedgeDisneyGRUTSPLakeland NS Boca Raton OCUD S P-SourcesCumulative P Uptake (mg P) 112 kg P ha-1224 kg P ha-156 kg P ha-1b bc cd cd d de e a Figure 3-10. Cumulative P uptake differences among P-sources. Letters indicate statistical differences across application rates. 63

PAGE 64

0 25 50 75 100 125 150 175 200 225 0 56 112 168 224 P-Source Application rate (kg P ha-1)P Uptake (mg ) OCUDS Boca Raton Lakeland TSP GRU Disney Greenedge Milorganite Figure 3-11. Relative P phytoavailability curves for P-sources. C ontrol P uptake was subtracted from treatment cumulative P uptake values, and the regression lines were forced through zero. Table 3-3. Slope-ratio estimates of biosolids relative P phytoav ailability for harvest 4 and harvest 12. Determined after 4 harvests Determined after 12 harvests Regress. RPP RPP Regress. RPP RPP P Source r2 coeff. (%) category r2coeff. (%) category TSP 0.96 0.472 100 High 0.98 0.736 100 High Milorganite 0.87 0.122 26 Mod. 0.92 0.342 46 Mod. Greenedge 0.96 0.222 47 Mod. 0.96 0.498 68 Mod. Disney 0.91 0.265 56 Mod. 0.95 0.585 79 High GRU 0.97 0.439 93 High 0.98 0.722 98 High Boca Raton 0.95 0.429 91 High 0.97 0.854 116 High Lakeland NS 0.85 0.439 93 High 0.93 0.737 100 High OCUD S 0.97 0.519 110 High 0.98 0.966 131 High RPP estimates for 56 kg P ha-1 rate amended columns were determined after 11 harvests 64

PAGE 65

Table 3-4. Point estimates of biosolids relati ve P phytoavailability following harvest 4 and harvest 12. Determined after 4 harvests Determined after 12 harvests P Source Rate Point RPP Average Category Point RPP Average Category TSP 56 100 100 High 100 100 High 112 100 100 224 100 100 Milorganite 56 38 31 Moderate 58 52 Moderate 112 31 55 224 23 43 Greenedge 56 54 49 Moderate 78 71 Moderate 112 46 67 224 47 67 Disney 56 69 62 Moderate 85 83 High 112 67 88 224 51 76 GRU 56 90 92 High 94 96 High 112 94 96 224 93 99 Boca Raton 56 88 92 High 111 115 High 112 98 120 224 89 115 Lakeland NS 56 109 96 High 119 109 High 112 106 113 224 73 94 OCUD S 56 108 110 High 128 130 High 112 113 128 224 109 133 RPP estimates for 56 kg P ha-1 rate amended columns were determined after 11 harvests 65

PAGE 66

y = 0.14x + 36.43 r2 = 0.85 0 50 100 150 200 250 0200400600800100012001400 Labile P Load (Biosolids PSI*Total-P Load)Cumulative P Uptake (mg) Figure 3-12. Cumulative P uptake plotted as a func tion of the labile P load (biosolids PSI*totalP load). y = 30.78x + 39.22 r2 = 0.76 0 20 40 60 80 100 120 140 0.00.51.01.52.02.53.03 Biosolids PSI RPP (%).5 Figure 3-13. Biosolids relative P phytoavailability (RPP) values plotted as a function of biosolids phosphorus saturation index (PSI) values. 66

PAGE 67

y = 0.08x + 23.88 r2 = 0.82 0 20 40 60 80 100 120 140 0200400600800100012001400 Labile P Load (Biosolids PSI*Total-P Load) Cumulative P Uptake (mg) Chinault (2007) O'Connor et al. (2004) Figure 3-14. Relationship between cumulative P uptake and the labile P load for two previously conducted short-term studies greenhouse studies (OConnor et al., 2004; Chinault, 2007). y = 38.10x + 11.98 r2 = 0.70 0 20 40 60 80 100 120 140 0.00.51.01.52.02.53.03.5 Biosolids PSIRPP (%) Short-term RPP estimates Long-term RPP estimates Figure 3-15. Long-term estimates (Oladeji, 2006; current greenhouse study) and short-term estimates (OConnor et al., 2004; Chinault, 2007) of biosolids relative P phytoavailability (RPP) plotted as a function of biosolids phosphorus saturation index (PSI) values. 67

