<%BANNER%>

Fate, Transport, and Risk of Biosolids-Borne Triclocarban

Permanent Link: http://ufdc.ufl.edu/UFE0024341/00001

Material Information

Title: Fate, Transport, and Risk of Biosolids-Borne Triclocarban
Physical Description: 1 online resource (237 p.)
Language: english
Creator: Hodges, Elizabeth
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

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

Notes

Abstract: FATE, TRANSPORT, AND RISK ASSESSMENT OF BIOSOLIDS-BORNE TRICLOCARBAN (TCC) Triclocarban (TCC) is an active ingredient in antibacterial bar soaps, a common constituent of domestic wastewater, and the subject of recent criticism by consumer advocate groups and academic researchers alike. Activated sludge treatment readily removes TCC from the liquid waste stream and concentrates the antimicrobial in the solid fraction, which is often processed to produce biosolids intended for land-application. Greater than half of the biosolids generated in the US are land-applied, resulting in a systematic release of biosolids-borne TCC into the terrestrial and, potentially, the aquatic environment. Despite widespread use of antimicrobial personal care products, the propensity of TCC to partition into biosolids at parts-per-million concentrations, and potential endocrine effects of TCC (Chen et al., 2008), human and ecological health risk assessments for TCC in land-applied biosolids have not been conducted. A project funded by USEPA, and in collaboration with the Procter & Gamble Company (P & G), was designed to fill multiple TCC data gaps and to facilitate an integrated human/ecological health risk assessment. Data developed under the project include conclusive solubility, partitioning, biosolids-borne TCC concentrations, environmental transport, persistence, and soil organism impacts. The data were used to characterize human/ecological exposure hazards and estimate potential adverse effects associated with TCC in land-applied biosolids. No biosolids-borne TCC exposure pathways resulted in estimates of adverse human health effects, even under worst-case land-application scenarios. Two exposure pathways of concern (i.e. biosolids?soil?soil organism?soil predator, and biosolids?soil?surface water?aquatic organism) were identified in the resulting integrated risk assessment, and preliminary TCC pollution limits protective of the most sensitive test species were calculated. The preliminary TCC pollutant limits, could have important implications for current biosolids land-application practices, but need to be reassessed in light of available measured TCC concentrations in environmental matrices and additional environmental effects data before it is appropriate to suggest modifications to current land-application regulations.
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 Elizabeth Hodges.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: O'Connor, George A.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-02-28

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024341:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024341/00001

Material Information

Title: Fate, Transport, and Risk of Biosolids-Borne Triclocarban
Physical Description: 1 online resource (237 p.)
Language: english
Creator: Hodges, Elizabeth
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

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

Notes

Abstract: FATE, TRANSPORT, AND RISK ASSESSMENT OF BIOSOLIDS-BORNE TRICLOCARBAN (TCC) Triclocarban (TCC) is an active ingredient in antibacterial bar soaps, a common constituent of domestic wastewater, and the subject of recent criticism by consumer advocate groups and academic researchers alike. Activated sludge treatment readily removes TCC from the liquid waste stream and concentrates the antimicrobial in the solid fraction, which is often processed to produce biosolids intended for land-application. Greater than half of the biosolids generated in the US are land-applied, resulting in a systematic release of biosolids-borne TCC into the terrestrial and, potentially, the aquatic environment. Despite widespread use of antimicrobial personal care products, the propensity of TCC to partition into biosolids at parts-per-million concentrations, and potential endocrine effects of TCC (Chen et al., 2008), human and ecological health risk assessments for TCC in land-applied biosolids have not been conducted. A project funded by USEPA, and in collaboration with the Procter & Gamble Company (P & G), was designed to fill multiple TCC data gaps and to facilitate an integrated human/ecological health risk assessment. Data developed under the project include conclusive solubility, partitioning, biosolids-borne TCC concentrations, environmental transport, persistence, and soil organism impacts. The data were used to characterize human/ecological exposure hazards and estimate potential adverse effects associated with TCC in land-applied biosolids. No biosolids-borne TCC exposure pathways resulted in estimates of adverse human health effects, even under worst-case land-application scenarios. Two exposure pathways of concern (i.e. biosolids?soil?soil organism?soil predator, and biosolids?soil?surface water?aquatic organism) were identified in the resulting integrated risk assessment, and preliminary TCC pollution limits protective of the most sensitive test species were calculated. The preliminary TCC pollutant limits, could have important implications for current biosolids land-application practices, but need to be reassessed in light of available measured TCC concentrations in environmental matrices and additional environmental effects data before it is appropriate to suggest modifications to current land-application regulations.
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 Elizabeth Hodges.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: O'Connor, George A.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-02-28

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024341:00001


This item has the following downloads:


Full Text

PAGE 1

FATE, TRANSPORT, AND RISK ASSESSMENT OF BIOSOLIDS-BORNE TRICLOCARBAN (TCC) By ELIZABETH ALLENE HODGES SNYDER A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009 1

PAGE 2

2009 Elizabeth Allene Hodges Snyder 2

PAGE 3

To my husband, Sam, and my mom and pops, Lisa and Mark We are products of our heredity and environment, and parents provide all of one and most of the other. Glenn W. Suter II 3

PAGE 4

ACKNOWLEDGMENTS Completion of this dissertation was made possible through the efforts of countless individuals who were generous with their knowledge, time, and re sources. The lions share of my gratitude goes to my major advisor, Dr. George OConnor. I remember the day he agreed to be my mentor. We discussed our respective expect ations for my term as a doctoral student as I sipped a cup of coffee and Dr. OConnor enjoyed a particular brand of cigar that I would soon come to always associate with him. At the e nd of our conversation, he looked squarely at me and said, I dont tolerate a ny bull. Truer words were ne ver spoken. Dr. OConnor holds himself to the highest standard of excellence, and expects the same from his students. But with that expectation comes the guidance, support, and advocacy that makes excellence achievable. Dr. OConnor, who pushed me, and believed in me even when I did not, is the reason I have any business calling myself a scientist. I would like to thank my committee member s, Drs. Joe Delfino, Jim Jawitz, Drew McAvoy, and Andy Ogram. In addition to individually contributing advice, encouragement, and/or access to instrumentation, they collectively formed a formidable group of scientists. The prospect of sitting acro ss the table from them during my qualif ication exam and defense kept me motivated throughout my doctoral studies. An extra thank you goes to Dr. McAvoy for securing the incredible opportuni ty to work as a summer intern at Procter & Gamble (P&G). During my internship, I learned from a group of people in the P&G Environmental Science Department who shared smiles and great ideas wi th equal fervor. Special thanks go to Brad Price (who introduced me to the exciting world of mass spectrometry), Erin Schwab (who took me into her home), and Nina Itrich (who introduced me to radiochemistry). Thank you to Drs. Margaret James, Bob Querns, and John Thomas for providing access to, and guidance regarding, analytical equipment that enabled me to generate the data presented in 4

PAGE 5

the following chapters. Thank you to Dawn Lu cas, Yu Wang, Gavin Wilson, and the members of the Soil Ecology and Molecular Biology Lab, who helped me when I felt most out of my element. I also need to thank some former and current members of the Soil Chemistry Lab, including Sampson Agyin-Birikora ng, Scott Brinton, Sarah Chinault, Matt Miller, Daniel Moura, Wale Oladeji, and Manmeet Waria, for their unwavering support and camaraderie. To my officemates in the Incubation Room, I will miss you. I am indebted to the USEPA Office of Wastew ater Management and the Procter & Gamble Company for funding all of my re search and much of my education. My experience with both organizations was overwhelmingly positive. Finally, it was the folks at home that helped me keep it all together. Thank you to my Mom and Dad for always insisting that I climb upon their shoulders so I could reach just a little further, and for being a sounding board that has never steered me wrong. Thank you to my husband, Sam, for loving me through all kinds of weather, and keeping my mind, heart, and belly well fed. 5

PAGE 6

TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ........10 LIST OF FIGURES.......................................................................................................................12 LIST OF ABBREVIATIONS........................................................................................................14 ABSTRACT...................................................................................................................................16 CHAPTER 1 INTRODUCTION AND PROJECT OBJECTIVES..............................................................18 Introduction................................................................................................................... ..........18 Intermediate Objectives........................................................................................................ ..20 Intermediate Objective 1: Confirm Physicochemical Properties of TCC......................20 Intermediate Objective 2: Characterize TCC Concentrations in Multiple Biosolids and Partitioning of Indigenous TC C in the Biosolids Matrix......................................21 Intermediate Objective 3: Characterize the Leachability of Biosolids-Borne TCC in Amended Soils.........................................................................................................23 Intermediate Objective 4: Characterize Biodegradation of Biosolids-Borne TCC in Amended Soils.............................................................................................................23 Intermediate Objective 5: Characterize Biosolids-Borne TCC Toxicity to Terrestrial Organisms...................................................................................................24 Intermediate Objective 6: Characteri ze Biosolids-Borne TCC Bioaccumulation..........25 Ultimate Objective: Conduct an Integrated Human and Ecologica l Risk Assessment of Biosolids-Borne TCC.....................................................................................................................................26 2 WATER SOLUBILITY AND OCTANOL-WATER PARTITIONING OF TCC................29 Introduction................................................................................................................... ..........29 Materials and Methods...........................................................................................................30 Chemicals........................................................................................................................30 Water Solubility Determination......................................................................................30 Octanol-Water Partition Coefficient (Kow) Determination..............................................31 HPLC/MS Analyses........................................................................................................31 Results and Discussion......................................................................................................... ..32 3 BIOSOLIDS-BORNE TCC CONCEN TRATIONS AND PARTITIONING........................40 Introduction .............................................................................................................................40 Materials and Methods...........................................................................................................42 6

PAGE 7

Chemicals........................................................................................................................42 Biosolids Extraction Method Development....................................................................43 First biosolids-borne TCC ex traction method validation.........................................43 Second biosolids-borne TCC extraction method validation....................................45 Third biosolids-borne TCC extraction method validation.......................................46 Fourth biosolids-borne TCC extraction method validation......................................47 Extraction and Analysis of Total Biosolids-Borne TCC.................................................48 Partitioning Coefficient Determination...........................................................................49 Indigenous TCC partitioning method.......................................................................49 Spiked TCC partitioning method.............................................................................49 Results and Discussion......................................................................................................... ..50 TCC Concentrations in Biosolids....................................................................................50 Partitioning of Indigenous and Spiked Biosolids-Borne TCC........................................55 4 LEACHABILITY OF BIOSOLIDS-BORNE TCC IN AMENDED SOIL COLUMNS.......65 Introduction................................................................................................................... ..........65 Materials and Methods...........................................................................................................66 Chemicals, Biosolids, and Soils......................................................................................66 Study Design................................................................................................................... 66 Leachate Preparation and Analysis.................................................................................67 Results and Discussion......................................................................................................... ..67 5 PLANT UPTAKE OF BIOSOLIDS-BORNE TCC...............................................................76 Introduction................................................................................................................... ..........76 Materials and Methods...........................................................................................................80 Chemicals, Biosolids, and Soils......................................................................................80 Study Design................................................................................................................... 80 Plant Tissue Extraction Met hod Validation and Analysis...............................................82 Results and Discussion......................................................................................................... ..83 TCC Concentrations in Plant Tissue...............................................................................83 Model-Predicted TCC Concentrations in Plant Tissue...................................................86 6 EARTHWORM TOXICITY AND BIOACCU MULATION OF BIOSOLIDS-BORNE TCC.........................................................................................................................................90 Introduction................................................................................................................... ..........90 Materials and Methods...........................................................................................................91 Chemicals, Biosolids, and Soils......................................................................................91 Range-Finding Test Design.............................................................................................91 Range-Finding Test Results............................................................................................92 Definitive Test Design.....................................................................................................93 Biosolids-Borne TCC Bioaccumulation Study Design...................................................94 Results and Discussion......................................................................................................... ..95 Toxicity of TCC to Earthworms in Biosolids-Amended Soils........................................95 Bioaccumulation of TCC by Earthworms in Biosolids-Amended Soils.........................96 7

PAGE 8

Measured TCC concentrations in earthworm tissue................................................96 Model-predicted TCC concentrations in earthworm tissue......................................96 7 SOIL MICROBIAL TOXICITY OF BIOSOLIDS-BORNE TCC......................................106 Introduction................................................................................................................... ........106 Triclocarban Mechanism of Action...............................................................................106 Potential Impacts of Biosolids-Bo rne TCC on Soil Microorganisms...........................107 Materials and Methods.........................................................................................................111 Chemicals, Biosolids, and Soils....................................................................................111 Range-Finding Test Design...........................................................................................111 Range-Finding Test Results..........................................................................................113 Definitive Test Design...................................................................................................114 Results and Discussion......................................................................................................... 115 Effects of Biosolids-Borne TCC on Soil Respiration....................................................115 Effects of Biosolids-Borne TCC on Ammonification and Nitrification.......................116 8 AEROBIC 14C-TCC BIODEGRADATION IN BIOSOLIDS-AMENDED SOILS............128 Introduction................................................................................................................... ........128 Materials and Methods.........................................................................................................137 Chemicals, Biosolids, and Soils....................................................................................137 Study Design.................................................................................................................13 8 Sample Analyses...........................................................................................................139 Results and Discussion......................................................................................................... 141 Explanation of the Sequential Extraction Scheme........................................................141 Parent Compound Loss and Changes in Extractability.................................................143 Differences between 14C recoveries in sequential extracts of biotic and inhibited-biotic samples......................................................................................143 Differences between 14C recoveries within treatment extracts over time..............144 Mineralization of 14C-TCC.....................................................................................144 Inferences made from measured mine ralization rates and differences in 14C extractability with time and between treatments................................................145 Confirmation of Biotic Conditions................................................................................148 Extract Speciation..........................................................................................................149 Estimation of TCC Half-Life in Soil.............................................................................149 9 RISK ASSESSMENT OF BIOSOLIDS-BORNE TCC.......................................................162 Introduction................................................................................................................... ........162 Integrated Risk Assessment..................................................................................................164 Assessment Step 1: Problem Formulation.....................................................................164 Assessment Step 2: Analysis........................................................................................166 Characterization of effects supporting studies....................................................166 Characterization of effects human and animal studies........................................167 Reference dose (RfD) calculation..........................................................................168 Screening-level exposure concentration calculation..............................................170 8

PAGE 9

Screening-level hazard index calculation...............................................................172 Risk Characterization of Critical Pathways...................................................................172 Pathway 10: Biosolids soil soil organism predator......................................172 Pathway 16: Biosolids soil surface water aquatic organism..........................175 Calculation of Preliminary Biosolids-Borne TCC Pollutant Limits..............................178 Cumulative pollutant loading rates (CPLRs).........................................................178 Annual pollutant loading rate (APLR)...................................................................180 Ceiling concentration limit.....................................................................................181 Pollutant concentration limit..................................................................................181 10 SUMMARY, CONCLUSIONS AND RECOMMENDATIONS.......................................206 Introduction................................................................................................................... ........206 Summaries of Intermediate Objective Results and Future Research Needs.........................208 Intermediate Objective 1: Confirm Physicochemical Properties of TCC.....................208 Intermediate Objective 2: Characterize TCC Concentrations in Multiple Biosolids and Partitioning of Indigenous TC C in the Biosolids Matrix....................................208 Intermediate Objective 3: Characterize Leachability of Biosolids-Borne TCC in Amended Soils...........................................................................................................210 Intermediate Objective 4: Characterize Biodegradation of Biosolids-Borne TCC in Amended Soils...........................................................................................................212 Intermediate Objective 5: Characterize Biosolids-Borne TCC Toxicity to Terrestrial Organisms.................................................................................................213 Intermediate Objective 6: Characterize Biosolids-Borne TCC Bioaccumulation.........214 Implications of the Preliminary TCC Pollutant Limits and Recommended TCC Research Priorities...........................................................................215 Research Priority #1: Improve Characte rization of the Most Sensitive Exposure Pathways....................................................................................................................216 Pathway 10-specific research priorities..................................................................216 Pathway 16-specific research priorities..................................................................219 Research needs relevant to bot h critical exposure pathways.................................219 Research Priority #2: Fill Remaining TCC Toxicity Gaps..........................................220 Applications to the USEPA Data Requirements for the Antimicrobial Pesticides Proposed Rule...................................................................................................................222 WORKS CITED.................................................................................................................... ......224 BIOGRAPHICAL SKETCH.......................................................................................................237 9

PAGE 10

LIST OF TABLES Table page 2-1 Commonly reported physicochemical properties of triclocarban (TCC)..........................38 2-2 Triclocarban (TCC) solubility measurements....................................................................38 2-3 Triclocarban (TCC) Kow measurements.............................................................................39 3-1 Batch 1 biosolids and select physicochemical properties..................................................60 3-2 Batch 2 biosolids and select physicochemical properties..................................................61 3-3 Triclocarban (TCC) concentr ations in 23 wastewater trea tment plant materials (n=3).....62 3-4 Indigenous triclo carban (TCC) log Kd and log Koc values for 16 biosolids......................63 4-1 Triclocarban (TCC) toxicity endpoints for aquatic indicator organisms (TCC Consortium, 2002a; Chalew and Halden, 2009)................................................................73 4-2 Triclocarban (TCC) in biosolids-amended soil column leachates.....................................74 5-1 Triclocarban (TCC) concentrations and lo ading rates in the biosolids-amended soil plant uptake study..............................................................................................................87 5-2 Measured and predicted triclocarban (TCC) plant tissue concentrations and bioaccumulation factors (BAF)..........................................................................................87 6-1 Measured (means; n = 3 and SE) and estim ated triclocarban (TCC) concentrations and bioaccumulation factor (BAF) values in earthworm tissue.......................................103 7-1 Day 31 NO3 --NO2 --N statistical groupings by treatment (definitive test)........................127 8-1 Select physicochemical properties of the soils used in the biosolids-borne triclocarban (TCC) study.................................................................................................153 8-2 Biodegradation experime nt sample treatments................................................................154 8-3 Radiolabel percent recove ries in the fine sand soil..........................................................155 8-4 Radiolabel percent recoveries in the silty clay loam soil.................................................156 9-1 Steps of human, ecological, and inte grated health risk assessments...............................183 9-2 Human and ecological exposure path ways for land-applied biosolids............................184 9-3 Toxicity and carcinogenicity of triclocarban (TCC) in rodents.......................................185 10

PAGE 11

9-4 Mutagenicity and clastoge nicity of triclocarban (TCC)..................................................186 9-5 Toxicity of triclocarban (TCC) to aquatic organisms......................................................187 9-6 Parameters and assumptions for calculating screening level hazard indices (HI values)..............................................................................................................................188 9-6 Continued.................................................................................................................. .......189 9-6 Continued.................................................................................................................. .......190 9-6 Continued.................................................................................................................. .......191 9-6 Continued.................................................................................................................. .......193 9-7 Equations used to calculate screen ing-level hazard indices (HI values).........................196 9-7 Continued.................................................................................................................. .......197 9-7 Continued.................................................................................................................. .......198 9-7 Continued.................................................................................................................. .......199 9-8 Parameters and assumptions for calculati ng adjusted hazard indices (HI values)...........200 9-8 Continued.................................................................................................................. .......201 9-9 Adjusted hazard indices (HI values) for exposure pathway 10.......................................202 9-10 Adjusted hazard indices (HI values) for exposure pathway 16.......................................203 11

PAGE 12

LIST OF FIGURES Figure page 1-1 Chemical structure of tricloca rban (TCC; N-(4-chlorophenyl)-N`-(3,4dichlorophenyl) urea).........................................................................................................28 2-1 Triclocarban (TCC) solubility dete rmination: column elution diagram............................39 3-1 Reported biosolids-borne triclocarban (TCC) concentration distributions, showing the mean, 25th and 75th percentile, maximum, and minimum values ( = biosolids; = sludge; = biosolids and sludge) (assuming normal distributions)...........................64 5-1 Relationship between triclocarban (TCC) concentrations in biosolids-amended soil and TCC concentrations in Bahia grass plant tissue..........................................................88 5-2 Mean bioaccumulation of triclocarban (T CC) in Bahia grass ti ssue as a function of the rate of organic carbon applied to biosolids-amended soil columns.............................88 5-3 Mean bioaccumulation of triclocarban (T CC) in Bahia grass ti ssue as a function of biosolids-amended soil concentration................................................................................89 6-1 Mean percent of living earthworms rema ining as a function of biosolids-borne triclocarban (TCC) concentration and e xposure duration in soils amended with CHCC biosolids (unreplicat ed range-finding test)..........................................................104 6-2 Mean percent of living earthworms (n=3) remaining as a function of biosolids-borne triclocarban (TCC) concentration and exposure duration in a sandy soil amended with CHCC biosolids (definitive test)..............................................................................105 7-1 Milligrams of CO2 evolved as a function of tric locarban (TCC) concentration and time (unreplicated range-finding test)..............................................................................122 7-2 Concentration of NH4 +-NH3 in soil and biosolids-ame nded soil as a function of triclocarban (TCC) concentration and time (unreplicated range-finding test).................122 7-3 Concentration of NO3--NO2--N in soil a nd biosolids-amended so il as a function of triclocarban (TCC) concentration and time (unreplicated range-finding test).................123 7-4 Milligrams of CO2 evolved as a function of tric locarban (TCC) concentration and time (definitive test) (like letters and colors indicate no significant difference between treatments).........................................................................................................124 7-5 Cumulative CO2 evolved from biosolids-amended sand as a function of triclocarban (TCC) concentration and time (definitiv e test)................................................................125 12

PAGE 13

7-6 Concentration of NH4 +-NH3-N in soil and biosolids-amended soil as a function of triclocarban (TCC) concentration and time (definitive test) (like letters and colors indicate no significant difference between treatments)...................................................126 7-7 Concentration of NO3 --NO2 --N in soil and biosolids-am ended soil as a function of triclocarban (TCC) concentration and time (definitive test) (like letters and colors indicate no significant difference between treatments; see Ta ble 7-1 for Day 31 statistical groupings)........................................................................................................127 8-1 Diagram of biodegradat ion experiment design................................................................157 8-2 Radiolabeled triclocarban (TCC) spike rec overies as a function of treatment, fraction (water, methanol, sodium hydroxide, 14carbon dioxide, and combusted), and time (normalized for 100% total spike recovery)....................................................................158 8-3 Cumulative CO2 production by biosolids-amended soil samples in the 14Ctriclocarban (14C-TCC) biodegradation experiment........................................................159 8-4 Radio-thin layer chromatography (RADTLC) standard (a.) and sample (b.; representative of T0-T7) chromatogr aphs (bottom) and corresponding RAD-TLC fingerprints (top)...........................................................................................................160 8-5 Percent radiolabel remaining (100%-%14CO2) in biosolids-amended sand over time....161 8-6 Percent radiolabel remaining (100%-%14CO2) in biosolids-amended silty clay loam over time..........................................................................................................................161 9-1 Depiction of the integrated health risk assessment framework (WHO, 2001)................204 9-2 Conceptual model of human and ecological exposures to biosolid s-borne triclocarban (TCC)...............................................................................................................................205 10-1 Summary of the biosolid s-borne triclocarban (TCC) ri sk assessment process................223 13

PAGE 14

LIST OF ABBREVIATIONS BASL4 Biosolids-amended soil level IV (computer model) CLPP Community-level physiological profiling DOM Dissolved organic matter d.w. Dry weight equivalent EC50 Effective concentration in 50% of the test population ECOSAR Ecological structur al activity relationships EPI Suite Estimation Programs Interface Suite (computer model) HEI Highly exposed individual HI Hazard index HPLC/MS High performance liquid chromatography, mass spectrometry HPLC/MS-MS High performance liquid chro matography, tandem mass spectrometry HPV High production volume chemical LC50 Concentration lethal to 50% of the test population LFER Linear free energy relationships LOEL Lowest observable effect level LOAEL Lowest observable adverse effect level LOD Limit of detection LOEL Lowest observable effect level LOQ Limit of quantitation LSC Liquid scintillation counting MIC Minimum inhibitory concentration MEI Most exposed individual NOEL No observable effect level NOAEL No observable adverse effect level 14

PAGE 15

NOEL No observable effect level OPPTS Office of Prevention, Pesticides, and Toxic Substances P&G Procter & Gamble PBT Profiler Persistent, Bioaccumulati ve, Toxic Profiler (computer model) POTW Publicly owned treatment works PTFE Polytetrafluoroethylene QSAR Quantitative structure activity relationship RAD-TLC Radio thin layer chromatography RfD Reference dose TCC Triclocarban TCS Triclosan TNSSS Targeted National Sewage Sludge Survey USEPA United States Environmental Protection Agency w.w. Wet weight equivalent WWTP Wastewater treatment plant 15

PAGE 16

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FATE, TRANSPORT, AND RISK ASSESSMENT OF BIOSOLIDS-BORNE TRICLOCARBAN (TCC) By Elizabeth Allene Hodges Snyder Aug. 2009 Chair: George A. OConnor Major: Soil and Water Science Triclocarban (TCC) is an active ingredient in antibacterial bar soaps, a common constituent of domestic wastewater, and the subject of recent criticism by consumer advocate groups and academic researchers alike. Activated sludge treatment readily removes TCC from the liquid waste stream and concentrates the antimicrobial in the solid fraction, which is often processed to produce biosolids intended for landapplication. Greater than half of the biosolids generated in the US are land-applied, resul ting in a systematic release of biosolids-borne TCC into the terrestrial and, potentially, the a quatic environment. Despite wi despread use of antimicrobial personal care products, the propensity of TCC to partition into biosolids at parts-per-million concentrations, and potential endocrine effects of TCC (Chen et al., 2008), human and ecological health risk assessments for TCC in land-applie d biosolids have not b een conducted. A project funded by USEPA, and in collaboration with the Procter & Gamble Company (P&G), was designed to fill multiple TCC data gaps and to fa cilitate an integrated human/ecological health risk assessment. Data developed under the proj ect include conclusive solubility, partitioning, biosolids-borne TCC concentratio ns, environmental transport, pe rsistence, and soil organism impacts. The data were used to characteriz e human/ecological exposure hazards and estimate potential adverse effects associated with TCC in land-applied biosolids. No biosolids-borne TCC 16

PAGE 17

17 exposure pathways resulted in estimates of adve rse human health effects, even under worst-case land-application scenarios. Two exposure pathways of concern (i.e. biosolids soil soil organism soil predator, and biosolids soil surface water aquatic organism) were identified in the resulting integrated risk assessment, a nd preliminary TCC pollution limits protective of the most sensitive test species were calculated. The preliminary TCC pollutant limits could have important implications for current biosolids land-a pplication practices, but need to be reassessed in light of available measured TCC concentra tions in environmental matrices and additional environmental effects data before it is appropr iate to suggest modifications to current landapplication regulations.

PAGE 18

CHAPTER 1 INTRODUCTION AND PROJECT OBJECTIVES Introduction Triclocarban (TCC; Figure 1-1) is an active ingredient in antibacterial bar soaps, a USEPA High Production Volume (HPV) chemi cal (i.e. 227-454 Mg consumed y-1), and the subject of recent criticism by consumer advocate groups an d academic researchers alike. Criticisms primarily center on uncertainties regarding TCC e nvironmental concentrations and fate (Heidler et al., 2006), persistence (Ying et al., 2007; Miller et al., 2008), toxicity (Heidler et al., 2006), bioaccumulation potential (Coogan et al., 2007; Darbre, 2006; Daughton and Ternes, 1999), endocrine effects (Chen et al ., 2008), and potential for antiba cterial resistan ce development (Walsh et al., 2003; Suller and Russell, 1999). Following typical use in personal care products, TCC is washed down the drain and becomes a co mmon constituent of domestic wastewater at concentrations of 0.4-50 ug L-1 (TCC Consortium, 2002a; Halden and Paull, 2004; Heidler et al., 2006). Activated sludge wastewater treatment [~75% of wastewater treatment plants (WWTPs) in the US; USEPA, 1989] readily removes TCC fr om the liquid waste stream (88-97% removal; TCC Consortium, 2002a) and concentrates the an timicrobial in the solid fraction (e.g. 76+ 30% sorbed; Heidler et al., 2006). The sludge accumu lated within WWTPs is often processed to produce biosolids intended for landapplication. About half of the biosolids generated in the US are land-applied (NRC, 2002), resulting in a systematic release of biosolids-borne TCC into the terrestrial environment. Despite widespread use of antimicrobial pe rsonal care products, the propensity of TCC to partition into sludge during wastewater treatment, and potential adverse health effects of TCC, human and ecological health risk assessments for TCC in land-applied biosolids have not been conducted. 18

PAGE 19

Assessing the risk of biosolids-borne TCC re quires information on the toxicity of the compound, as well as data on the degree of exposu re. The presence of TCC in the environment is not a hazard unless humans or other organism s are exposed to concentrations sufficient to induce an adverse effect. The environmental fa te of biosolids-borne TCC is particularly important to identifying the relevant exposed populations. Following en vironmental release, biosolids-borne TCC could become an immobile sorbed component of the receiving soil, be taken up by biota, leach through the soil profile, chemically or biologically degrade, and/or become relocated via a host of natural or anth ropogenic processes. Multiple components of biosolids-borne TCC human and ecological ri sk assessments are incomplete, including conclusive solubility, partitioning, biosolids-borne TCC concentra tions, environmental transport, persistence, and soil organism impact data. Beha vior and effects of biosolids-borne TCC in the environment have been hypothesized based on esti mated chemical properties, but little is known about the fate of TCC in biosolids-amended soils. Also missing are characterizations of important exposure routes associated with bi osolids-amended soils and identification of potentially susceptible human and ecological populations. The absence of critical environmental data precludes effective risk assessments. The United States Environmental Protection Agency (USEPA) collected screening-level health and environmental effects data (TCC Consortium, 2002a and 2002b) voluntarily submitted by TCC sponsor companies under the High Production Volume (HPV) Challenge Program, and performed a screening-level TCC hazard characterization (USEPA, 2008). The hazard characterization will eventually be coup led with 2007 Inventory Update Reporting (IUR) exposure potential data for a sc reening-level risk characteriza tion (USEPA, 2008). Considerable 19

PAGE 20

effort was expended in compiling the HPV repor ts and multiple researchers are now studying TCC, but no one has characterized the risk specifi c to land-application of biosolids-borne TCC. The ultimate objective of the present study wa s to perform an integrated human and ecological health risk assessment of biosolid s-borne TCC that could support one of two hypotheses: 1) the risk associated with the environmental presence of TCC in land-applied biosolids is negligible or 2) the risk is sufficient to warrant further regulatory attention and research to reduce environmental contamination a nd potential adverse effects. The intermediate objectives were designed to provide data required to fulfill the ultimate objective, and included confirming physicochemical properties of TCC, and characterizing biosolids-borne TCC concentrations, partitioning behavi ors, degradation, leachability, toxi city to terrestrial organisms, and bioaccumulation potential. Intermediate Objectives Intermediate Objective 1: Confir m Physicochemical Properties of TCC Basic physicochemical properties, such as water solubility an d the octanol-water partitioning coefficient (Kow) (Equation 1-1), are key risk a ssessment input parameters. The characteristics greatly affect environmental concentrations, exposure pathways, susceptible populations, and the potential for subsequent huma n/ecological health outcomes. Triclocarban solubility data are scarce, and are limited to inadequately described methods (0.11 mg L-1, Roman et al., 1957; 0.11 or 11 mg L-1, TCC Consortium, 2002b) or are estimated using quantitative structure activity relatio nship (QSAR) analyses (0.65-1.55 mg L-1; Halden and Paull, 2005). The same is true for TCC log Kow data (4.2-6.0; TCC Consortium, 2002a; Halden and Paull, 2005; Heidler et al., 2006). Kow = Co / Cw (1-1) 20

PAGE 21

where: Co = concentration in octanol (mg L-1) Cw = concentration in water (mg L-1) Given the importance of basic physicochemical properties to accurate predictions of organic contaminant fate, transport, and effects, measured values determined using appropriate standardized methods are pref erred. The USEPA Office of Prevention, Pesticides, and Toxic Substances (OPPTS) Harmonized Test Guidelines comprise a set of standardized methods particularly useful to the present study. The OPPTS guidelines are the result of blended test guidance from the USEPA Office of Pollution Prevention and Toxics (OPPT), the USEPA Office of Pesticide Programs (OPP), and the Organization for Economic Cooperation and Development (OECD). The relevant OPPTS gui delines were used to determine TCC water solubility and the octanol-wate r partitioning coefficient, and the detailed methods, results, and implications are presented in Chapter 2. Intermediate Objective 2: Characterize TC C Concentrations in Multiple Biosolids and Partitioning of Indigenous TCC in the Biosolids Matrix No published data pertaining to the concentr ations of TCC in land-applied biosolids were available at the project initiation in 2005. Thus a reconnaissance study of end products from 23 WWTPs utilizing a variety of biosolids proces sing methods was conducted (Chapter 3). The data helped identify WWTP treatment methods that might result in the greatest and least biosolids-borne TCC concentra tions, and facilitated estimates of environmentally-relevant biosolids-borne TCC concentrations for subseque nt experimentation (Cha pters 6, 7, and 8) and risk assessment (Chapter 9). The USEPA Targeted National Sewage Sl udge Survey, initiated in 2006, also characterized TCC (and other contaminant) c oncentrations in WWTP end products (USEPA, 2009a and 2009b). Materials from >70 WWTPs were analyzed duri ng one of the nations most 21

PAGE 22

extensive sewage sludge/biosolids assessments, but the data were not stratified to allow correlations between treatment methods and resu lting TCC concentrations. Nor did the TNSSS report distinguish between sewage sludge and processed biosolids intended for land-application. Nevertheless, the TNSSS data were useful for assessing the representative ness of the biosolids analyzed herein and for characterizing the range of potential environmental concentrations in the biosolids-borne TCC risk assessment calculations. Equally important as biosolids-borne TCC c oncentrations is the partitioning of TCC between the solid and aqueous phases of biosolids. The Kow is a good first estimate of organic compound solid-phase partitioning and predicto r of bioaccumulation potential, but the Kd (Equation 1-2) and the Koc (Equation 1-3) describe partitioning in a specific environmentally relevant matrix and facilitate improved estimates of environmental mobility and bioavailability. Partitioning of biosolids-borne TCC was measured in select biosolids, and the data were used to support discussions of measured TCC leachability (Chapter 4), biotic uptake (Chapters 5 and 6), toxicity (Chapters 6 and 7), and biodegradation (Chapter 8). Kd = Cs / Cpw (1-2) where: Cs = concentration sorbed to soil or biosolids (mg kg-1) Cpw = concentration in pore water (mg L-1) Koc = Kd / foc (1-3) where: foc = fraction organic carbon 22

PAGE 23

Intermediate Objective 3: Characterize the Leachability of Biosolids-Borne TCC in Amended Soils The potential exposure pathways associated with biosolids-borne TCC in amended soil require an understanding of how the chemical moves through, or over, the soil profile. If TCC remains in the surface horizons following la nd-application, the most important exposure pathways might include inhalation of air-borne pa rticulates, ingestion of contaminated soil, and ingestion of exposed terrestrial bi ota. Alternatively, if TCC migrat es vertically with infiltrating water or a fluctuating water tabl e, the most relevant exposure pa thways might include chemical ingestion in contaminated groundwater, ingestio n/absorption in subsequently contaminated surface water, or ingestion of e xposed aquatic biota and their predators. The downward mobility of TCC in a sandy soil amended with a variety of biosolids was assessed in leachates generated by periodic water applications over 5.5 months (Chapter 4). The study was originally designed for assessment of biosolids-borne phosphorous environmental transport, and was not based on a specific OPPTS guideline. Intermediate Objective 4: Characterize Biodegradation of Biosolids-Borne TCC in Amended Soils Biodegradation is one of the most important parameters affecting the toxicity, persistence, and fate of soil contaminants. Biodegradation can be influenced by such factors as temperature, microbial populations, degree of acclimation, accessibility of nutrients cellular transport properties, and chemical partitio ning (Singh and Ward, 2004). The rate of biosolids-borne TCC degradation in amended soil logically influences the extent and duration of exposure by human and ecosystem populations. Rapid degradation is expected to reduce the potential for transport of the parent compound between environmental compartments, uptake by biota, and adverse health effects. Conversely, persistence could promote TCC migration th rough soils to ground or surface waters (either as a dissolv ed constituent or bound to suspe nded particulates), entry into 23

PAGE 24

the food chain, and accumulation to critical toxicity levels. Persistence, in turn, can be affected by the interrelated characteristics of compound so lubility, degree of so lid-phase sorption, and bioavailability. No data quan tifying the persistence of TCC in land-applied biosolids were available in the published literat ure as of 2005, making it impossible to verify mathematical predictions or models of biosolids-borne T CC fate in the terrestrial environment. Data on the persistenc e of biosolids-borne 14C-labeled TCC (and potential metabolites) were collected during a 7.5-month study based on the OPPTS Guideline 835.3300 Soil Biodegradation (Chapter 8). Radiolabeled T CC was spiked to biosolids and subsequently amended to soil. During the study, 14CO2 evolution from the incubated amended-soil samples was monitored as a measure of parent com pound mineralization. Soil samples were also periodically destructivel y sampled and subjected to sequentia l extractions. Methanol extracts were examined with radio-thin-layer-c hromatography (RAD-TLC) for speciation of 14C moieties. Intermediate Objective 5: Characterize Biosolids-Borne TCC Toxi city to Terrestrial Organisms The majority of TCC toxicity data pertai n to aquatic indicator species, and great uncertainty is introduced by ex trapolating the information to larger and/or non-aquatic organisms. Triclocarban toxicity data for terr estrial organisms are primarily limited to direct human and animal dermal exposures, and anim al ingestion exposures (TCC Consortium, 2002a; European Commission, 2004). No data are availa ble for the toxicity of TCC to terrestrial organisms (plant, animal, or microbial) in biosol ids-amended systems. Estimating the ecological risks of biosolids-borne TCC re quires an understanding of TCC toxicity to the organisms most immediately exposed. Reducing risks to non-human organisms will often also reduce the risk to human health. Two non-human organism groups an ticipated to be the mo st highly exposed are 24