PAGE 68

0 10 20 30 40 50 60 70 80 90 100 ControlMilorganiteGreenedgeDisneyGRUTSPBoca Raton Lakeland NS OCUD S P-SourcesOverall P Lability (mg P) b bc cd cde de f fg g a efg A 0 50 100 150 200 250 300 ControlMilorganiteGreenedgeDisneyLakeland NS GRUBoca Raton TSPOCUD S P-SourcesOverall P Lability (mg P) b c c d de de d a eB Figure 3-16. Overall P lability (agronomic a nd environmental) of various P-sources for A) P-based rate (56 kg P ha-1) treatments and B) N-based rate (224 kg P ha-1) treatments. Letters indicate significant differences among P-sources. 68

PAGE 69

0 50 100 150 200 250 300 MilorganiteGreenedgeDisneyGRULakeland NS TSPBoca Raton OCUD S P-SourcesOverall P Lability (mg P) 112 kg P ha-1224 kg P ha-156 kg P ha-1ab b c cd cd cd d a Figure 3-17. Overall P lability (agronomic and environmental) differences among P-sources. Letters indicate statistical differences acro ss application rates. 69

PAGE 70

0 50 100 150 200 250 0 56 112 168 224 P-Source Application rate (kg P ha-1)Overall P Lability (mg) OCUD S Boca Raton TSP Lakeland GRU Disney Greenedge Milorganite Figure 3-18. Relative P overall lability curves for P-sources. Control P lab ility was subtracted from treatment overall P labili ty values, and the regression lines were forced through zero. Table 3-5. Slope-ratio estimates of biosolids relative P overall labili ty for harvest 4 and harvest 12. Determined after 4 harvests Determined after 12 harvests Regress. RPL RPL Regress. RPL RPL P Source r2 coeff. (%) category r2coeff. (%) category TSP 0.99 0.587 100 High 1.00 0.917 100 High Milorganite 0.87 0.122 21 Low 0.92 0.343 37 Mod. Greenedge 0.96 0.223 38 Mod. 0.96 0.499 54 Mod. Disney 0.91 0.265 45 Mod. 0.95 0.589 64 Mod. GRU 0.98 0.486 83 High 0.99 0.835 91 High Boca Raton 0.96 0.447 76 High 0.98 0.927 101 High Lakeland NS 0.87 0.450 77 High 0.97 0.857 94 High OCUD S 0.97 0.538 92 High 0.99 1.043 114 High RPL estimates for 56 kg P ha-1 rate amended columns were determined after 11 harvests 70

PAGE 71

Table 3-6. Point estimates of biosolids relative P overall lability for harvest 4 and harvest 12. Determined after 4 harvests Determined after 12 harvests P Source Rate Point RPL Average Category Point RPL Average Category TSP 56 100 100 High 100 100 High 112 100 100 224 100 100 Milorganite 56 38 28 Moderate 58 48 Moderate 112 30 54 224 17 32 Greenedge 56 54 45 Moderate 78 65 Moderate 112 45 65 224 35 50 Disney 56 69 58 Moderate 85 76 High 112 66 86 224 38 57 GRU 56 90 88 High 94 93 High 112 93 95 224 80 90 Boca Raton 56 88 85 High 111 108 High 112 97 118 224 70 96 Lakeland NS 56 110 91 High 119 106 High 112 106 112 224 57 87 OCUD S 56 109 102 High 128 122 High 112 113 127 224 85 109 RPL estimates for 56 kg P ha-1 rate amended columns were determined after 11 harvests. 71

PAGE 72

y = 0.17x + 29.57 r2 = 0.90 0 50 100 150 200 250 0200400600800100012001400 Labile P Load (Biosolids PSI*Total-P Load)Overall P Lability (mg)' Figure 3-19. Overall P labi lity plotted as a function of the labile P load. y = 30.30x + 27.98 r2 = 0.79 0 20 40 60 80 100 120 140 0.00.51.01.52.02.53.03 Biosolids PSI RPL (%).5 Figure 3-20. Biosolids relative P overall lability (R PL) plotted as a function of biosolids phosphorus saturation index (PSI). 72

PAGE 73

y = 0.10x + 21.33 r2 = 0.80 0 20 40 60 80 100 120 140 160 0200400600800100012001400 Labile P Load (Biosolids PSI*Total-P Load) Overall P Lability (mg) Chinault (2007) O'Connor et al. (2002) Figure 3-21. Relationship between overall P labil ity and the labile P load for two previously conducted short-term studies gr eenhouse studies (OConnor et al., 2002; Chinault, 2007). y = 0.16x + 4.41 r2 = 0.73 0 5 10 15 20 25 30 35 40 0 50 100 150 200 250 Labile P Load (Biosolids PSI*Total-P Load)Cumulative P Leached (mg) Figure 3-22. Cumulative P leaching from a 5.5 laboratory incubation study plotted as a function of the environmentally effective P load. Data taken from Chinault (2007). 73

PAGE 74

CHAPTER 4 CONCLUSIONS A major challenge to biosolids land app lication under P-based management was the absence of a long-term study to validate the P la bility characteristics of biosolids suggested by short-term studies. The 16-month greenhouse study described herein represents a substantial effort to characterize the long-term environmental lability and phytoavailability of biosolids-P. Measures of bahiagrass P uptake at the P-based bi osolids application rate continued until tissue P concentrations decreased and remained below cri tical levels. At the Nbased application rate, bahiagrass P uptake was quantified until tissue P concentrations were below a conservative critical threshold (1 g P kg-1), and tissue N concentrations declined below the minimum recommended concentration. Eventually, tissue nut rient concentrations at both application rates suggested additional P fertilization was necessary; indicating the 16-month greenhouse study fully characterized the long-term agronomic valu e of biosolids-P. Less soluble-P (high Fe+Al, conventiona lly processed) biosolids clearly pose significantly less environmental P risk than high so luble-P inorganic fertiliz ers. Most states do not explicitly acknowledge biosolids-P solubility in their P-indices desp ite the growing number of studies demonstrating the lower lability of conven tional biosolids-P compared to inorganic fertilizer-P. Nutrient management plans, including P-Index consider ations, could restrict biosolids application rates to P-based fertiliz er recommendations and increase land application costs to municipalities. Yet, applying less sol uble-P biosolids at rate s greater than crop P demands did not appreciably increase the risk of P leaching. Even at the N-based application rate, P leaching from less soluble-P biosolids was 0.5% of the total-P applied. Data from the greenhouse study suggest that regula tory distinctions in state Pindices are justified for less soluble-P conventional biosolids. However, state P-indices should not utilize a single, 74

PAGE 75

umbrella factor that suggests all biosolids-P is less environmentally labile. The risk of P leaching is similar for BPR biosolids, BPR-like bi osolids, and TSP, and no regulatory distinction is warranted for BPR and BPR-like biosolids. State P-indices should di stinguish between more soluble-P biosolids (BPR and BPR-like biosolids) and less soluble-P conventional biosolids. The distinction will become increasing important as more municipalities utilize BPR techniques to meet effluent P standards. Relative agronomic effectiveness also differs among biosolids. Le ss soluble biosolids-P is less phytoavailable than TSP-P. Estimates of biosolids RPP suggest application rates of conventional, less soluble-P biosolids should increase 25-50% to supply phytoavailable P in quantities equal to inorganic fe rtilizer-P. Very low, P-based biosolids application rates are becoming increasingly mandated under enviro nmental P management regulations, and accounting for reduced phytoavailability of less sol uble-P biosolids is necessary to ensure that crops receive sufficient P. The long-term phytoa vailability of BPR and BPR-like biosolids-P is equal to or greater than inorganic fertilizer, and these more soluble-P biosolids are very effective P fertilizers. No agronomic application rate adjustment is warranted for BPR and BPR-like biosolids. We hypothesized (hypothesis # 1) that the long-term phytoavailability and environmental lability of conventional biosolids-P is less than fertilizer-P (TSP-P), and that the long-term phytoavailability and environmental lability of BP R and BPR-like biosolids-P is similar to TSPP. Data from the prolonged (16-month) greenho use study confirm the expectations, and the first hypothesis is accepted. Combining P uptake an d P leaching into an overall P lability characterization provided an indication of the ulti mate fate of biosolids-P. Data from the greenhouse study suggest that less soluble biosolidsP is ultimately about 40-65% as labile as 75