PAGE 25

soil microbes and earthworms, i.e. terrestrial indi cator organisms. The two ecological receptors are expected to be at greater risk than humans due to unique modes of exposure (e.g. constant, direct contact with soil, consumption of soil an d pore water), inherently greater sensitivity to environmental contaminants (e.g. increased body burdens, permeable membranes), and susceptibility to minute changes in the soil enviro nment (e.g. alterations in microbial community structure, pH, organic matter degradation). The toxicity of biosolids-borne TCC to earthworms and soil microbial communities was characterized using the OPPTS Guidelines 850.6200 Earthworm Subchronic Toxicity Test (Chapter 6) and 850.5100 Soil Microbial Community Toxicity Test (Chapt er 7), respectively (USEPA, 1996a; USEPA, 1996b). Test organism s were incubated in soils amended with biosolids containing varying concentrations of TCC, and the e ffects on earthworm mortality, or soil microbial processes, were measured over approximately 30 d. Intermediate Objective 6: Characteri ze Biosolids-Borne TCC Bioaccumulation Bioaccumulation is the general te rm for the net accumulation of a substance in an organism due to uptake from all environmental media, irre spective of the route of exposure (Suter, 2007). A bioaccumulation factor (BAF) is calculat ed by dividing the bioaccumulated substance concentration in the organism by the concentratio n in the relevant environmental medium, which can include food, water, dust, and vapors. Bioconcentration is a specific form of bioaccumulation that addresses uptake only th rough non-dietary exposures via the aqueous solution (including the so il solution), and is often determined experimentally (Suter, 2007). Bioconcentration factors (BCFs) are calculated in a similar fashion as BAFs, but only the fraction of the concentration in the organism attri butable to aqueous exposure is considered. The difficulty associated with distinguishing animal uptake via the non-dietary aqueous route from all other routes typically makes cal culation of BAFs more practical than BCFs, particularly for 25

PAGE 26

terrestrial organisms. In the cas e of plant accumulation, the term BCF is often used to describe combined uptake from the aqueous phase, so rption, particle-phase deposition and vapors, although the term BAF is more appropr iate (and is used herein). No data characterizing the bioaccumulation of biosolids-borne TCC by plants or terrestrial fauna are available, but data describing concentrations in aq uatic organisms are relatively abundant. Triclocarban bioconcentrati on factors in algae (predominantly Cladophora spp.) collected adjacent to, and dow nstream of, a WWTP outfall si te reached 1600-2700 (Coogan et al., 2008), and were similar to mathematical estim ates for fish (1186; Ying et al., 2007). The substantial bioconcentration of T CC in aquatic organisms suggests bioaccumulation in terrestrial organisms might also be an important envir onmental transport mechanism for land-applied biosolids-borne TCC. Bioaccumulation in plants was assessed by analysis of Bahia grass tissue ( Paspalum notatum ) harvested from soil columns amended with a variety of biosolids of known indigenous TCC contents (Chapter 5). Bahia grass is used extensively in the Southern states as a crop rotation and general forage pasture grass. Soil dwelling organisms low on the terrestrial food chain, Eisenia fetida worms were also grown in biosolidsamended soil and analyzed for TCC accumulation (Chapter 6). Ultimate Objective: Conduct an Integrated Human and Ecological Risk Assessment of Biosolids-Borne TCC Approximately 30% of bar soaps contain antim icrobial agents, and represent an annual expenditure of around $240 million, or 15% of the so ap, bath, and shower product market total in the US. Eighty-four percent of antibacterial bar soaps contain TCC (Perencevich et al., 2001). Market research shows that antimicrobial agen t additives in soaps promote sales increases (Mintel International Group, 2004), despite growing concerns by some (Daughton, 2004) that the 26

PAGE 27

compounds may contribute to antimicrobial resistance development and environmental degradation, and despite scientif ic studies demonstrating the ineffectiveness of TCC in bar soaps to reduce the incidence of common diseases (Lu by et al., 2005). Nevert heless, the addition of TCC to soaps is a popular and lucrative business, and will continue to cont ribute the bulk of the antimicrobial in WWTPs and biosolids destined for land-application. A biosolids-borne TCC integrated human a nd ecological health risk assessment was conducted utilizing data and conclusions resul ting from Intermediate Objectives 1-6. The assessment sought to identify human/ecologi cal exposure hazards and to estimate TCCassociated environmental impacts resulting from current biosolids land-application practices. Characterized hazards and the integrated risk assessment were used to identify the most susceptible populations (Chapter 9), and to suggest additiona l research to guide future environmental regulation of TCC for the protection of human and environmental health (Chapter 10). 27

PAGE 28

28 p-chloroaniline ring 3,4-dichloroaniline ring Figure 1-1. Chemical structure of tric locarban (TCC; N-(4-chlorophenyl)-N`-(3,4dichlorophenyl) urea)

PAGE 29

CHAPTER 2 WATER SOLUBILITY AND OCTANOL-WATER PARTITIONING OF TCC Introduction Water solubility and Kow are important factors for character izing the fate and transport of environmental contaminants. The two factors critically influence the extent to which a compound leaches through the soil profile or moves laterally with surface runoff, and the amount of compound available for microbial degradation, plant uptake, and bioaccumulation. Solubility and Kow tend to be inversely related and are commonly used to estimate soil/sediment adsorption coefficients and the bioaccumulation potential of a given chemical (Lyman et al., 1982). Compounds that preferentia lly partition from the aqueous phase to the octanol phase (i.e. log Kow > 0) are considered lipophilic (Guy and Hadgraft, 2002), and are expected to also partition preferentially to soil organic matter or the fa tty tissue of living organisms (Calabrese, 1992). When an organism is unable to rapidly metabolize and/or eliminate a compound with a high Kow (i.e. high lipophilicity), the chemical tends to bioaccumulate. However, the definition of a high log Kow in the literature often varies from 4.5-6 (Boese et al., 1999; Kah and Brown, 2008), and bioaccumulation is affected by addi tional factors, including environmental concentrations of the contaminant, duration of exposure, contaminant bioavailability, and growth stage of the exposed organism. Contaminant bi odegradation is also influenced by chemical solubility and partitioning behavior. Hydrophobi c compounds are often strongly associated with other hydrophobic components of th e soil, resulting in limited bioavailability and inhibited biodegradation rates (S emple et al., 2003). Water solubility and/or the Kow can be estimated with various computer models (e.g. ECOSAR, KOWWIN, WATERNT, and WSKOWWIN) (USEPA, 2009c ), but measured values 29

PAGE 30

are preferred. Thus, the water solubility and Kow of TCC were measured according to OPPTS Guidelines (USEPA, 1996c; USEPA, 1996d). Materials and Methods Chemicals Solvents of HPLC-grade or greater were purch ased from Alderich or Fisher Scientific. Water was from a Milli-Q System (Milipore; Milford, MA USA). Analytical-grade n-octanol was purchased from Alderich. Ammonium acetate was purchased from JT Baker. Triclocarban (CAS No. 101-20-2) was obtained from United St ates Pharmacopeia ( 99.9% purity) and the Procter & Gamble Company (98+% purity). Deuterated Triclocarban (TCC-d7; internal standard for HPLC-MS analyses) was also supplied by the Procter & Gamble Company. Water Solubility Determination Triclocarban water solubility was determin ed according to the EP A Product Properties Test Guideline OPPTS 830.7840 Water Solubility ; Column Elution Method (USEPA, 1996c). The column elution method is prescribed for co mpounds with expected water solubilities below 10 mg L-1. An excess of TCC was dissolved in acetone, poured onto glass beads (100/120 mesh), and dried by Rotavap. Two grams of the TCC-coated beads were loaded into a glass column (185 mm x 10 mm) plugged with 10 mm of glass wool. The glass beads occupied approximately 1.5 cm3 (one bed-volume). The loaded column was filled with eight bed-volumes of Milli-Q water, inserted into a water jacket (25C), and allowed to equilibrate for two h. Following equilibration, the column was connect ed to a circulating pump via HPLC tubing (Figure 2-1) and five bed-volumes of water were diverted to re move potential impurities. The divert valve was then switched off and water was allowed to circulate through the column at a rate of 10 bed-volumes h-1. Approximately 500 uL was diverted and collected at 1-h intervals, spiked with TCC-d7 internal standard, and analyzed by HPLC/MS. The OPPTS Guideline 30

PAGE 31

defines equilibrium as five successive samples whose concentrations do not differ by more than 30% in random fashion. In the second of tw o runs, water flow was allowed to continue overnight and a final sample was taken after a 14-h interval to confirm saturation. Octanol-Water Partition Coefficient (Kow) Determination The TCC Kow was determined according to the EPA Product Properties Test Guideline OPPTS 830.7550 Partition Coefficient (n-octano l/water); Shake Flask Method (USEPA, 1996d). A standard solution was created by dissolving 6.7 5 mg TCC in 135 mL of n-octanol saturated with Milli-Q water. The following proportions of standard solution and M illi-Q water (saturated with n-octanol) were combined, in duplicate, in 35 mL glass centrifuge tubes: 15 mL standard solution/15 mL water (samples A and B), 7.5 mL standard solution/15 mL water (samples C and D), and 30 mL standard solution/15mL water (sam ples E and F). The centrifuge tubes were placed on a wrist-action shaker for five minutes a nd then centrifuged at 800 x g to separate the phases. One mL aliquots of the aqueous phase and the diluted n-octanol phase were prepared with TCC-d7 internal standard and analyzed by HPLC/MS. HPLC/MS Analyses Analytical methodology was developed according to TCC research previously conducted at Procter & Gamble and guidance provided in Halden and Paull (2005) and Chu and Metcalfe (2007). Liquid chromatography was performed on a Phenomenex Luna C18 column (3 or 5 m particle size, 2 x 100mm; Phenomenex, Inc., To rrance, CA). All HPLC/MS analyses were carried out at the Procter & Gamble laborat ories on a Waters Alliance 2795 HPLC system (Waters Corporation, Milford, MA) coupled to a Waters mass spectrophotometer controlled by MassLynx 4.0 software. Analyte separation wa s achieved using an elution gradient of water:MeOH with 1 mM ammonium acetate: 25 :75 water:MeOH (held 1 minute), increasing to 0:100 water:MeOH (over 5 minutes, and held 3 minutes), and decreasing back to 25:75 31

PAGE 32

water:MeOH (over 0.5 minutes). Mass spectromet ry was performed in negative electrospray ionization (ESI) mode with sel ect ion recording (SIR). Prec ursor and product ions monitored were m/z 313, 315 (TCC), 320 (TCC-d7) and m/z 160 (TCC), 163 (TCC-d7), respectively. A linear calibration curve us ing the response ratios of TCC to the TCC-d7 internal standard, and comprised of > 7 standard levels spanning expected samp le analyte concentrations, was used for quantification. Results and Discussion Triclocarban solubility, as measured in duplicate determinations, was 0.044 + 0.001 mg L-1 and 0.046 + 0.004 mg L-1, averaging 0.045 mg L-1 (Table 2-1). The average measured solubility, 0.045 mg L-1, is less than the most commonl y reported measured (0.11 mg L-1, Roman et al., 1957) and estimated (0.65-1.55 mg L-1, Halden and Paull, 2005) values. In light of a measured TCC water solubility less than previously published values, the measured TCC log Kow was expected to be equal to, or greater than, the commonly reported log Kow values (4.2-6.0; TCC Consortium, 2002a; Halden and Paull, 2005). However, the measured log Kow of TCC was 3.5 + 0.06 (Table 2-2). Mass balance of TCC in the octanol and water fractions was 101 + 1.8%, indicating minimal loss of TCC during the partitioning test. The unexpected relationship betw een TCC solubility and log Kow values highlights the importance of measured data collected by sta ndardized methodology. Physicochemical property estimates are acceptable for first approximations (i .e. non-validated predictions), but an accurate quantitative risk assessment requires definitive m easurements (preferably utilizing standardized laboratory methods). Chemical property estimation tools such as linear free-energy relationships (LFERs) and EPI Suite (i.e. a co llection of chemical property and environmental fate estimation computer programs developed by the USEPA Office of Pollution Prevention Toxics and Syracuse Research Corporation) (USEPA, 2009c) ar e often utilized withou t considering whether 32

PAGE 33

the training dataset is appropriate for the com pound of interest, and prediction output requires substantiation. Free-energy relationships equations usually assume a nega tive linear relationship between two chemical parameters (e.g. solubility and Kow), and are derived using a group of compounds for which the two parameters are known. Multiple Kow/solubility relationships are available in the literature (Lyman et al., 1982), but few are appropriate for assessing TCC solubility or Kow. An ideal LFER is one developed using multiple compounds similar to the compound of interest and ch aracterized by a large R2 value (i.e. the regression coefficient), indicating good parameter predictability for th e compounds on which the LFER is based. In some publications, the data on which a LF ER is based are only described by the r2 value (i.e. the coefficient of determination), which simply re lates the strength of association between the independent and dependent variables. The LFER developed by Hansch et al. (1968) incorporates 156 mixed-class compounds with an R2 value of 0.874 (Equation 2-1). log(1/S) = (1.339*log Kow) 0.978 + 0.0095(tm 25) (2-1) where: S is measured in mol L-1 tm = melting point = 250C Using the measured log Kow value (3.5), Equation 2-1 estima tes TCC solubility as 0.45 mg L-1, ten-fold greater than the value measured here in. Using the measured TCC solubility value (0.045 mg L-1), Equation 2-1 overestimates the TCC log Kow as 4.2. Various estimation programs within EPI Suite (i.e. ECOSAR, KOWWIN, WATERNT, and WSKOWWIN) (USEPA, 2009c) can also err oneously predict TCC solubility and log Kow. When none of the three physicochemical paramete rs utilized in ECOSAR (i.e. solubility, Kow, and melting point) are manually entered, the predicted TCC solubility and log Kow values are 33

PAGE 34

1.55 mg L-1 and 4.9, respectively (Halden and Paull, 2005). Solubility is estimated in ECOSAR using LFERs developed by Meylan and Howard (1994a, 1994b, 1996) (Equations 2-2 and 2-3), and the Kow is copied from output of KOWWIN. The KOWWIN program estimates Kow using the Atom/Fragment Contribution (AFC) method, by which the structure of a compound is divided into fragments that are each given coe fficient values derived by multiple regression of >2400 measured log Kow values. The fragment coefficients are summed to estimate the log Kow. Regardless of the solubility and melting point entered into ECOSAR, the log Kow remains 4.9. log(S) = -0.312 1.02(log Kow) (2-2) where: S is measured in mol L-1 n = 1450 mixed-class compounds r2 = 0.786 log(S) = 0.2236 1.009(log Kow) 0.00956(tm 25) (2-3) where: S is measured in mol L-1 tm = melting point = 250C n = 1450 mixed-class compounds r2 not available The ECOSAR program does not allo w back-calculation of the log Kow using the LFERs employed in estimating solubility, but the calc ulations can be performed independently. Equations 2-2 and 2-3 predict TCC log Kow values of 6.4 and 4.9, respectively (when using 0.045 mg L-1 solubility) When the measured TCC log Kow is entered into ECOSAR as a known parameter (and solubility and the melting point are not entered) the model over-predicts TCC solubility as 41.41 34

PAGE 35

mg L-1. However, the predicted TCC solubility (0.04242 mg L-1) is similar to the measured solubility when only the measured melting poi nt (250 C; TCC Consortium, 2002a) is manually entered. A host of other solubility values are pr edicted when various combinations of measured and estimated parameters are entered into ECOSAR, and range from 0.20 to 5.1 mg L-1. The WATERNT program in EPI Suite estimates solubility by utilizing a fragment constant method similar to that used in KOWW IN. Unlike most predictions by ECOSAR that overestimate TCC solubility WATERNT undere stimates TCC solubility as 0.024 mg L-1. Unlike WATERNT, which does not inte grate the melting point and log Kow into solubility estimates, WSKOWWIN allows manual entry of the physicochemical parameters. Like ECOSAR, WSKOWWIN utilizes LFERS (Equations 2-4 and 2-5) (Meylan and Howard, 1994a; 1994b; 1996) modified from those used by ECOSAR, and the log Kow predicted by KOWWIN. The WSKOWWIN program predicts TCC solubility as 0.59 and 0.23 mg L-1 with and without manual entry of the measured log Kow, respectively, and a melting point of 250C. Predicted TCC solubility values range from 0.12 to 10.2 when a variety of additional measured and estimated parameter combinations are entered into WSKOWWIN. Backcalculation of the log Kow using the measured TCC solubility in Equations 2-4 and 2-5 result in predicted values of 6.3 and 4.7 (tm = 250C), respectively. log(S) = 0.796 0.854(log Kow) 0.00728(MW) (2-4) where: S is measured in mol L-1 MW = molecular weight = 313 g mol-1 n = 1450 mixed-class compounds r2 = 0.934 35

PAGE 36

log(S) = 0.693 0.96(log Kow) 0.0092(tm 25) 0.00314(MW) (2-5) where: S is measured in mol L-1 tm = melting point = 250C MW = molecular weight = 313 g mol-1 n = 1450 mixed-class compounds r2 = 0.970 The differences in predicted TCC solubilities and log Kow values as a function of estimation program and use of measured/estimated values highlight the importance of confirmed physicochemical parameters. Chemical solubility and partitioning behavior are key input parameters in most organic contaminant environmen tal fate and transport models used to screen chemicals for human and environmental health ri sks. Use of inaccurate physicochemical data could contribute to erroneous predictions of fa te, toxicity, and risk, or direct focus to inappropriate environmental compartments. Measured values of TCC solubility and Kow are expected to yield more accurate predictions of chemical fate. However, discrepancies can also occur be tween measured physicochemical parameters when inconsistent or inappropriate determina tion methodologies are applied. For example, the solubility of hydrophobic compounds (i.e. <10-2 g L-1; USEPA, 1996c) should not be measured using a shake-flask method (as opposed to the co lumn elution method herein, or the generator column method), as relativ e errors in weighing solute mass prio r to dissolution can be large. Similarly, the log Kow can be inaccurately measured if th e water and octanol are not saturated with each other prior to addition of the compound of interest, as emulsions that exaggerate solubility can develop. 36

PAGE 37

The USEPA Office of Prevention, Pesticides, and Toxic Substances (OPPTS) Harmonized Test Guidelines provide guidan ce for appropriate and consistent determinations of chemical properties, and were used herein to contribute to an accurate and uniform TCC database. The OPPTS Guidelines are the result of blending te st guidance from the USEPA Office of Pollution Prevention and Toxics (OPPT), the USEPA Of fice of Pesticide Programs (OPP), and the Organization for Economic Cooperation and Deve lopment (OECD). Harmonization of multiple guidelines reduces inconsistencies in testing methodology and helps standardize data required under the Toxic Substances Control Act (TSCA) ( 15 U.S.C. 2601) and the Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) (7 U.S. C. 136 et seq.). The Product Properties OPPTS Guidelines simplify the process of meas uring organic contaminant physicochemical characteristics, and determinations can be quick and inexpensive. Triclocarban solubility (0.045 mg L-1) and log Kow (3.5) measured according to OPPTS Guidelines were both less than previously predic ted values, and were used to support accurately informed discussions of biosolids-borne TCC environmental transport (Chapter 4), bioaccumulation (Chapters 5 and 6), toxicity (Cha pters 6 and 7), and persistence (Chapter 8). 37

PAGE 38

Table 2-1. Commonly reporte d physicochemical propertie s of triclocarban (TCC) Property Value Measured or estimated Source Boiling point (C) 434.57 Estimated by EPI Suitea Ying et al., 2007 Melting point (C) 182.04 255.3 Estimated by EPI Suitea Measured Ying et al., 2007 TCC Consortium, 2002a Vapor pressure (mm Hg at 25C) 3.61 x 10-9 Estimated by EPI Suitea Ying et al., 2007 Water solubility (mg L-1 at C) 0.6479 0.65 0.65-1.55 0.11 11 Estimated by EPI Suitea Estimated using Solaris V4.67b Estimated using PBT Profilerc and ECOSARd Measured Misreported Ying et al., 2007 Sapkota et al., 2007 Halden and Paul, 2005 Roman et al., 1957 TCC Consortium, 2002a Log Kow 4.9 4.2-6.0 5.74 Estimated by EPI Suitea Various methods Estimated using Solaris V4.67b Ying et al., 2007 TCC Consortium, 2002a Sapkota et al., 2007 Log Koc 3.732 Estimated by EPI Suitea Ying et al., 2007 pKa 12.77+ 0.7 Estimated by ECOSARd Sapkota et al., 2007 a: Estimation Programs Interface Suite (computer model); b: computer model; c: Persistent, Bioaccumulative, and Toxic Profiler (computer model); d: Ecological Structural Activity Relationships (computer model) Table 2-2. Triclocarban (TCC) solubility measurements Sampling interval Run 1 (mg TCC L-1) Run 2 (mg TCC L-1) 1 0.038 0.039 2 0.041 0.049 3 0.042 0.049 4 0.045 0.053 5 0.044 0.045 6 0.043 0.043 7 0.045 0.045 8 0.045 0.044 Mean of last 5 intervals 0.044+ 0.001 0.046+ 0.004 Mean of duplicate runs 0.045 38

PAGE 39

Table 2-3. Triclocarban (TCC) Kow measurements n-octanol/water ratio Log Kow 1:1 3.6 1:1 3.5 1:2 3.4 1:2 3.6 2:1 3.4 2:1 3.5 Average 3.5 + 0.06 TCC-coated glass beads Glass wool Figure 2-1. Triclocarban (TCC) solubility determination: column elution diagram 39

PAGE 40

CHAPTER 3 BIOSOLIDS-BORNE TCC CONC ENTRATIONS AND PARTITIONING Introduction Heidler et al. (2006) published the first meas ured concentration of TCC in processed biosolids approved for land application, 51 mg kg-1, almost 50 y after the compound was first added to bar soaps. The biosolids analyzed by Heidler et al. (2006) was collected from a single activated sludge WWTP, a nd since publication, the c oncentration of 51 mg kg-1 has been used in the peer-reviewed literature (Chu and Mecalfe, 2007, Ying et al., 2007) and the popular media to describe expected TCC concentrations in bi osolids across the US. A single concentration measurement, however, is insufficient to base estimates of TCC loading in biosolids-amended soils and the associated risks, and prompted in clusion of TCC in the recent Targeted National Sewage Sludge Survey (USEPA, 2009a and 2009b). The great variabilit y between treatment plants (e.g. population served, operating capacity, tr eatment method, product form, disposal practices, and location) in the US could be reasonably expected to result in a range of biosolidsborne TCC concentrations, and differ from the single value reported by Heidler et al. (2006). Of the factors that might influence biosolids-borne TCC concentrations, WWTP designers and operators have the greates t control over the wastewater treatment method and the product form. Understanding the effects that sludge tr eatment processes and product form have on TCC concentrations in biosolids could contribute to better informed waste ma nagement decisions that reduce environmental contamination and human/ ecological exposure. For example, lime stabilization of biosolids might impact bioso lids-borne TCC concentrat ions or environmental behavior, as high pH levels (11 or 12) could result in ionizatio n and a subsequent increase in solubility (pKa = 12.8; estimated Sapkota et al., 2007), and/ or changes in the hydrolysis half-life predicted at circumneutral pH levels (>1 y; HYDROWIN; USEPA, 2009c). Processes and forms 40

PAGE 41

that significantly reduce TCC con centrations could then be prefer entially selected for biosolids intended for land application. Twenty-three biosolids and one sludge, s upplied via the USEPA TNSSS and the USEPA City Survey of publicly owned treatment wo rks (POTWs), or collected by the author, were analyzed to supply data that support three s ub-objectives: 1. identify wastewater treatment processes that could reduce the risk of potential adverse effect s on soil organisms, vegetation quality, and water quality by reduc ing biosolids-borne TCC concen trations; 2. facilitate the design of environmentally relevant scientific st udies (Chapters 5-8), as biodegradation rates, plant uptake, and toxic effects are likely affected by bios olids-borne TCC concentrations in amended systems; and 3. calculate environmenta l concentrations in estimates of human and ecological health risk s (Chapter 9). In addition, the partitioning of biosolids-bor ne TCC between the solid and liquid phases (i.e. the Kd value; Equation 1-2) and between the or ganic carbon and liquid phases (i.e. the Koc value; Equation 1-3) were characterized for se lect biosolids. The parameters describe partitioning in specific environmentally relevant matrices and facilitate improved estimates of environmental fate, transport, and impacts. As with the measured TCC solubility and log Kow values (Chapter 2), the Kd and Koc data were used to support discussions of measured TCC leachability (Chapter 4), biotic uptake (Chapter s 5 and 6), toxicity (Chapters 6 and 7), and biodegradation (Chapter 8). Characterization of biosolids-borne TCC partit ioning was further divided into assessments of indigenous compound (i.e. TCC that entered wa stewater treatment as a constituent of the influent) and compound spiked to the final biosolids product. Indigenous TCC entering wastewater treatment is pres ent throughout the entire biosolids production process. The 41

PAGE 42

compound ages as an inherent component of the solid fraction from the time TCC enters the waste stream to eventual bios olids land-application. Thus, TCC is expected to become thoroughly incorporated and evenly distributed within the organic co mponent of the biosolids. A spike, alternatively, is added to the exterior of the matrix and allowed to equilibrate after the biosolids are produced. A spike might not sorb as completely or evenly as an indigenous compound within the timefr ame of a typical partitioning experiment ( 24 h), or even in multimonth incubation studies (e.g. degradati on study described in Chapter 8). A Kd or Koc value estimated using a spiked matrix could, thus, potentially underestimate the true solid-phase partitioning of a hydrophobic compound such as T CC. Underestimating partitioning and/or sorption could lead to erroneous predictions of environmental mobility, persistence, and bioavailability. Uncertainties associated with estimating par titioning with spike additions are avoided by measuring the indigenous concentrations of a biosolids-borne compound in the biosolids liquid and solid phases. The partitioning of indigenous TCC was quantified in 16 of the 23 materials previously analyzed for total TCC content, a nd compared to spiked TCC data collected by colleagues in the Soil Chemistry Laboratory in the Soil and Water Science Department at the University of Florida. The comparisons were ma de to assess the validity of generalizing results of lab-based studies using spiked compound to real world scenarios. Materials and Methods Chemicals Solvents of HPLC-grade or greater were purchased from Alderich, JT Baker, or Fisher Scientific. Ammonium acetate wa s purchased from JT Baker. Tr iclocarban (CAS No. 101-20-2) was obtained from United States Pharmacopeia (99.9% purity) and the Procter & Gamble Company (98+% purity). Deuterated TCC (TCC-d7) and 14C-TCC (specific activity: 75 42

PAGE 43

mCi/mmol) were also supplied by the Procter & Gamble Compa ny. All tested wastewater treatment materials were the produc ts of activated sludge treatmen t and represented a variety of biosolids production methods. Bi osolids analyzed for indigenous and spiked TCC partitioning were selected for dry matter content less than or equal to 35%, which permitted collection of the indigenous (equilibrium) liquid phase for analysis following centrifugation. Biosolids Extraction Method Development Heidler et al. (2006) reported 91+ 8% and 93+ 17% recoveries from a single digested sludge that was spiked with 50% and 100% of indige nous TCC content, respectively. The spiked sludge was allowed to equilibrate overnight (exact time not reported) and extracted with acetone by accelerated solvent extraction (A SE). Similar percent recoveries were obtained by extracting with 50:50 MeOH:acetone on a mechanical shaker for 18 h (Heidler et al., 2006). Although subsequent samples were dried prior to analysis, Heidler et al. (2006) did not specify whether the spike recovery experiments were performed with wet or dry material. Prior to adoption in the present study, the method published by Heidler et al. (2006) was assessed with the intent of determining the necessity of drying samples, re ducing co-extraction of other organic compounds (by use of less aggressive solv ents) that could interfere with future LC/MS analysis, and reducing sample preparation time. All preliminary method valida tion work was conducted using 14C-TCC to facilitate rapid sample analys is by liquid scintilla tion counting (LSC). First biosolids-borne TCC extraction method validation Three biosolids representing a range of solids contents (low : 6%, medium: 13%, and high: 72%), were selected for initial extraction method development. Two batches of one gram (dry wt. equivalent) subsamples of each biosolids (n = 4 for each biosolids in each batch) were transferred to 35 mL glass, r ound-bottom centrifuge tubes and wette d to ~94% water content (by weight) to promote even distribution of the 14C-TCC spike (~100,000 dpm or 0.189 ug 14C43

PAGE 44

TCC). Each spiked sample was equilibrated on a rocking platform for 4 h. The Batch 1 samples were assigned to ASE extraction and immediately placed in a freezer (-20C), frozen for 1 h, and lyophilized (3 d). Following lyophi lization, the dried samples were mixed with approximately 1 g diatomaceous earth and loaded into 11 mL ASE vials. Vials were extracted twice at 60C and 1500 psi with 100% MeOH. Methanol, as opposed to acetone, was the solvent selected in an effort to reduce the expected co-extraction of compounds that could interf ere with future LC/MS analyses of cold TCC extracts. Extracts were dried under nitrogen and reconstituted in 5 mL methanol. Duplicate aliquots of each sample were analyzed by LSC and percent recoveries calculated. Batch 2 samples were assigned to an 18 h shake-flask extrac tion with 100% methanol (20 mL) immediately following spike equilibration. Samples were centrifuged at 800 x g and the supernatant was collected. A second extracti on (20 mL MeOH) was performed on a multi-tube vortex for 1 h and samples were again centrif uged. Combined extracts were dried under nitrogen, reconstituted, and analyzed by LSC. The average percent recoveries of the ra diolabeled spike by ASE and shake-flask extraction were 86+ 4% and 95+ 3%, respectively. The ASE method recovered approximately 80% of the radiolabel in the fi rst extraction and required less solv ent (6 mL per extraction) than the shake-flask method (20 mL per extraction). Us e of additional solvent is unlikely to improve percent recovery by ASE, as average recovery in the second extraction was only 5%. Shakeflask extraction recovered approximately 70% of th e radiolabel in the firs t and 20% in the second extraction. The lower ASE percent recovery, as compared to the total shake-flask recovery, was hypothesized to be due to the effectively longer spike equilibration peri od (i.e. lyophilization) 44

PAGE 45

prior to ASE extraction. A second 14C-TCC extraction efficiency e xperiment was performed to assess the effects of moisture and a lengthened equilibration time on spike percent recovery. Second biosolids-borne TCC ex traction method validation Four batches (Batches 3-6) of six biosolids representative of various treatment processes (n = 4 for each biosolids in each batch) were brought to equal moisture contents as described above and spiked with 14C-TCC at a rate of 100,000 dpm g-1 (dry weight equivalent). Samples were vortexed for 4 h, and allowed to sit for an additional 20 h in the dark at room temperature. Batch 3 samples were lyophilized and extracted by ASE (100% MeOH, 100C, 1500 psi) and Batch 4 samples were lyophilized and extracted by shak e-flask (2 x 20 mL 100% MeOH; 18-h shake followed by 1-h vortex). The remaining samples (B atches 5 and 6) remained wet. Batch 5 was refrigerated during the freeze dr ying period of Batches 3 and 4, and subsequently extracted by shake-flask (2 x 20 mL 100% MeOH; 18-h shake followed by 1-h vortex ). Batch 6 was shakeflask extracted (100% MeOH; 18-h shake followe d by 1-h vortex) immediately following the 24h equilibration. Extracts from all batches were dried under nitrogen, reconstituted in a known volume, and analyzed by LSC. Percent recoveries in the second biosolids-borne TCC ex traction method validation study were 69+ 11% (Batch 3; 24 h equilibration, lyophilization, ASE with MeOH), 66+ 7% (Batch 4, 24 h equilibration, lyophilizati on, shake-flask with MeOH), 56+ 8% (Batch 5, 24 h equilibration, 3-d wet storage, shake-flask with MeOH), and 54+ 13% (Batch 6, 24 h equilibration, shake-flask with MeOH). The decreased percent recovery by ASE following a 24 h sp ike equilibration time (Batch 3), as compared to a 4 h equilibration time (Batch 1), indica tes the 3 d lyophilization process alone does not allow complete spike equi libration. The similar spike recoveries in Batches 5 (56+ 8%; 96 h effective equilibration) and 6 (54+ 13%; 24 h equilibration) suggest a 24 h spike equilibration in wet material is adequate, but extraction of wet material is inefficient. 45

PAGE 46

Previous 14C-TCC spike equilibration work with sludge at P&G (Kerr and McAvoy, 1999), and later work conducted with bios olids at the UF Soil Chemis try Lab (Agyin-Birikorang and OConnor, unpublished), also indicate spike equilibration in wet material is reached in 24 h. The spike recoveries in Batches 3 (69+ 11%) and 4 (66+ 7%) were much less than the recoveries reported by Heidler et al. (2006), but reflect the similar extracti on efficiencies of the ASE and shake-flask methods with dry material. The second biosolids-borne TCC extraction method validation confirmed that a 24 h minimum spike incubation period is required to reach equilibrium, and therefore better represents the extractab ility of indigenous compound. The insufficient percent recoveries achieved using MeOH, however, indicated the need for a more aggressive extraction solvent. The ASE and shake-flask methods used by Heidle r et al. (2006) were applied in the third biosolids-borne TCC extr action method validation. Third biosolids-borne TCC ex traction method validation Batches 7 and 8 each consisted of three bios olids representative of various treatment processes (n = 4 for each biosolids in each batch). Each biosolids was spiked and equilibrated (24 h) as described in the second method valid ation. Batch 7 was extracted by ASE with 100% acetone (100C, 1500 psi), and Batch 8 samples were extracted by shake-flask with a 50:50 MeOH:acetone solvent mixture (18-h shake and 1-h vortex). The extracts were collected, dried under nitrogen, reconstituted in a known volume of extr action solvent, and analyzed by LSC. Recoveries were 82+ 4% (Batch 7) and 84+ 8% (Batch 8), reflecting the similar extraction efficiencies of the ASE and shake-flask methods reported by (Heidler et al., 2006). The percent recoveries using the MeOH:acetone mixture were better than recoveries with 100% MeOH, but were approximately 10% less than the recove ries reported by Heidle r et al. (2006). The difference could be due to a s horter equilibration time used by Heidler et al. (2006) (time not 46

PAGE 47

reported), the small mass of 14C-TCC spike (< 3% indigenous TCC concen tration as compared to the 50% and 100% spikes used by Heidler et al.), or the variable nature of biosolids. All subsequent biosolids extracti ons were performed using the shake-flask method with 50:50 MeOH:acetone, as the method was most easily utili zed in future extractions in the UF Soil Chemistry Laboratory. Fourth biosolids-borne TCC extraction method validation The 14C-TCC spike recovery experiments used only trace levels of radiolabeled compound, in contrast to work by Heidler et al. (2006) in which a single bios olids material was spiked with 50% and 100% of indigenous TCC. To assess whether spike mass influe nces recovery, two biosolids from the same treatment plant (anaerobi cally digested centrifuged cake before and after air drying) were spiked w ith 200%, 100%, and 50% of the anticipated indigenous TCC concentration (30, 15, and 7.5 mg TCC kg-1 biosolids, respectively), and fortified with TCC-d7 surrogate standard. Biosolids were equilibrated, ex tracted by shake-flask, and dried as described in the third biosolids-borne T CC method validation. The dried ex tracts were reconstituted in 50:50 MeOH:Milli-Q water, and analyzed by HPLC /MS at the P&G laboratories in Cincinnati, OH. Subsequent analyses of non-spiked biosolid s revealed that the TCC spikes were actually approximately 400% (Batch 9), 200% (Batch 10), and 100% (Batch 11) of the indigenous TCC concentration. The percent spike recoveries were 98+ 9% (Batch 9), 91+ 12% (Batch 10), and 90+ 11% (Batch 11), and similar to the spike recoveries reported by Heidler et al. (2006). The improved spike recoveries, as compared to spike recove ries in the third biosolids-borne TCC method validation study, suggest the use of large TCC spik e masses facilitates aver age percent recoveries > 90%. 47

PAGE 48

The conservative approach, in terms of ra w biosolids-borne TCC concentration data correction and risk assessment, would be to a pply the 84% recovery determined in the third method validation study to future biosolids-borne TCC quantifications. However, if the isotope dilution method is applied prior to sample pr eparation using a mini mal mass of surrogate standard and an adequate equi libration period, no raw data corr ection for percent recovery is required. Extraction and Analysis of Total Biosolids-Borne TCC Biosolids were analyzed in two batches (Tables 3-1 and 3-2). For Ba tch 1 (8 biosolids + 1 unprocessed sludge), one gram (dry wt. equivalent) replicates (n = 4) were lyophilized and extracted (2 x 20 mL 50:50 MeOH:acetone) by shake-flask (18 h) on a platform shaker, followed by sonication (2 h). Extracts were dried unde r nitrogen, reconstituted in 50:50 MeOH:Milli-Q water, fortified with TCC-d7 internal standard, and analy zed by HPLC/MS at the Procter & Gamble laboratories in Cincinnati, OH. Triclocarban concentra tions in Batch 1 materials were corrected for extraction efficiency by applying the average percent recovery (84%) calculated from Part 3 of the biosolids-borne TCC extraction efficiency validation. Total TCC analysis was improved for Batch 2 (15 materials) by applying the isotope dilution technique (Heidler et al ., 2006) to automatically correct for percent recovery. One gram (dry wt. equivalent) replicates (n = 4) of Batc h 2 biosolids were wetted to ~94% water content and spiked with 1 ug TCC-d7 as the surrogate standard. Following equilibration (24 h), the biosolids were processed identically to Batch 1 materials. Extracts were analyzed by HPLC/MS using the same methodology employed in the water solubility and log Kow determinations (Chapter 2). The calculated limits of detection (LOD) a nd quantitation (LOQ) by HPLC/MS were 3 and 10 ng TCC mL extract-1, respectively (i.e. 0.12 and 0.4 mg TCC kg biosolids-1, 48

PAGE 49

respectively). The LOD and LOQ values were calculated as 3x and 10x, respectively, the standard deviation in the signal from multiple runs of the lowest calibration standard. Partitioning Coefficient Determination Indigenous TCC partitioning method Sixteen of the 23 biosolids analyzed for T CC content were selected for indigenous TCC partitioning assessment because the solids content (< 35%) facilitated collec tion of the aqueous fraction. Subsamples of the se lected biosolids (~1.5 grams wet weight) were transferred to microfuge tubes and centrifuged (1 0 minutes at ~2,000 x g) to separate the solid/liquid phases. The supernatant was loaded into HPLC/MS-MS autosampler vials, fortified with TCC-d7 internal standard, and analyzed for TCC. Indigenous TCC partitioning was calculated on both a solids and organic carbon content basis. Concentratio n of sorbed TCC was calculated by subtracting the mass of TCC in the aqueous phase in a give n unit of wet biosolids from the total mass of TCC in the same unit, and dividing by the dry weight. The log Kd was calculated as the log of the quotient: concentration of TCC in the dry so lid phase divided by the concentration in the analyzed supernatant. The Koc for each material was calculated by dividing the Kd value with the fraction of organic carbon, as de termined using the Walkley-Black method (Walkley and Black, 1934) and the assumption that 77% of the tota l organic carbon was oxidized (Nelson and Somers, 1996) (Tables 3-1 and 3-2). Spiked TCC partitioning method Colleagues in the Soil Chemistry Labor atory (Agyin-Birikor ang and OConnor, unpublished) characterized partitioning ( according to OPPTS Guideline 835.1220) of 14C-TCC spiked into the same 16 biosolids assesse d for indigenous TCC distribution, allowing comparisons between Kd and Koc values for spiked and indigenous biosolids-borne TCC. 49

PAGE 50

Following spike addition, samples were equilibra ted on an end-over-end shaker (24 h), and subsequently centrifuged. Aliquots of th e supernatant were analyzed by LSC. Results and Discussion TCC Concentrations in Biosolids The mean TCC concentration for the 23 assessed biosolids was 20+ 11 mg kg-1 (Table 3-3). Of the biosolids tested, the aerobically digested biosolids (GRBC) and the aerobically composted biosolids (DYSK and DYMK) contained the lowest concentrations of TCC (6-8 mg kg-1). The TCC concentrations in biosolids prepar ed under aerobic conditions (6-8 mg kg-1) suggest enhanced TCC biodegradation, as compared to anaerobic conditions (7-43 mg kg-1). However, additional aerobically treated bi osolids should be analyzed before a conclusive correlation between aerobic conditions and enhanced TCC bi odegradation is established. No other TCC concentrations in biosolids identified as products of aerobic treatment processes are available in the current literature. Anaerobically digested material s (n=9) exhibited a wide T CC concentration range (7-43 mg kg-1), but the greatest TCC concentration was 8 mg kg-1 less than the biosolids-borne TCC value (51 mg kg-1) reported by Heidler et al. (2006). The ORBC-BL material (activated sludge prior to lime addition; not approved for la nd-application) contained a moderate TCC concentration (25 mg kg-1), which decreased following lime addition to 18 mg kg-1 (sample ORBC-AL). Differences in TCC concentrations acr oss biosolids could be due to several factors, including: varying influent T CC concentrations, differing diges tion periods and methods, adding of other materials, or other treatment conditions unique to particular WWTPs. For example, the addition of lime, wood chips, or food waste (which are not expected to cont ain TCC) could act to simply dilute TCC concentrations. Conversely, extended anaerobic digestion periods often improve total solids reduction, and could, concentrate TCC in the final biosolids product. 50

PAGE 51

Multiple researchers have quantified TCC concentrations in a variety of biosolids and sludges (Heidler et al., 2006; Chu and Metc alfe, 2007; Halden, 2007; Sapkota et al., 2007; USEPA, 2009a and 2009b) since the start of the present project in 2005. Shortly after Heidler et al. (2006) reported a con centration of 51 mg kg-1 for a single biosolids produced by activated sludge treatment, Halden (2007) presented data for 19 additional activated sludge biosolids. The reported average TCC concentration (~22 mg kg-1) is similar to the value in the 23 biosolids characterized herein (Figure 3-1). Sapkota et al (2007) also reported an average concentration of 19 mg kg-1 in five primary sludges. The lowest av erage biosolids-borne TCC concentration (5 mg kg-1) was reported by Chu and Metcalfe (2007) for four materials collected from WWTPs in Ontario. The low concentrations might be a function of differences in antib acterial soap use in Canada as compared to the United States. In Figure 3-1, the box-plot whiskers indicate the minimum and maximum reported bi osolids/sludge-borne TCC concentrations. The box bottoms and tops represent the 25th and 75th percentiles (assuming a normal distribution), respectively, and the diamonds indicate the mean TCC concentrations. The largest study of biosolids-borne TCC concentrations to date was performed as a component of the Targeted National Sewage Sludge Survey (USEPA, 2009a), in which biosolids/sewage sludge from WWT Ps in 35 states were analyzed for a host of organic and inorganic contaminants. Treatment facilities had to meet multiple criteria for inclusion in the TNSSS, including employment of secondary tr eatment or better, and treatment of > 106 million gallons of wastewater d-1 (MGD). Plants meeting the latter cr iterion are responsible for treating ~94% of wastewater in the United States. The inclusion criteria app lied to 3,337 WWTPs, from which 74 were statistically selected to represen t the distribution of plants receiving 1-10, 10-100, and >100 MGD. Targeted compounds in the TN SSS were selected in reaction to the National 51

PAGE 52

Academy of Sciences National Research Counc ils report Biosolids Applied to the Land: Advancing Standards and Practices (NRC, 200 2) and to EPAs 2003 review of biosolids regulations. The purpose of the survey was to evaluate sewa ge sludge characteristics on a national scale and to identify emerging contamin ants of concern (ECCs), including TCC. Triclocarban was detected in each of th e 84 samples tested in the TNSSS (range: 0.187441 mg kg-1), with a mean and 95th percentile concentration of 39+ 60 mg kg-1 and 131 mg kg-1, respectively (USEPA, 2009a). After removing statistic al outliers (e.g. 441 mg kg-1), removing TCC concentrations measured in materials not labeled as final products, and using the average concentration from duplicate samples, the mean concentration of the remaining 70 TNSSS materials was 32+ 32 mg kg-1. The distribution of TCC concentrations in the 70 TNSSS materials is highly skewed, with the majority of measurements < 30 mg kg-1. Nine of the materials provided to the UF Soil Chemistry Lab we re received via the TNSSS, and were part of split samples, thus allowing direct WWTP-specific comparisons of data analyzed in different labs without concerns about sampling time differences. The differences between TCC concentrations measured in the P&G labor atories and TNSSS contract labs were < 15%. The reasonableness of the range of reported average biosolid s-borne TCC concentrations following activated sludge treatment can be assessed using the total mass of TCC utilized in the US (2.27 x 105 4.74 x 105 kg yr-1; TCC Consortium, 2002a), the mass of biosolids produced in the US (5.4 x 106 dry metric tons yr-1) (NRC, 2002), the percent of activated sludge treatment plants in the US (75%; USEPA, 1989)), and the percent of infl uent TCC sorbed to the solid phase during activated sludge treatment (76%; He idler et al., 2006). If all TCC produced is disposed to municipal wastewater, if 133 Mg of TCC is sorbed to the solid phase, and if 4.2 x 106 dry metric Mg of biosolids are produced by activated sludge treatment each y, the expected 52

PAGE 53

average biosolids-borne TCC concentration is 30-60 mg kg-1. Adjusting the calculation for the percent of TCC not entering municipal wastewat er treatment (e.g. septic systems, sewage leaks/overflows) and potential effects of biosolids preparation processes could alter the estimate. For example, the concentration of TCC in bi osolids produced by trickling filter plants is expected to be less, given the lower aqueous-phase removal efficiency as compared to activated sludge treatment (65-93% and 88-97%, respectively; TCC Consortium, 2002b). Nevertheless, the calculations confirm the reas onableness of the biosolids-borne TCC concentrations measured herein, as well as most of the concentrations reported in the TNSSS report (USEPA, 2009a and 2009b) and by Heidler et al. (2006), Halden (2007), and Sapkota et al. (2007). The strength of the TNSSS report is the large a nd statistically significant sa mple population. However, the materials sampled during the TNSSS project were simply final products of wastewater treatment, and not necessarily processed biosolid s. The USEPA considered whether to stratify the results by disposal type (t hus land-applied materials could be distinguished as biosolids processed to meet Part 503 requi rements), but decided that ther e was no reason to suspect that different disposal practices w ould lead to different treatment mechanisms resulting in significantly different final products (USEPA, 2009b). Therefore, the TCC concentrations reported for final products do no t specify if the material is biosolids or sewage sludge. The validity of the assumption made by USEPA can not be verified by the available data, and the appropriateness of applying the mean TCC c oncentration in final products (32 mg kg-1) to biosolids is drawn into questi on. The relatively low TCC concen trations in the aerobically digested and composted biosolids analyzed here in suggest the processing method is important, despite the USEPA conclusion otherwise. 53

PAGE 54

Alternatively, the mean biosolids concentration of ~20 mg TCC kg-1 reported herein and by Halden (2007) suggest that the more appropriate mean biosolids-borne TCC concentration is approximately 20 mg kg-1 less than the mean calculated from the TNSSS data. The factors confounding use of the 20 mg kg-1 mean, however, include potential WWTP sampling overlap, small sampling sizes, and a similar mean in prim ary sludges analyzed by Sapkota et al. (2007). The 19 mg kg-1 mean in primary sludges lends stre ngth to the USEPA assumption that the method used during treatment of the final pr oduct is inconsequential, with the potential exception of aerobic processing methods (e.g. aer obic digestion, composting, and air-drying) and lime stabilization (Table 3-3). Accordi ngly, the TNSSS data (i.e. mean: 39 mg TCC kg biosolids-1, 95th percentile: 131 mg TCC kg biosolids-1) are first used in subsequent environmental concentration calcul ations herein, and to create e xposure scenarios (Chapter 9). When useful, the TNSSS-based calculations are co mpared to calculations using the data from the biosolids characterized herein. Expected TCC concentrations in biosolids-ame nded agricultural soils can be estimated in the same manner as expected biosolids concentr ations. Assuming typical agronomic biosolids loading rates of 5-10 Mg ha-1 yr-1 (OConnor et al., 2001) incorporated into the top 15 cm of the soil (bulk density = 1.3 g cm-3), and an average biosolids-borne TCC content of 39 mg kg-1 (USEPA, 2009a), the calculated biosolids-amended soil TCC con centration is 0.09-0.18 mg TCC kg amended soil-1, or a ~400x-200x reduction in the biosolids-borne TCC concentration. The estimated soil concentration is low, but of uncer tain significance because data characterizing the toxicity or bioaccumulation potential of TCC for soil-dwelling organisms are scarce (see Chapters 7 and 8). Additional exposure and risk assessment scenarios are explored in Chapter 9. 54

PAGE 55

Partitioning of Indigenous and Spiked Biosolids-Borne TCC The log Kd values for indigenous biosolids-borne TCC were 1.8-3.9 (Table 3-4) and, with the exception of the ORBC-AL and ORBC-BL mate rials, indicated extensive TCC sorption to the solid-phase. The ORBC-AL (Class B biosolids) log Kd of 1.8 is likely the result of lime stabilization. Lime stabilization is used to reduce pathogens, and Class B lime stabilized biosolids must reach a pH of 12 for two h following addition. The pKa of TCC is 12.8 (Table 21) and the pH of ORBC-AL biosolids at the time of partitioni ng analysis was 12.0. Thus, much of TCC in ORBC-AL biosolids likely exists in th e more soluble dissociated form that is less extensively sorbed (smaller log Kd value). The pH of the ORBC-AL material would also likely be adjusted to a value significantly lower than the pKa of TCC prior to la nd application, thus reducing the fraction of di ssociated TCC. The ORBC-BL material (log Kd = 2.9) is undigested sludge (ORBC-AL biosolids before lime addition) and is not approved for land-application. The explanation for the relatively low log Kd in the ORBC-BL material is not known, but might be a function of the sludge treatment process. Th e ORBC material was the only one of the 16 materials assessed that was a pr oduct of the Bardenpho proce ss (a method that utilizes a combination of anaerobic, anoxic, and aerobic tr eatment conditions to process wastewater), which might result in a material with unique prop erties that affect TCC partitioning. Removing both ORBC materials from considera tion, the measured indigenous log Kd values for the remaining biosolids were within the narro wer range 3.1-3.9, with an average of 3.4 + 0.2 (Table 3-4). Previous 14C-TCC partitioning work (16 h equilib ration time) with an unprocessed activated sludge (Kerr and Mc Avoy, 1999) resulted in a log Kd of 4.2. Normalizing indigenous Kd values for organic carbon cont ents of the biosolids did not narrow the range of log Koc values (2.3-4.6), suggesting differe ntial influence of organic carbon on TCC sorption across biosolids an d/or that OC was not the only sorbing solid. After removing 55

PAGE 56

the ORBC materials from consideration and converting to logs, the measured indigenous log Koc values for the remaining biosolids are 3.6-4.6, with an average of 3.9 + 0.2 (Table 3-4). Mean log Kd and log Koc values in 14C-TCC spiked biosolids were 3.3 0.1 and 3.9 0.1, respectively. There were minimal differences between the means of the log Kd values for the indigenous and spiked TCC, and between the mean log Koc values. Results indicate that partitioning of spiked TCC following 24 h of equilibration can be considered to closely represent partitioning of indigenous compound. Similar partitioning studies using soils, and characterizations of TCC sorption/desorption kine tics, were also performed by Agyin-Birikorang and OConnor (unpublished). The re sults and implications of the data to TCC leachability and biodegradation are discussed in Ch apters 4 and 6, respectively. The preference for measured over estimated physicochemical parameters applies to Koc, as well as solubility and Kow (Chapter 2). The LFER describe d by Di Toro et al. (1991) uses Kow to estimate Koc (Equation 3-1), but underestimates the meas ured indigenous (3.9) and spiked (3.9) TCC log Koc values with a prediction of 3.4 (using log Kow = 3.5). The average TCC log Koc is overestimated (4.1-5.6) when the range of commonly reported estimated log Kow values (4.2-6.0) (TCC Consortium, 2002a; Halden and Paull, 2005; Ying et al., 2007) is entered into Equation 31, although an estimated log Koc of 4.1 falls within the rang e of measured values. The PCKOCWIN computer estimation program in EPI Suite (USEPA, 2009c) also underestimates the average measured log Koc, but predicts a log Koc (3.7) that falls within the range of measured values. The PCKOCWIN program employs a me thodology that incorporates the first order molecular connectivity index (i.e represents the degree of br anching or connectivity in a molecule) and multiple, compound-specific correcti on factors (Equation 3-2). However, without 56

PAGE 57

validation of the estimated log Koc values, it is not possible to id entify which predicted value is closest to the true log Koc. log Koc = 0.983(log Kow) + 0.00028 (3-1) where: n = 129 mix-class compounds r2 = 0.81 log Koc = 0.53*MCI + 0.62 + Summation (Pf) (3-2) where: MCI = molecular connectivity index Summation (Pf) = summation product of all applicable correction factors. Even the measured TCC partitioning data shou ld be interpreted w ith caution, however, as the extractability and bioavailab ility of indigenous and spiked TCC may differ despite similar Kd/Koc values. If the extracta bility of spiked compound is gr eater than indigenous TCC, the bioavailability might also be expected to be greater, and measured TCC concentrations in biosolids extracts would underestimate the true total biosolids-borne TCC concentrations. Results of the second biosolids-borne TCC extraction method validation study indicated no difference between spiked TCC extractability following 24 h and 96 h equilibration periods, but even 96 h might not allow a spike association wi th the biosolids as thorough and resistant to extraction as indigenous compound. A closed-system, bench-top, simulated wastewater treatment study that allows comp lete mass balance of radiolabeled TCC would be required to confirm the extractability of indigenous compou nd. Sequential extraction of fresh biosolids and analysis by HPLC/MS would begin to address th e issue of indigenous TCC extractability and bioavailability, but given the large fraction of influent TCC identif ied as transformed or lost by 57

PAGE 58

Heidler et al. (2006) (21 + 30%), mass balance (and confirma tion of complete recovery of biosolids-borne TCC) is problema tic. Previous calculations of expected biosolids-borne TCC concentrations based on the masses of TCC used and biosolids produced in the United States, however, suggest that the current method of biosolids extraction (50:50 MeOH:acetone by shake-flask) is adequate for quantifyi ng biosolids-borne TCC concentrations. The role of dissolved organic matter (DOM) in the partitioning of indigenous and spiked TCC in biosolids remains unclear. The indigenous Kd/Koc values might, or might not, reflect the fraction of TCC that could be associated with dissolved organic matter (DOM). Partitioning studies that utilize a liquid-liqui d extraction technique, in which a solvent is used to sorb the compound of interest from the isolated liquid phase of a matrix, can remove the freely dissolved compound, as well as separate the fraction of compound sorbed to DOM (Lee et al., 2003). Subsequent analysis of the extracting solvent would underest imate the true distribution coefficients for freely dissolved compound (Lee et al., 2003). The method used to isolate the aqueous-phase indigenous TCC in biosolids (i.e. centrifugation of fresh biosolids and collection of supernatant) likely also recovered both freely dissolved TCC and DOM-associated TCC. However, the ability of the an alytical method (i.e. HPLC/MS-MS analysis of the non-filtered biosolids supernatant in nega tive ESI mode) to quantify the DOM-associated fraction is unknown. If the DOM-associated fraction of TCC is not detected by HPLC/MS-MS, the indigenous Kd/Koc values represent only the freely dissolved TCC fraction and could underestimate the amount of TCC which might be environmentally mobile in the aqueous phase. Alternatively, if the DOM-associated fraction of TCC is isolated during chromatography, or efficiently ionized to the target identificati on ions during MS analysis, the indigenous Kd/Koc values represent the freely dissolved TCC plus DOM-associated TCC, and could overestimate 58

PAGE 59

the amount of bioavailable TCC. Compounds a ssociated with DOM are often considered nonbioavailable (Klaus et al., 2000). A study of DOM-bound anilazine (solubility: 8 mg L-1; log Koc: 0.34) residues determined that MS analysis in the ESI mode does not re lease the bound chemical (Klaus et al., 2000), but the generalizability of the finding is unknown. Further, th e results of the spiked 14C-TCC biosolids partitioning experiments (Agyin-Birikorang and OConnor, unpublished) suggest that the indigenous Kd/Koc values do represent the total TCC in the aqueous phase of biosolids (i.e. freely dissolved plus DOM-associated TCC). Li quid scintillation counting of the supernatant from centrifuged biosolids spiked with 14C-TCC quantifies the total radioactivity in the aqueous phase. Given the identical average Koc values for indigenous TCC and spiked 14C-TCC (3.9), the indigenous TCC quantified by HPLC/MS-MS likely al so represents total TCC. Therefore, the partitioning coefficients associated with indi genous and spiked compound are hypothesized to underestimate the true Kd/Koc values (i.e. partitioning of the free compound, only) and to overestimate bioavailability. 59

PAGE 60

Table 3-1. Batch 1 biosolids and select physicochemical properties Batch 1 biosolids Treatment process End-product form dry matter (g kg-1) OC (g kg-1) pH EC (S m-1) Compost 640 n/d n/d n/d DYMK Mixed compost (food waste and mulch) Compost 720 n/d n/d n/d DYSK Compost (mulch) GRBC Aerobic digestion Cake 58 460 8.2 0.11 Cake 160 410 6.1 5.3 ORBC-BL Untreated (before lime stabilization) Cake 180 340 12 40 ORBC-AL Lime stabilization (following lime addition) CFBC Anaerobic digestion Cake 160 410 7.2 4.1 GEPZ Anaerobic digestion Heat-dried 930 n/d n/d n/d RCKF Anaerobic digestion Cake 210 390 8.1 0.97 OSBC Anaerobic digestion Cake 130 420 7.7 1.9 60

PAGE 61

Table 3-2. Batch 2 biosolids and select physicochemical properties Batch 2 biosolids Treatment process End-product form dry matter (g kg-1) OC (g kg-1) pH EC (S m-1) UNKD Anaerobic digestion Cake 280 310 7.5 0.56 UNKG Anaerobic digestion Cake 270 420 8.4 0.69 UNKH Anaerobic digestion Cake 220 300 8.1 0.74 UNKB Anaerobic digestion Cake 250 380 7.8 0.58 UNKC Anaerobic digestion Cake 220 370 7.6 1.6 UNKF Anaerobic digestion Cake 350 210 8.0 0.34 CHST-AD Anaerobic digestion Air-dried 570 n/d n/d n/d CHST-CC Anaerobic digestion Cake 260 280 8.0 0.83 UNKE Anaerobic digestion Cake 240 320 8.3 1.1 UNKI Anaerobic digestion Cake 230 240 8.2 0.46 CHCM-AD Anaerobic digestion Air-dried 890 n/d n/d n/d CHCM-CC Anaerobic digestion Cake 240 n/d n/d n/d UNKJ Unknown Air-dried 880 n/d n/d n/d UNKK Unknown Cake 260 280 8.1 0.71 UNKL Unknown Air-dried 840 n/d n/d n/d 61

PAGE 62

Table 3-3. Triclocarban (TCC) c oncentrations in 23 wastewater treatment plant materials (n=3) Batch 1 biosolids Treatment process TCC content (mg kg-1) Batch 2 biosolids Treatment process TCC content (mg kg-1) DYMK Mixed compost 8 + 2 UNKD Anaerobic digestion 43 + 3 DYSK Compost 6 + 1 UNKG Anaerobic digestion 35 + 1 GRBC Aerobic digestion 7 + 1 UNKH Anaerobic digestion 31 + 0.7 ORBC-BL Untreated 25 + 1 UNKB Anaerobic digestion 25 + 1 ORBC-AL Lime stabilization 18 + 1 UNKC Anaerobic digestion 24 + 1 CFBC Anaerobic digestion 40 + 2 UNKF Anaerobic digestion 23 + 0.3 GEPZ Anaerobic digestion 29 + 3 CHST-AD Anaerobic digestion 14 + 0.8 RCKF Anaerobic digestion 21 + 3 CHST-CC Anaerobic digestion 13 + 0.9 OSBC Anaerobic digestion 14 + 2 UNKE Anaerobic digestion 10 + 0.3 UNKI Anaerobic digestion 8 + 0.4 CHCMAD Anaerobic digestion 8 + 0.8 CHCMCC Anaerobic digestion 7 + 0.9 UNKJ Unknown 31 + 0.4 UNKK Unknown 31 + 1 UNKL Unknown 12 + 0.5 Overall average (omitting ORBC materials) 20 + 11 62

PAGE 63

Table 3-4. Indigenous tr iclocarban (TCC) log Kd and log Koc values for 16 biosolids Biosolids Treatment process Average log Kd Fraction OC (WalkleyBlack) Average log Koc CALC Anaerobic digestion 3.3 + 0.2 0.28 3.8 + 0.2 CFBC Anaerobic digestion 3.2 + 0.1 0.41 3.6 + 0.1 GRBC Aerobic digestion 3.5 + 0.1 0.46 3.9 + 0.1 ORBC-AL Lime stabilization 1.8 + 0 0.34 2.3 + 0 ORBC-BL Prior to lim e stabilization 2.9 + 0.1 0.41 3.3 + 0.1 OSBC Anaerobic digestion 3.7 + 0.1 0.42 4.1 + 0.1 RCKF Anaerobic digestion 3.3 + 0 0.39 3.7 + 0 UNKB Anaerobic digestion 3.3 + 0.1 0.38 3.8 + 0.1 UNKC Anaerobic digestion 3.5 + 0.6 0.37 3.7 + 0.6 UNKD Anaerobic digestion 3.4 + 0.1 0.31 3.9 + 0.1 UNKE Anaerobic digestion 3.1 + 0 0.32 3.6 + 0 UNKF Anaerobic digestion 3.9 + 0.1 0.21 4.6 + 0.1 UNKG Anaerobic digestion 3.5 + 0.8 0.42 3.9 + 0.8 UNKH Anaerobic digestion 3.2 + 0.1 0.30 3.7 + 0.1 UNKI Anaerobic digestion 3.3 + 0 0.24 3.9 + 0.1 UNKK Unknown 3.4 + 0.1 0.28 4.0 + 0.1 Overall average (omitting ORBC materials) 3.4 + 0.2 3.9 + 0.2 63

PAGE 64

0 50 100 150 200 250 300 350 400 450S ny d er et al. (2009) n= 2 3 T NSSS ( 2 00 9 ) n =8 4 Sapk o t a et al. (20 0 7) n =5 H alden (200 7 ) n = 19 Ch u and M etcal fe (20 07 ) n= 4 Hei d ler e t al (200 6 ) n =1Biosolids-borne or sludge-borne TCC (mg kg -1) Figure 3-1. Reported biosolids-bor ne triclocarban (TCC) concentration distributions, showing the mean, 25th and 75th percentile, maximum, and minimum values ( = biosolids; = sludge; = biosolids and sludge) (assuming normal distributions) 64

PAGE 65

CHAPTER 4 LEACHABILITY OF BIOSOLIDS-BORNE TCC IN AMENDED SOIL COLUMNS Introduction The function of soils as a filter, reservoir, and recycler of environmental contaminants like TCC plays a critical role in reducing human and ecological exposures, and mitigating toxic health effects that might be a ssociated with land-applied bioso lids. Myriad biosolids:TCC:soil interactions including physic ochemical processes, adsorp tion, chemical reactions, and mechanical trapping are suspected to influence chem ical retention in or on the soil (Yaron et al., 1996). Such processes affect the extent of bi oaccumulation, biodegrada tion, off-site transport via particulates suspended in water or air, and groundwater contamination. Sandy soils with low organic matter and clay contents are partic ularly prone to leaching of surface applied contaminants in the dissolved and/or soil colloid s-associated form, as vertical transport is facilitated by the minimal soil retention capacity and large pore spaces (Yaron et al., 1996). Because organic matter is a principal component affecting retention of neutral organic compounds in soils, the introduc tion of TCC into the environment as a component of landapplied biosolids is expected to minimize leaching potential. The sparingly soluble nature and the large partitioning coe fficients of TCC (log Koc = 3.3-4.6), combined with the likely thorough indigenous TCC penetration and trapping in the internal stru cture of biosolids, should significantly restrict vertical mobility associated with surface infiltration and water table fluctuations. However, association of TCC with soil organic matter suspended or dissolved in pore water could act to increase appare nt compound solubility (see Chapter 3). Triclocarban leachability from biosolids-am ended soils under a worst-case application scenario (i.e. sandy soil with mini mal OM content) was assessed in a laboratory incubation study to facilitate predictions of biosolids-borne TCC contributions to aquatic contamination, the 65

PAGE 66

propensity of the compound to remain at the site of application, and the risk associated with leachate-related exposure pathways (Chapter 9). The study utilized unplanted biosolidsamended soil columns prepared in an unrel ated phosphorus leaching study conducted by Chinault and OConnor (2008). Multiple biosol ids (11) were applied to samples of an Immokalee fine sand at high rate s (varying from 18 to 52 Mg ha-1) and the columns were periodically leached, which facili tated initial assessmen ts of the influence of biosolids type, biosolids loading rate, and total TCC applie d on the mass and fractions of TCC leached. Materials and Methods Chemicals, Biosolids, and Soils Solvents of HPLC-grade or greater were purch ased from JT Baker or Fisher Scientific. Ammonium acetate was purchased from Fisher Scie ntific. Triclocarban (CAS No. 101-20-2) was obtained from Sigma Alderich ( 99.9% purity). Deuterated TCC (TCC-d7) was supplied by the Procter & Gamble Company. A Florida Immokalee fine sand (sandy, siliceous, hyperthermic Arenic Alaquods) (clay: <10 g kg-1, OM: 7 g kg-1) (OConnor et al., 2004; Chinault, 2007) was collected from a site with no known history of receiving land-applied biosolids or sludge. Study Design Eleven biosolids and triple super phosphate we re selected as the amendments in the soil column leaching study (Chinault and OConnor, 2008). The fertilizer amended treatment served as the control. Biosolids were amended to 400 g (d.w.) samples of Immokalee fine sand Ahorizon at an N-based rate, resulting in equiva lent amendment rates of 18-52 Mg biosolids ha-1. Amended soils were wet with 40 mL water to simu late field capacity and incubated for two wks in zip-lock bags. The bags were opened each d and the soils mixed to promote aerobic conditions. Following incubation, amended soils were loaded into 17 cm x 5 cm PVC leaching columns (3 replicates for each biosolids), fitted with a screen to prevent loss of soil, to a depth of 66

PAGE 67

13 cm and a bulk density of approximately 1.51 g cm-3. The loaded columns (11 materials x 3 reps + 3 TSP controls = 36) were mounted vertic ally on a rack and leached bi-weekly for 14 wks (total 3.5 months). A final leaching was perfor med at 5.5 months. All leachates were frozen until TCC analysis. Each leaching event produced ~60 mL (1/2 por e volume) of leachate, resulting in ~4 total pore volumes of leachate by the conclusion of the study. As each pore volume represented approximately 5.6 cm of drainage, a total of ~22.4 cm of leachate was collected in 5.5 months. Average yearly rainfall in Florida is 140 cm yr-1 (Obeysekera et al., 2004 ) and evapotranspiration is ~70% (Nachabe et al., 2005), so about 40 cm of drainage, or the equivalent of 7 surface soil pore volumes, can be expected each y. Thus, th e 4 pore volumes collected in the lab incubation study represented approximately 7 months of surface soil drainage in the field. Leachate Preparation and Analysis One mL aliquots of unfiltered, thawed leachate samples were transferred to 1.5 mL PTFE microfuge tubes, spiked with internal standard (TCC-d7) and centrifuged at ~2,000 x g for 15 minutes to remove suspended particles. The presence of particulates in, and coloration of, leachates were visually assessed and noted. Samples were analyzed via HPLC/MS/MS (ThermoFisher Quantum Discovery MS with th e ThermoFisher Surveyor HPLC, LS/MS/MS LC-Triple Quad MS), with an LOQ of 0.03-0.06 ng mL-1 (LOD = 0.009-0.018 ng mL-1). The LOD was the no-observable-effect-concentration (NOE C) for sensitive a quatic invertebrate indicator species (0.06-49 ng mL-1; TCC Consortium, 2002b). Results and Discussion Eight of the 11 biosolids amended to the soil columns leached detectable TCC during the first three leaching events (Table 4-1). Only MLPZ (anaerobic digestion, pelletized), GEPZ (anaerobic digestion, pelletized), and OCEC (not stabilized) continued to leach detectable TCC 67

PAGE 68

through events four and five. No biosolids-am ended columns leached detectable amounts of TCC during leaching events six through eight Columns amended with DYSK (compost), GRBC (aerobic digestion), and THAC (anaerobic digestion) leached no detectable TCC during the entire experiment. There was no apparent correlation between ma ss of TCC leached and mass of biosolids applied, biosolids treatment proce ss, particulate matter in leachate or leachate coloration. For the four biosolids for which initial TCC concen trations were known (G EPZ, DYSK, OSBC, and GRBC), however, there was a positive trend between total TCC applied and total TCC leached. Columns amended with GEPZ (58,080 ng applied) leached an average of 105 total ng, or 0.18% of total TCC applied (Figure 4-1). Columns amended with OSBC (12,000 ng TCC g biosolids-1, 21,240 ng applied) leached an average of 11 to tal ng, or 0.05% of total TCC applied. The DYSK(5,000 ng TCC g biosolids-1, 26,320 ng applied) and GRBC-amended (6,000 ng TCC g biosolids-1, 7,380 ng applied) columns leached < 0.02% and 0.06% of TCC applied, respectively. Samples with non-detectable levels of TCC were assigned concentrations of one-half the LOD. The mean biosolids-borne TCC concentration (39 ug g-1) from the TNSSS report (USEPA, 2009a) was used to estimate the amount of TCC applied and percen t leached for biosolids with unknown TCC contents. Estimates of a pplied TCC (using 39 ug TCC g biosolids-1) leached ranged from 0.01%-0.69%. Log Kd values for the biosolids-amended soil treatments in the leaching study were estimated to evaluate the e ffects of TCC partitioning on le achability. The mean log Koc of TCC in biosolids-amended soils (3.82 + 0.34) (Agyin-Birikorang and OConnor, unpublished) was divided by the fraction of OC in each of the biosolids-amended sand treatments to estimate the effective Kd values (Table 4-1). The estimated log Kd values ranged from 1.60 to 1.80, but there 68

PAGE 69

was no relationship between TCC partitioning in bi osolids-amended soil and percents, masses, or concentrations of TCC leached. The estimated percentage of TCC leached was greatest for the ATAD biosolids (0.69%), and might be related to the unique ATAD (autothermal thermophilic aerobic digestion) treatment process, or the very low solids content (3%) of the material. A leachi ng study characterizing the transport of a similar compound [triclosan, (TCS); log Koc = 4.4 (Agyin-Birikorang and OConnor, unpublished)] delivered to a clay loam as a component of biosolids slurry found that 3.4% of applied TCS leached into tile drains (Lapen et al., 2008) Given a TCC log Koc less than that of TCS, and the use of a sandy soil in the st udy herein, the estimate of TCC leached from the ATAD material is not surprising. The OCEC (0.36% leached) and OCED (0.035 %) materials are cake and dried undigested sewage sludge, respectively, from the same WWTP. The cake material is not approved for land application and does not represent a realistic TCC environm ental contamination source. Indigenous and spiked TCC Kd/Koc values were measured for only two of the biosolids (OSBC and GRBC) used in the l eaching study, due to ava ilability of archived materials and/or the requirement of a dry matter content > 35% (Chapter 3). The OSBC and GRBC materials were characterized by the sec ond and fourth greatest measured indigenous TCC log Koc values to date (4.1 and 3.9, respectively), and leached 0.05 and < 0.06%, respectively, of the total TCC applied in the biosolids-amended soil columns. Multiple other biosolid s utilized in the soil column study (e.g. MLPZ, DYSK, BRBC, THAC, BC BC) leached lower percentages of applied TCC than the OSBC and GRBC biosolids. The column study likely repr esented a worst-case-scenario for biosolids-borne contaminants and the percentages of TCC leached are expected to decrease under common field 69

PAGE 70

conditions characterized by vegetative cover, or in soils with greater organic matter and clay contents. All biosolids for which initial T CC content was known leached <<1% (0.02-0.18%) of the TCC applied. Results of spiked 14C-TCC sorption/desorption experiments conducted by colleagues in the Soil Chemistry Laborator y (Agyin-Birikorang and OConnor, unpublished) suggest percent applied TCC loss from finer textured soils will be less than loss from sandy soils, when amended at similar biosolid s application rates. In the so rption/desorption study, biosolids, soils (2 sands, 1 silty clay loam, and 1 clay loam), and biosolids-amended soils were spiked with 14C-TCC and shaken until sorption equilibration wa s reached (24 h, based on preliminary spike equilibration assessments). Immediately followi ng equilibration, a seri es of 24-h desorption steps (in CaCl2) was carried out. Sorption in the sa ndy soils and biosolids-amended sandy soils was similar (65 and 70%, respectively), as was sorption in the loamy soils and biosolidsamended loamy soils (78%). Sorption was quick er and more complete in the biosolids-only samples (98% in 18 h). The resultant mean log Koc (log Kd) values (Chapter 3) were 3.79+ 0.38 (1.71+ 0.09), 3.82+ 0.34 (1.9+ 0.16), and 3.88+ 0.14 (3.34+ 0.13) in the spiked soils, amended soils, and biosolids, respectively. The differences between the mean log Koc values for the spiked soils, amended soils, and biosolids were not statis tically significant, and indicated the influence of OC on spike sorption was similar in all materials. The differences in Kd values between different soil textures amended with biosolids, however, is expected to be one of the most important factors influencing the percent of appl ied TCC that leaches from amended soils, as the most readily leachable fraction will be the TCC present in the aqueous phase. The log Kd values in the sandy soils were 1.5-2.0 x less than the log Kd values in the loamy soils (i.e. Kd values in the sandy soils were ~5-16 x less than the Kd values in the loamy soils), meaning a greater 70

PAGE 71

fraction of TCC in biosolids-amended soils will be initially available for leaching in the sandy soils as compared to the loamy soils. The rate and extent of TCC desorption from bi osolids-amended soils is also expected to affect leaching (and bioavaila bility, Chapter 6). The 14C-TCC desorption data (AgyinBirikorang and OConnor, unpublished; data not shown) indicate sl ow, hysteretic release of the compound, and that leaching of TCC decreases as the ratio of silt and/or clay to sand increases. Desorption from the sandy amended soils plateaued at approximately 80% after five d, but desorption from the loamy amended soil plateaued at approximately 70% after 6 d. Therefore, not only is TCC sorption less complete in sandy amended soils, but desorption is faster and greater as compared to loamy amended soils. For most treatments, the total mass of TCC leached exceeded the amount in itially present in the aqueous phase, suggesting desorption contributed to total TCC loss. All leachates were centrifuged at ~2,000 x g for 15 minutes to remove particulates prior to analysis. Removing particulates could remove solid-bound TCC from the leachate fraction, and underestimate total TCC mass leached. However, adding the TCC-d7 surrogate standard prior to sample centrifugation corrects recovery for the frac tion of TCC present in the leachate as sorbed compound. Therefore, the measured concentrati ons likely represent total TCC leached (i.e. dissolved + sorbed), and could ac tually overestimate the amount of water soluble, and potentially bioavailable, TCC in leachate from biosolids-amended sandy soil. Quantified TCC concentrations were generally greatest in leachates from leaching events one and two (0.08-3.3 ng mL-1), and decreased in leaching even ts three through five (0.05-1.1 ng mL-1). Leachate concentrations were comparable to TCC concentrations detected in treatment plant effluent (0.24-0.83 ng mL-1, activated sludge; 4.8 ng mL-1, trickling filter) (Kerr and 71