PAGE 76

TSP-P, and much of conventiona l (high total Fe + Al) biosolid s-P is ultimately unavailable. Conversely, the ultimate lability of BPR and BPR-like biosolids-P is similar to TSP-P. The second study hypothesis was that shortterm greenhouse and lab incubation studies are good approximations of the relative phytoavailabi lity and environmental lability of biosolidsP. Little change in relative P lability differences among biosolid s occurred from harvest 4 to harvest 12, suggesting that shortterm studies adequately approxi mate the relative P lability differences among P-sources. Thus, the second hypothesis is also ac cepted. However, short-term characterizations tend to underestim ate the ultimate relative labili ty of biosolids-P. Biosolids RPP and RPL estimates increased roughly 10-20% from harvest 4 to harvest 12. Short-term studies underestimated the RPP and RPL values of less soluble-P biosolids in particular, because these materials tend to act as a slow release fert ilizer. Ideally, P-source coefficients that are based on long-term relative lability values shou ld be incorporated in state P-indices. The third study hypothesis was th at some measure of biosolidsP is a useful indicator of ultimate biosolids-P lability and biosolids-P phytoavailability. The correlations of overall biosolids-P lability to labile P load (biosolids PSI*total-P load) suggest that biosolids PSI is a useful a priori indicator of biosolids-P lability. Bioso lids PSI values also correlate well with biosolids RPL values. Values of PWEP are good indicators of biosolids-P runoff and leaching potential, but PWEP values tended to underes timate the quantity of biosolids-P ultimately available to plant roots. The P uptake vs. labile P load correlati ons suggest that biosolids PSI is also a useful a priori indicator of biosolids-P agronomic effectiveness, and adequately predicts biosolids relative P phytoavailability (RPP) values. Thus, th e third hypothesis is accepted. Several caveats to using biosolids PSI as an a priori P-lability measure exist. The greenhouse study utilized a P-defici ent sandy soil with little P sorp tion capacity so that the soil 76

PAGE 77

would not mask P-source lability differences. An a priori P-source P lability measure is especially useful in areas such as Florida where low P sorbing soils dominate and where P lability is expected to be controlled by P-sour ce characteristics. Soils with appreciable P sorption capacity dominate in the majority of the US, and an a priori biosolids-P lability measure is of little value because P lability is controlled by the soil. The PSI concept is not applicable to all biosolids. Reactive (oxalate-extractable) Fe+A l is expected to control P lability in most biosolids; however, some municipalities utili ze advanced alkaline stab ilization techniques and the PSI concept is not applicable for such materials. Lastly, few data exist to show that the PSI concept adequately predicts bi osolids-P lability in field conditions. A field study, utilizing biosolids with a wide-range of P solubilities, could further confirm that bioso lids PSI is a useful a priori measure of biosolids-P lability. 77

PAGE 78

APPENDIX A R SQUARED VALUES FOR VARIOUS CORRELATIONS Table A-1. Relationship between va rious measures of biosolids-P with cumulative P uptake and overall P lability in the 16-month greenhouse study. Correlation with Correlation with Measure of biosolids-P P uptake overall P lability ---------------r2 value -------------Total-P load 0.45 0.47 Iron strip-P load 0.03 0.04 Oxalate extractab le-P load 0.64 0.62 Environmentally effective P load 0.47 0.56 Labile P load 0.85 0.90 Determined by Chinault (2007) The product of the biosolids percent water extr actable-P value (PWEP) and the total-P load The product of the biosolids phosphorus satu ration index value (PSI) and the total-P load Table A-2. Relationship between va rious measures of biosolids-P with estimates of biosolids relative P phytoavailability (RPP) a nd biosolids relative P lability (RPL) in the 16month greenhouse study. Correlation with Correlation with Measure of biosolids-P RPP RPL ---------------r2 value ------------Total-P 0.21 0.33 Iron strip-P 0.00 0.01 Oxalate extractable-P 0.50 0.53 Percent water extractable-P (PWEP) 0.34 0.46 Phosphorus saturation index (PSI) 0.76 0.79 Determined by Chinault (2007) The molar ratio of oxalate extracta ble-P to oxalate extractable Fe+Al 78