PAGE 72

McAvoy, 1999; Heidler et al., 2006). All leachate concentrat ions were below documented acute and chronic (i.e. 90 d 2 y; IRIS, 2009) toxicity endpoints for fish and aqua tic plants (Tables 4-1 and 4-2). However, most quantified TCC concen trations from leaching events one through three (with the exception of OSBC Leachate 3) were above the chronic no-observable-effectconcentrations (NOEC) and concentrations effective in 50% of test populations (EC50) for aquatic invertebrates. Concentrations of TCC from leaching event one in columns amended with ATAD were also within the rang e of acute toxicity endpoints fo r aquatic invertebrates. The concentrations of TCC in biosolids-amended soil leachate are expect ed to be diluted upon interception by groundwater or surf ace water, reducing TCC impact However, the implications of the TCC concentrations measured in the soil column leachates should be interpreted with caution, as concentrations can be expected to vary as a function of leachate volume. One scenario of concern, however, is the pr esence of TCC in con centrations toxic to aquatic invertebrates in field drainage ditches. The aquatic invertebrates (Table 4-2) Ceriodaphnia dubia Daphnia magna, and Mysidopsis bahia are common aquatic indicator organisms for water quality and represent importa nt organisms at the bottom of the aquatic food web. Adverse effects on indicator organisms could adversely affect greater trophic level organisms through disruption of the food chain. Results of the soil column study, however, indicate that acutely toxic levels of TCC in biosolids-amended soil leachate will likely be rare, and that both acutely and chr onically toxic levels are likely short-lived following a single application of biosolids. The sandy soil columns were also designed to simulate worst case leaching conditions, and TCC leachability in finer te xtured soils is expected to be much less. 72

PAGE 73

73 Table 4-1. Triclocarban (TCC) toxicity endpoints for aquatic indicator organisms (TCC Consortium, 2002a; Chalew and Halden, 2009) a: no observable effect concentration; b: lethal concentra tion to 50% of population; c: effective concentration in 50% of population; d: lowest observable effect concentration Endpoint Concentration range (ng mL-1) Acute toxicity Fish NOECa LC50 b 49 97-180 Aquatic invertebrates NOECa EC50 c 2-9 3-13 Chronic toxicity Fish NOECa 5 Aquatic invertebrates NOECa EC50 c 0.06-3 0.2-10 Aquatic plants LOECd EC50 10-10,000 20-36,000

PAGE 74

Table 4-2. Triclocarban (TCC) in bios olids-amended soil column leachates Leaching event Biosolids Average leachate TCC concentration (ng mL-1) Average mass of TCC leached (ng) Total Biosolids Applied (g, dry wt.) Biosolidsborne TCC (ug g-1) TCC Applied (ng) % Solids 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 Total TCC leached per gram of biosolids (ng/g) Estimated % applied TCC leached Estimated log Kd MLPZ 1.96 39 76,440 98 1.7 + 0.29 0.44 + 0.38 0.47 + 0.11 0.05 + 0.04 ND* ND* ND* ND* 112 + 22 26 + 22 25 + 4.6 3.2 + 3.0 ND* ND* ND* ND* 84.7 0.22 1.70 ATAD 1.35 39 52,650 3 3.3 + 0.42 2.0 + 0.50 0.63 + 0.05 ND* ND* ND* ND* ND* 205 + 24 125 + 26 33 + 1.7 ND* ND* ND* ND* ND* 269 0.69 1.66 OCEC 2.06 39 80,340 16 1.4 + 0.47 2.3 + 0.70 1.1 + 0.62 ND* 0.05 + 0.01 ND* ND* ND* 83 + 29 145 + 52 62 + 33 ND* 2.6 + 0.96 ND* ND* ND* 141 0.36 1.72 GEPZ 2.42 24 58,080 93 1.1 + 0.40 0.28 + 0.12 0.19 + 0.02 0.05 + 0.02 ND* ND* ND* ND* 76 + 27 17 + 7.0 9 + 0.74 3.3 + 1.3 ND* ND* ND* ND* 43.4 0.18 1.72 DYSK 3.76 7 26,320 66 ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* < 1.14 < 0.020 1.80 OSBC 1.77 12 21,240 10 ND* 0.13 + 0.01 0.05 ND* ND* ND* ND* ND* ND* 8 + 0.92 3 ND* ND* ND* ND* ND* 6.21 0.050 1.65 BRBC 1.59 39 62,010 13 0.08 + 0.03 0.17 + 0.02 0.07 + 0.01 ND* ND* ND* ND* ND* 8.9 + 0.11 12 + 1.1 4 + 0.77 ND* ND* ND* ND* ND* 15.7 0.040 1.68 GRBC 1.23 6 7,380 5 ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* < 3.50 < 0.060 1.66 OCED 1.81 39 70,590 95 0.12 + 0.04 0.22 + 0.03 0.08 + 0.02 ND* ND* ND* ND* ND* 7.3 + 2.6 14 + 2.3 4 + 0.78 ND* ND* ND* ND* ND* 13.8 0.035 1.70 THAC 1.96 39 76,440 98 ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* ND* < 2.21 < 0.00571.60 BCBC 2.05 39 79,950 13 0.15 + 0 0.16 + 0.08 ND* ND* ND* ND* ND* ND* 3.2 + 2.1 10 + 4.8 ND* ND* ND* ND* ND* ND* 6.34 0.016 1.70 *ND: not detected; Samples with non-detectable levels of TCC were assigned concentrations of one-half the LOD : estimate based on mean sludgeborne TCC concentration (39 ug g-1) in the TNSSS (2009a) 74

PAGE 75

0 1 2 3ATADMLPZOCECGEPZBCBCOCEDBRBCOSBCDYSKGRBCTHAC BiosolidsTCC Concentration in Leachate (ng mL-1) Leaching 1 Leaching 2 Leaching 3 Leaching 4 Leaching 5 Figure 4-1. Triclocarban (TCC) concentrati ons in leachates 1-5, and percent of tota l TCC leached from biosolids of known TCC content (i.e. GEPZ, OSBC, DYSK, and GRBC) 0.18% < 0.02% < 0.06% 0.05% 75

PAGE 76

CHAPTER 5 PLANT UPTAKE OF BIOSOLIDS-BORNE TCC Introduction Plant uptake of organic contaminants introduced in land-applied biosolids, and subsequent human and/or animal exposure, is a major c oncern of governmental regulators and the public (Laturnus et al., 2007). Compounds, lik e TCC, that remain in the soil at the origin of application (Chapters 4 and 6) are of partic ular interest, given the effectiv ely longer plant root exposure periods and the potential for increa sing soil concentrations with re peated biosolids applications (Laturnus et al., 2007). Accumulation of other bi osolids-borne organic ch emicals in plant tissue [e.g. polycyclic aromatic hydrocar bons (PAHs), chlorobenzenes, and polychlorinated biphenyls (PCBs)] has been documented, but contaminatio n is generally restricted to the roots (DuarteDavidson and Jones, 1996). With the exception of root vegetables, contaminant accumulation in above ground biomass that is consumed by animals and/or humans poses the greatest risk to human and ecological health. The chemical concentration and physicochemic al properties, plant species, soil type and conditions, and time of exposure can all affect pl ant uptake of soil contaminants (Chiou, 2002). Plant roots take up subsurface organic contaminan ts via the water and vapor phases in the soil (Trapp and McFarlane, 1995). Given the extremely low vapor pre ssure of TCC, the rate at which the aqueous-phase of the compound crosse s plant root membranes will likely be the critical factor affecting uptake of biosolids-borne TCC in amended soil (Trapp and McFarlane, 1995). The process of TCC plant uptake, if po ssible, is likely passive, and influenced by diffusion, water solubility, and plant membrane sol ubility, as there is no evidence for active plant uptake of anthropogenic organic chemicals (with the exception of some hormone-like compounds) (Trapp and McFarlane, 1995; Collins et al., 2006). Passive plant uptake of 76

PAGE 77

nonionized compounds, such as TCC at typical soil pH ranges, is primarily a function of lipophilicity (Trapp and McFarlan e, 1995), which is positively correlated with the log Kow and log Koc. Uptake of strongly lipophilic compound s is generally limited, due to preferential sorption to soil organic matter (Trapp and McFa rlane, 1995). Nonionized chemicals like TCC are expected to be taken up by plant roots from the soil solution through equilibration of the soil aqueous phase concentration with the interior root concentra tion, and/or sorption onto the lipophilic components of the root s themselves (Trapp and McFarlane, 1995). However, the assumption that root uptake is a partitioning process independent of chemical concentration can be incorrect for some compounds. Research on plant uptake of naphthalene (log Kow = 3.3) suggests roots can become saturated with the co mpound before concentrati ons inside and outside of the root reach equilibrium (Cousins and MacKay, 2001). Over 80% of nonionized compound uptake by plants is typically attributed to the process of aqueous phase equilibration when the log Kow of a chemical of interest is less than 1.5, but the process of lipophilic sorption becomes important with compounds characterized by greater log Kow values (Trapp and McFarlane, 1995). Cousin s and McKay (2001) argue that the aqueous phase in plants represents a compartment distin ct from the plant lipid phase only for chemicals with a log Kow < 2 and the Henrys law constant < 100. Wild and Jones (1992) screened biosolids-borne organic compounds for plant uptake and classified nonionized compounds with a log Kow > 4 as prone to remain in plant roots a nd resist transfer to aboveground biomass. The partitioning of TCC between the soil solids and soil water is a critical factor for estimating accumulation in plant ti ssue, given that plant uptake of soil-borne contaminants with low vapor pressures occurs via th e soil solution. As described in Chapter 3, the concentration of TCC in the soil solution can be estimated by the so il/soil water partition distribution coefficient, 77

PAGE 78

Kd. Nonionized compounds primarily sorb to the organic components of th e soil, so expressing partitioning in terms the so il organic carbon fraction (Koc = Kd/foc) more accurately describes the distribution of a compound between the solid and aqueous phases than the Kd alone. The Koc, in turn, can be related to the mo re readily available term Kow. The positive linear relationship of Koc to Kow indicates that as the organic carbon fracti on of the soil increases, so does the sorption of TCC to the soil solids, t hus reducing the availability of TCC for plant uptake. Nonionized compounds taken up by plant roots m ove predominately via water flow in the xylem in a process much like column chromat ography, i.e. the transport of lipophilic compounds is retarded by reversible partitioning onto the lip id-like plant solids, and tend to accumulate in roots, in the interveinal spaces and at the leaf margins (Trapp and McFarlane, 1995). To reach the xylem, however, a compound must pass through multiple layers, including the epidermis, cortex, endodermis, and the pericycle (Trapp an d McFarlane, 1995). The combined effects of chemical solubility in water and in lipid-rich membranes influences the degree to which a compound moves from the roots to the shoots of the plant (Collins et al., 2006). Movement via the phloem is typically negligible for non-polar, nonionized compounds (Trapp and McFarlane, 1995). Bromilow and Chamberlain (1989, in Tra pp and McFarlane, 1995) designed a figure incorporating data from multiple studies of organic chemical uptake that relates chemical pKa, log Kow, and mode of chemical movement in plants. Given the estimated pKa (12.8) and measured log Kow of TCC (3.5), movement of the antibacterial is predicted to occur only via the xylem, as opposed to a combination of the xylem and the phloem. The typically reported estimated log Kow (4.2), alternatively, places TCC in the non-systemic category (i.e. no movement within the plant). 78

PAGE 79

Relationships have also been identified between log Kow and uptake factors for various plant compartments, including the roots, xylem, and stem (Trapp and McFarlane, 1995; Suter et al., 2000). The relationship between log Kow and an uptake factor can be positive or negative, depending upon whether the plant uptake factor incorporates soil-wa ter or whole-soil contaminant concentrations. Soil-water c oncentrations result in positive log Kow relationships, but whole-soil concentrations, in which a two-step partitioning process (i.e soil-soil water-plant root) occurs, can result in correlati ons that are either positive or negative (Suter et al., 2000). Models relating soil properties, exposure time, an d chemical concentration in plants are more rare (Suter et al., 2000). Ma thematical relationships derive d from laboratory data, often generated in hydroponics systems, however, should be cautiously applied to biosolids-amended soil systems (Trapp and McFarlane, 1995). Cons ideration to bioavailablity in the soil and potential chemical metabolism within plants mu st be given. For example, researchers have hypothesized that the relatively hi gh rates of polychlorinated di benzodioxin and dibenzofuran (PCDD/F) accumulation in pumpkin ( Cucurbitu pepo L. cv. Gelber Zentner) are due to the ability of the plants to increase chemical bioavailability with root exudates (Hulster et al., 1994). In addition, plants, like microbes, are capable of metabolizing many different classes of organic compounds, ranging from highly polar (e.g. glyp hosate) to highly nonpolar [e.g. dichlorodiphenyl-trichloroethane (DDT) and hexachloro benzene] chemicals (Trapp and McFarlane, 1995). In some cases, the parent compound or metabolite is incorporated into components of the plant cell and made resistant to extraction by typical solvents (Tra pp and McFarlane, 1995). Plant uptake of biosolids-borne TCC and translocation into aboveground biomass was characterized using greenhouse-incu bated soil columns amended with one of four biosolids and planted with Bahia grass ( Paspalum notatum ) seed. Bahia grass is a warm season lawn and 79

PAGE 80

pasture grass that requires little maintenance, is suitable for dry sa ndy soils, and is used extensively in the Southern states as a crop rotation and general forage pasture grass. Triclocarban concentrations measured in harveste d tissue were used to calculate bioaccumulation factors (BAFs) and compared to computer model estimates of uptake. In the case of plant accumulation, the term BCF is often used to desc ribe combined uptake from the aqueous phase, sorption, particle-phase deposition and vapors, ev en though the term BAF is more appropriate. The term BAF is used from this point forward to describe Bahia grass accumulation of biosolidsborne TCC, as the mechanisms of uptake have not yet been confirmed and total amended-soil TCC concentrations (as opposed to soil pore water concentrations) are used in accumulation calculations. Materials and Methods Chemicals, Biosolids, and Soils Solvents of HPLC-grade or greater were purch ased from JT Baker or Fisher Scientific. Ammonium acetate was purchased from Fisher Scie ntific. Triclocarban (CAS No. 101-20-2) was obtained from Sigma Alderich ( 99.9% purity). Deuterated TCC (TCC-d7) was supplied by the Procter & Gamble Company. A Florida Immokalee fine sand (sandy, siliceous, hyperthermic Arenic Alaquods) (clay: <10 g kg-1, OM: 7 g kg-1) (Chinault, 2007) was collected from a site with no known history of receiving land-applied biosolids or sludge. Study Design A greenhouse-based planted soil column study, originally designed to characterize the environmental lability and phytoavailability of biosolids phosphorus in a typical sandy Florida soil, was conducted (Chinault, 2007). The st udy was a complete randomized block design arranged in four blocks, each of which contained a single treatment replicate. For the purposes of TCC plant uptake assessment, a control co lumn, which received triple super phosphate 80

PAGE 81

amendment, was placed in each of the four bloc ks. Bahia grass tissue harvested from columns amended with one of four biosolids (GEPZ, OS BC, DYSK, and GRBC) previously analyzed for TCC content (Table 3-3) was used to characte rize plant uptake of biosolids-borne TCC. Biosolids were mixed with 4 kg (dry wt. equiva lent) of the A horizon of an Immokalee fine sand at a 224 kg P ha-1 equivalent rate, resulting in variab le total biosolids and TCC application rates (Table 5-1). The amended soils were wett ed to approximately field capacity in zip-lock bags and equilibrated for two wks prior to colu mn loading. Soil columns (45 cm x 15 cm) were constructed of polyvinyl chloride (PVC) pipes f itted with a screen at the bottom to support the overlying soil and sealed with a punctured cap to allow periodic drainage events (leachings). The columns were first filled with 30 cm (~ 8 kg) of base sand, simulating a native Spodosol E horizon (bulk density 1.5 g cm-3) with adequate depth to allo w unrestricted rooting of the Bahia grass. Readily soluble c onstituents of the artificial E horizon were removed by saturating the base sand and allowing free drainage. The equilibrated amended Immokalee A horizon soil (bulk density 1.5 g cm-3) was subsequently loaded on top of the E horizon in each column, and the columns were planted with 5 grams of Bahia grass seed (~ 29 Mg seed ha-1 equivalent) in June 2006. The seeding rate was ~2.5 times the 11 Mg seed ha-1 rate recommended by Chambliss et al. (2001), thus ensuring thorough Bahia grass so il cover and maximized plant uptake of TCC. The seed was overlain by a thin layer of soil and kept moist by misting every three to four h. Once germina tion occurred, the grass was watered from the tap (pH = 5.0) on a daily to semi-daily basis for 498 d. The columns were weighed twice a week, and soil moisture content was maintained at ~80% (by mass) of column pot-holding capacity by adding tap water as necessary. Potential greenhouse position effects were minimized by rotating columns within their respective blocks each wk. 81

PAGE 82

Bahia grass tissue was harvested every 4-8 wks by cutting the blades to a height of 3.8 cm above the soil surface and the cl ippings were placed in pre-wei ghed paper bags. The tissue was dried at 68C to a constant wei ght to determine the mass yield at each harvest. Immediately following each harvest, columns were leached with tap water (to yield ~500 mL or pore volume of drainage) and supplemented with addi tional nitrogen and other essential nutrients (Chinault, 2007). Plant Tissue Extraction Method Validation and Analysis A plant extraction method for TCC was develo ped and was modeled after the biosolids extraction method described by Heidler et al. (2006) One gram samples of archived dried Bahia grass tissue previously ground in a Wiley mill to pass through a #20 (0.85 mm) sieve were ballmilled in plastic scintillation vials for 20 minutes. Approximately 0.75 g samples of milled tissue were then weighed into 15 mL disposable glass centrifuge tubes and divided (in triplicate) into one of three treatments: 1.) TCC-d7 surrogate standard add ition (15 uL of 1000 ng TCC-d7 mL MeOH-1), 2.) 24.4 ng TCC addition (60 uL of 405 ng TCC mL MeOH-1), or 3.) no additional compound addition. The tubes were loosely cappe d, placed in the laboratory hood, and allowed to dry for 24 h. Following solvent evaporation, 10 mL of 50:50 MeOH:acetone was added to each centrifuge tube. The samples were agitated on a platform rocker for 18 h and centrifuged for 30 min at 800 x g. The supernatant was tran sferred to 20 mL glass scintillation vials and dried under a steady stream of nitrogen. Dried extracts were reconst ituted in 1 mL 50:50 MeOH:acetone, sonicated 30 min to facilitate ex tract re-suspension, and passed through 0.2 um polytetrafluoroethylene (PTFE) syringe filters. Extracts from treatment groups (2) and (3), and three method blanks, were spiked with 15 ng TCC-d7 mL-1 internal standard, and all extracts were analyzed by HPLC/MS/MS. 82

PAGE 83

The plant tissue extr action method validati on yielded a spike recovery of 74+ 8.2% and the technique was applied, unchanged, toward the analysis of tissue samples collected. Tissue collected in Harvests 1 and 2 (44 and 72 d afte r planting, respectively) from columns amended with biosolids GEPZ, OSBC, DYSK, and GRBC at the greatest biosolids application rate (i.e. Nbased rate) were chosen for analysis. Uptake of TCC was expected to be maximized in early harvests from soil columns receiving the greatest bi osolids loads. For each treatment replicate, a yield-weight proportioned composite sample from Harvests 1 and 2 was loaded into a 15 mL disposable glass centrifuge t ube, spiked with 15 ng TCC-d7 (15 uL of 1000 ng TCC-d7 mL-1) surrogate standard, and processed as described above. Triclocarban concentrations in the plant extracts were automatically corrected for any loss during extraction us ing the isotope dilution method. Measured TCC concentrations in plant extracts were adjusted for the mass of tissue extracted, and final values were expressed as ng TCC g oven dry tissue-1 (or ppb). Results and Discussion TCC Concentrations in Plant Tissue Triclocarban concentrations in the composite d Harvests 1 and 2 tissues from columns amended with biosolids GEPZ, OSBC, DYS K, and GRBC were, respectively, 1.2 + 1.2, 0.20 + 0.086, 0.051 + 0.051, and 0.010 + 0.010 ng TCC g oven dry tissue-1 (Table 5-2). The corresponding whole-soil BAF va lues, assuming no change in biosolids-amended soil TCC concentrations with time, were 0.008 + 0.008, 0.003 + 0.001, 0.001 + 0.001, and 0.00041 + 0.00041, indicating 0.041-0.82% of the concentration in the biosolids-amended soils was present in the aboveground grass biomass. Bioaccumulation factors < 0.01 in plants are typically considered inconsequential (OConnor, 1996). Th e large standard errors for the DYSK and GRBC treatments are due to TCC being detected in only one of the four treatment replicates. Triclocarban was detected in two of the GEPZ and three of the OSBC treatment replicates. 83

PAGE 84

The use of uptake factors (e.g. BAFs) to predic t tissue concentrations relies on the implicit assumption that uptake is a simple linear function of environmenta l media concentrations, with a y-intercept of zero (Suter, 2007), i.e. the BAF is constant. However, uptake is often non-linear with respect to soil concentration (as the plant data presented herein il lustrate), and can be influenced by such factors as pH, soil organi c matter content, and saturation kinetics (Suter, 2007). The relationship between the mean soil T CC concentrations and the corresponding mean Bahia grass tissue concentrations is best described by a power function (R2 = 0.9529) (Figure 51), but would be expected to eventua lly reach some unknown uptake threshold. There was no correlation between the mass of organic carbon added to the columns and the mean BAF values (Figure 5-2). The partitioning (i.e. Kd and Koc) of TCC in OSBC and GRBC biosolids was known, but also did not explain th e difference between the BAF values for plant tissue harvested from the two treatments. An inverse relationship between TCC partitioning in the OSBC and GRBC biosolids (log Koc(log Kd) = 4.1 + 0.1 (3.7 + 0.1) and 3.9 + 0.1 (3.5 + 0.1), respectively) and the BAFs (0.003 and 0.00041, respectiv ely) was not observed, but the lack of a relationship was not surprising given the mi nor difference between the measured log Koc values. The BAF values, instead, seem to be better predicted by the amended soil TCC concentration using a linear function that does not intercept at zero (R2 = 0.944) (Figure 5-3). The TCC partitioning coefficients suggest onl y a small percentage (<<1%) of biosolidsborne TCC will be present in soil solution (and thus bioavailable), but desorption over time could increase the effective bioavail ability of the compound. Reversib le desorption could, in theory, allow all of the previously sorbed compound to even tually enter the soil solution and be available for plant uptake. However, Agyin-Birikorang an d OConnor (Chapter 4) documented extensive, but incomplete, desorption of spiked 14C-TCC from biosolids-amended sandy soils. Desorption 84

PAGE 85

of the spike plateaued at approximately 80% af ter 5 d, suggesting 20% of biosolids-borne TCC will remain irreversibly sorbed to the solid fraction. Further, sorption back onto the solid fraction of soil was also rapid (amended sandy soil: 70% sorption in 24 h), and would act to reduce the net amount of aqueous-phase TCC. One other group of researchers quantified TCC concentrations in plant tissue harvested from field plots that received annual applicati ons of biosolids for th ree y (Higgins et al., unpublished). The first plot was a silty clay loam (clay: 370 g kg-1, OC: 32 g kg-1) that received 50 Mg biosolids ha-1 y-1 in the winter of 2004, 2005, and 2006. The second plot was a sandy soil (clay: 30 g kg-1, OC: 3.2 g kg-1) that received 50 Mg biosolids ha-1 in the spring of 2005, 2006, and 2007. The mean biosolids-borne TCC concentrations in duplicate samples of biosolids applied to the loamy and sandy soils were 10 mg kg-1 and 6.5 mg kg-1, resulting in measured amended soil concentrations of 0.51 mg kg-1 and 0.20 mg kg-1, respectively, at the time of sampling in 2007. Duplicate corn stover samples also collected from each soil in 2007 contained 0.002 (loamy soil) and 0.003 (s andy soil) mg TCC kg tissue-1 (d.w.) (range not available). The resulting calculated BAF values were 0.004 (loamy soil) and 0.015 (sandy soil), and were similar to the BAF values calculated for Bahia grass. The regression equation calculated using the TCC concentration in the amended Immokalee soil (sa nd) and the Bahia grass BAF predicts a BAF for the corn stover grown in sandy soil (0.012) similar to the measured value, which further supports the observed positive correlation between soil TCC concentration and BAF. However, the lack of correlation in the lo amy soil suggests soil texture also affects TCC phytoavailability, and/or that an additional y of equilibration in the field (the last biosolids application to loamy soil was in 2006, versus 2007 in the sandy soil) reduces the amount of TCC available for plant uptake. 85

PAGE 86

Model-Predicted TCC Concentrations in Plant Tissue The measured leaf tissue concentrations can be compared to the leaf tissue concentrations predicted by the Biosolids Amended Soil Level IV (BASL4) comput er model (Trent University, 2009). Values for the model input parameters were selected to represen t the greenhouse column growing conditions and duration as closely as possible. As with the measured concentrations, the predicted environmental concentrations were adjusted for percent cumulative yield from Harvests 1 and 2. The predicted leaf tissue concentrations for the GEPZ, OSBC, DYSK, and GRBC treatments were 1330, 449, 353, and 269 ng TCC g-1 (d.w.), or 3-4 orders of magnitude greater than the measured Bahia grass concen trations (Table 5-2). The corresponding BAFs independently calculated using the BASL-4 predicted soil and leaf tissue concentrations were 8.8, 8.4, 7.4, and 13.6, suggesting significant plant bioa ccumulation of biosolids-borne TCC, and greatly overestimating bioaccumulation of bios olids-borne TCC by Bahia grass. The only agreement in the measured and predicted values is in the relative tre nd between increasing BAF values and TCC concentration in the soil. The BASL4 model also predicts the same TCC concentrations in the roots as in the aboveground biomass. The tissue concentrations and uptake factor s predicted by BASL4 using measured TCC physicochemical parameters (solubility: 0.045 mg L-1 ; log Kow: 3.5) imply that bioaccumulation of TCC in plants represents an important expo sure pathway for human and ecological endpoints. However, measured concentrations of TCC in Bahia grass suggest that plant uptake of biosolidsborne TCC, and translocation to the aboveground biomass, is extremely limited. The measured BAFs were well below 0.01, indicating TCC uptake by humans and wildlife via consumption of plant tissue grown in biosolids-amended soil is no t a significant exposure pathway. The range of potential TCC exposures linked to plant ingestio n, and the associated risks, are explored in Chapter 9. 86

PAGE 87

Table 5-1. Triclocarban (TCC) c oncentrations and loading rates in the biosolids-amended soil plant uptake study Biosolids Treatment process Biosolids-borne TCC concentration (mg kg-1) Biosolids loading rate (mg kg-1) Calculated TCC concentration in amended soil (ng g-1) GEPZ Anaerobic digestion 29 6060 145 OSBC Anaerobic digestion 14 4480 62.7 DYSK Compost 6 9360 56.2 GRBC Aerobic digestion 7 3320 23.2 Table 5-2. Measured and pred icted triclocarban (TCC) plant tissue concentrations and bioaccumulation factors (BAF) Measured Biosolids Amended Soil, Level IV model Biosolids Leaf tissue concentration (ng g-1 d.w.) BAF (whole soil) Predicted leaf tissue concentration (ng g-1 d.w.) Predicted BAF (whole soil) GEPZ 1.2 + 1.2 0.008 + 0.008 1330 8.8 OSBC 0.20 + 0.086 0.003 + 0.001 449 8.4 DYSK 0.051 + 0.051 0.001 + 0.001 353 7.4 GRBC 0.010 + 0.010 0.00041 + 0.00041 269 13.6 87

PAGE 88

y = 2E-06x2.6694R2 = 0.9529 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 020406080100120140160 Biosolids-amended soil TCC concentration (ng g-1)Plant tissue TCC concentration (ng g-1) Figure 5-1. Relationship between triclocarban (T CC) concentrations in biosolids-amended soil and TCC concentrations in Bahia grass plant tissue 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 01234567 Organic carbon application rate (g kg-1)BAF8 Figure 5-2. Mean bioaccumulati on of triclocarban (TCC) in Bahi a grass tissue as a function of the rate of organic carbon applied to biosolids-amended soil columns 88

PAGE 89

y = 7E-05x 0.0016 R2 = 0.944 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 020406080100120140160 Biosolids-amended soil TCC concentration (ng g-1)BAF Figure 5-3. Mean bioaccumulati on of triclocarban (TCC) in Bahi a grass tissue as a function of biosolids-amended soil concentration 89

PAGE 90

CHAPTER 6 EARTHWORM TOXICITY AND BIOACCUMULA TION OF BIOSOLIDS-BORNE TCC Introduction Toxicity tests are valuable tools for assessing ecological and human health risks (Singh and Ward, 2004). The general design of toxicity test s involves exposure of te st organisms to the chemical of interest at increasing concentra tions, and subsequent monitoring of relevant biological endpoints (e.g. mortal ity, reproduction, growth, and be havioral changes) (Singh and Ward, 2004). Earthworms are commonly used in soil toxicity assessments because the organisms live in intimate contact with the soil through particle inges tion and sorption of soil solution (in a manner analogous to dermal exposure in vertebrates) and can represent a significant fraction of the diet of some terrestrial vertebrates (Suter et al., 2000). Further, earthworms play an important role in eco logical function by mitigating soil compaction, facilitating seed germination, a nd releasing nutrients for plant growth (Suter et al., 2000). Contaminant concentrations in the soil, soil characteristics, interactions with other contaminants, earthworm activity patterns, material ingested, elimination rates, and earthworm burrowing depth can all influence uptake and bioaccumulation. Ea rthworm exposure to biosolids-borne TCC is especially relevant given the apparent immobility and persiste nce of the compound in the surface layer of amended soil, and the pr opensity of earthworms to consume organic matter near the soil surface (i.e. the top 30 cm) (Suter, 2007). Eisenia fetida earthworms are considered to have an average sensitivity to toxicants as compared to other earthworm species. They pr eferentially feed on soil organic matter (Suter, 2007), and are the worms specified in the USEP A OPPTS 28-d earthworm subchronic toxicity test (OPPTS Guideline 850.6200) (USEPA, 1996a). The guideline requi res two sequential assessments. The first assessment is the range-fi nding test, during which earthworms are exposed 90

PAGE 91

to a wide range of contaminant concentrations to identify a more appropriate, narrower range of concentrations for use in the subsequent definitiv e test. The original guideline prescribes direct addition of the chemical interest to an artifi cial soil matrix. However, the effects of TCC delivered to real soil as a component of bios olids are of interest, so the guideline was thus modified to include TCC-spiked biosolids amended to the artificial soil, as well as a Florida sand and an Ohio silty clay loam. Toxicity and bioaccumulation of biosolids-borne TCC was characterized in terms of earthworm mortalit y and tissue concentrations, respectively. Materials and Methods Chemicals, Biosolids, and Soils Solvents of HPLC-grade or greater were purchased from Alderich, JT Baker, or Fisher Scientific. Ammonium acetate wa s purchased from JT Baker. Tr iclocarban (CAS No. 101-20-2) was obtained from Sigma Alderich ( purity: 99.9%). Deuterated TCC (TCC-d7) was supplied by the Procter & Gamble Company. Anaerobically digested, Class B dewatered cake biosolids (CHCC; indigenous TCC concentration: 7 mg kg-1) were collected from the Calumet Water Reclamation Plant in Chicago, Illinois. Two so ils, the Florida Immokal ee sand (sandy, siliceous, hyperthermic Arenic Alaquods) and an Ohio Genes ee silty loam (fine-loam y, mixed, superactive, mesic Fluventic Eutrudepts) were collected from sites with no known history of receiving landapplied biosolids or sludge. Range-Finding Test Design An earthworm toxicity range-finding test (OPPTS Guideline 850.6200) was conducted to identify the appropriate range of biosolids-borne TCC concentrations for a subsequent definitive earthworm toxicity assessment. Two-gram sa mples of oven-dry, CHCC biosolids were spiked with 0, 10, 100, 1000, or 10,000 mg TCC kg biosolids-1 using acetone as the carrier solvent, and were subsequently dried, re-wetted, and equi librated for 48 h. Spikes added to the CHCC 91

PAGE 92

indigenous TCC concentration, so the actual final TCC concentrati ons in the biosolids samples were 7, 17, 107, 1007, and 10,007 mg TCC kg biosolids-1. Biosolids were amended to 200 g (dry wt.) of an artificial soil (68% silica sand, 20% kaolin clay, 10% sphagnum peat moss, 2% calci um carbonate, by weight), the sand (clay: <10 g kg-1, OC: 5.5 g kg-1), and the silty clay loam (clay: 330 g kg-1, OC: 23 g kg-1) at a 22 Mg ha-1 rate in 800 mL glass Mason jars. The soils were brought to field capacit y (artificial, 35%; sand, 10%; silty clay loam, 24% by wt.) and ten Eisenia fetida earthworms were added to each sample. Also included in the study were soil-only controls and carrier solvent-free biosolids-amended soil controls to quantify potenti al effects of biosolids addition and the carrier solvent, respectively. Replicates were not required by the OPPTS range-fi nding guideline. The lids were placed loosely on top of the incubation jars to reduce moisture loss and prevent earthworm escape. Dead earthworms at the soil surface were counted and removed as necessary each d, and the number of living earthworms were tallied each wk for four wks. Range-Finding Test Results TCC concentrations in the amended sandy so il affected earthworm mortality only at 10,007 mg TCC kg biosolids-1 (equivalent to ~100 mg TCC kg amended soil-1), and at the end of Week 4, all worms in the 10,007 mg TCC kg biosolids-1 treatment were dead (Figure 6-1). All worms were also dead in the sandy soil only contro l at the end of Week 4. Die-off in the sandysoil only control was hypothesized to be attribut able to the lack of a supplemental food source (i.e. biosolids) in a sandy soil, as die-off was not observed in the sandy-soil-solv ent-free control and the sandy soil 7 mg kg-1 treatments (biosolids added; without spike). Earthworms in the silty clay loam soil we re not adversely affected by TCC at any concentration, and earthworm die-o ff at 7 and 10,007 mg TCC kg biosolids-1 in the artificial soil (2 dead by Week 4) was the same. Results of th e range-finding test suggested that TCC present 92

PAGE 93

in biosolids at concentrations many fold grea ter than documented nationally does not adversely affect earthworm survival in so ils amended with biosolids at ag ronomic rates. Results also suggested that other potential cont aminants in the biosolids did not adversely affect earthworm survival. Definitive Test Design Results of the range-finding te st identified a narrower ra nge of biosolids-borne TCC concentrations required for the definitive test. On ly the sandy soil was used in the definitive test, as the biosolids-borne TCC in the silty clay loam and artificial soils had no effect on earthworm mortality at the greatest treatment level of 10,007 mg TCC kg biosolids-1. Spiking TCC at concentrations greater th an 10,000 mg TCC kg biosolids-1 is not recommended, as TCC in the spiking solvent crystallizes on the surface of the biosolids and creates an unrealistic TCC loading scenario. The definitive earthworm toxicity test w ith the sandy soil included six spiking levels between 1000 and 10,000 mg TCC kg biosolids-1 in 1.5x increments, triplicate treatments, and controls. Total TCC concentrations in spiked biosolids for the definitive test were 1317, 1975, 2963, 4444, 6666, and 10,000 mg kg-1 (i.e. indigenous plus spiked TCC). Amended soils were prepared, incubated, and monitored as in the range-finding toxicity test. Spiking levels were confirmed by analyzing both the spiking so lvents and spiked bi osolids extracts by HPLC/MS/MS. Earthworm survival data were analyzed using SAS soft ware, version 9.1 (SAS Institute, 2002). The Tukeys Studentized Range Test was used to determine the wk in which earthworm survival across treatments differe d from Week 0. Single degree-of-freedom orthoganol contrasts were used to assess treatment effects at each wk. Alpha was set to 0.05 in both statistical analyses. 93

PAGE 94

Biosolids-Borne TCC Bioaccumulation Study Design An earthworm TCC bioaccumulation test wa s run concurrently with the definitive earthworm toxicity test. Two-gram samples of CHCC biosolids were spiked with 0x, 10x (~70 mg TCC kg-1 biosolids) and 100x (~700 mg TCC kg-1 biosolids) the indigenous TCC concentration and amended to the fine sand, silty clay loam, and artificial soils in triplicate as in the definitive toxicity test. Surviving worm s at the end of Week 4 were removed, counted, washed, and weighed. The worms were allowed to depurate for 24 h in Petri dishes lined with moistened filter paper (Banks et al., 2006), a nd were subsequently frozen. Depuration is necessary to allow worms to excrete any TCC-c ontaminated soil or organic matter that might remain in the gut, so that measured TCC concentrations in earthworms reflect accumulation in the tissue, rather than TCC sorbed to gut contents. Thawed worm s were transferred to aluminum weigh boats, dried at 50C to a constant weight and ground. The dried tissue was loaded into 25 mL glass centrifuge tubes, spiked with 10 ng TCC-d7 surrogate standard in MeOH, and incubated for 24 h to allow carrier solvent evaporation. A solvent mixture of 50:50 MeOH:acetone (10 mL) was added to each centrifuge tube and the extraction was performed on a platform shaker for 18 h. Extraction was concluded with an a dditional 60 minutes of sonification in a water bath. Tubes were centrifuged at 800 x g, and the supernatant was transferred to 20 mL glass scin tillation vials and dr ied under a gentle nitrogen stream. The extracts were reconstituted in 1 mL of the extraction solvent and filtered through 0.2 um PTFE filters. Filtered extracts were analyz ed by HPLC/MS-MS (LOQ = 1.2 ng g dry tissue-1; LOD = 0.36 ng g dry tissue-1). Triclocarban concentrations in the earthworm tissue were used to calculate TCC bioaccumulation factors (BAFs) for each soil. 94

PAGE 95

Results and Discussion Toxicity of TCC to Earthworms in Biosolids-Amended Soils The sandy-soil-only and sandy-soil-solvent-f ree controls met the OPPTS Guideline requirement that earthworm survival be > 80% at the conclusion of the test, as did the 7, 77, and 707 mg TCC kg biosolids-1 treatments. The 100% survival in the sandy-soil-only treatment (Figure 6-2) verified that sufficient nutrients for Eisenia fetida worms were available in the soil without additional feed amendment. Biosolids addition resulted in a small negative effect on earthworm survival, and the carrier-solvent ha d no impact. No TCC treatment effect on earthworm survival occurred until Week 2. The most striking feature of Figure 6-2 is the large standard errors associated with the sandy soil 1317 mg TCC kg biosolids-1 and greater treatments, due the large variability in survival of earthworms exposed to the highest TCC concentrations. The variability within treatments resulted in no statistical differences across the sandy soil 77-10,000 mg TCC kg biosolids-1 treatments. Mean percent survival in the sandy soil 77 and 707 mg TCC kg biosolids1 treatments was only 10% and 7% less, respectiv ely, than survival in the sandy soil 7 mg TCC kg biosolids-1 treatment. In the 1317 and 1975 mg TCC kg biosolids-1 treatments, however, mean percent survival decreased to 29% and 33%, respectively. The Week 4 mean percent survivals suggest a minimal adverse treatment ef fect on earthworm surviv al in the sandy soil up to the 707 mg TCC kg biosolids-1 treatment, a large adverse tr eatment effect in the 1317 and 1975 mg TCC kg biosolids-1 treatments, and a moderate adverse treatment effect at the 2963 mg TCC kg biosolids-1 treatment and greater. For an initial toxicity endpoint approximati on, the mean earthworm mortality data were entered into the USEPA statistical program PROBIT, version 1.5 (USEPA, 2009e). The program provided a rough estimate of an LC50 of ~4000 mg TCC kg biosolids-1 for the biosolids 95

PAGE 96

application conditions tested (i.e. on e-time application at a 22 Mg ha-1 rate). Application of biosolids at a 22 Mg ha-1 rate, followed by incorporation into the top 15 cm of soil, dilutes the biosolids-borne TCC by ~100-fold, a nd results in an equivalent LC50 of ~40 mg TCC kg amended soil-1. The latter LC50 is 17x less than the earthworm LC50 predicted by the USEPA Ecological Structure Activity Relationships (ECOSAR) computer program (14-d LC50 = 670 mg TCC kg soil-1). However, biosolids with a typical TCC concentrati on (39 mg TCC kg biosolids1`) would have to be land-applied at a 22 Mg ha-1 rate for ~80 y before the LC50 of 40 mg TCC kg amended soil-1 would be reached, assuming no TCC degradation or off-site loss. Bioaccumulation of TCC by Earthworms in Biosolids-Amended Soils Measured TCC concentrations in earthworm tissue The measured concentrations of TCC in earthworm tissue prepared from the fine sand, silty clay loam, and artificial soils amended with 707 mg TCC kg biosolids-1 at a 22 Mg ha-1 rate were 127 + 14, 142 + 8.4, and 36.5 + 0.89 mg TCC kg worm-1 (d.w.), corresponding to whole soil BAF values of 18 + 3.5, 20 + 2.1, and 2.2 + 0.22, respectively (Table 6-1). The measured BAF values are based on whole bioso lids-amended soil TCC concentrations. Model-predicted TCC concentrations in earthworm tissue First estimates of soil-borne organic contamin ant uptake by earthworms typically consider only partitioning of the contamin ant between soil pore water and the organism, but should be in agreement with soil-earthworm uptake models wh en the three phases (i.e. earthworm, water, and soil) are in equilibrium (Suter et al., 2000). Soil pore water-earthworm models have been criticized for potentially underestimating contaminant uptake in cases where absorption via the gut is the primary mechanism, but the models have the advantage of utilizing QSAR-generated earthworm bioconcentration factor s (BCFs, i.e. water-biota part itioning factors) or measured aquatic invertebrate BCF values, which are commonly available (Suter et al., 2000). Measured 96

PAGE 97

aquatic invertebrate BCFs are commonly used as a surrogate for earthworm BCFs (Suter et al., 2000). It is important to note that BCFs calcula ted using soil pore-water concentrations will be greater than BAFs calculated using whole-soil concen trations (unless non e of the compound of interest in sorbed to the solid fraction; in which case the BAF and the BCF would be equal). The Kow of an organic compound can facilitate a first approximation of earthworm uptake and a chemicals tendency to bioconcentrate when measured BAFs/BCFs ar e not available. The concentration of a chemical in ear thworms, as related to the concen tration of the chemical in soil pore water, is calculated by: Cworm = Cpw Kpww (6-1) where Cworm is the chemical concentration in the worm (mg kg fresh weight-1), Cpw is the chemical concentration in the soil pore water (mg L-1), and Kpww is the soil pore water-worm partitioning coefficient (L kg fresh weight-1), or the BCF. Connell and Markwell (1990, in Suter, 2007) developed a model based on 32 lipophilic organic chemicals that relates the Kow to the Kpww: log Kpww = log Kow 0.60 (n = 60, r = 0.91) (6-2) The calculated Kpww, or BCF, using the measured log Kow (3.5) is 790, and, as expected, is much greater (~90-400x) than the measured BAFs. The Cpw is calculated using the formula: Cpw = Cs / Kd (6-3) 97

PAGE 98

where Cs is the chemical concentration in th e solid fraction of the soil (mg kg soil-1). The Kd for a soil with a known organic carbon content (foc) can also be estimated from the Kow, using the model developed by Di Toro (1985): log (Kd/foc) = 0.983*log Kow + 0.00028 (n = 129, r = 0.90) (6-4) The log Kd values estimated by Equation 8-4 for the fine sand, silty clay loam, and artificial soils used in the bioaccumulati on study are 1.15, 1.79, and 2.20, respectively. The estimated log Kd values for the fine sand and the silty clay loam soils are slightly less than the measured values for the soils alone (1.41 + 0.52 and 1.95 + 0.18, respectively), and for the biosolids-amended soils (1.53 + 0.37 and 2.22 + 0.66, respectively) (A gyin-Birikorang and OConnor, unpublished; see Chapter 4). The difference between the estimated and measured Kd values, although small, provides an example of why measured valu es are preferred in calculations of exposure, uptake, and risk. The measured Kd values for the biosolids-amended fine sand and silty clay loam soils, and the estimated Kd values for the artificial soil, are used in the following earthworm uptake calculations. It should be noted that the hysteretic desorption of TCC from biosolids-amended soils (80% max. desorption at 5 d; Agyin-Birikorang and OConnor, unpublished) suggests that the Kd can be expected to eff ectively increase once the desorption plateau has been reached. The result could be reduced or halted earthworm uptake of TCC in the aqueous phase, and subsequent reductions in accumulation. The remaining values required to calculate Cworm are Cs and Cpw. The measured Kd values, although relatively small, are large enough to estimate Cs as equal to the total TCC concentration 98

PAGE 99

in the soil, without in troducing significant error by not acc ounting for soil water content. If biosolids containing 707 mg TCC kg-1 are applied and incorporated into the top 15 cm of each soil at a 22 Mg ha-1 rate (as in the earthworm bioava ilability test), the predicted Cpw values are 0.21 (sand), 0.043 (silty clay loam), and 0.045 (artificial) mg TCC L-1. In the case of the sand, however, the predicted Cpw is greater than the measured water solubility of TCC (0.045 mg L-1; Chapter 2). Assuming the Cpw is approximately the maximum TCC solubility in each of the 707 mg kg-1 application scenarios, the resultant estima ted earthworm concentration in the fine sand, silty clay loam, and artificial soils is the same [i.e. 44 mg TCC kg worm-1 (d.w.)] (earthworm water content = 0.80; BASL4 assumption), and is mu ch less than the concen trations measured in the earthworms grown in the fine sand and silty clay loam soils. The whole soil BAF corresponding to earthworm concen trations of 44 mg TCC kg worm-1 (d.w.) in all three soils is 6.2, and underestimates the extent of measured TCC uptake by earthworms in the fine sand and silty clay loam soils. The computer modeling program BASL4 can also be used to make first approximations of earthworm bioaccumulation. The BASL4 program applies the concept of fugacity (i.e. a measure of chemical potential that describes the escaping tendency of a substance from a heterogeneous system), and models chemical distribution in soils using three levels of simplifying assumptions, i.e. equilibrium, steadystate, and non-steady st ate. Logically, the equilibrium level assumes equilibrium has been ach ieved in all compartments, and thus fugacity is also constant. Chemical con centration in earthworm tissue is calculated as a function of soil fugacity, the Zworm factor (i.e. the capacity of the worm phase for a given chemical), and the bioavailability. The steady state level accounts for earthworm uptake a nd elimination of the chemical of interest, but does not allow removal of the compo und from any soil layer, whereas 99

PAGE 100

the dynamic level accounts for compound loss by incorporating chemical flux into the earthworm concentration calculations. In the earthworm bioaccumulation test, worms were incubated in a jar with a single layer of soil and there was no e xpected chemical loss, so the steady state level was selected. However, by sele cting parameters to simulate a single-layered soil, the differences between predicted earthworm concentrations by each le vel (i.e. equilibrium, steady-state, and non-steady stat e) of BASL4 was less than 2% The steady state earthworm concentration equation is: Cworm = Zworm FugWormSS MolMass 1000 / Densityworm (6-5) where: Zworm = ((VolFractWormlipid + 0.035 VolFractWormnolm) Kow + VolFractWormwater ) Zwater (6-6) FugWormSS calculates the fugacity in the worm at steady state by incorporating the fugacity in the soil, earthworm uptake and elim ination diffusion, and bioa vailability. Earthworm concentration BASL4 output in wet weight was converted to a dry weight basis for easy comparison to the measured concentrations and the aforementioned calculated concentrations. The predicted TCC concentrations in the fine sand, silty clay loam, and artificial soils amended with biosolids contai ning 707 mg TCC kg-1 applied at a 22 Mg ha-1 rate, and assuming 100% bioavailability, were 33, 14, and 7.1 mg TCC kg tissue-1 (d.w.), respective ly. The corresponding BAFs were 5.1, 2.2, 1.0, and were closer to the BAF of 6.2 calculated using Equations 6-1 through 6-4. 100

PAGE 101

The earthworm concentrations and correspond ing BAF values calcula ted with BASL4 and Equations 6-1 through 6-4 greatly underestim ated earthworm bioaccumulation of TCC in biosolids-amended fine sand and s ilty clay loam soils. The diffe rence is hypothesized to be due to preferential consumption of biosolids (containing TCC), as so il organic matter is the primary component of Eisenia fetida worm diet. The organic matter c ontents of the fine sand and silty clay loam soils were ~10 g kg-1 and ~40 g kg-1, respectively, and were less than the organic matter content of the ar tificial soil (~100 g kg-1). Thus, the worms incubated in the artificial soil were exposed to a larger pool of food (i.e. Spha gnum peat moss) that was not contaminated with TCC, as compared to the worms incubated in th e fine sand and silty clay loam soils. If the hypothesis is correct, the increas ed earthworm uptake of biosolids-borne TCC in the sandy and silty clay loam soils, as compared to the artifi cial soil, highlights the necessity of considering earthworm ingestion of the compound, as well as so rption via the soil pore water. The apparent contribution of food ingestion to earthworm upt ake underscores the appr opriateness of labeling biosolids-borne TCC uptake factors by earthworm s BAFs, rather than BCFs. Bioconcentration factors (BCFs) are calculated by dividing the compound concentration in the organism by the concentration in the aqueous phase, and assu mes uptake occurs primarily from solution. Alternatively, bioaccumulation factors (BAF s) are calculated by dividing the compound concentration in the organism by the concentr ation in a select environmental medium, and assumes multiple uptake routes contri bute to accumulation (Suter, 2007). The differences between the measured and estimated values for earthworm bioaccumulation of biosolids-borne TCC provide another example of why the use of measured values in calculations of environmental exposure and risk, when possibl e, is preferred. The measured BAFs presented herein (~2-20) were greater than estimates based on partitioning 101

PAGE 102

102 coefficients (6.2) and fugacity (~1-5). The greatest, and theref ore most conservative, BAF based on measured concentrations (i.e. 20) was applie d to calculations of w ildlife exposures to TCC originating in biosolids, and associated risks (Chapter 9).

PAGE 103

Table 6-1. Measured (means; n = 3 and SE) and estimated triclocarban (TCC) concentr ations and bioaccumul ation factor (BAF) values in earthworm tissue Measured tissue concentration (mg kg-1 d.w.) Measured BAF (d.w.) Calculated tissue concentration (mg kg-1 d.w.) Calculated BAF d.w. BASL4 tissue concentration (mg kg-1 d.w.) BASL4 BAF (d.w.) Sand 127 + 14 18 + 3.5 44 6.2 33 5.1 Silty clay loam 142 + 8 20 + 2.1 44 6.2 14 2.2 Artificial 37 + 1 5.2 + 0.22 44 6.2 7.1 1.0 : Using Equations 8-1 through 8-4; : Using the Biosolids Amended Soil Level IV computer model 103

PAGE 104

Sandy soil Silty clay loam soil Artificial soil 0 10 20 30 40 50 60 70 80 90 100s oil on l y 7 (solvent-free ) 7 17 107 100 7 10,0 07 s o i l onl y 7 (sol ve nt-f r e e ) 7 17 107 10 07 10,0 0 7 soil onl y 7 (so lve nt-f r e e ) 7 1 7 107 1007 10, 0 07mg TCC kg biosolids-1Percent of living worms remainin g Week 1 Week 2 Week 3 Week 4 Figure 6-1. Mean percent of living earthworm s remaining as a function of biosolids-bor ne triclocarban (TCC) concentration and exposure duration in soils ame nded with CHCC biosolids (unrep licated range-finding test) 104

PAGE 105

0 10 20 30 40 50 60 70 80 90 100soil onl y 7 ( s olve nt-f r ee ) 7 77 70 7 1317 1975 2963 4444 6666 1 0,00 0mg TCC kg biosolids-1Percent of living worms remainin g Week 1 Week 2 Week 3 Week 4 Figure 6-2. Mean percent of living earthworms (n=3) remaining as a function of biosolids-borne triclocarban (TCC) concentratio n and exposure duration in a sandy soil amended with CHCC biosolids (definitive test) 105

PAGE 106

CHAPTER 7 SOIL MICROBIAL TOXICITY OF BIOSOLIDS-BORNE TCC Introduction Triclocarban Mechanism of Action Triclocarban is primarily effective as a bacteriostatic against gram positive microorganisms at dilutions up to 1:30 million (Roman et al., 1957; Beaver et al., 1956; Walsh et al., 2003), although some have reported effectiveness agains t gram negative microbes and fungi at much greater concentrations of 1:1000, or 0.1% (Gledhill, 1975). The average biosolids-amended soil TCC concentration calculated in Chapter 3 (0.09-0.18 mg kg-1, or ~1:11 to 1:5.5 million) falls within the effective concentration for some gram positive microorganisms. However, if TCC must be in soil solution (as opposed to the whole biosolids-amended soil) to be effective, the biosolids-amended soil partitioning data co llected by Agyin-Birikorang and OConnor (unpublished) (Chapters 3 and 4) indicates the average TCC concentration in soil solution will be ~0.001-0.002 mg L-1, or ~15-30x less than the greatest di lution shown to be effective against gram positive microorganisms of 1:30 million (assumes a log Kd of 1.9 in amended soils, a biosolids-borne TCC con centration of 39 mg kg-1, and a 5-10 Mg ha-1 application rate). Wastewater treatment products containing the gr eatest documented TCC concentration (441 mg kg-1) (USEPA, 2009b), even prior to land application, are unlikely to significantly affect gram negative microbes and fungi. Very few studies examining the TCC mode of action have been performed, and a definitive answer for how TCC functions as an antimic robial compound has yet to be determined. Triclocarban is an anilide (i.e. contains a C6H5NH2 group, Figure 1-1), a group of compounds shown to induce cell death by adsorbing to and destroying the semipermeable nature of the cytoplasmic membrane (McDonnell, 2007). Anilide interference with proteins or the membrane 106

PAGE 107

phospholipid bilayer could disrup t the proton motive force across the bacterial su rface and/or interrupt active transport and energy metabo lism (McDonnell, 2007). Walsh et al. (2003) specifically investigated TCC activity and mechanisms of action against both gram positive and gram negative bacteria. Escherichia coli ATCC (gram negative), Staphylococcus aureus (gram positive), and Pseudomonas aeruginosa (gram negative) were selected as the test organisms. The authors examined the activity of TCC usi ng agar minimum inhibitory concentrations (MICs), the potentiation of TCC by EDTA and Nerolidol, potassium leakage as an indicator of cell membrane damage, and sub-MIC effects on bact eria growth rates. As expected, TCC only inhibited the gram positive S. aureus Although EDTA is generally used to compromise the integrity of the outer membrane of gram negativ e bacteria (by binding divalent cations required to stabilize the lipopolysaccharide molecules of the cell membrane), EDTA was also found to potentiate TCC activity against S. aureus. This finding suggests EDTA may also affect the cell wall of gram positive bacteria. Triclocarban did not cause potassium leakage in any of the test organisms. Interestingly, sub-MIC concentrations of TCC (0.1-5 ug mL-1) slowed the 18-h growth rate of P. aeruginosa and E. coli ATCC, suggesting possible TCC penetration of gram negative outer membranes. The average concen tration of TCC in biosolids-amended soil pore water, however, is likely ~2 orders of magnit ude less than the lowest sub-MIC concentration tested (0.1 ug mL-1). Potential Impacts of Biosolids-Borne TCC on Soil Microorganisms Acidobacteria and Proteobacteria are typically the most numerically dominant phyla in the soil environment (Fierer et al., 2005). Relatively little is known about th e distribution, diversity, or function of Acidobacteria (Kielak et al., 2009), which are t hought to be gram negative like the more extensively studied Protoeobacteria (Barns et al., 1999). The seven other bacteria phyla (out of ~100 phyla recognized according to the Hugenholtz taxonomic framework in the 107

PAGE 108

Greengenes database) also most commonly found in soil include the Actinobacteria (gram positive), Chloroflexi (gram negative), Bacteroidetes (gram negative), Firmicutes (gram positive), Planctomycetes (gram negative), Verrucomicrobia (gram negative), and Gemmatimonadetes (gram negative) (Youssef and Elshahed, 2009). A soil microbial community performs numerous processes that affect ecosystem dynamics, and the processes are typical microbiological toxi city assessment endpoints (Suter et al., 2007). Three such microbially-mediated processes are re spiration, ammonification, and nitrification, all of which play critical roles in global nutrient cycles by facilitating nutrient flow and decomposition of organic residues (Boyle and Pa ul, 1989). Assessment of biosolids-borne TCC impacts on respiration, ammonifica tion, and nitrification allows an initial characterization of broad soil microbial community effects. Resu lts could confirm outright that biosolids-borne TCC is a hazard to soil microbial communities, and/or direct additional research into more subtle potential adverse effects. Microbial respirati on converts organic substrates (e.g. biodegradable organic contaminants) into useable energy. When degradation is complete (ultima te), the substrate is converted to inorganic end-products. Ultimate degradation of carbon-containing compounds releases water and carbon dioxide, which can be m onitored as a measure of the respiration rate. Potential adverse impacts of TCC on soil respirat ion could interfere with the carbon balance and nutrient flows (via decomposition a nd mineralization) within an ecosystem, in turn affecting soil fertility, crop production (Luo and Zhou, 2006), and the habitat of soil-dwelling organisms. The impact of TCC in land-applied biosolids on the respiration rate of various soil microorganisms will likely differ not only as a function of susceptibility to TCC (i.e. classification as gram negative or gram positive), but also as a functi on of time. Acute and chronic exposures to 108

PAGE 109

differing concentrations of bioavailable TCC can be expected to affect different microbial populations (Ogram and OConnor, unpublished). For example, the well-studied and ubiquitous soil bacteria with bioremediation capabilities, Pseudomonas sp. (Cornelis, 2008), are Proteobacteria generally considered copiotrophs (Bas tian et al., 2009), and their growth is expected to increase rapidly follo wing addition of a substrate-rich amendment like biosolids. As gram negative microorganisms that are very highly versatile with regard us e of carbon substrates (Rehm, 2008), Pseudomonas sp. will unlikely be affected by TCC and, together with other unaffected copiotrophs, might be responsible for much of the initial increase in soil respiration following biosolids additions to soil. The incr ease in soil respiration attributable to Pseudomonas sp. and other copiotrophs could ma sk impacts on other slower growing microorganisms, particularly gram positiv es (Ogram and OConnor, unpublished). The limitations of monitoring whole soil respiration as a measure of biosolids-borne TCC microbial toxicity should be considered when characteri zing the environmental hazard of the antibacterial compound. Two other microbial toxicity endpoints, ammo nification and nitrification, are together referred to as nitrogen mineralization (Yuan et al., 2005). Ammonifi cation is the process by which organic nitrogen compounds are converted to ammonia (NH3) or ammonium (NH4 +) compounds, which may then be utilized as plant nutr ients or serve as substrates for nitrification (Er et al., 2004). Nitrification is a two step process ini tiated by ammonia-oxidizing bacteria (AOB) that convert ammonium to nitrite (NO2 -), and concluded by nitrite-oxidizing bacteria (NOB) that convert ni trite to nitrate (NO3 -) (Yuan et al., 2005). Nitrate is also an inorganic form of nitrogen available for plan t uptake. The gram negative Nitrosomonas and Nitrosospira spp are commonly found in soil and WWTPs, and can be differentially selected according to 109

PAGE 110

environmental conditions and nutrient av ailability (Dytczak, 2008). In soils, Nitrosospira spp. are more likely to be found under natural conditions, as opposed to Nitrosomonas spp., which are more commonly found in soils with large inputs of nitrogen. However, both may be present under either circumstance (Dytczak, 2008). The gram negative Nitrobacter, Nitrococcus, Nitrospira and Nitrospina spp. comprise the NOB group, but most studies on the role of NOB in soils and WWTPs focus on Nitrobacter and Nitrospira spp. (Dytczak, 2008). As with Nitrosomonas spp., Nitrobacter spp. do best in soils with la rge nitrogen inputs and are also present in WWTPs. Nitrospira can also be found in a variety of soils, but is the dominant NOB in most wastewater treatment systems. Nevertheless, Nitrobacter spp. can utilize a wider array of organic substrates than Nitrospira and can grow under aerobic as well anaerobic conditions (Dytczak, 2008). The nitrogen cycle is likely also in fluenced by Archaea (Jetten, 2008), a group of microorganisms comprising a domain separate from Bacteria and Eukaryota. The contribution of Archaea to the processes of ammonification and nitrification is not as well understood as that made by the aforementioned bacteria, but resear ch suggests that archaea might be equally, or more, important than bacteria. An assessment of potential biosolids-borne TCC-induced changes in ammonification and nitrification could identify wher e in the nitrogen mineralization process impacts occur, and thus, identify which groups of microorganisms are likel y affected. The effectiveness of TCC against gram positive microorganisms, and the relatively high TCC concentrations required to inhibit gram negative microorganisms (Walsh et al., 2003b), however, suggest biosolids-borne TCC will not adversely affect the rate of nitrogen minera lization performed by bacter ia in soils amended at agronomic rates with biosolids containing the range of TCC concentrations documented in 110

PAGE 111

Chapter 3. Nevertheless, it is prudent to as sess the impacts of biosolids-borne TCC on soil processes as important as nitr ogen mineralization and respirat ion, with a particular focus on impacts that might occur following y of poten tial TCC accumulation in repeatedly amended soils. Toxicity of biosolids-borne TCC was asse ssed according to the USEPA OPPTS Soil Microbial Community Toxicity Test (Guideline 850.5100) (USEPA, 1996b) in terms of impacts on the soil microbial processes of ammonification, nitrification, and respiration. The guideline requires both a range-finding and a definitive test, and prescribes direct addition of the chemical of interest to a natural soil. The protocol was modified herein to deliver TCC as a component of biosolids in an effort to best simulate the primary mechanism of TCC transfer to the soil. Materials and Methods Chemicals, Biosolids, and Soils Solvents of HPLC-grade or greater were purchased from Alderich, JT Baker, or Fisher Scientific. Ammonium acetate was purchased from JT Baker. Laboratory chemicals were purchased from Fisher Scientific. Triclocarban (CAS No. 101-20-2) wa s obtained from Sigma Alderich (purity: 99.9%). Deuterated TCC (TCC-d7) was supplied by the Procter & Gamble Company. Anaerobically digeste d, Class B dewatered cake bioso lids (biosolids ID: CHCC; TCC content: 24 mg TCC kg biosolids-1) were collected from the Calumet Water Reclamation Plant in Chicago, Illinois. The A-horizon of Immokal ee fine sand (a sandy, siliceous, hyperthermic Arenic Alaquods) was collected from a site in Fl orida with no known history of receiving landapplied biosolids or sludge. Range-Finding Test Design The soil microbial community toxicity range-finding test (OPPTS Guideline 850.5100) was initiated to identify an appropriate range of biosolids-borne TCC concentrations for a 111

PAGE 112

subsequent definitive toxicity assessment. One-gram samples of oven-dry, CHCC biosolids were spiked with 0, 10, 100, 1000, or 10,000 mg TCC kg biosolids-1 (in addition to 24 mg TCC kg biosolids-1 indigenous TCC) using acetone as the carri er solvent. Samples were subsequently dried, re-wetted, and equilibrated for 48 h in 300 mL glass Mason jars. Biosolids were mixed with 100 g (dry wt.) of Immokalee sand at a 22 Mg ha-1 rate and brought to field capacity (10% by wt.). Also included was a soil-only control, a carrier solvent-free biosolids-amended soil control [24 mg TCC kg biosolids-1 (solvent-free)], and an empty jar (to confirm efficacy of the CO2 scrubber system). Replicates were not re quired by the OPPTS range-finding guideline. The jars were placed on an aeration rack and aerated with CO2-free humidified air at approximately 22C. Incoming air was stripped of CO2 and humidified by pumping ambient air first through 2 M KOH, followed by CO2-free water, a column of soda lime chips, additional 2 M KOH, and once more through CO2-free water. Carbon dioxide evolved from each treatment jar was deemed a measure of microbial respiration, and was collected in a series of two base traps containing 100 mL of 0.15 M KOH. Base traps were removed and replaced on Days 5 and 28, and analyzed for CO2 (Anderson, 1982). A solution of BaCl2 was used to first precipitate the carbonates (representing trapped CO2) in a known volume of base tr ap, which was subsequently centrifuged at ~2,000 x g for 10 minutes. One mL of the supernatant was transferred to a glass 20 mL scintillation vial, treated w ith phenolphthalein indicator, and titrated to neutrality with 0.1 M HCl. The 1:1 relationship between HCl re quired to achieve neutrality and the unused KOH remaining in the base trap was used to calcu late the moles of KOH ne utralized by evolved CO2. Also on Days 0, 5 and 28, subsets of amended so ils were destructively sampled and shaken with 100 mL 1 M KCl to extract NO3 --NO2 and NH4 +-NH3. Extracts were filtered through 112

PAGE 113

No.42 Whatman filter paper and analyzed for NO3 --NO2 nitrogen (USEPA, 1993a, Method 353.2) and NH4 +-NH3 nitrogen (USEPA, 1993b, Method 350.1) concentrations to assess TCC impacts on nitrification a nd ammonification, respectively. Method 353.2 accomplishes NO3 reduction by passing soil extrac t through a copper-cadmium coil, and the resultant total NO2 is reacted with multiple reagents to form an azo die for colorimetric analysis. Results were reported as mg NO3 --NO2 -_N per kg soil or biosolids-amended soil. Method EPA 350.1 results in the conversion of extracted NH4 + to NH3, which then reacts with phenol to give indophenol for colorimetric analysis. Results were reported as mg NH4 +-NH3-N per kg soil or biosolidsamended soil. Range-Finding Test Results Biosolids addition was the primar y cause of differences in CO2 evolution between the treatments during Days 0-5 of th e range-finding test (Figure 7-1) Biosolids addition enhanced CO2 production throughout the study, as compared to the soil-only control. Approximately 5070 mg (~3.5-4.5x) more CO2 evolved from the biosolids-amende d samples than in the soil-only control at Day 5, and 100-130 mg (~4-5x) more at Day 28. There were minimal differences in CO2 evolution between the TCC treatments durin g Days 0-5, but the Day 28 data suggested moderately inhibited microbial respir ation at 1,024-10,024 mg TCC kg biosolids-1. Biosolids addition also increased NH4 +-NH3-N production (Figure 7-2). Biosolids amendment at Day 0 increased NH4 +-NH3-N concentrations by ~5x the soil-only control. By Day 5, NH4 +-NH3-N concentrations in the biosolids-amended treatments were ~6-7x greater than in the soil alone, regardless of the TCC spike le vel. A similar trend occurred at Day 28, with minimal differences between TCC spike levels, and NH4 +-NH3-N concentrations greater (~6-7x) in the biosolids-amended treatments than in the soil-only control. The exception was the 24 mg TCC kg biosolids-1 treatment at Day 28 (10 mg NH4 +-NH3-N kg amended soil-1). The large 113

PAGE 114

reduction in the NH4 +-NH3-N concentration in the 24 mg TCC kg biosolids-1 treatment was particularly difficult to explain in the absence of treatment replicates, but was likely due to an analytical error. The definitive test desi gn, however, also included a 24 mg TCC kg biosolids-1 treatment, and allowed reassessment of the TCC level with replicates. Triclocarban concentration and biosolids addition appeared to affect nitrification (Figure 73). Nitrate-nitrite production was reduced in the biosolids-amended treatments at Day 5, as compared to the soil-only control. The Day 5 reduction in NO3 --NO2 --N was most pronounced in the 1024 and 10,024 mg TCC kg biosolids-1 treatment, suggesting an ad verse effect on nitrifying microorganisms at the two greatest TCC spik e levels. At Day 28, a reduction in NO3 --NO2 --N production was evident even at the second lowe st TCC concentration, 34 mg TCC kg biosolids-1, and the NO3 --NO2 --N concentration in the 24 mg TCC kg biosolids-1 treatment returned to a level similar to that of the soil-only control. Definitive Test Design In response to the reduction in NO3 --NO2 --N production at the second lowest TCC concentration tested (34 mg TCC kg biosolids-1), a concentration range of ~24-75 mg TCC kg biosolids-1 was selected for the definitive (replicated) soil microbial community toxicity test. Biosolids were spiked to 24, 31, 45, and 73 mg TCC kg biosolids-1. In addition, a 717 mg TCC kg biosolids-1 treatment was included to simulate potenti al effects after 30 y of annual biosolids application, assuming no physical, chemical, or bi oavailability loss of TC C with time. A soilonly control and a 24 mg TCC kg biosolids-1 (solvent-free) control we re also included. Five replicates were prepared for each treatment and control, and for each sampling period. Samples were removed and analyzed in the same ma nner as in the range-finding soil microbial community toxicity test. A scheduling conflict re quired the definitive test to run 31 d, 3 d longer 114

PAGE 115

than the range-finding test. Data were analyzed using SAS software, version 9.1 (SAS institute, 2002). The Tukeys Studentized Range Test was us ed to assess statistical differences at each sampling period across treatments. Regression analysis with the ESTIMATE procedure was used to assess statistical differences be tween sampling periods within treatments. Results and Discussion Effects of Biosolids-Borne TCC on Soil Respiration Triclocarban concentrations di d not significantly affect CO2 evolution at any time period (Figure 7-4). Only bioso lids addition affected CO2 evolution at Days 0-5, 5-31, and 0-31. Approximately 60-90 mg (~2.5-3x) more total CO2 evolved from the biosolids-amended samples than from the soil-only sample by Day 31 (Figur e 7-5), and the increase was attributed to the addition of carbon and other nutrient sources as a component of the biosolids. Total CO2 evolved was consistent with CO2 production in the range-finding te st. Respiration, as assessed according to the OPPTS Soil Microbial Community Toxicity Test (850.5100), was not affected by TCC concentrations up to 717 mg TCC kg biosolids-1 (or ~30 y of land-applying biosolids containing 24 mg TCC kg biosolids-1 at a 22 Mg ha-1 rate), assuming no loss of TCC or decrease in bioavailability. The CO2 evolution data herein, interpreted wi th respect to a study of sewage sludge impacts on respiration and the number of oligot rophs and copiotrophs in soil (Wolna-Maruwka et al., 2007), suggest differences in biosolidsborne TCC impacts on the respiration of broad groups of copiotrophs and oligot rophs might not be significant. Wolna-Maruwka et al. (2007) amended a grey-brown podzolic soil at three rates (2, 4, and 8 Mg biosolids ha-1), immediately planted barley ( Hordeum vulgare L. ), and cultured oligotrophs and copiotrophs using the plate method at 0, 36, 50, 69, 121, and 135 d. Oligotrophic and copiotrophic mi croorganism numbers were lowest at Day 0 in all treatments, and o ligotrophic microorganism numbers were >10x that 115

PAGE 116

of copiotrophic organisms. Both oligotrophs and copiotrophs in creased in number up to Day 69. Although copiotrophs more than doubl ed in the first 36 d, the number of oligotrophs were always greater than the number of copiotrophs by at least a factor of ~6, re gardless of the sludge application rate. The ratio of o ligotrophs to copiotrophs in all treatments was greatest at Day 0, dipped between Day 50 and Day 69 (depending on the treatment), and finally recovered to match the ratio in the control at Day 135. Most pe rtinent to the assessm ent of TCC impacts on respiration is the correlation between CO2 evolution and the number of oligotrophs or copiotrophs as measured by Wolna-Moruwka et al. (2007). Positive correlations of similar significance were found between CO2 evolution and the number of oligotrophs, and CO2 evolution and the number of copiotro phs at all sludge application rate s. If the correlations also apply to the amended soils (22 Mg ha-1) in the OPPTS Soil Microbial Community Toxicity Study, then the relatively consistent rates of CO2 evolution between all TCC treatments would indicate no differential impact on oligotrophs and copiotrophs as a function TCC concentration. The OPPTS Soil Microbial Community Toxi city Test, however, does not prescribe characterization of contaminant impacts on resp iration beyond 30 d, so longer term differential impacts of TCC on groups of mi croorganisms would not be detected. Future studies should extend the OPPTS guideline well beyond 30 d (or util ize long-term field studies) and incorporate an assessment of oligotroph and copiotroph numbe rs. The studies could confirm the long-term influence of increasing TCC concentrations on total CO2 evolution and on the correlations between CO2 production and oligotroph and copiotroph numbers. Effects of Biosolids-Borne TCC on Ammonification and Nitrification Ammonium-ammonia-N concentrations at Days 0, 5, and 31 were affected only by biosolids addition (Figur e 7-6). The Day 0 NH4 +-NH3-N concentrations, and subsequent NH4 +NH3-N increases up to Day 31 at all TCC treatment levels, were consistent with previous 116

PAGE 117

research on nitrogen mineralizati on rates in biosolids (He et al., 2000). There was no significant difference between NH4 +-NH3-N concentrations across TCC treatment levels at Days 5 and 31. The absence of a TCC treatment effect on am monification was expected given the limited effectiveness of TCC on gram negative organisms, and the gram negative classification of bacteria capable of converting organic nitrogen compounds to NH4 + and NH3. Unlike the effect of biosolids addition on CO2 evolution and NH4 +-NH3-N concentration, biosolids did not increase NO3 --NO2 --N concentrations in amended soils at Day 0 (Figure 7-7). There was no significant difference in Day 5 NO3 --NO2 -N concentrations up to the 45 mg TCC kg biosolids-1 treatment. The Day 5 concentrations in the 73 and 717 mg TCC kg biosolids-1 treatments were statistically greater than in the control and the 24 and 31 mg TCC kg biosolids-1 treatments, but were not significantly different from the 45 mg TCC kg biosolids-1 treatment. The two greatest TCC spike concentrat ions (73 and 717 mg TCC kg biosolids-1) were the only treatments to increase in NO3 --NO2 -N concentration between Day 0 and Day 5. Nitrate-nitrite concentrations ultimately de creased in all treatments over the 31-d test, regardless of biosolids and/or TCC add ition (Table 7-1). Differences in NO3 --NO2 --N concentrations between treatments at Day 31 were much more variable than differences at Days 0 and 5. Day 31 NO3 --NO2 --N concentrations in the cont rol and 45 mg TCC kg biosolids-1 treatments were statistically greater than concentrations in 73 mg TCC kg biosolids-1 treatment, and the concentration in the 73 mg TCC kg biosolids-1 treatment was not statistically different from that in the 717 mg TCC kg biosolids-1 treatment. The trend, considered alone, suggests a TCC concentration effect on nitrate production/consumption at the two greatest TCC spike levels. However, Day 31 NO3 --NO2 --N concentrations in the 717 mg TCC kg biosolids-1 treatment were not significantly different from con centrations in the solvent-free controls, or the 117

PAGE 118

31 and 45 mg TCC kg biosolids-1 treatments, and concentrations in the 24 mg TCC kg biosolids-1 treatment were not statistically different from concentrations in the solvent-free control. Thus, it cannot be definitively concluded that th ere was a TCC treatment effect on NO3 --NO2 --N concentrations beginning at the 73 mg TCC kg biosolids-1 treatment level. As with CO2 evolution and NH4 +-NH3-N concentrations, re sults of the OPPTS Soil Microbial Community Toxicity Test suggest no T CC treatment effect on NO3 --NO2 -N concentrations, up to 717 mg TCC kg biosolids-1. Future work utilizing an extended (i.e. >31 d) incubation of biosolids-amended soils, or a long-term field study, could be useful to assessing whether the modest, non-statistically significant differences in NO3 --NO2 -N concentrations at 73 and 717 mg TCC kg biosolids-1 are indicative of a TCC effect on nitrifiers. Work is necessary to assess th e implications of the nonstatistically significant differences in NO3 --NO2 -N concentrations at the 73 and 717 mg TCC kg biosolids-1 treatment levels, and to potentially id entify TCC-induced effects that the OPPTS guideline is too insensitive to detect. The work could employ t echniques to characterize the structures of nitrifying populations. One appr oach is to extract DNA, using a commercially available kit (e.g. MoBio Power Soil Kit), from bi osolids-amended soils spiked with a range of TCC concentrations, and perform PCR amplification of Bacterial amoA (ammonia monooxygenase gene) and Archaeal amoA (Ogram and OConnor, unpublished). Ammonia monooxygenase (AMO) is the primary enzyme involve d in the first step of nitrification (i.e. oxidation of ammonium to nitrite), and data on the more extensively studied, chlorinated antibacterial triclosan (TCS) indi cate TCS (and, therefore, possibly TCC) inhibits nitrification by competitive inhibition of AMO (Zhao, 2006). Following amplification, the products could be cloned, sequenced, and placed into a phylogeny using MEGA (Ogram and OConnor, 118