PAGE 79

APPENDIX B ADDITIONAL BIOSOLIDS PSI CORRELATIONS y = 32.09x + 36.43 r2 = 0.82 0 20 40 60 80 100 120 140 0.00.51.01.52.02.53.03.5 Biosolids PSIRPP (%) Current greenhouse study Oladeji (2006) Figure B-1. Long-term estimates of biosolids re lative P phytoavailability (RPP) plotted as a function of bios olids PSI values. The long-term RPP estimates are from the current greenhouse study and Oladeji (2006). y = 39.83x + 0.08 r2 = 0.81 0 20 40 60 80 100 120 140 0.00.51.01.52.02.53.03.5 Biosolids PSIRPP (%) O'Connor et al. (2004) Chinault (2007) Figure B-2. Short-term estimates of biosolids relative P phytoavailability (RPP) plotted as a function of biosolids PSI values Short-term RPP values were determined by OConnor et al. (2004) and Chinault (2007). 79

PAGE 80

y = 0.03x + 25.47 r2 = 0.62 0 10 20 30 40 50 60 70 020040060080010001200 Labile P Load (Biosolids PSI*Total-P Load)P Uptake (mg ) Figure B-3. Cumulative P uptak e from a short-term greenhouse study (OConnor et al., 2004) plotted as a function of the labile P load. The short-term greenhouse study utilized Candler soils with a relative P adsorption value of ~15%. 80

PAGE 81

LIST OF REFERENCES Adjei, M.B., C.S. Gardner, D. Mayo, T. Seaw right, and E. Jennings. 200 0. Fertilizer treatment effects on forage yield and quality of tropical pasture grasses. Soil and Crop Sci. Soc. Florida Proc. 59:32-37. Agyin-Birikorang, S., G.A. OConnor, and S.R. Brinton. 2008. Evaluating phosphorus loss from a Florida spodosol as affected by P-source application. J. Environ. Qual. 37:1180-1189. Andersen, J.M. 1976. An ignition method for de termination of total phosphorus in lake sediments. Water Research 10:329-331. Attai, B., O. Moghaddam, M. Bruno, and J. Young. 2008. One mile under. Water Environ. and Tech. 20(3): 44-49. Box, G.E.P. and D.R. Cox. 1964. An analysis of transformations. J. Royal Stat. Soc. 26:211-243. Brandt, R.C., H.A. Elliott, and G.A. OConnor. 2004. Water-extractable phosphorus in biosolids: implications for land-based recy cling. Water Environ. Res. 76:121-129. Chambliss, C., P. Miller, and E. Lo rd. 2001. Florida Cow-Calf Management, 2nd Ed.-Forages. UF/IFAS Publication AN118. Univ. FL, Gainesville. Champagne, P. Feasibility of producing bioethanol from waste residues: a Canadian perspective, Feasibility of producing bio-ethanol from wast e residues in Canada. Resour. Conserv. Recycl. 50:211-230. Chinault, S.L. 2007. The agronomic and enviro nmental characterization of phosphorus in biosolids produced and/or marketed in Florida. Masters thesis presented to the University of Florida Graduate School, Gainesvile, FL. De Haan, S. 1980. Sewage sludge as a phosphate fertilizer. Phosphorus in Agric. 34:33-41. Elliott, H.A., G.A. OConnor, and S. Brinton. 2002. Phosphorus leaching of biosolids-amended sandy soils. J. Environ. Qual. 31:681-689. Elliott, H.A, R.C. Brandt, and J.S. Shortle. 2007. Biosolids disposal in Pennsylvania. Final report to the Center for Rural Pennsylvania, Harrisburg, PA. Elliott, H.A., and G.A OConnor. Phosphorus manage ment for sustainable biosolids recycling in the United States. 2007. Soil Biol. and Biochem. 39:1318-1327. Epstein, E. 2003. Land application of sewage slud ge and biosolids. Lewis Publ., Boca Raton, FL. 81