PAGE 119

unpublished). Libshuff and Unifrac could be used to statistically assess differences between clone libraries. A similar approach could be taken to assess impacts on microorganisms involved in the second step of nitrification (i.e. oxidation of nitrite to nitrate), as TCS has been shown to inhibit oxidation of nitrite in ba tch reactor experiments using nitr ifying bacteria isolated from mixed liquor in the aerobic stage at wastewater treatment plant (Dokian akis et al., 2004). Differences in the structures of nitrif ying communities exposed to increasing TCC concentrations would indicate a TCC effect and potential selection of TCC-resistant ammonia oxidizers (Ogram and OConnor, unpublished). The relationship between NH4 +-NH3-N and NO3 --NO2 --N concentrations in the soil microbial community toxicity test appears to be a function of biosolids addition, rather than a TCC effect. In well-drained, non-biosolids-ame nded soils above 10C, ammonium is typically oxidized to nitrate within 3-4 wks, and the pr ocess of ammonification is much slower (Wild, 1993). The relationship between nitrification and ammonifica tion rates can be reversed, however, in biosolids-amended soils. He et al .(2000) incubated biosolids under field conditions in Immokalee sand for one y and periodically monitored changes in nitrogen mineralization rates. Initial NH4 +-NH3-N concentrations were much greater than NO3 --NO2 --N concentrations, and the rate of mineralization to NH4 +-NH3-N far exceeded the mineralization rate to NO3 --NO2 -N in the first 30 days. For the first six months of the He et al. study, NH4 +-NH3-N was the dominant form of mineralized nitrogen. Taken together, results from a study characteri zing the effects of TC S on nitrification in soils (Waller and Kookana, 2009), and TCC soil sorption/desorption data (Agyin-Birikorang, unpublished), suggest soil textur e and resulting sorption/desorption kinetics affect the bioavailability, and thus potential impacts, of the antimicrobial compounds on nitrifying 119

PAGE 120

microorganisms. Waller and Kookana ( 2007) spiked a sandy soil (clay: 100 g kg-1, OC: 8.5 g kg1) and a clayey soil (clay: 480 g kg-1, OC: 18.5 g kg-1) to 0, 1, 5, 10, 50, and 100 mg TCS kg soil1, and monitored changes in nitrogen mineralizati on products with time. The concentration of TCS required to affect nitrification was ten times greater in the clayey soil (50 mg kg-1) than in the sandy soil (5 mg kg-1). The difference in nitrification e ffects could not be attributed to a greater sorptive capacity of the clayey soil as compared to the sandy soil, given that the measured Kd values in the two soils were nearly identical, and the Koc in the sand was greater than the Koc in the clay. Waller and Kookana (2007) hypothesized the difference in nitrification effects between the two soils were due to unique soil microbial community structures, but work by Agyin-Birikorang and OConnor (unpublished) suggests that differences in TCS desorption could also contribute to varying bioavailability in the soils. Triclosan (and TCC) partitions similarly onto sandy (~60%) and clayey soils (~75 %) after 24 h of equili bration, but desorption is less complete from the clayey (70%) soil as compared to the sandy soil (80%). The limited desorption from clayey soils, coupled with reduced impacts on nitrification in a clayey soil as compared to a sandy soil at a given TCS soil concentration, suggests that TCS is more effective in the aqueous phase. The same is expected to be true for TCC. Additional research addressing the toxicity of TCC to gram negative organisms, and the bioavailability of biosolids-borne TCC to microorganisms involved in nitrogen minera lization, would help to validate the findings presented herein. It is hypothesized, however, th at the effects on gram negative microorganisms involved in nitrogen mineralization are negligible and that bioavailab ility to all soil organisms is limited (Chapter 8). Materials containing the greatest confirme d concentration of TCC (441 mg TCC kg sludge/biosolids-1) (USEPA, 2009b) would have to be la nd-applied annually at a 22 Mg ha-1 rate 120

PAGE 121

for ~225 y to reach a whole-soil TCC concentration (1000 mg TCC kg amended soil-1) equal to the concentration found to inhibit gram negativ e organisms (Gledhill, 1975). The estimate assumes no loss of TCC or decrease of TCC bioava ilability with time, and that all TCC in the system is bioavailable. The average log Kd in biosolids-amended soils (3.82; Chapter 4) suggests the 225 y period of annual bioso lids applications (22 Mg ha-1) containing 441 mg TCC kg -1 would have to be increased by an additional ~6600x if only TCC in the aqueous phase is bioavailable to gram negatives involved in nitr ogen mineralization. The estimated application periods would increase for biosol ids containing the mean TCC concentration determined in the USEPA (2000b) (Chapter 3) and applied at common agronomic rates (5-10 Mg ha-1). 121

PAGE 122

0 20 40 60 80 100 120 140 160 180s o il o nl y 2 4 (s ol ven t f ree ) 2 4 34 1 24 102 4 1 0,0 24mg TCC kg biosolids-1CO2 evolved (mg) Day 5 Day 28 Total Figure 7-1. Milligrams of CO2 evolved as a function of tric locarban (TCC) concentration and time (unreplicated range-finding test) 0 20 40 60 80 100 120 140s oil o n l y 24 ( s olvent-fre e ) 24 3 4 1 2 4 1024 1 0 ,0 2 4mg TCC kg biosolids-1NH4 +-NH3-N concentration (mg/kg ) Day 0 Day 5 Day 28 Figure 7-2. Concentration of NH4 +-NH3 in soil and biosolids-ame nded soil as a function of triclocarban (TCC) concentration and time (unreplicated range-finding test) 122

PAGE 123

0 0.1 0.2 0.3 0.4 0.5 0.6so i l o n ly 24 (solv e nt-free) 2 4 3 4 1 2 4 1 02 4 10 ,0 24mg TCC kg biosolids-1NO2 --NO3 --N concentration (mg/kg ) Day 0 Day 5 Day 28 Figure 7-3. Concentration of NO3--NO2--N in so il and biosolids-amended soil as a function of triclocarban (TCC) concentration and time (unreplicated range-finding test) 123

PAGE 124

124 0 25 50 75 100 125 150soi l only 24 (solvent-free) 24 31 45 73 71 7mg TCC kg biosolids-1CO2 evolved (mg) Days 0-5 Days 5-31 Days 0-31 f f f f f f d b b d cd d b d b cd b b e c a Figure 7-4. Milligrams of CO2 evolved as a function of tric locarban (TCC) concentration and time (definitive test) (like letters and colo rs indicate no signifi cant difference between treatments)

PAGE 125

0 20 40 60 80 100 120 14005101520253035 DaysCummulative evolved CO 2 (mg) soil only 24 (solvent-free) 24 31 45 73 717 Figure 7-5. Cumulative CO2 evolved from biosolids-amended sand as a functi on of triclocarban (TCC) concentration and time (definitive test) 125

PAGE 126

0 20 40 60 80 100 120 140soil only 24 ( s o lventf ree) 24 31 45 73 717mg TCC kg biosolids-1NH4 +-NH3-N concentration (mg/kg ) Day 0 Day 5 Day 31 g g g g g g e e e e e e bc b bc bc bc c f d a Figure 7-6. Concentration of NH4 +-NH3-N in soil and biosolids-amended soil as a function of triclocarban (TCC) concentration and time (definitive test) (like letters and colors indicate no significant diffe rence between treatments) 126

PAGE 127

0 3 6 9 12 15s o il only 24 (solvent-free) 2 4 3 1 4 5 73 717mg TCC kg biosolids-1NO3 --NO2 --N concentration (mg/kg ) Day 0 Day 5 Day 31 a e e de ab cd b b b b b cd cd c Figure 7-7. Concentration of NO3 --NO2 --N in soil and biosolids-am ended soil as a function of triclocarban (TCC) concentration and time (definitive test) (like letters and colors indicate no significant difference between treatments; see Ta ble 7-1 for Day 31 statistical groupings) Table 7-1. Day 31 NO3 --NO2 --N statistical groupings by treatment (definitive test) Treatment Day 31 NO3 -NO2 --N concentration (mg kg-1) Day 31 statistical grouping soil only 12 a 24 (solvent-free) 9.8 bcd 24 11 acd 31 9.9 bd 45 9.6 bd 71 8.3 e 717 9.0 be 127

PAGE 128

CHAPTER 8 AEROBIC 14C-TCC BIODEGRADATION IN BIOSOLIDS-AMENDED SOILS Introduction Microbial reactions with TCC in the aerobic soil environment could be as simple as singlestep transformations to intermediate metabolites, or as complete as progression to ultimate degradation (i.e. mineralization) of the com pound, yielding water and carbon dioxide (Singh and Ward, 2004). Efficient biodegradation of any organic contaminant re quires that the compound be bioavailable, with the rate of degradation often being a func tion of microbial uptake and mass transfer from the solid to the aqueous phase of the soil (Singh and War d, 2004). The extent and rate of degradation might also be affected by en vironmental factors such as soil pH, temperature, oxygen levels, distribution of microbial popula tions, degree of microbe acclimation to the antibacterial, and accessibility of nutrients, or chemical-specific properties such as molecular structure, hydrophobicity, concentration, and t oxicity (Singh and Ward, 2004). In biosolidsamended soils, bioavailability is expected to be heavily influenced by the processes of TCC adsorption/desorption and diffusion, as desorption, or diffusion out of the inaccessible pores, is generally believed to be necessa ry for the biodegradation of most compounds to occur, but exceptions exist (Singh and Ward, 2004). Indigenous TCC likely preferentially partitions to the biosolids matrix, but migration into or onto othe r soil fractions following land-application could alter bioavailability over time. Aging typically refers to the increasi ng contact time between a compound and soil (Gevao et al., 2000), but can also refer to increasing compound:bios olids interaction over time. Increasing contact time can allow multiple proce sses to occur that lead to strengthened associations between the compound and the solid matr ix of interest, and subs equent reductions in bioavailability and extractability (Gevao et al., 2000). The associa tions can include the 128

PAGE 129

formation of covalent bonds, or passive processes such as sorption to mineral or organic matter, and diffusion into micropores that are too narrow for microbial access (Gevao et al., 2000). In some cases, strongly bound and well-aged or ganic compounds resist removal even when subjected to vigorous, exhaustiv e sequential extraction methods using polar and/or non-polar solvents (Alexander, 1999). Such non-extr actable compounds are commonly labeled bound residues, and were originally de fined by the Internati onal Union of Pure and Applied Chemistry (IUPAC) as chemical speciesthat are unextracted by methods which do not significantly change the chemical nature of the residues (R oberts, 1984). The definition was later modified by Fuhr and Ophoff (1998) to read compounds in so ils, plants or animals which persist in the matrix in the form of the parent substance or its metabolite(s) after ex traction. The extraction method must not substantially change the compounds themselves or the struct ure of the matrix. Thus, the term bound residue is operationally de fined (Chilom et al., 2004) and depends on the selected extraction method and solvent(s). Th e formation of bound residues can occur through a variety of mechanisms and sites of binding that can act independently or in concert, and can limit bioavailability, inhibit biodegrada tion, and reduce toxicity (Gevao et al., 2000; Alexander, 1999). Bound residues can form through the same mechan isms involved in aging, but can also develop via binding to the humic fraction by ionic a nd hydrogen bonding, charge-transfer or electron donor-acceptor mechanisms, van der Waal forces and ligand exchange (Gevao et al., 2000); mineral surface catalyzed reactions (Chilom et al., 2004); and mi crobially mediated reactions (e.g. oxidative coupling via enzyme catalysis) (Chilom et al., 2004). One might hypothesize that TCC exists as a boun d residue in the bios olids matrix, given that TCC: 1) preferentially partitions to the solid fraction; and 2) is present at the very start of a biosolids production process that can take up to many wks to conclude and utilizes the 129

PAGE 130

degradative capabilities of mi crobes. Recovery experiments employing sequential solvent extractions of indigenous or spiked biosolids-bor ne TCC have not been performed, nor has work characterizing how the extractability of indigenous or spiked TCC with multiple solvents might change with time. Although TCC spiked into biosolids and equilibrated 24 h distributes between the aqueous and solid phases in the same pr oportion as indigenous compound (Chapter 3), the difference between extractability and/or bioavailability of spiked and indigenous TCC associated with the solid fraction could be large. Any process that acts to lower the effectiv e solution concentration of a soil contaminant could be expected to affect the rate of biodegradation (Alexander 1999). Failure to consider the kinetics of diffusion or sorption/desorption in so ils might lead to gross misestimates of true biodegradation rates (Alexander, 1999). Compound desorption from biosolids, and compound sorption to the amended soil solid fraction ar e two processes expected to affect TCC bioavailability. For some compounds, sorption and desorption occur as rapid-equilibrium, reversible processes, or as two-phase processes comprised of an initial fast stage of sorption followed by a longer slow phase (Alexander, 1999), or some variation of the two. 14Carbonlabeled TCC sorption and desorption studies in biosolids, soils, and biosolids-amended soils (Agyin-Birikorang and OConnor, unpub lished) suggest desorption of TCC from biosolids could be the rate limiting step for subsequent sorption to the soil and be the most important factor affecting bioavailability. Sixty to ninety percent of spiked 14C-TCC partitioned to biosolids in 8 h, and partitioning plateaued at >99% at 24 h. After 10 d of shaking with 0.01 M CaCl2 solution, desorption of the spike plateaued at ju st 50%. Less than 5% of the spiked compound desorbed in the first 24 h, but the rate of desorption was nearly constant up to Day 10. The results suggest that ~50% of biosolids-borne TC C will remain associated with the biosolids in 130

PAGE 131

amended soil, at least until othe r potential chemical or biologica l processes (e.g. degradation of organic matter) affect the exte nt of sorption. Sorption of 14C-TCC to sandy and clay loam soils was slower and less extensive than sorption to biosolids. The fastes t rate of sorption occurred in the first 6 h following spike addition, and reached 35% and 45% in the sandy and clay loam soils, respectively. Sorption plat eaued at 60% (sands) and 75% (c lay loams) sorption at 24 h. Subsequent desorption from the sands was slow and incomplete in the sands (60% of sorbed compound in 24 h; 80% plateau at Day 5) and cl ay loams (45% of sorbed compound in 24 h; 70% plateau at Day 6). Taken together, the 14C-TCC sorption/desorption data suggest that equilibration of biosolids-borne TCC in amended soils could occur in as little as 10 d (i.e. the time required for TCC desorption from biosolids), thus allowing little time for biodegradation or environmental transfer. As a result of time-dependent changes in bioavailability, the biodegradation rate is generally the greatest imme diately following introduction of a contaminant, but slows appreciably, or plateaus, with time due to ag ing and bound-residue formation mechanisms. The result is often a hockey stick-shaped plot of chemical loss with time (Alexander, 1999). Observation of the latter process with TCC requires that a portion of the biosolids-borne contaminant is readily bioavailable, and that the remaining fraction is either sequestered in the biosolids matrix or forms an association with the soil that lim its bioavailability. Even if TCC in biosolids-amended soils is to tally bioavailable, the exposed soil microbial community might not have the metabolic capabi lity to degrade the compound. The accumulation of TCC in biosolids suggests that microbial communities in full-scale WWTPs may have limited ability to completely degrade TCC under typical treatment conditions (21+ 30% degraded during activated sludge treatment; Heid ler et al., 2006). However, aer obic degradation is likely not 131

PAGE 132

optimized during full-scale activated sludge trea tment. Removal of solids (to which TCC is sorbed) from primary clarifiers promotes bypass ing hours of aerobic microbial contact time in activated sludge tanks (Heidler et al., 2006), and might result in less degradation than would otherwise be expected. If degradability, not sorption, is the limiting factor for bi odegradation in soils, prolonged exposure (due to persistence and/or repeated applications of bios olids) of soil microbes to TCC could facilitate the evolution of required metabolic systems, as documented with other organic environmental contaminants (Singh and Ward, 2004) A delay in the development of metabolic pathways would be observed graphically as a lag time or acclimation period (Alexander, 1999), during which no biodegradation occurs. The lag could be followed by an increasing rate of degradation as the microbial community ad apts to utilize the new compound. Organisms with the capability to degrade a new chemical could increase from essentially non-detectable levels to dominance of the microbial commun ity that may last long after the compound is degraded or dissipates (Singh and Ward, 2004). Lag times, which can range from 1 h to many months, have particular signi ficance to estimates of human and ecological health risks (Alexander, 1999). A long lag time prior to in itiation of TCC biodegradation, for example, would increase the opportunity for environmental transport and exposure. Other factors found to induce a lag time in organic contaminant biodegr adation include gradual proliferation of the degrading population, inhibiting toxic effects, pr edation by protozoa, slow appearance of new genotypes, and diauxie (i.e. seque ntial metabolism of two sugars in a mixture) (Alexander, 1999); any of which might apply to biosolids-borne TCC. If the lag time overlaps with the period of decreasing compound bi oavailability, however, an in crease in the rate of biodegradation might not be observed or could be truncated. 132

PAGE 133

In addition to the rate of ultimate TCC degr adation, the metabolites of TCC, which could be of equal, lesser, or greater toxicity than the parent compound, are also of great interest. The transformation products could be characterized by different physic ochemical properties, bioavailability, and mobility (Si ngh and Ward, 2004), thus requiring an independent, or at least amended, risk assessment. Detoxication might occur because metabolites are less toxic, are more susceptible to degradation, or because of an increased affinity to form soil-bound complexes (Singh and Ward, 2004). For exampl e, reactions catalyzed by such inorganic substrates as iron, silica, clay oxides, and metal oxyhydroxides, or enzymes such as peroxidases, can promote covalent linkage of chlorinated phe nols and anilines to humic substances or soil phenols (Singh and Ward, 2004) and reduce the risk of immediate toxic eff ects. Alternatively, organics could be converted into more toxi c, bioavailable compounds in a process called activation (Alexander, 1999). The pr ospect of activation is strong justification fo r characterizing not only the rate of TCC biodegr adation, but also the pathways. The earliest work addressing TCC degradati on in environmental matrices involved raw sewage and activated sludge, two 14C-labeled TCC analogs (uniformly labeled in the pchloroaniline ring or in the 3,4-di chloroaniline ring), and two benchtop apparatus: a shake-flask and a continuous flow activated sludge (CFAS) system (Gledhill, 1975). Complete degradation of the 14C-p-chloroaniline ring (14C-PCA-TCC) plateaued at 80 -90% over 12 wks in the raw sewage and activated sludge shake-flask studies (initial 14C-PCA-TCC concentration: 200 ug L1). When the 14C-PCA-TCC concentration was increased to 2000 ug L-1, there was a two wk lag time before degradation began, and mineralizati on was only ~70% over 12 wks. Degradation of the 3,4-dichloroaniline ring (14C-DCA-TCC) in activated sludge (200 ug L-1 spike level) was significantly slower than 14C-PCA-TCC, yielding only 7% mineralization over 12 wks (Gledhill, 133

PAGE 134

1975). Weekly additions of PCA, through apparent co-oxidation, increased the degradation of 14C-DCA-TCC to 44%. In the 10-h rete ntion-time CFAS system, 56% of 14C-PCAand 15% of 14C-DCA-TCC mineralized in ~6 wks. Aerobic degradation products identified in the CFAS effluent included DCA, and to a lesser extent, PCA, aniline, and TCC condensation products (i.e. 4-chloro-4`-(4-chloroaniline)-azobenzene a nd 4,4`-dichloro-azobenzene) (Gledhill, 1975). The postulated anaerobic degradation products dichlorocarbanilide (DCC), monochlorocarbanilide (MCC), and nonchlorinated carbanilide (NCC), were identified in sediment cores from the Back River (a Chesap eake Bay tributary that receives WWTP effluent) (Miller et al., 2008). The mechanism of dehalogenation has not been confirmed, but dehalorespiring microorganisms could be involved (Miller et al., 2008). The anaerobic process of reductive dechlorination is consistent with the presence of minimal quantities of the dechlorinated postulated TCC degradation pr oducts in the partially oxygenated surface sediments. The greatest concentrations of the dechlorinated compounds were instead present in the deepest (hypothesized to be anoxic) secti on of the cores (30-40 cm; y of deposition: 19701999) (DCC > MCC > TCC > NCC). The authors estimated that 94% of the TCC deposited during 1970-1999 was transformed by 2006, whereas less than 40% was transformed in the upper sediment layers correspondi ng to 2000-2006. Similar patterns in the presence of potential TCC degradation products were not observed in sediment core s taken from Jamaica Bay, an urbanized estuary on the southwes tern shore of Long Island. In the Jamaica Bay cores, TCC was the predominant compound, with virtually no evidence of TCC degradation. The authors hypothesized site-specific factors including micr obial community composition, competition for electron donors, or chemical i nhibition might explain the di fferences in TCC degradation between the Back River and Jamaica Bay sediments. 134

PAGE 135

No one has attempted to isolate or identif y TCC biodegradation products in soils or biosolids-amended soils under controlled labora tory conditions. The two known studies of TCC biodegradation in soils focused primarily on rates of degradation, quantified as decreases in the amount of parent compound extractable at multiple time intervals (Ying et al., 2007; Xia et al., 2008). Ying et al. (2007) applied TCC directly to an unamended loam (1 mg TCC kg soil-1) and incubated the samples under biotic or abiotic (i.e. autoclaved so ils), and aerobic or anaerobic, conditions for 70 d. A subset of samples we re extracted weekly with acetone for TCC quantification. No statistically significant T CC loss was detected in the abiotic treatments, regardless of the aeration condi tions, suggesting abiotic (chemi cal) degradation of TCC was insignificant and that there was no change in extractability with time. Similarly, TCC degradation under biotic anaerobic conditions was negligible and extractability was unaffected by time. Cumulative TCC loss in the biotic aerobic treatment reached 41% by Day 70, corresponding to an estimated ha lf-life of 108 d, based on assume d first-order reaction kinetics. Xia et al. (2008) assessed TCC biodegradati on in soils (a sandy loam and a fine sandy loam; sterilized by autoclave and non-sterilized;) and the same soils amended with biosolids (73 Mg ha-1 rate). Soils and amended soils were spiked with TCC to concentrations of 1 mg TCC kg soil-1, and incubated for 100 d under aerobic cond itions. Sub-samples were periodically extracted with acetone [using accelerated solv ent extraction (ASE)] to quantify TCC loss. Cumulative TCC loss at 100 d in the non-sterilized soil-only tr eatments (47%, sandy loam; 29%, fine sandy loam) was similar to that reported by Ying et al. (2007), and corresponded to halflives of 72 and 244 d, respectively. However, TCC losses in the sterilized sandy loam and fine sandy loam (30% and 26%, respectively) were only slightly less than th ose in the non-sterile soils. The data suggest that the decreasing reco veries of TCC with time in the sterile and non135

PAGE 136

sterile soils were due to decrea ses in TCC extractability rather than degradation. Triclocarban disappearance was slowed in the biosolids-am ended soils, as evidenced by a reduction in cumulative TCC loss to 13% (amended sandy loam) and 14% (amended fine sandy loam). The methods of TCC loss quantification em ployed by Ying et al. ( 2007) and Xia et al. (2008) make it difficult to confirm that the d ecrease in extractable TCC was due to actual degradation (primary or ultimate), biologically-m ediated decreased extractability of the parent compound with time, or a combination of the two. Results of the study performed by Ying et al. (2007) suggest TCC spiked into soil is degrad ed only under aerobic, bi otic conditions, and/or that decreases in extractability are mediated by aerobic microbes. Results of the Xia et al. (2008) study suggest that aerobic degradation of TCC spiked into soils can occur biotically and abiotically, and/or that decrease s in extractability are both micr obially and physically/chemically mediated (assuming the sterile treatments we re not contaminated during incubation). Triclocarban concentrations in samples collected from biosolids-amended field plots (Higgins, unpublished) further c onfound interpretation of the Ying et al. (2007) and Xia et al. (2008) data. After three y of annual biosolidsborne TCC additions to one sand field plot and two silty clay loam field plots, almost all of the total applied TCC remained in parent form (Higgens, personal communication). Triclocarban loss in a second sand field plot, however, was consistent with TCC degradation rates calculate d by Ying et al. (2007) an d Xia et al. (2008). The field data suggest that extractability (u sing the method developed by Heidler et al., 2006) does not change with time. The details of the acetone extraction methods (e.g. extraction duration or ASE parameters, and clean-up methods) utilized by Ying et al. (2007) and Xia et al. (2008) are not available. 136

PAGE 137

A study that allowed isolation of metabolites and calculation of mass balance was needed to confirm and characterize TCC biodegradation and changes in extractability in soils. Further, the preferential partitioning of TCC onto organi c matter and the potential impacts of partitioning on bioavailability necessitated th e inclusion of biosolids in a realistic assessment of TCC degradation following introduction in to the terrestrial environment. The primary mechanism of TCC transfer to the soil is land-application of biosolids, and a study that used biosolids as the TCC delivery agent was needed to best characterize TCC persistence in soil systems. Therefore, biodegradation of TCC was charac terized using a respirometry test, in which the microbial metabolism of 14C-labeled TCC was quantified by monitoring the evolution of 14CO2 from spiked biosolids-amended soils (USEPA, 1998a). The test was coupled with methods to characterize changes in chemical extractability/bio availability with time and to isolate potential degradation products. Results were used to estimate the ultimate half-life of T CC in biosolidsamended soils and characterize the influence of so il type on degradation rate s and extractability. Materials and Methods Chemicals, Biosolids, and Soils 14C-Triclocarban uniformly labeled on the 4-chloroaniline ring (specific activity: 75 mCi/mmol; 98.5% purity; most rapidly degraded of the two rings) was synthesized by Amersham Life Science and supplied by th e Procter & Gamble Company. Methanol, chloroform, NaOH, HCl, and sodium azide (NaN3) were purchased from Fisher Scientific. Anaerobically digested, Cla ss B dewatered cake biosolids (CFBC; indigenous TCC concentration: 40 mg kg-1, Table 3-3) was collected from the belt-press of a WWTP serving primarily single family homes and a dye factory in Fairfield, Ohio. Samples of two soils, a Florida Immokalee fine sand (sandy, siliceous, hyperthermic Arenic Alaquods) and an Ohio 137

PAGE 138

Genesee silty clay loam (fine-lo amy, mixed, superactive, mesic Fluventic Eutrudepts) (Table 81) were collected from sites with no known history of receiving land-applied biosolids or sludge. Study Design Glass, round-bottom, 30 mL centrifuge tubes were used to prepare 146 samples (2 soils x 1 biosolids x 1 rate x 4 replicates x 8 sampling periods x 2 treatments (biotic/inhibited-biotic) + 6 acidified samples + 12 controls) (T able 8-2). Biosolids samples (0.10 g d.w.) were loaded into the centrifuge tubes, spiked with 1.3 x 106 dpm 14C-TCC/g (d.w.), allowed to equilibrate 24 h, and amended to 10 g (d.w.) of soil to simula te a realistic field loading rate (22 Mg ha-1 or ~10 tons acre-1). The radioisotope spike increased the effective biosolids-borne TCC concentration to ~65 mg TCC per kg, and the total TCC load in amended soil was ~0.65 mg TCC per kg amended soil. The inhibited-biotic treatment was included in the study design to faci litate differentiation between microbially-induced and non-microbi ally-induced effects on TCC degradation and extractability with time. The inhibited-bi otic samples were prepared by mixing 1000 ug g-1 NaN3 (0.1%) in MeOH carrier solvent into a subset of the biosolids-amended soils (Ou, personal communication, 2007). The term inhibi ted-biotic, rather than abioti c, is used to describe the treatment because of the di fferential effects of NaN3 on gram negative and gram positive microorganisms. Sodium azide primarily acts as a bacteriostatic against gram negative microoorganisms, and is less effective against gram positives (Lichstein and Soule, 1944). Given the typical predominance of gram negatives in the soil environmen t (Fierer et al., 2005) and the expectation that gram negatives will be responsible for most TCC biodegradation (See Chapter 7), the addition of NaN3 was expected to inhibit much but perhaps not all, of the microbial activity and TCC biodegradation in th e treated biosolids-amended samples. Soil microbial community tolerance of the 14C-TCC spike and the effectiveness of the NaN3 was 138

PAGE 139

deemed adequate based on total CO2 evolution measurements th roughout the duration of the study (see Sample Analysis, and Results and Discussion sections below) Each centrifuge tube was fitted with a si licon/Teflon septum and connected, via Dow Corning Silastic laboratory tubi ng, to a series of glass trappi ng vials containing 5 mL 0.2 M KOH (Figure 8-1). The first vial remained empty to prevent backflow into the centrifuge tubes if pump failure were to occur. An oilless pump aerated samples with humidified, CO2-free air by first pumping air through 1 M KOH, and then deionized H2O. Although not prescribed by the OPPTS guideline, carbon dioxide was removed from in coming to air to facil itate calculations of total CO2 evolved from the amended soil samples (as a measure of biotic activity; additional OPPTS guidelines for the assessment of anaerobic degradation are also available). The experiment was conducted in the dark at ~23C. Samples were weighed pe riodically and sterile water was added, as needed, to maintain a soil moisture content of 10% w/w (sand) and 20% w/w (silty clay loam). The six acidified samples were incubated identically to all other samples, but were acidified with HCl at the final sampling period to determine extent of 14Clabeled CO2 dissolution in soil water. Sample Analyses Once a week, position-1 base traps were removed, remaining traps were moved forward, and a fresh trap was added to the newly open position-3 (Figure 8-1). Four replicates of each biosolids-amended soil treatment were peri odically removed and sequentially extracted (sonication, 1 h) with 20 mL of deionized H2O (twice), MeOH (twice), and 1 M NaOH (once). The water and MeOH extractions served as first estimates of TCC bioavailability in biosolidsamended soils (Kelsey et al., 1997; Jussara et al., 2006), and the NaOH extraction estimated the extent of TCC incorporation in to the humic fraction (Schnitzer 1982). Further explanation of the extraction scheme is provided in the Chap ter 8 Results and Discussion section. Sample 139

PAGE 140

extracts and base traps were aliquoted into 15 mL Ecoscint A liquid sc intillation cocktail and stored 24 h prior to LSC analysis. A subsampl e of each sequentially ex tracted replicate was removed, weighed, and combusted on a sample oxidizer to complete mass balance of the 14CTCC spike. The Tukeys Studentized Range Test was used to assess differences in 14C extractability and CO2 with time and across treatments. Alpha was set to 0.05. Extracts containing adequate radioactivit y to allow radio-thin -layer-chromatography (RAD-TLC) analysis were concentrated to a known volume under nitrogen and spotted (30-60 uL) onto Whatman thin layer chromatography plates (Partisil LK5D, Silica Gel 150 20x20 cm). Plates were developed in a seal ed glass chamber containing 100 mL of 98:2 chloroform:MeOH and read on an automated RAD-TLC scanner. The initial sampling schedule prescribed bi osolids-replicate removal at 0, 3, 6, 9, 12, 15, 18, and 21 wks (T0-T7, respectively). However, because CO2 evolution rates and RAD-TLC extract analyses of initial sa mplings indicated slow TCC degr adation, sampling periods T5, T6, and T7 were extended to 16, 24, and 30 wks, re spectively. At wk 19, the remaining samples received a second biosolids amendment of 0.1 g dry wt. equivalent, raising the effective TCC concentration to ~1 mg kg-1. Biosolids were added to the sa mples to compensate for potential microbial die-off due to depletion of nutrients over the course of the study and as an attempt to increase 14C-TCC biodegradation rates. Furt her, in an attempt to confirm 14C-TCC addition at study initiation did not inhibit biotic samples, total CO2 (i.e. cold plus ra dioactive) evolution was determined for represen tative replicates (Anderson, 1982). A solution of BaCl2 was used to first precipitate the carbonates (representing trapped CO2) in a known volume of base trap, which was subsequently centrifuged at ~2,000 x g for 10 minutes. One mL of the supernatant was transferred to a glass 20 mL scin tillation vial, treated with phenolpht halein indicator, and titrated 140

PAGE 141

to neutrality with 0.1 M HCl. The 1:1 relationship between HCl required to achieve neutrality and the unused KOH remaining in the base trap was used to calcula te the moles of KOH neutralized by evolved CO2. Results and Discussion Explanation of the Sequential Extraction Scheme The terms labile and bioavailable are commonly used interchangeably, a practice explained by the similar popular definitions of the two terms, as they relate to the field of soil science. The Soil Science Society of America defines labile as readily transformed by microorganisms or readily available to plants (SSSA, 2009). The USEPA defines bioavailability as the degree of ability to be absorbed and ready to interact in organism metabolism (USEPA, 2006). Some authors atte mpt to tease apart the two terms by defining labile as the fraction of cont aminants that is most chemica lly active, e.g., more soluble or most easily extracted (i.e. the fraction in so il solution plus the fraction that is readily desorbable), and bioavailability as the portion of the contaminant that can be assimilated or otherwise exposed to a target organism (Cofield et al., 2008). The Cofield et al. (2008) definitions imply that the labile fraction will be more likely to be environmentally mobile than the non-labile fraction, and that the bioavailable fraction of a contam inant is likely to be a subset of the labile fraction. Multiple researchers have explored the relatio nship between spiked chemical extractability and bioavailability, as measured by microbial mineralization or earthworm accumulation (Chung and Alexander, 1999; White et al., 1999; Reid et al., 2000; Liste and Alexander, 2002; Tang et al., 2002). Water extractions are commonly used to estimate or ganic compound leachability in soils, but do not always correlate well with bioavailability (Hickman and Reid, 2005). The bioavailability of hydrophobic compounds, in part icular, can be underest imated using water 141

PAGE 142

extractions, as the aqueous solubility can limit the amount of bioavailab le chemical extracted (Semple et al., 2003). The use of MeOH as an extractant is one approach to circumvent the problem with compounds characterized by low water solubility. A combination of water and MeOH has been used to successfully estimate the bi oavailability of atrazi ne (Kelsey et al., 1997) and simazine (in conjunction with CaCl2) (Jussara et al., 2006), but can underestimate the bioavailability of other compounds, such as pyrene (Macleod and Semple, 2003). In other instances, a relationship between chemical extractability with water and/or MeOH and bioavailability is established in a single soil, but the relationship does not persist when additional soils or compounds are assessed (White et al ., 1999; Johnson et al., 2002). A sequential extraction method beginning with individual wate r and MeOH extracts was selected here to provide a first estimate of TCC bioavailability to microorganisms in biosolids-amended soil. The final extractant, NaOH, was selected to release 14C-TCC (or 14C-metabolites) that might be associated with the humic fraction, as humic mate rials are soluble in alkali solution (Schnitzer, 1982). The humic fraction is extremely resistant to microbial degradation, can form stable water-soluble and insoluble complexes with meta l ions and hydrous oxide s, and can interact with clay minerals (Schnitzer, 1982). Dilute alkaline extracts (e.g. < 0.5 M NaOH) are not thought to alter the physical and ch emical properties of the soil or ganic matter matrix (Schnitzer, 1982), so compounds present in the extractant woul d not qualify as bound residue. However, a more concentrated NaOH solution (1 M ) was used in the 14C-TCC biodegradation experiment sequential extraction scheme, and could have rel eased bound residues if the organic matter was, in fact, altered. The analyti cal methodologies employed in the 14C-TCC biodegradation study did not allow for an assessment of organic prope rties before and after extraction, so the proper categorization of the NaOH-extractable radioactivity is uncertain. 142

PAGE 143

Parent Compound Loss and Changes in Extractability Biotic sand (BSF) total percent recoveri es for T0-T7 (Table 8-3) averaged 91+ 5%. However, recovery in one of the four replicat es for BSF T3 was unusually and inexplicably low (62%). Ignoring the low repli cate yielded a more representative BSF T3 average percent recovery of 86+ 5%, and an overall average recovery percentage for all times of 92+ 4%. Inhibited-biotic sand (ASF) total percent recoveries for T0-T7 were averaged 109+ 6%. Percent recoveries in the biotic (BSO) a nd inhibited-biotic (ASO) silty cl ay loam for T0-T7 (Table 8-4) averaged 107+ 8% and 109+ 9%, respectively. There was no decreasing (or increasing) trend in total percent recoveries over time in any treatmen t. Percent recoveries were consistently greater in the inhibited-biotic fine sand samples than in the biotic fine sand samples, but the same trend was not observed for the silty clay loam samples. Differences between 14C recoveries in sequential extracts of biotic and inhibited-biotic samples Across sampling times, water extractability of 14C was significantly less in the BSF samples than in the ASF samples (Table 8-3) (F igures 8-2a and 8-2b: normalized for 100% total spike recovery). The same statistical trend was true in the MeOH extracts, but, graphically, was less obvious. 14C-carbon dioxide evoluti on was significantly greater in the BSF samples as compared to the ASF samples. Recoveries in the NaOH and combusted fractions were not statistically different between the two sand treatments. At every sampling, radioactivity recovered by water and MeOH extr action was significantly greater in the sand than in the silty clay loam (biotic and inhibited-biotic). The diffe rence is likely associated with the greater clay and OC content of the silty clay loam (clay: 330 g kg-1, OC: 23 g kg-1) as compared to the sand ( clay: 10 g kg-1, OC: 5.5 g kg-1). Conversely, recovery in th e combusted fraction was always greater in the silty clay loam than in the sand. 143