PAGE 82

Florida Department of Environmental Protection. 2007. Summary of Class AA residuals 2006. http://www.dep.state.fl.us/water/wastewater/dom/docs/Cla ssAA_residuals_yr_report_200 6.pdf Florida Department of Environmental Protection. Tallahassee, Florida. June 2007. Harris, W.G., R.D. Rhue, G. Kidder, R.B. Brow n, and R. Little. 1996. Phosphorus retention as related to morphology and sandy coastal plain soil materials. Florida Coop. Ext. Serv. Circ. 817. Univ. of Florida, Gainesville Havlin, J.L., J.D. Beaton, S. L. Tisdale, and W.L. Nelson. 1999. Soil fertility and fertilizers: an introduction to nutrient management. 6th edition. Prentice-Hall Inc., New Jersey. He, Z.L., A.K. Alva, and Y.C. Li. 1999. Sorptio n-desorption and solution concentration of phosphorus in fertilized sandy so il. J. Environ. Qual. 28:1804-1810. Hocking, R. R. 1976. The analysis and selection of variables in linear regression. Biometrics 2:149. Kennedy, V.H., A.P. Rowland, and J. Parringt on. 1994. Quality assurance for soil nutrient analysis. Commun. Soil Sci. Plant Anal. 25:1605. Kleinman, P.J.A., A.N. Sharpley, A.M. Wolf, D. B. Beegle, and P.A. Moore Jr. 2002. Measuring water extractable phosphorus in manur e. Soil Sci. Soc. Amer. J. 66:2009-2015. Lemunyon, J.L., and R.G. Gilbert. 1993. The con cept and need for a phosphorus assessment tool. J. Prod. Agric. 6:483-496. Lu, P., and G.A. OConnor. 2001. Biosolids effects on P retention and release in some sandy Florida soils. J. Environ. Qual. 30:1059-1063. Maguire, R.O., J.T. Sims, S.K. Dentel, F.J. Coale, and J.T. Mah. 2001. Relationship between biosolids treatment process and soil phosphor us availability. J. Environ. Qual. 30:10231033. Maguire, R.O., and J.T. Sims. 2002. Soil testing to predict phosphorus leaching. J. Environ. Qual. 31:1601. McLaughlin, M.J., and L. Champion. 1987. Sewage sludge as a phosphorus amendment for sequioxic soils. Soil Sci. 143:113-119. Moss, L.H., E. Epstein, and T.L. Logan. 2002. Evaluating risks and benefits of soil amendments used in agriculture. Final Report Proj ect 99-PUM-1, Water Environment Research Foundation, Alexandria, VA. Murphy, J., and J.P. Riley. 1962. A modified sing le solution method for the determination of phosphate in natural water. Anal. Chim. Acta 27:31. 82

PAGE 83

Nair, V.D., and W.G. Harris. 2004. A capacity factor as an alternative to soil test phosphorus in phosphorus risk assessment. N. Z. J. Agric. Res. 47:491-497. National Research Council (NRC). 1996. Nutrie nts requirements of beef cattle, 7th ed. Natl. Acad. Press., Washington, DC. National Research Council (NRC). 2002. Biosolids applied to land: Adva ncing standards and practice. Natl. Acad. Press., Washington, DC. Nelson, D.W., and L.E. Sommers. 1996. Total carbon, organic carbon, and organic matter. p. 961. In D.L. Sparks et al. (ed.) Methods of soil analysis. Part 3. SSSA Book Ser. 5. SSSA, Madison, WI. OConnor, G.A., and D. Sarkar. 1999. Fate of land applied residuals-bound phosphorus. DEP WM 661. Florida Environmental Protection Agency, Tallahassee. OConnor, G.A., D. Sarkar, S.R. Brinton, H.A. E lliott, and F.G. Martin. 2004. Phytoavailability of biosolids phosphorus. J. Environ. Qual. 33:703-712. OConnor, G.A., D. Sarkar, D.A. Graetz, and H.A. Elliott. 2002. Characterizing the forms, solubilities, bioavailabili ties, and mineralization rates of phosphorus in biosolids, commercial fertilizers, and manures (Phase I). Final Report Project 99-PUM-2T, Water Environment Research Foundation, Alexandria, VA. OConnor, G.A., and H.A. Elliott. 2006. The agronomic and environmental availability of biosolids-P (Phase II). Final Report Proj ect 99-PUM-2T, Water Environment Research Foundation, Alexandria, VA. Oladeji, O.O. 2006. Management of phosphorus s ources and water treatm ent residuals (WTR) for environmental and agronomic benefits. Di ssertation presented to the University of Florida Graduate School, Gainesville, FL. Sarkar, D., and G.A. OConnor. 2004. Plant and so il responses to biosolids-phosphorus in two Florida soils with high phosphorus conten t. Commun. Soil Sci. Plt. Anal. 35:1569-1589. Schatman, D.P, R. J. Reid, and S.M. Ayling. 1998. Phosphorus uptake by plants: from soil to cell. Plt. Physiol. 116:447-452. Sharpley, A.N., S.C. Chapra, R. Wedepohl, J. T. Sims, T.C. Daniel and K.R. Reddy. 1994. Managing agricultural phosphorus for protection of surface waters-Issues and options. J. Environ. Qual. 23:437-451. Sharpley, A.N., T.C. Daniel, J.T. Sims, and D.H. Pote. 1996. Determining environmentally sound phosphorus levels. J. Soil Water Conserv. 51:160-166. 83

PAGE 84

Sharpley, A.N., and B. Moyer. 2000. Phosphorus fo rms in manure and compost and their release during simulated rainfall. J. Environ. Qual. 29:1462-1469. Sharpley, A.N., J.L. Weld, D.B. Beegle, P.J.A. Kleinman, W.J. Gburek, P.A. Moore Jr., and G. Mullins. 2003. Development of phosphorus indices for nutrient management planning strategies in the United States J. Soil Water Conserv. 58:137-152. Shober, A.L., and J.T. Sims. 2003. Phosphorus rest rictions for land application of biosolids: Current status and future tr ends. J. Environ. Qual. 32:1955-1964. Sims, J.T., R.R. Simard, and B.C. Joern. 1998. Phosphorus loss in agricultural drainage: historical perspective and current research. J. Envi ron. Qual. 27:277-293. Sinclair, T.R., J.D. Ray, P. Mislevy, and L. M. Premazzi. 2003. Growth of subtropical forage grasses under extended photoperiod duri ng short-daylength months. Crop Sci. 43:618623. Stehouwer, R.C., A.M. Wolf, and W.T. Doty. 2000. Chemical monitoring of sewage sludge in Pennsylvania: variability and application uncertainty. J. Environ. Qual. 29:1686-1695. Ternouth, J.H, G. Bortolussi, D.B. Coates, R.E. Hendricksen, and R.W. McLean. 1996. The phosphorus requirements of growing cattle c onsuming forage diets. J. Agric. Sci. 126:503-510. Tukey, J.W. 1949. Comparing individual means in the analysis of variance. Biometrics 5:99-114. USEPA. 1993a. Standards for the use and dis posal of sewage sludge. 40 CFR Part 257, 403, and 503. Federal Register 58:9248-9415. USEPA. 1993b. Methods for determination of inor ganic substances in environmental samples. Revision 2.0 365.1. Washington, D.C. USEPA. 1995. Process design manual: Land app lication of sewage sludge and domestic septage. EPA/625/R-95/001. Office of Res. and Dev. Cincinnati, OH. USEPA. 1999. Biosolids generation, use and disposal in the United States. http://www.epa.gov/epaoswer/non-hw/compost/biosolid.pdf USEPA. Washington, D.C. March 2007. June 2007. Vance, C.P., C. Uhde-Stone, and D.L. Allan. 2003. Phosphorus acquisiti on and use: critical adaptations by plants for securing a nonr enewable resource. New Phytol. 157:423-447. Wen, G., T.E. Bates, R.P. Voroney, J.P Winter, and M.P. Schellenbert. 1997. Comparison of phosphorus availability with application of sewage sludge, sludge compost, and manure compost. Commun. Soil Sci. Plt. Anal. 28:1481-1497. 84

PAGE 85

85 BIOGRAPHICAL SKETCH Matthew (Matt) L. Miller was born in Athens, Georgia, in 1983. Matt is the son of Lane and Sheila Miller and has one older sister, Shannon. Matt spen t much of his youth on a small farm in Bishop, Georgia, where he learned to appreciate agriculture and the environment. Matt received a B.S. in forest resources from the Un iversity of Georgia in May 2006. He enjoyed the undergraduate research he performed under Dr. La rry Morris at the Univer sity of Georgia and chose to attend graduate school. Matt joined the Soil and Wate r Science Department at the University of Florida and began his graduate st udy in soil chemistry under the supervision of Dr. George A. OConnor in August 2006. Matt is sche duled to graduate with his M.S. degree in August 2008.