PAGE 144

With the exception of no significant difference between water extractability in the BSO and ASO samples, relationships between the recove ries in the sequential ex tracts of the biotic and inhibited-biotic silty clay loam samples were similar to the relationships between the biotic and inhibited-biotic sand samples (Table 8-4, Figures 8-2c and 8-2d). Wate r extractability in the BSO samples, although not statistica lly significant, appeared to be less than in the ASO samples. Differences between 14C recoveries within treatment extracts over time Water extractability of the radiolabel in the BSF and BSO samples decreased significantly with time (14% to 3%, and 3% to ~0.6%, respectively), and CO2 evolution increased (Figures 82a and 8-2c, respectively). A lthough no statistically significant ch anges occurred over time in the remaining fractions of either biotic treatment, MeOH extractability tended to decrease, and the combusted fraction tended to increase. No si gnificant changes occurred in any fraction of the ASF or ASO samples, but as with the biotic tr eatments, MeOH extractability tended to decrease and the combusted fraction tended to increase. Mineralization of 14C-TCC Total 14CO2 production was significantly greater in th e biotic silty clay loam than in the sandy soil (Figure 8-3). Despite addition of bi osolids to samples remaining after the T5 sampling period, no increase in 14CO2 production occurred in the biotic amended sand between T5 and T6, and only a small increase (from 2% to 4%) occurred in the biotic silty clay loam samples (Table 8-4). The greatest 14CO2 production (prior to second bi osolids addition) occurred between T0 and T1 in the silty clay loam and be tween T0 and T1 or T2 in the sand. The percent of added 14C recovered as 14CO2 increased slightly in subsequent samplings, but most mineralization occurred within the first two wks of the experiment. All weekly base traps contained radioactivity on the order of fractions of a percent, suggesting that 14C-TCC continued to slowly degrade. After 30 wks of incubati on (T7 samples), approximately 3% and 5% of 144

PAGE 145

applied 14C-TCC was captured in biotic sand (Table 83) and silty clay loam (Table 8-4) base traps, respectively. Correcting for 14C-TCC captured by control base traps, approximately 2% (biotic sand) and 4% (biotic silty clay loam) of 14C-TCC was mineralized in 7.5 months. Nonspiked, biosolids-amended sand and silty clay loam control base traps contained an equivalent of ~1% and 0.8% of radioactivity applied to spiked treatments, respectivel y. Cross-contamination most likely occurred through in advertent changes in air-flow direction during periodic CO2 scrubber refreshments. Inferences made from measured mineralization rates and differences in 14C extractability with time and between treatments The large amount of radioactivity extractable in the water a nd MeOH fractions, paired with minimal CO2 evolution, suggests three possible conc lusions (none of which are mutually exclusive): 1. the soil microorga nisms did not have the capabili ties to degrade the radiolabel before bioavailability was reduced, 2. water and methanol are not good predictors of TCC bioavailability to soil microorgani sms, or 3. the radiolabel spike ki lled or otherwise inhibited the potential degraders. It is clea r that some fraction of the radiol abel was bioavailable, given the production of even small amounts of 14CO2 in both biotic treatments. Therefore, conclusion number one is false. Decreasing bioavailability with time did appear to affect degradation, as the majority of cumulative evolved 14CO2 was collected in the first three wks of the experiment (until additional biosolids were added at Week 19). Conclusion number two is only pa rtially correct. Water seemed to be a better predictor of bioavailability in the silty clay lo am than in the sand. In the first three wks in the silty clay loam samples, the fraction of water extractable radi oactivity decreased by approximately the same percentage as the amount mineralized. However, water less accurately predicted mineralization in the sandy soil. The large percent recoveries and the absence of statistically significant 145

PAGE 146

decreases in extractability with time in the MeOH fractions of both so ils did not reflect the limited rates of radiolabel mineralization. Conclusion number three is thought to be unlikely. Soil resp iration results of the soil microbial community toxicity study suggested minimal TCC impacts on the composition of soil microbial communities in soils amended with bi osolids containing up to 717 mg TCC kg applied at the same 22 Mg ha-1 rate (Chapter 7). Further, the microorganisms found to be capable of degrading TCC in activated sludge are related to the genus Burkholderia (Miller et al., 2006), gram negative organisms not be expected to be adversely affected by TCC at the concentrations present in the 14C-TCC biodegradation study. Sodium hydroxide solution extracted little of the radioactivity from the amended soils in all treatments, and extractability did not significantly change with time. The limited extractability of 14C-TCC by NaOH indicates only a small fraction of the potentially bound fraction was attributable to T CC incorporation into the humic fraction. The apparent rapid irreversible sorption of 14C-TCC to the silty clay loam (~30% in combusted fraction even at T0) is hypothesized to represent conversion of TCC in biosolids extractable by H2O and MeOH to recalcitrant forms associated with the co lloidal fraction in the amended soil. In all soils, recovery by combustion appeared to increase with time, suggesting the formation of bound residues. Combustion and 14CO2 recoveries for the silty clay loam were consistently (and significantly) greater than fo r the sand, and suggested that the microbial communities in the silty clay loam might be better equipped to degrade the fraction of bioavailable 14C-TCC. Results of the biosolids-borne 14C-TCC biodegradation study shar e some similarities with the Ying et al. (2007) and Xia et al. (2008) data, but also di ffer in some important ways. The 146

PAGE 147

lack of radiolabel mineralization in the inhibited biotic treatments is consistent with the findings of Ying et al. (2007), but is in contrast to th e 30% and 26% loss in the two sterile soils assessed by Xia et al. (2008). It is hypothesized that th e sterile soils were contaminated during incubation, as supported by the only slightly grea ter TCC losses reported in the non-sterile soils (47% and 29%, respectively) (Xia et al., 2008). The most significant difference between the findings reported herein and the Ying et al. (2007) and Xi a et al. (2008) data is the extent of degradation. Between 26 and 41% of TCC was re ported degraded in the Ying et al. (2007) and Xia et al. (2008) studies, but only 24% was mineralized in the current 14C-TCC biodegradation experiment. One explanation for the discrepancy is the measure by which TCC loss was quantified in the three st udies. Ying et al. (2007) and Xia et al. (2008) quantified degradation as loss of the extractable parent compound, whic h could occur through primary and/or ultimate degradation. The 14C-TCC study, however, quantified only ultimate degradation by measuring 14CO2 evolution. Although no radiolabeled metabolites were identified in the MeOH extracts, it is possible that 14C-TCC degradation products were presen t in the other extr act or combusted fractions. The presence of 14C-TCC metabolites in the non-MeOH fractions would indicate more degradation than first calculated, a nd half-lives that might be closer to those predicted by Ying et al. (2007) and Xia et al. (2008). However, the similarities in sequential extraction 14C recoveries across biotic and inhibited-biot ic treatments within soils suggest the calculated 2% and 4% 14CTCC degradation is accurate. It seems unlikely th at metabolites formed in the biotic treatments would be sequentially extracted in a pattern nearly identical to the sequential extrac tion of parent compound in the inhibited-biotic treatments (Figures 8-2). Another possible explanation fo r the differences in measured TCC degradation is the use of spiked soils and spiked biosolids amended to soils. The greatest degradation reported by Ying 147

PAGE 148

et al. (2007) and Xia et al. (2008) occurred in spiked, non-am ended soils. Significantly less degradation occurred in spiked biosolids-amended soils (~13%) (Xia et al., 2008). Sorption of TCC spike to the organic biosolids matrix likely reduced bioavailability, and thus, biodegradation. The 13% degradation in biosolid s-amended soils reported by Xia et al. (2008) is only 9-11% greater than the de gradation reported herein. Th e difference could be easily explained by soil texture differences, or variations in the ability of soil communities to degrade TCC, as evidenced by the 18% difference in TCC degradation in the sandy loam and fine sandy loam used by Xia et al. (2008). Confirmation of Biotic Conditions The increasing cumulative CO2 evolution from biotic samples with time confirmed that the addition of TCC did not sterilize the system, nor did it depress total resp iration (Figure 8-3). However, the impact on soil microbial community structure could not be assessed, as base traps from the non-spiked control samples were comp romised (due to contamination by ambient air during prolonged storage) before total CO2 analyses could be performed. The circles in Figure 8-3 indica te the first base trap sample d following the second biosolids addition. As expected, CO2 production increased with the second addition of organic material. Biotic sample CO2 evolution rates throughout the study were comparable to CO2 evolution rates measured from poultry litter amended-soils (~0.5 mg CO2 g-1 soil) (Khalil et al., 2005), but were less than rates measured from a bioso lids-amended sandy clay loam (~9 mg CO2 evolved 1 g-1 soil) (Franco-Hernandez and Dendooven, 2006). Th e biosolids used in the Franco-Hernandez and Dendooven study, however, were processed in a bioreactor and treated with slake-lime (biosolids pH = 12). Lime treatment can solubilize organic matter and increase bioavailability, leading to increased mineralization rates (Kemmitt et al., 2006). 148

PAGE 149

Despite the addition of NaN3 to inhibited-biotic samples at T0 and again at the time of biosolids re-application, microbial activity continued in the inhibited-biotic samples throughout the course of the study, albeit at a reduced rate compared to the biotic samples. Extract Speciation Figures 8-4a (bottom) illustrates a typical RAD-TLC chromatograph for the 14C-TCC standard. Figure 8-4b (bottom) illustrates a typical RAD-TLC chromatograph for MeOH extracts obtained from both soils during T0 -T7 sampling periods. Chromatographs of MeOH extracts of both soils from T0-T7 were sim ilar, indicating no change in the composition of MeOH extracts with time, i.e. no MeOH-extractable degradation products formed. Distinct regions of radioactivity above background are nu mbered in Figure 8-4. Regions 3, 4, 7, and 9 (i.e. the origin) in Figures 8-4a and 8-4b are clear in all chromatographs of stock radiolabel solution and of MeOH extracts from T0-T7. The additional peaks present in the MeOH extracts are consistent from T0-T7, suggesti ng regions not associated with pure 14C-TCC are caused by complexes with the biosolids matrix formed immediately following spike addition or by impurities in the spike itself, and not by 14C-TCC break-down products during the course of the study. Estimation of TCC Half-Life in Soil Figures 8-5 and 8-6 illustrate the percent of radiolabel remaining in the biotic biosolidsamended soil samples with time, and reflect the extremely slow mineralization of the 14C-TCC spike. A one-phase linear model was fit to the da ta, resulting in estimated TCC half-lives of 20 y and 8 y in the sand and silty clay loam, respectively. However, because the 7.5 month degradation experiment did not r each even the first TCC half-life, predicting TCC half-life in the biosolids-amended soils is problematic and the estim ated half-lives should be interpreted as first approximations. An exponential model could also be fit to the data (w ith no change in the R2 149

PAGE 150

values), resulting in 14C-TCC half-lives of 26 and 6.6 y in the sand and silty clay loam, respectively. The half-life of biosolids-borne TCC in amended soils might be even less than predicted if anaerobic degradation occurs in non-aerated microor macro-sites. The study described herein, however, scrubbed CO2 from the incoming air and could have prevented the conditions necessary for anaerobic degradation to proceed. Additional research using adapted OPPTS guidelines (i.e. 835.3400 Anaerobic Biodeg radability of Organic Chemicals; 835.3420 Anaerobic Biodegradability of Organic Compounds in Digested Sludge; 835.5154 Anaerobic Biodegradation in the Subsurface) would be requir ed to characterize the influence of anaerobic conditions on TCC loss in biosolids-amended soils It should be noted, however, that minimal TCC degradation has been measured in anaerobi cally digested biosolid s (Heidler et al., 2006). The data collected here might represent such early stages in the TCC degradation process (no more than 4% degraded over 7.5 months) that the true shape of the degradation curve cannot be accurately predicted. Although a linear, first order regression lin e was fit to the data, a curve that extends to a time period >>7.5 months mi ght be better characterized with a more complicated model(s). Further, given the purity of the radiolabel (98.5%), cross-contamination of radioactivity into control base traps (~1/4 1/2 of evolved 14CO2), and the variation in total percent recoveries following seque ntial extraction, trends in degr adation could be artifacts of measurement variability. Of greater importance, may be the relative persistence measured vs. half-life values generated by others. The half-lives estimated from the biodegradatio n experiment are considerably greater than the 120 d soil half-life predicted by Halden and Paull (2005) using the QSAR analyses software EPA PBT Profiler, and the weeks to months predicted by BIOWIN in EPI Suite (which incorporates the same estimation software p ackages as PBT Profile r). BIOWIN and PBT 150

PAGE 151

Profiler estimate half-lives using a fragment constant appro ach, by which the structure of a compound is used to assess degradability based on a dataset of previously analyzed compounds. Given the research findings of Gouin et al. (2004) and Aronson et al. (2006), the difference between the laboratory-based and computer model prediction is not surp rising. Both groups of researchers compared half-life estimates based on laboratoryand field-based measurements for a variety of persistent organic pollutants (POP s) to predictions calculated in the estimation software programs PBT Profiler and BIOWIN. Both studies concluded that POP half-life predictions by the computer models are significantly shorter than estimates based on laboratory or field studies. Gouin et al. (2004) found BIOWIN estimated half -lives 100-1000 times less than measured values for multiple polychlorinated biphenyls (PCBs) and polychlorinated dibenzo-p-dioxins and furans (PCCD/Fs). The authors argue that the use of multiple universal extrapolation factors, and the failure to in corporate the influence of partitioning and bioavailability, in calculating half-lives can und erestimate the persistence of some compounds. In the case of TCC, the 120 d predicted half-l ife is 60 times less than the maximum half-life estimated from the laboratory study herein. The influence of the measured TCC solubility and Kow values on the half-lives predicted by PBT Profiler and BIOWIN cannot be assessed, as the programs do not permit manual entry of the physicochemical characteristics. According to the PBT Profiler, TCC is persi stent, bioaccumulativ e, and toxic. If TCC were a new chemical entering the market today, the persistent label would allow the compound to be commercialized un der the current regulations, but Toxic Release Inventory-type reports on environmental releases would be requir ed, and specific limits on exposures, releases, and uses would be applied. However, the ver y persistent classifica tion suggested by the measured data herein would completely ban use of the compound, pending additional testing. 151

PAGE 152

The differences between the software-based a nd laboratory-based chemi cal property/behavior estimates illustrate the value of measured data for accurately estimating human and environmental exposure risks and regulatory impact. 152

PAGE 153

Table 8-1. Select physicochemical properties of the soils used in the biosolids-borne triclocarban (TCC) study Immokalee sand Genesee silty clay loam pH 4.5 6.4 Electrical conductivity (S m-1) 6.6 9.9 Organic carbon (g kg-1) 5.5 23 Sand (%) 99 9 Silt (%) <1 59 Clay (%) <1 33 153

PAGE 154

154 Table 8-2. Biodegradation e xperiment sample treatments Soil Spike (dpm) Treatment Abbreviation Replicates Total Samples Florida 1.3 x 106 Biotic BSF 32 Florida 1.3 x 106 Inhibitedbiotic ASF 32 Ohio 1.3 x 106 Biotic BSO 32 Ohio 1.3 x 106 Inhibitedbiotic ASO 32 Acidified samples Florida 1.3 x 106 Biotic acidification BSF 3 Ohio 1.3 x 106 Biotic acidification BSO 3 Controls Ohio No spike Biotic BNO 2 Ohio No spike Inhibitedbiotic ANO 2 Florida No spike Biotic BNF 2 Florida No spike Inhibitedbiotic ANF 2 Soil-less Florida 1.3 x 106 Autoclaved water SSWF 2 Soil-less Ohio 1.3 x 106 Autoclaved water SSWO 2 TOTAL 146

PAGE 155

Table 8-3. Radiolabel percent r ecoveries in the fine sand soil Percent recovery in fine sand soil Fraction T0 Biotic T0 Inhibitedbiotic T1 Biotic (3 wks) T1 Inhibitedbiotic (3 wks) T2 Biotic (6 wks) T2 Inhibitedbiotic (6 wks) T3 Biotic (9 wks) T3 Inhibitedbiotic (9 wks) T4 Biotic (12 wks) T4 Inhibitedbiotic (12 wks) T5 Biotic (16 wks) T5 Inhibitedbiotic (16 wks) T6 Biotic (24 wks) T6 Inhibitedbiotic (24 wks) T7 Biotic (30 wks) T7 Inhibitedbiotic (30 wks) H20 14 + 3 11 + 1 4 + 0.9 7 + 0.3 6 + 2 14 + 1 2 + 0.7 7 + 0.5 3 + 0.6 7.2 + 0.4 2 + 0.5 8 + 4 1 + 0.2 12 + 0.3 3 + 0.3 9 + 2 MeOH 74 + 16 81 + 8 73 + 4 87 + 3 68 + 7 87 + 5 56 + 9 79 + 2 72 + 2 90 + 3 75 + 6 87 + 9 64 + 2 70 + 2 63 + 6 80 + 3 NaOH 2 + 0.3 4 + 1 5 + .7 6 + 0.5 1 + 0.3 1 + 0.5 1 + 0.5 2 + 0.7 1 + 0.2 0.4 + 0.3 2 + 0.3 2 + 1 3 + 1 3 + 2 5 + 2 4 + 0.3 CO2 0 0 1 + 0.6 0.8 + 0.1 2 + 0.1 0.5 + 0.1 1 + .04 0.4 + 0.02 2 + 0.3 0.4 + 0.01 3 + 0.1 0.7 + 0.1 2 + 0.4 0.8 + .04 3 + 0.4 0.9 + .03 Combustion 6 + 1 6 + 0.3 9 + 0.2 10 + 0.6 12 + 2 11 + 0.5 20 + 2 19 + 2 15 + 2 13 + 1 16 + 1 14 + 1 21 + 1 15 + 2 18 + 6 22 + .08 Total 96 + 16 102 + 7 93 + 5 111 + 4 88 + 9 114 + 4 80 + 12 107 + 4 93 + 3 110 + 10 97 + 7 111 + 6 92 + 1 98 + 5 91 + 6 115 + 4 155

PAGE 156

156 Percent recovery in silty clay loam soil Fraction T0 Biotic T0 Inhibitedbiotic T1 Biotic (3 wks) T1 Inhibitedbiotic (3 wks) T2 Biotic (6 wks) T2 Inhibitedbiotic (6 wks) T3 Biotic (9 wks) T3 Inhibitedbiotic (9 wks) T4 Biotic (12 wks) T4 Inhibitedbiotic (12 wks) T5 Biotic (16 wks) T5 Inhibitedbiotic (16 wks) T6 Biotic (24 wks) T6 Inhibitedbiotic (24 wks) T7 Biotic (30 wks) T7 Inhibitedbiotic (30 wks) H20 3 + 0.6 3 + 0.9 0.7 + 0.2 2 + 0.6 0.9 + 0.1 2 + 0.1 0.8 + 0.1 1 + 0.1 0.6 + 0.1 1 + 0.1 0.8 + 0.1 2 + 0.4 0.8 + 0.1 2 + 0.5 0.7 + 0.1 2 + 0.5 MeOH 56 + 3 73 + 7 68 + 2 57 + 3 59 + 41 50 + 6 50 + 1 47 + 2 48 + 5 51 + 3 58 + 4 47 + 19 42 + 21 45 + 7 52 + 0.6 37 + 11 NaOH 3 + 3 4 + 3 3 + 0.7 4 + 1 3 + 0.4 5 + 1 0.5 + 0.4 0 0.4 0 0.5 + 0.8 .03 + 0.1 3 + 0.2 0.3 + 0.3 1 + 2 2 + 0.4 CO2 0 0 2 + 0.1 0.8 + 0.3 2 + 0.0 0.6 + 0.1 2 + 0.4 0.4 + .04 2 + 0.7 0.5 + 0.1 2 + 0.04 0.6 + .01 4 + 0.1 09 + 0.2 5 + 0.6 1 + 0.3 Combustion 36 + 4 40 + 3 44 + 7 54 + 4 46 + 3 52 + 3 46 + 3 54 + 7 50 + 4 46 + 2 47 + 4 57 + 9 80 + 25 63 + 14 56 + 26 47 + 9 Total 98 + 4 120 + 6 118 + 8 118 + 6 111 + 2 110 + 2 99 + 5 102 + 6 101 + 3 99 + 3 106 + 2 115 + 7 116 + 1 110 + 10 104 + 17 98 + 12 Table 8-4. Radiolabel percent recove ries in the silty clay loam soil

PAGE 157

air flow Multi-sample rack Air pump H2O 1 M KOH air flow Position-1 Positi on-2 Position-3 Amended soil sample Base traps Figure 8-1. Diagram of biodegradation experiment design 157

PAGE 158

T0 (0 wks) T1 (3 wks) T2 (6 wks) T3 (9 wks) T4 (12 wks ) T5 (16 wks ) T6 (24 wks ) T7 (30 wks ) Biotic sand 80 100 0 20 40 60 H20MeOHNaOHCO2Comb. Biotic silty clay loam 0 20 40 60 80 100 H20MeOHNaOHCO2Comb. Inhibited-biotic sandy clay loam 0 20 40 60 80 100 H20MeOHNaOHCO2Comb. Inhibited-biotic sand 80 100 0 20 40 60 H20MeOHNaOHCO2Comb. Figure 8-2. Radiolabeled tricloca rban (TCC) spike recoveries as a function of treatment, fraction (water, methanol, sodium hydroxide, 14carbon dioxide, and combusted), and time (normalized for 100% total spike recovery) Fraction Percent recover y a. b. d. c. 158

PAGE 159

0 20 40 60 80 100 120 140 160147101316192225 Week NumberCumulative CO 2 (mg) Biotic silty clay loam Biotic find sand Inhibited-biotic silty clay loam Inhibited-biotic fine sand First week following 2nd biosolids addition Figure 8-3. Cumulative CO2 production by biosolids-amended soil samples in the 14Ctriclocarban (14C-TCC) biodegradation experiment 159

PAGE 160

160 b. a. Figure 8-4. Radio-thin layer chromatography (RAD-TLC) standard (a.) and sample (b.; representative of T0-T7) chromatogr aphs (bottom) and corresponding RAD-TLC fingerprints (top)

PAGE 161

y = -0.047x + 99.257 R2 = 0.6316 95 96 97 98 99 100 05101520253035 Time (weeks)Percent radiolabel remainin Figure 8-5. Percent radi olabel remaining (100%-%14CO2) in biosolids-amended sand over time y = -0.1183x + 99.455 R2 = 0.875 95 96 97 98 99 100 05101520253035 Time (weeks)Percent radiolabel remainin y = -0.12x + 99.455 Figure 8-6. Percent radi olabel remaining (100%-%14CO2) in biosolids-amended silty clay loam over time 161

PAGE 162

CHAPTER 9 RISK ASSESSMENT OF BIOSOLIDS-BORNE TCC Introduction Independently assessing human and ecological health is a common approach to risk assessment. Human and ecological health are as sessed and protected independently, and overlap occurs only when threats to human health are environmentally mediated or transmitted (Suter, 2007). Justifications for independent assessments can include differing data needs, methods (e.g. epidemiological versus modeling), acceptable leve ls of uncertainty, and levels of perceived importance (Suter et al., 2000). A second approac h, integrated risk assessment, responds to the notion that human and ecological health are in co ntinual, dynamic associ ation, and that adverse health impacts from a single hazard can cy cle between the two s pheres (Suter, 2007). The World Health Organization (WHO, 2001) pub lished a framework for integrated risk assessment and cited coherence of assessment results interdependence, quality, and efficiency as arguments for the approach. When a single deci sion must be made regarding an environmental hazard that can affect both human and environmenta l health, consideration of both types of risk assessment data is required. However, the re sults of independent human and ecological risk assessments are often inconsistent in terms of spatial and temporal scales, variance and uncertainty, degrees of conservatism, and a ssumptions, and makes weighing and prioritizing risks difficult or impossible (Suter, 2007; WHO, 2001). Integration of risk assessments use consistent parameters can provide a much strong er basis for decision making, as compared to a piecemeal approach that use independently produ ced data sets. The integrated data set recognizes the interdependence of human and eco logical health, and iden tifies modes of action that affect both human and non-human receptors. The mechanisms of environmental contaminant release, transport, transformation, and exposure affect all receptors, so sharing 162

PAGE 163

resources, information, and techniques in an inte grated approach can improve the quality and the efficiency of a comprehensive risk assessment (Suter, 2007; WHO, 2001). The integrated risk assessment framewor k devised by WHO (2001) evolved from the USEPA ecological risk assessment (ERA) mode l (USEPA 1992, 1998b) that evolved from the National Research Council (NRC, 1983) framewor k for human health risk assessment The three frameworks share many common step s or components, a lthough the names and organization of objectives differ (Table 9-1). The Hazard identification and Planning steps in the human and ecological risk assessments, resp ectively, are combined to create the first step (i.e. Problem formulation) of the integrated he alth risk assessment (Fi gure 9-1) (WHO, 2001). The second (i.e. Dose-response assessment and P roblem formulation, respectively) and third (i.e. Exposure assessment and Analysis, respec tively) steps of the hu man and ecological risk assessments are integrated to create the single second step (i.e. Analysi s) of the integrated assessment. The third and final step of the integrated assessment, Risk characterization, combines the risk characterization processes in the human and ecological assessments. Where applicable, stakeholder participation and risk mana gement run parallel to the entire integrated risk assessment process and can occur at any time it is appropriate (WHO, 2001). The 10-y (1983-1993) biosolids risk assessment for the EPA Part 503 Rule (1995) applied an integrated framework to establish biosolids ma nagement practices, pollutant limits for select metals, and pathogen elimination standards de signed to protect both human and ecological health. The assessment initiated the first Na tional Sewage Sludge Survey and addressed 200 pollutants of concern (including multiple or ganic compounds). The integrated human and ecological biosolids-borne TCC risk assessment used herein includes many of the parameters and assumptions used in the development of the Part 503 Rule and coupled them with the integrated 163

PAGE 164

risk assessment framework developed by WHO (2001). Some assumptions and parameters (e.g. identification/definition of the hi ghly exposed individual, and target endpoints) were modified to increase the scope and/or rele vancy of the risk assessment. Integrated Risk Assessment Assessment Step 1: Problem Formulation The grant proposal submitted to the USEPA Office of Wast ewater Management (2005) that supported our research fulfilled the first component of problem formulation by identifying the goals, objectives, and activities of the biosolids-borne TCC risk assessment. The second component was addressed by the scope of the pr oject: identification of assessment endpoints and development of a conceptual model. Assessm ent endpoints were the species (human and nonhuman), life stages, and responses (e.g. no-observable-effect) releva nt to the risk assessment (Suter, 2007). The ultimate goal of performing the land-applied biosolids TCC integrated risk assessment is to identify bios olids-borne TCC limits to guid e sustainable (long-term) landapplication practices. Therefore, the most appropriate and sensitive endpoints for adult and child humans were those for chronic (i.e. long term; 10% of human life span; 90 d 2 y in laboratory animals; IRIS, 2009) exposures to biosolids-borne TCC. Adverse health imp acts associated with chronic exposures to a contaminan t typically occur at lower environmental concentrations than adverse health impacts associated with an acute exposure(s). Thus, biosolids-borne TCC regulations based on chronic exposur e toxicity data should also be protective of health impacts due to acute biosolids-borne TCC exposures. The most appr opriate response endpoints were, thus, the chronic exposure no-observable-adve rse-effect-level (NOAEL) and the lowestobservable-adverse-eff ect-level (LOAEL). Alternatively, both acute and chronic exposur e endpoints can be relevant for non-human receptors, as wildlife is often much more suscepti ble than humans to xenobiotic exposure (Suter, 164

PAGE 165

2007) Domesticated and wild anim al species data were selected to best represent the animal endpoint for each relevant pathway. Beef cattle pigs, poultry, and fish were selected as components of exposure pathways concluding with human consumption of animal meat, as the animals produce the most commonly consumed meat/fish products in the US (USEPA, 1997). Specific wild animal species were selected based on exposure behaviors relevant to the biosolids-borne TCC risk assessment, and on data availability. The cow ( Bos primigenius ), short-tailed shrew ( Blarina brevicauda ), American woodcock (bird; Scolopax minor) were selected as terrestrial animal endpoints exposed to TCC via soil consumption. The short-tailed shrew and the American woodcock were also sele cted as predators of soil-dwelling organisms (i.e. earthworms). The cow ( Bos primigenius ) and the herbivorous Ea stern cottontail rabbit ( Sylvilagus floridanus) were selected as the animal endpoi nts for plant consumption. A variety of aquatic indicator organisms, osprey ( Pandion haliaetus ), and the river otter ( Lutra canadensis ) were selected as aquatic species endpoints. Osprey and the ri ver otter are good indicators of bioaccumulation potential, given their piscivorous diet and high trophic level (USEPA, 1993c). The biosolids-borne TCC conceptual exposur e model delineates pote ntial relationships between the point of biosolid s-borne TCC environmental entry and the identified exposure receptors. The relationships between the assess ment endpoints and the source of TCC were used to inform decisions made in the subsequent Analysis phase. The conceptual model designed for the biosolids-borne TCC integrated risk assessm ent (Figure 9-2) was adapted from the site conceptual model applied to exposure analysis for dioxins, dibenzofurans, and coplanar polychlorinated biphenyls in sewage sludge (USEPA, 2002) and land-applied biosolids exposure pathways outlined in the Part 503 Biosolids Rule (1995). The model is specific to human and ecological exposures following land application of biosolids. The 14 exposure pathways used in 165

PAGE 166

the risk assessment process for the Part 503 Bioso lids Rule were included in the biosolids-borne TCC conceptual model (Table 9-2). The origin al pathway numbers are unchanged, to allow for easy cross reference to the Part 503 Biosolids Rule documents. Two new pathways (identified as Pathways 15 and 16) were included to incorporat e terrestrial animal exposures to contaminated fish, and aquatic organism exposure to contamin ated surface water. The receptor for each pathway was a highly exposed individual (HEI), defined using a combination of high-end and mid-range assumptions. The HEI is considered to be representative of the subset of the population with the greatest, but rea listic, exposures, as compared to the most exposed individual (MEI), who is unrealistic and does not actually exis t (Epstein, 2003). Assessment Step 2: Analysis The Analysis step was divided into four components: Characterization of effects, Reference dose calculation, Screening-level exposure concentration calculation, and Screeninglevel hazard index calculation. Characterization of effects supporting studies Recent publications (Halden and Paull, 2004; Heid ler et al., 2006) cited the ability of TCC to induce methemoglobinemia in exposed humans as a cause for concern over the presence of TCC in the environment. The two most commonly cited outbreaks of TCC-induced methemglobinemia were due to either direct appl ication of a TCC soap solution onto the skin of newborn babies, or the administ ration of heated soap-sud enemas to adult hospital patients (Johnson et al., 1963). The heating of the soap gel produces a primary aromatic amine, a class of compounds known to be capable of induc ing methemoglobinemia (Johnson et al., 1963). Another outbreak of methemoglobinemia occurred in infants following the use of an antibacterial diaper rinse containing 3.25% TCC (Johnson et al., 1963). 166

PAGE 167

Biosolids-borne TCC-induced methemglobinemia due to dermal exposures is not of great concern, as direct application of biosolids to infant skin and in the delivery of enemas are not realistic exposure pathways, and mean TCC con centrations in biosolids are at least 450x less than the typical TCC concentration in antibacterial bar so ap (1.5%; TCC Consortium, 2002b), and at least 1000x less than the TCC concentrati on reported in diaper rinse (3.25% by Johnson et al. (1963). However, because human metabolism of TCC absorbed via the dermal and oral routes is similar (SCCP, 2004), ingestion of T CC is a potential, but un likely, exposure pathway resulting in methemglobinemia. Characterization of effects human and animal studies An assembly of TCC toxicity data relevant to humans and other terr estrial vertebrates is available in the European Co mmission Scientific Committee on Consumer Products Opinion on Triclocarban for Uses Other than a Preser vative (2004). Included are various endpoint results for acute and chronic oral and dermal/m embrane exposures, such as oral toxicity LD50s (lethal dose to 50% of test popul ation), subchronic ( up to 10% of human lif e span; 30-90 d in laboratory animals; IRIS, 2009) toxicity NOAELs and LOAELs, chronic no-observable-effectlevels (NOELs) and lowest-observable-effect-levels (LOELs), mutagenicity, clastogenicity, carcinogenicity, and reproductive an d developmental toxicity NOELs and LOELs (Table 9-3). There is no evidence indicating TCC is a carcinog en, mutagen, or clastogen (i.e. an agent that causes breaks in chromosomes). The human toxicity data are limited to derm al patch exposures with skin irritation and sensitivity endpoints. The majority of studies re vealed no skin irritation or sensitivity to TCC concentrations up to 10% in the delivery medium (i.e. soap or petrolatum). In two studies, slight to very mild skin irri tation occurred in select par ticipants following exposures to < 1.5% TCC. 167

PAGE 168

The majority of toxicity studies we re performed on Sprague-Dawley rats (Rattus norvegicus ) and NMRI (Swiss) mice ( Mus musculus ). Two acute toxicity studies (Marty and Wepierre, 1979 and Bayer AG, 1991, in SCCP, 2004) administered 5000 or 2000 mg TCC kg bw-1 (neat product) to mice and rats, respectivel y. No mortalities o ccurred over 7-14 d of observation, and there were no ot her signs of toxicity. The LD50 was set to >5000 mg kg bw-1 for mice, and >2000 mg kg bw-1 for rats.. Data are also available for multiple subchronic or chronic dietary feeding studies with rats. The lowest concentration at which no observable effects (NOEL) were detected was 25 mg TCC kg bw-1 d-1, which was administered to rats in dietary feed for two y (Monsanto, 1981 in SCCP, 2004). Reference dose (RfD) calculation Reference doses (RfD) for humans, animals, and aquatic organisms were calculated or identified from the literature. The USEPA ( 1993d) defines the human reference dose as an estimate of a daily exposure to the human popul ation (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects dur ing a lifetime, a nd is typically expressed in units of milligrams per kilogram of bodyweight per d (mg kg bw-1 d-1). Reference doses can also be calculated for non-human endpoints. The RfD is calculated as: RfD = NOAEL / (UF x MF) (9-1) where: UF is one or more uncertainty factors, and MF is a modifying factor based on professional judgme nt (default = 1) For human RfD calculations, ten-fold uncertain ty factors are applied when extrapolating animal study data to humans (10a), when extrapolating prolonge d exposure data to average healthy humans (10b) (accounts for variations in sensitiv ity among people), when extrapolating 168

PAGE 169

less than chronic results in animal s to chronic human exposures (10s), and when deriving an RfD from a LOAEL rather than a NOAEL (10l) (USEPA, 1993d). The NOAEL (or LOAEL) is typically selected over the NOEL (or LOEL), as the latter does not specif y if the effects of interest were adverse (USEPA, 1993d). The NOA EL determined in a subchronic (8 w) dietary feeding rat study conducted by Monsanto (NOAEL = 75 mg kg bw-1 d-1; 1985) first appears to be the most appropriate endpoint for calculating th e RfD, but the lack of a control group, histology of tissues, and blood chemistry an alyses critically weakened the credibility of the results. Alternatively, a chronic (2 y) dietary feeding rat study conducted by Monsanto (1981) resulted in a NOEL of 25 mg kg bw-1 d-1. The observed effects at greater doses in the study (anemia, reduced body weight, and increased liver and spleen weights), although not labeled as such in the report, can be reasonably considered advers e. The RfD for humans was calculated as: RfD = (25 mg kg bw-1 d-1) / (10a x 10l) = 0.25 mg kg bw-1 d-1 (9-2) A separate human RfD for methemglobinemi a was estimated using data supplied by Johnson et al. (1963) and the SCCP (2004). Me themglobinemia was induced in infants at a hospital following exposure to a soap solution containing TCC (Johnson et al., 1963). The TCC concentration in the soap so lution was unknown, but was 3.25% in diaper rinse that led to a second outbreak during the same time period (Johnson et al., 1963). The SCCP (2004) calculated 0.51 mg TCC is absorbed by adults over a single day of typical antibacterial product use (1.5% TCC content). Assuming that 0.51 mg T CC would also be absorbed by an infant, and adjusting for an assumed soap solution TCC c ontent of 3.25%, an infant would absorb 1.1 mg TCC over one day of exposure. Dividing 1.1 mg TCC by the average body weight of a 1-month169

PAGE 170

old infant (4.1 kg; USEPA, 1997) results in a methemglobinemia RfD of 0.27 mg TCC kg bw-1 d-1. The estimated methemglobinemia RfD was slightly greater than the chronic toxicity RfD, so the latter value (0.25 mg TCC kg bw-1 d-1) was selected as the more conservative, and thus more protective, dose for use in subsequent risk calculations. The RfD for terrestrial animals was calculated from the chronic rat study in a similar fashion, but incorporated uncerta inty factors specific for ecosy stem risk assessment (Suter, 2007). An uncertainty factor of 5 is applied if the test species and endpoi nt species of interest are in the same class but different orders (5a). Extrapolation from acute to chronic exposures, and differences in toxicity when extrapolating from NOELs to lethal effects, are accounted for with an uncertainty factor up to 15 (not required in the present ex ample). Differences in species sensitivity, laboratory-field extrapolation, and intr aspecific variability can each be accounted for with an uncertainty factor of two (2b, 2c, 2d, respectively) (Suter, 2007). The RfD for terrestrial animals was calculated as: RfD = (25 mg kg bw-1 d-1) / (5s x 2b x 2c x 2d) = 0.625 mg kg bw-1 d-1 (9-3) Many more toxicity data are av ailable for aquatic species than for terrestrial organisms, and have been assembled elsewhere (TCC Cons ortium, 2002; Chalew and Halden, 2009). The lowest toxicity endpoints for each category of aqua tic organism (i.e. fish, crustaceans, and algae) (Table 9-4) were selected as RfDs for the su rface water component of the biosolids-borne TCC risk assessment. Screening-level exposure concentration calculation A tiered approach was applied to calculations of exposure concentrations. Simple and conservative estimates were used to first calcula te screening-level exposure concentrations for 170

PAGE 171

each exposure pathway in the conceptual model (Tab les 9-6 and 9-7). If the resulting calculated risk was low (i.e. HI<1; see below), additional assessment was not performed. High risk exposure calculations prompted higher tier eval uations (e.g. refine environmental concentrations and fate parameters) and reassessment of risk. The basic concept for calculating TCC exposur es involved multiplying the concentration in the relevant environmental medium by the corresponding human/animal intake rate. Exposures to fish and aquatic organisms were taken as 100% of the surface water concentration. Where possible, TCC exposure concentration calculations used measured environmental data available in the literature, or were predic ted using the laboratory results pres ented in the previous chapters. Structure activity relationships (SARs) develo ped from studies of multiple organic compounds were used in conjunction with measured TCC solubility and Kow values to estimate bioaccumulation factors when measured enviro nmental concentrations were missing. The screening-level exposure concentrations assumed the same worst-case (50 Mg ha-1, one-time application) and -year (5 Mg ha-1, annual applications for 100 y) biosolids land-application scenarios applied in the Part 503 Biosolids Rule risk assessment (NRC, 2002). Typical agronomic land-application rates, ho wever, are closer to 5-10 Mg ha-1 y-1 (OConnor et al., 2001) and are often restricted by regula tions limiting nutrient loads. Further, land-application at a single location is unlikely to o ccur for 100 consecutive y. Thus, for exposure pathways in which the screening-level risk calculati ons indicated a potential hazard to human or ecological health, the default loading rates were reexamined. The Part 503 Biosolids Rule risk assessment also used 95th percentile contaminant concentrations in the screening-level exposure concentration calculations. The 95th percentile concentration from the TNSSS study (assuming a lognormal distributi on) was 131 mg TCC kg-1 171

PAGE 172

d.w. (USEPA, 2009a). The worst-case, screenin g-level surface water concentration was the greatest concentration meas ured during a survey of US streams (6750 ng L-1; Halden and Paull, 2005), likely the result of raw sewage overflow, rather than legal and sustainable landapplication of biosolids. As with the biosolids land-applicat ion scenarios, the surface water concentration assumption was reexamined for each pathway in which the screening-level risk calculations indicated a po tential hazard to human or ecological health. Screening-level hazard index calculation The hazard indices (HI) for each exposure pathway were calculated by dividing the screening-level exposure concentration by the RfD for the corresponding speci es endpoint (Table 9-7). The screening-level HI values for the ma jority of pathways (i.e. Pathways 1-9, 11-15) under both land-application scenario s were well below the critical level of one, suggesting that the corresponding HEIs would not be exposed to problematic concentrations of biosolids-borne TCC over a lifetime. Pathways in which the screen ing-level HI values were greater than one (i.e. Pathways 10 and 16) were selected for more detail ed risk evaluations and calculations of healthprotective biosolids loading rate s in the Risk characterization component of the biosolidsborne TCC assessment. Reassessment of Pathwa ys 10 and 16 followed a tiered approach in which estimates of TCC environmental concentrati ons and fate were refine d using relevant data and risk was recalculated. The need for addi tional risk estimate refinement using measured environmental concentrations is discussed in Chapter 10. Risk Characterization of Critical Pathways Pathway 10: Biosolids soil soil organism predator The worst-case scenario screening-level HI values for the short-tailed shrew (Blarina brevicauda ) and the American woodcock (Scolopax minor) were 10 and 7.6, respectively, and are 10-fold greater in the 100-year application scenario (assuming no TCC loss with time). If the 172

PAGE 173

biosolids application rate is adjusted to the typical ag ronomic rate of 5-10 Mg ha-1, the HI values for the shrew and the woodcock after a single biosolids application are significantly reduced to 1.0-2.1 and 0.76-1.5, respectively, but (with the exception of the lower bound for the woodcock) remain equal to or greater than the critical HI value of one. The HI values under a 100-year application scenario, assumi ng the most conservative halflife calculated in Chapter 6 (20 y) and the most conservative estimated percent leached in Chapter 4 (0.001%), can also be reduced from 100 to 16 (shrew) and 76 to 11 (woodcock), but remain well above one. The HI values can be further reduced if the 95th percentile TCC concentration in the biosolids charac terized in Chapter 3 (39 mg TCC kg-1) is applied, but given the potential overlap with the TNSSS data (i.e. some of the biosolids analy zed herein were split samples from the TNSSS study) and the small sample size, the approach does not seem appropriate. The assumption that 100% of the American w oodcock diet consists of earthworms was not reduced to 33% as assumed in the Part 503 Rule risk assessment. Various studies of American woodcock feeding habits indicate that the mean percent of earth worms in the diet ranges from 58-99% (USEPA, 1993c). Using a mean percent in take less than 100% no longer constitutes an HEI. The home range of the American woodc ock is 3.1-74 ha, and the birds frequent open pastures and fields to feed (USE PA, 1993c), so it is feasible that the entire diet could consist of earthworms growing in biosolids-amended soil. True bioavailability of earthworm-borne TCC might not be 100%, but no species-specific data exis t to prove otherwise. Bioavailability to the American woodcock would have to be <13% to reduce the worst-case scenario HI to <1. Alternatively, bioavailability would have to be < 67% at the agronomic bios olids application rate, 173

PAGE 174

or <9% in the 100-year scenario ad justed for TCC loss to reduce the HI values to <1. To date, there are no reports linking adverse e ffects in wildlife to TCC exposure. Much less than 100% of the s hort-tailed shrews diet consis ts of earthworms. Shrew ingestion habit studie s suggest a realistic range of 31-42% (USEPA, 1993c). However, a significant portion of the shrew di et can also include slugs and sn ails, which might be expected to bioaccumulate TCC to a degree similar to earth worms. The slug and snail fraction of the diet is 5-27% (USEPA, 1993c). The gr eatest earthworm/slug/snail comb ined percentage in a single shrew feeding study is 59% (USEPA, 1993c). Adjusting the worst-case, agronomic, and 100year (with TCC loss) scenarios for a reduced earthworm/slug/snail fraction of the diet results in HI values of 5.9, 0.59-1.2, and 2.8, respectively. As with the American woodcock, the bioavailability of earthwormborne TCC might not be 100%, but there are no species-specific studies that support the use of a reduced percentage. Bioavailabili ty to the short-tailed shrew would have to be <17% to reduce the worst-case scenario HI to <1. Alternatively, bioavailability would have to be <83% at the agronomic biosolids application ra te, or <36% in the 100-year scenario adjusted for TCC loss to reduce the HI values to <1. Bioavailability is influenced by the combined effects of digestion efficiency, absorption efficiency, and assimilation efficiency (Newman and Clements, 2007). The efficiencies, in turn, can be affected by chemical pKa and pH conditions in the various regions of the gut, gastric emptying rates, ingestion behavi ors, and/or in some cases, Kow (Newman and Unger, 2003). Uptake efficiency tends to declin e at a predictable rate at log Kow values >6, but appears to be independent of log Kow at values <6 (Gobas et al., 1989). Typical dietary assimilation of many halogenated hydrophobic organics is 40-70% (Newman and Unger, 2003), and a value of 50% has been proposed for general modeling purpos es (Opperhuizen and Sc hrap, 1988). However, 174

PAGE 175

the studies cited are based on fish assessments as are most studies characterizing total bioavailability of organic com pounds in wildlife. The USEPA Part 503 risk assessment also assumed 50% bioavailability for biosolids-borne PCBs. If a value of 50% is assu med for biosolids-borne TCC bioavailability following contaminated earthworm consumption by the American woodcock and the shrew, the adjusted HI values can be reduced once more. The fina l adjusted worst-case, agronomic, and 100-year scenario HI values for the American woodc ock are 3.8, 0.38-0.75, and 5.7. The final adjusted worst-case, agronomic, and 100-year scenario HI values for the short-tailed shrew are 3.0, 0.300.60, and 4.5 (Table 9-8). The adjusted HI values for the American woodcock and the shorttailed shrew suggest that curre nt, typical (i.e. 5-10 Mg ha-1) one-time land-application practices do not pose an ecological health risk. The calcula tion of HI values >1 under the worst-case and 100-year application scenarios, however, indicat e biosolids-borne TCC pollutant limits are needed to guide sustainable long-term land-application of biosolids (see Calculation of Pollutant Limits section). Pathway 16: Biosolids soil surface water aquatic organism An HI of one was exceeded for acute toxicity (NOEC) to Ceriodaphnia sp., chronic toxicity (NOEC, hatchability and survival) to Pimephales promelas and chronic toxicity (NOEC, reproduction) to Mysidopsis bahia. The HI values, however, were calculated using the greatest recorded TCC concentratio n in US surface waters (6750 ng L-1), which likely represents contamination from a raw sewage overflow rather than from run-off or sedimentation from biosolids-amended fields (Halden and Paull, 2005). The surface water concentration of 6750 ng L-1 falls within the range of undiluted TCC concentr ations in wastewater treatment plant effluent (240-12,000 ng L-1; TCC Consortium, 2002a), as opposed to other documented surface water 175

PAGE 176

concentrations (0.00001-190 ng L-1; TCC Consortium, 2002a), or maximum concentrations measured in biosolids-amended soil runoff (3.4 ng L-1; Topp, personal communication). When a surface water concentration of 190 ng TCC L-1 is used to calculate aquatic organism HI values, Mysidopsis bahia is the single organism with an HI >1 (3.17). The biosolids-amended soil runoff concentration (3.4 ng L-1; Topp, personal communication) provides a better approximation of TCC surface water concentrations attributable only to landapplied biosolids, and results in HI values well below one for all sensitive organisms. The concentration of TCC in wastewater treatment ef fluent and soil runoff will be diluted following entry into surface waters, and equilibrium con centrations will likely be reduced following sorption to sediment, resulting in further reduced HI values. The concentration of TCC in surface waters attributable only to land-application of biosolids under the worstcase and 100-year scenarios can also be calculated using a series of equations from the Part 503 Biosolids Rule risk assessment that account for sedimentation from biosolids-amended fields and TCC partitioning into the aqueous phase (Table 9-9). The concentration of TCC in soil eroding from amended land (Csma) is calculated as: Csma = (Pa fero) / MEsms CF (9-4) where: Pa = annual application rate of TCC (mg ha-1 y-1) fero = fraction of total TCC loss by erosion MEsma = estimated rate of soil loss for the SMA SMA = biosolids management area CF = unit conversion factor 176

PAGE 177

Under a worst-case erosion scenario within the worst-case or the 100year land-application of biosolids-borne TCC scenario, 100% of the er oded material is biosolids containing 131 mg TCC kg-1, resulting in a Csma of the same concentration. Only a fraction of eroded soil in a given watershed will originate from a biosolids management area as eroded biosolids will be dilu ted with non-contaminated soil prior to entering a surface water body. Multiplying Csma by an appropriate dilution factor (DF) will yield the concentration of TCC in sedi ment entering surface water (Csed). The DF is calculated as: DF = (Asma Ssma) / [(Asma Ssma) + ((Aws Asma) Sws)] (9-5) where: Asma = area affected by land application of biosolids (SMA) Ssma = sediment delivery ratio for the SMA Aws = area of the watershed (ha) Sws = sediment delivery ratio for the watershed The rates of soil erosion from the SMA and th e remainder of the watershed are assumed to be the same. The values assumed for Asma, Ssma, Aws, and Sws in the USEPA Part 503 risk assessment (1995) are 1074 ha, 0.46, 440,300 ha, and 0.17, respectively. The corresponding calculated DF is 0.0066 and Csed is 0.61 mg TCC kg-1. The expected surface water concentration (Csw) associated with the calculated Csed is estimated by: Csw = Csed / [KDsw + (Pl / Ps) (1/w)] (9-6) where: KDsw = partition coefficient between solid s and liquids with in the stream (assumes reversible partitioning) 177

PAGE 178

Pl = percent liquid in the water column Ps = percent solids in the water column w = density of water The KDsw is estimated as 2512 L kg-1, i.e. the average Kd of indigenous TCC in biosolids calculated in Chapter 3. The ratio of Pl to Ps used in the USEPA Part 503 risk assessment is 62,500, and the density of the surface water is assumed to be 1 kg L-1. The concentration of TCC surface water attributable to the erosion of land-applied bi osolids containing 131 mg TCC kg-1 is 1.3 x 10-5 mg TCC L-1. The adjusted aquatic organism hazard indices calculated using the calculated Csw in the worst-case biosolids application sc enario are all <1 (Table 9-10). Clearly, the major potential threats to the health of a quatic organisms by surface water contamination of TCC will be due primarily to wastewater treatm ent plant effluent, and more importantly, raw sewage overflows and leaking sewer lines (neither of which apply to sustainable and legal land application of biosolids). Calculation of Preliminary Biosolids-Borne TCC Pollutant Limits The USEPA Park 503 Biosolids Rule risk assess ment established four types of pollutant limits for land-applied biosolids: Cumulative po llutant loading rates (C PLRs), Annual pollutant loading rates (APLRs), Ceiling concentration lim its, and Pollutant concentration limits. Each limit addresses specific approaches and considerations involved in land-application of biosolids. The implications and limitations of the preliminary biosolidsborne TCC pollutant limits are discussed in Chapter 10. Cumulative pollutant loading rates (CPLRs) The TCC CPLRs are taken directly from ri sk assessment results and represent the maximum cumulative concentration of TCC in amended soils (kg pollutant ha-1) that remains protective of human and ecological health. A CP LR pertains to biosol ids land-applied in bulk, 178

PAGE 179

and can be calculated by setting the HI for a re levant exposure pathway to one and solving for the TCC concentration in the soil. Land-application of biosolids does not have to cease once the CPLR is reached at a given site, but biosolids containing the pollutant concentration limit (see below) can no longer be applied. The American woodcock is the more sensitive of the two soil organism predators assessed (due primarily to the fact that 100% of the woodcock diet is earthworms), and the TCC CPLR should be protective of the bird. The amount of TCC that can be land-applied without expectation of adverse effects to the American woodcock (i.e. the CPLR) can be calculated by setting the Pathway 10 HI equation to 1, adding a multiplication factor of 0.5 (to account for the assumed 50% bioavailability) and solvi ng for TCC concentration in the soil: 1 = SC BAFdw FI / BW / RfD 0.5 (9-7) where: SC = TCC concentration in soil (mg TCC kg amended soil-1) BAFdw = bioaccumulation factor in worm (d.w.) FI = food ingestion rate of predator (g d-1 d.w.) The calculated CPLR for the American woodcock is 0.56 mg TCC kg amended soil-1, or 1.2 kg TCC ha-1. Mysidopsis bahia is the most sensitive aquatic organi sm indicator species, with a chronic toxicity reproduction rate NOAEL of 0.00006 mg TCC L-1. The CPLR protective of Mysidopsis bahia can be calculated for surface water TCC con centrations resulting from erosion and from runoff. The most stringent limit will be protective of aquatic organism health in both exposure pathways. 179

PAGE 180

The CPLR under the erosion scenario can be ba ck-calculated using a target HI of one and Equations 9-5 through 9-7. The calcula ted CPLR is large (591 mg TCC kg-1 or 1300 kg ha-1) and suggests erosion of biosolids from amended soil poses little risk to aquatic organisms. A CPLR protective of Mysidopsis bahia through the surface water contamination via the land-applied biosolids runoff pathway was calcula ted using data from the single known study of TCC concentrations in biosolids-amended soil runoff (Topp, personal communication). The study was performed using dewatered biosolids inco rporated to a depth of 15 cm in a silt loam (7 slope) at an 8 Mg ha-1 rate. Triclocarban content of the biosolids was 8.2 mg kg-1, and the resulting soil concentration was ~0.030 mg TCC kg amended soil-1. The maximum TCC concentration in amended-soil runoff followi ng multiple rainfall simulations was 3.4 ng L-1. Assuming an increase in runoff concentration pr oportional to an increase in soil concentration, the amended-soil concentration required to resu lt in runoff reaching the value critical to Mysidopsis bahia (0.00006 mg L-1 or 60 ng L-1) is 0.53 mg TCC kg amended soil-1, or 1.2 kg TCC ha-1. Coincidentally, the CPLR protective of aquatic organisms is similar to the CPLR protective of terrestrial vertebrates. The calculated CPLRs translate to a one-tim e application of bios olids containing the TNSSS 95th percentile TCC concentration (131 mg TCC kg-1) at a rate of 8.9 Mg ha-1, or a onetime application of biosolids containing the TNSSS mean TCC concentration (39 mg TCC kg-1) at a rate of 30 Mg ha-1 (i.e. ~3-6 annual appl ications at a 5-10 Mg ha-1 rate). Annual pollutant loading rate (APLR) Annual pollutant loading rates (A PLRs) apply to biosolids that are sold or given away in a bag or container, and were established due to the impracticality of imposing CPLRs at small land-application sites (e .g. home gardens, home lawns, or public contact site). The APLR identifies the maximum mass of pollutant that can be land-applied in one y, and is calculated by 180

PAGE 181

dividing the CPLR by 20 (i.e. assumes 20 y of annual applica tions at a given site and no chemical loss with time). The TCC APLR protec tive of both terrestrial vertebrates and aquatic organisms is 0.6 kg TCC ha-1 y-1. The bag or container must be labeled with application instructions that ensure the biosolids are us ed properly and the APLR is not exceeded. Ceiling concentration limit The ceiling concentration limit identifies the maximum pollutant concentration allowable in land-applied biosolids, and prevents the lo west quality biosolids (in terms of pollutant concentration) from being land-applied. Ceiling limits are perceived by some risk assessors to increase public confidence in the safety of land-applied biosolids. A ceiling limit is defined by the Part 503 Rule as the least stringent of the 99th percentile pollutant concentration in biosolids, and the biosolids pollutant limit ca lculated from the risk assessment. The risk assessment-based soil concentration limit for the most sensitive sp ecies (the American woodcock) is 0.53 mg TCC kg soil-1, and translates to biosolids-borne TCC concentration limits of 23.32 and 15.4 mg TCC kg biosolids-1 in the worst-case and the TCC loss-adjusted 100-year application scenarios, respectively. Alternatively, the 99th percentile biosolids-borne T CC concentration calculated using the TCC contents in final products an alyzed in the TNSSS (2009a) is 277 mg TCC kg biosolids-1. The least stringent biosolids-borne TCC concentrati on, and thus, the designated preliminary ceiling concentration li mit, is 277 mg TCC kg biosolids-1 (d.w.). Pollutant concentration limit The pollutant concentration limit is the most st ringent biosolids-borne pollutant guideline. Biosolids meeting the pollutant concentration li mit (in combination with the Part 503 Rule pathogen and vector control requi rements) can be land-applied fr eely, without having to track pollutant loadings. The pollutant concentrati on limit assumes a total of 1000 Mg of biosolids will be land-applied to a given site over the course of 100 y or le ss, and is calculated by dividing 181

PAGE 182

182 the CPLR by 100 y of application at a 10 Mg ha-1 y-1 rate. The resulting risk-based pollutant concentration limit for biosolids-bor ne TCC is 1.2 mg TCC kg biosolids-1. Thus, the preliminary TCC concentration limit indicate s only biosolids containing < 1.2 mg TCC kg biosolids-1 can be land-applied as freely as any other type of fer tilizer or soil conditioner. The preliminary TCC concentration limit is significantly less than th e mean TCC concentration in biosolids assessed nationally (39 mg TC kg biosolids-1) (USEPA, 2009a).

PAGE 183

Table 9-1. Steps of human, ecological, a nd integrated health risk assessments Human health risk assessment (NRC, 1983) Ecological health risk assessment (USEPA, 2002, 2008; Suter, 2007) Integrated health risk assessment (WHO, 2001) Step 1 Hazard identification: determine whether exposure to an agent can cause an increase in the incidence of a health outcome Planning: consult with risk manager and stakeholders to identify management goals, management options, and scope/complexity of risk assessment Problem formulation: define goals, objectives, scope and activities of the assessment Step 2 Dose-response assessment: characterize relationship between dose and adverse health effect(s); estimate incidence of effect as a function of exposure Problem formulation: integrate available information on contaminant sources, effects, and receiving environment; define assessment endpoints; describe hypothesized relationships between the sources and receptors; develop a data collection plan Analysis: collect data and calculate estimates characterizing exposure and effects on human and ecological systems Step 3 Exposure assessment: measure or estimate the intensity, frequency, and duration of exposure Analysis: characterize the nature, extent, and distribution of exposure; characterize relationship between dose and adverse effect(s) Risk characterization: integrate information from the analysis phase to estimate risk Step 4 Risk characterization: combine information from the doseresponse and exposure assessments to estimate the incidence of a health effect in human populations Risk characterization: combine information from the analysis phase to estimate and describe risks 183

PAGE 184

Table 9-2. Human and ecol ogical exposure pathways for land-applied biosolids Pathway Description of HEI 1. Biosolids soil plant human Human (except for home gardener) lifetime ingestion of plants grown in biosolids-amended soil 2. Biosolids soil plant human Human (home gardener) lifetime ingestion of plants grown in biosolidsamended soil 3. Biosolids soil human Human (child) ingesting biosolids 4. Biosolids soil plant animal human Human lifetime ingestion of animal products (animals raised on forage grown on biosolids-amended soil) 5. Biosolids soil animalhuman Human lifetime ingestion of animal products (animals ingest biosolids directly) 6. Biosolids soil plant animal Animal lifetime ingestion of plants grown on biosolids-amended soil 7. Biosolids soil animal Animal lifetime ingestion of biosolids 8. Biosolids soil plant Plant toxicity due to taking up biosolids-borne TCC when grown in biosolids-amended soils 9. Biosolids soil soil organism Soil organism ingesting biosolids/soil mixture 10. Biosolids soil soil organism predator Predator of soil organisms that have been exposed to biosolids-amended soils 11. Biosolids soil airborne dust human Adult human lifetime inhalation of particles (dust) 12. Biosolids soil surface water human Human lifetime drinking surface water and ingesting fish containing TCC from biosolids 13. Biosolids soil air human Adult human lifetime inhalation of volatilized TCC from biosolidsamended soil 14. Biosolids soil groundwater human Human lifetime drinking well water containing TCC from biosolids that leached from soil to ground water 15. Biosolids soil surface water animal Animal lifetime drinking surface water and ingesting fish containing TCC from biosolids 16. Biosolids soil surface water aquatic organism Aquatic organism exposed to water containing TCC from biosolids 184

PAGE 185

Table 9-3. Toxicity and carcinogenicity of triclo carban (TCC) in rodents Study type Species Endpoint Exposure Result References Acute oral toxicity Rat Mouse LD50 LD50 Single oral, neat product Single oral, neat product LD50 >2000 mg kg bw-1 LD50 >5000 mg kg bw-1 Bayer AG, 1991 in SCCP, 2004 Marty and Wepierre, 1979 Subchronic (30 d) oral gavage Rat Subchronic toxicity 0, 500, 1000 mg kg bw-1 d-1 NOAEL > 1000 mg kg bw-1 d-1 Monsanto, 1960 in SCCP, 2004 Subchronic (8 wk) dietary feeding Rat Subchronic toxicity 25, 75, 250 mg kg bw-1 d-1 NOAEL = 75 mg kg bw-1 d-1 LOAEL = 250 mg kg bw-1 d-1 Monsanto, 1985 in SCCP, 2004 Chronic (2 y) dietary feeding Rat Carcinogenicity and chronic toxicity 0, 25, 75, 250 mg kg bw-1 d-1 NOEL = 25 mg kg bw-1 d-1 LOEL = 75 mg kg bw-1 d-1 Not carcinogenic Monsanto, 1981 in SCCP, 2004 Three generation dietary feeding Rat Reproductive and developmental toxicity 0, 25, 500, 1000, 3000 mg kg-1 NOAEL F0 = 280 mg kg bw-1 d-1 NOAEL F1 = 95 mg kg bw-1 d-1 NOAEL F2 = 300 mg kg bw-1 d-1 Monsanto, 1983 in SCCP, 2004 185

PAGE 186

Table 9-4. Mutagenicity and clas togenicity of triclocarban (TCC) Study type Test system Endpoint Result References Ames test S. typhimurium strains TA 98, TA 100, TA 1535, and TA 1537 Mutagenicity Not mutagenic with or without metabolic activation Bayer AG, 1982 in SCCP, 2004 Ames test S. typhimurium strains TA 98, TA 100, TA 1537, and TA 1538 Mutagenicity Not mutagenic with or without metabolic activation Bonin et al., 1982 in SCCP, 2004 Chromosomal aberration test Chinese hamster ovary cells Clastogeni city Negative for the induction of chromosomal aberrations with or without metabolic activation Soap and Detergent Association, 2002 in SCCP, 2004 186

PAGE 187

Table 9-5. Toxicity of triclocar ban (TCC) to aquatic organisms Study type Species Endpoint Exposure Result References Acute toxicity (96 h) Lepomis macrochirus (fish, fresh water) Salmos gairdneri (fish, estuary, fresh water) NOEC NOEC Unknown range NOEC = 0.049 mg L-1 NOEC < 0.049 mg L-1 Monsanto, 1976 in TCC Consortium, 2002 Monsanto, 1976 in TCC Consortium, 2002 Acute toxicity (48 h) Ceriodaphnia sp. (crustacean) NOEC Unknown range NOEC = 0.0019 mg L-1 Monsanto, 1987 in TCC Consortium, 2002 Toxicity to aquatic plants (72 h) Pseudokirchneriella subcapitata (algae) Growth rate Unknown range NOEC 0.01 mg L-1 Yang et al., 2008 in Chalew and Halden, 2009 Toxicity to aquatic plants (14 d) Microcystis aeruginosa (algae, blue, cyanobacteria) Growth rate Unknown range NOEC = 0.01 mg L-1 Monsanto, 1980 in TCC Consortium, 2002 Chronic toxicity (35 d) Pimephales promelas (fish, fresh water) Hatchability of eggs and growth and survival of fry Unknown range NOEC = 0.005 mg L-1 Monsanto, 1992, in TCC Consortium, 2002 Chronic toxicity (28 d) Mysidopsis bahia (crustacea) Reproduction rate Unknown range NOEC = 0.00006 mg L-1 Monsanto, 1992 in TCC Consortium, 2002 187

PAGE 188

Table 9-6. Parameters and assumptions for calcu lating screening level hazard indices (HI values) Abbreviation Parameter definition Va lue Assumptions/Explanation Pathway Reference BAF Bioaccumulation factor: the quotient of the concentration of an element or compound in an organism divided by the concentration in an environmental medium, when the concentrations are near steady state, and when multiple uptake routes may contribute (Suter, 2007) BAFp = 0.0070 (plant, w.w.) BAFww = 4.0 (worm, w.w.) BAFdw = 21 (worm, d.w.) BAFr = 0.15 (ruminant, fat tissue) BAFnr = 0.43 (non-ruminant, fat tissue) BAFb = 0.143 (bird, fat tissue) Greatest BAF values calculated in present studies of Bahia grass and Eisenia fetida earthworms logBAFr = 0.191 0.608*log(solubility) (r2 = 0.54) log BAFnr = 0.527 0.538*log(solubility) (r2 = 0.49) log BAFb -2.743 + 0.542*logKow (r2 = 0.54) 1, 2, 4, 6 10 5 5 5 Chapters 5 and 8 Garten and Trabalka, 1983 BAR Biosolids application rate 50 Mg ha-1 (d.w.) 5 Mg ha-1 y-1 (d.w.) x 100 y Worst-case scenario, one-time application rate Application rate; applied annually for 100 y 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 14 NRC, 2002 188

PAGE 189

Table 9-6. Continued. Abbreviation Parameter definition Va lue Assumptions/Explanation Pathway Reference BCF Bioconcentration factor: the quotient of the concentration of an element or compound in an organism divided by the concentration in water, when the concentrations are near steady state and when only direct uptake from solution contributes (Suter, 2007) BCFwfw = 140 (whole fish, w.w.) BCFmfw = 13 (fish muscle, w.w.) Whole fish BCF used for animal consumption calculations (entire organism consumed) Fish muscle BCF used for human consumption calculations (non-muscle components typically not eaten) 12, 15 12, 15 TCC Consortium, 2002a BW Body weight (live weight) Adult: 70 kg Child: 16 kg Cow: 590 kg Shrew: 15 g (short-tailed) American woodcock: 180 g Eastern cottontail rabbit: 1100 g Osprey: 1600 g River otter: 8100 g Mean Mean Mean Mean Mean Mean Mean Mean 1, 2, 3, 4, 12 7 7 7 6 15 15 USEPA, 1997 USEPA, 1993c 189

PAGE 190

Table 9-6. Continued. Abbreviation Parameter definition Va lue Assumptions/Explanation Pathway Reference CA TCC concentration in consumed animal CAf (animal fat) CAm (animal meat) CAwfw (whole fish; w.w.) CAmfw (fish muscle, w.w.) BAFf SC (or PC) CAm FF BCFwfw SWC BCFmfw SWC 5 5 5 5 USEPA, 1995 CAR TCC concentration in air 3.0 x 10-8 mg L-1 (worst-case) 3.0 x 10-7 mg L-1 (100 y) SC DC 13 13 USEPA, 1995 CB TCC concentration in biosolids 131 mg kg-1 (d.w.) 95th percentile concentration in 2009 TNSSS 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 14 TNSSS, 2009a CL TCC concentration in leachate 0.021 mg L-1 (worst-case) 0.21 mg L-1 (100 y) 0.18% of total TCC applied / cm3 leached from one ha in a single leaching event in a sandy soil Chapter 4 DC Dust concentration in air 0.010 mg L-1 Concentration used in Part 503 biosolids risk assessment 11 USEPA, 1995 190

PAGE 191

Table 9-6. Continued. Abbreviation Parameter definition Va lue Assumptions/Explanation Pathway Reference EC TCC concentration in earthworm 63 mg kg-1 (worst-case, d.w.) 626 mg kg-1 (100 y, d.w.) BAFdw SC 9, 10 9, 10 Chapter 5 FC Fish consumption Adult: 20 g d-1 Child: 4.2 g kg-1 d-1 Mean Mean 12 12 USEPA, 1997 USEPA, 2008 FF Fat fraction in meat Beef: 0.10 Pork: 0.090 Poultry: 0.060 Mean 4, 5 4, 5 4, 5 USEPA, 1997 FI Food ingestion rate Cow: 9100 g d-1 (d.w.) Short-tailed shrew: 2.2 g d-1 (d.w.) Eastern cottontail rabbit: 76 g d-1 (d.w.) American woodcock: 19 g d-1 (d.w.) FImammals = 0.235*BW0.82 FIbirds = 0.65*BW0.65 7 7 7 6 Nagy, 1987 in Suter, 2007 191

PAGE 192

Table 9-6. Continued. Abbreviation Parameter definition Va lue Assumptions/Explanation Pathway Reference Osprey: 78 g d-1 (d.w.) Osprey: 310 g d-1 (w.w.) River otter: 380 g d-1 (w.w.) Fraction water weight of bony fish: 0.75 15 15 15 USEPA, 1993c River otter: 1500 g d-1 (d.w.) Fraction water weight of bony fish: 0.75 15 FS Soil fraction of diet Cow: 0.025 Short-tailed shrew: 0.13 American woodcock: 0.10 Value used in Part 503 biosolids risk assessment 7 7 7 USEPA, 1995 Talmage and Walton, 1983 Beyer et al., 1994 FVC Combined fruit and vegetable consumption 7.7 g kg-1 d-1 (w.w.) Mean 1, 2 USEPA, 1997 HFS Hectare-furrow-slice mass 2.2 x 106 kg Soil bulk density = 1.3 g cm-3 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 14 Brady and Weil, 2002 192

PAGE 193

Table 9-6. Continued. Abbreviation Parameter definition Va lue Assumptions/Explanation Pathway Reference IR Inhalation rate Adult: 23 L min-1 Child: 11 m3 d-1 Mean Mean 11 11 USEPA, 1997 USEPA, 2008 LE Leaching event 5.6 cm Water leached at each leaching event in biosolids-amended soil column study Chapter 5 MC Meat consumption Adult Beef: 90 g d-1 Pork: 27 g d-1 Poultry: 67 g d-1 Child Beef: 1.8 g kg-1 d-1 Pork: 0.84 g kg-1 d-1 Poultry: 1.5 g kg-1 d-1 Mean Mean 4 4 4 4 4 4 USEPA, 1997 USEPA, 2008 OYC One-hundred y biosolidsamended soil TCC concentration 21 mg TCC kg soil1 (d.w.) Biosolids containing 131 mg TCC kg-1 applied at a 5 Mg ha-1 rate for 100 y; no TCC degradation or loss 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 14 USEPA, 2009a 193

PAGE 194

Table 9-6. Continued. Abbreviation Parameter definition Va lue Assumptions/Explanation Pathway Reference PC TCC concentration in plant tissue PCwcw = 0.021 (worst-case, w.w.) PCoyd = 0.21 (100 y, w.w.) PC = BAFp SC 1, 2, 4, 6, 8 Chapter 4 RfD Reference dose: daily intake of chemical that during an entire lifetime appears to be without appreciable risk on the basis of all the known facts at the time Humans: 0.25 mg kg-1 d-1 Animals: 1.3 mg kg bw-1 d-1 Humans: calculated using the NOEL (25 mg kg bw-1 d-1) from a 2 y chronic toxicity rat study; adjusted using an interspecies uncertainty factor (10) and a sensitivity uncertainty factor (10) Animals: calculated using the NOEL (25 mg kg bw-1 d-1) from a 2 y chronic toxicity rat study; adjusted using an interspecies uncertainty factor (5), a sensitivity factor (2), and a lab-to-field extrapolation factor (2) 1, 2, 3, 4, 5, 7, 11, 14 6, 10 Monsanto, 1981 in SCCP, 2004 Monsanto, 1981 in SCCP, 2004 Suter, 2007 SC TCC concentration in soil 2.1 mg kg-1 (worst-case, d.w.) 21 mg kg-1 (100 y, d.w.) Assuming a one-time application of biosolids containing 131 mg TCC kg-1 amended at a 50 Mg ha-1 rate to a depth of 15 cm Assuming 100 annual applications of biosolids containing 131 mg TCC kg-1 amended at a 5 Mg ha-1 rate to a depth of 15 cm USEPA, 1995; TNSSS, 2009b 194

PAGE 195

Table 9-6. Continued. SI Soil ingestion Adult: 0.05 g d-1 Child: 0.2 g d-1 Pica child: 10 g d-1 Mean 3 3 3 USEPA, 1997 SWC Surface water concentration of TCC 6750 ng L-1 Greatest documented TCC concentration in surface water; contains contamination from sewer overflows 12, 15, 16 Halden and Paull, 2005 WCC Worst-case biosolidsamended soil TCC concentration 2.1 mg TCC kg soil-1 (d.w.) Biosolids containing 131 mg TCC kg-1 applied at a 50 Mg ha-1 rate 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 14 TNSSS, 2009b WI Water intake Adult: 1.4 L d-1 Child: 0.87 L d-1 Mean 12, 14 12, 14 USEPA, 1997 195

PAGE 196

Table 9-7. Equations used to calculate sc reening-level hazard indices (HI values) Pathway Hazard Index Equation Worst-case Hazard Index 100-year Hazard Index Comments/assumptions 1. and 2. Biosolids soil plant human (PC FVC) / RfD CF Adult: 0.00064 Child: 0 .00064 Adult: 0 .0064 Child: 0.0064 All produce consumed grown in biosolidsamended soil 3. Biosolids soil human (SC SI / BW) / RfD CF Adult: 8.5 x 10-6 Child: 0.00015 Pica child: 0.0074 Adult: 8.5 x 10-5 Child: 0.0015 Pica Child: 0.074 4. Biosolids soil plant animal human (CAmp MC / BW) / RfD CF Adult Beef: 1.6 x 10-6 Pork: 1.2 x 10-6 Poultry: 6.9 x 10-7 Child Beef: 2.3 x 10-6 Pork: 2.7 x 10-6 Poultry: 1.1 x 10-6 Adult Beef: 1.6 x 10-5 Pork: 1.2 x 10-5 Poultry: 6.9 x 10-6 Child Beef: 2.3 x 10-5 Pork: 2.7 x 10-5 Poultry: 1.1 x 10-5 Individual HI values calculated assuming exclusive consumption of each meat product 100% of animal diet consists of plants grown in biosolidsamended soil 196

PAGE 197

Table 9-7. Continued Pathway Hazard Index Equation Worst-case Hazard Index 100-year Hazard Index Comments/assumptions 5. Biosolids soil animal human (CAms MC / BW) / RfD CF Adult Beef: 0.00023 Pork: 0.00018 Poultry: 9.8 x 10-5 Child Beef: 0.00033 Pork: 0.00039 Poultry: 0.00015 Adult Beef: 0.0023 Pork: 0.0018 Poultry: 0.00098 Child Beef: 0.0033 Pork: 0.0039 Poultry: 0.0015 Individual HI values calculated assuming exclusive consumption of each meat product 6. Biosolids soil plant animal (PC FI / BW) / RfD CF Cow: 0.00073 Cottontail rabbit: 0.0022 Cow: 0.0073 Cottontail rabbit: 0.022 100% of diet consists of plants growing biosolids-amended soil 7. Biosolids soil animal (SC FS FI / BW) / RfD CF Cow: 0.0018 Shrew: 0.090 American woodcock: 0.053 Cow: 0.018 Shrew: 0.90 American woodcock: 0.53 8. Biosolids soil plant SC / RfD CF Insufficient toxicity data Insufficient toxicity data 9. Biosolids soil soil organism SC / RfD CF Eisenia fe tida: 0.074 Eisenia fetida: 0.74 Based on the LC50 Lifetime spent in biosolids-amended soil 10. Biosolids soil soil organism predator (ECd FI / BW) / RfD CF Shrew: 15 American woodcock: 11 Shrew: 150 American woodcock: 110 100% of diet consists of earthworms growing in biosolids-amended soil 197

PAGE 198

Table 9-7. Continued. Pathway Hazard Index Equation Worst-case Hazard Index 100-year Hazard Index Comments/assumptions 11. Biosolids soil airborne dust human (CAR IR / BW) / RfD CF Adult: 5.6 x 10-5 Child: 8.1 x 10-5 Adult: 5.6 x 104 Child: 5.6 x 104 Exposed to maximum concentration of biosolids dusts for 24 h d-1 12. Biosolids soil surface water human ((SWC WI / BW) + (BCFmfw FC / BW)) / RfD CF Adult: 0.00064 Child: 0.0029 100% of water intake consists of maximally contaminated surface water 100% of fish intake consists of fish caught in maximally contaminated surface water Light activity levels of water intake 14. Biosolids soil groundwater human (CL WI / BW) / RfD CF Adult: 0.0017 Child: 0.0046 Adult: 0.017 Child: 0.046 100% of water intake consists of maximally contaminated groundwater Light activity levels of water intake 15. Biosolids soil surface water animal (CAwh FI / BW) / RfD CF Osprey: 0.29 River otter: 0.28 100% of diet consists of fish in maximally contaminated surface water 198

PAGE 199

Table 9-7. Continued. Pathway Hazard Index Equation Worst-case Hazard Index 100-year Hazard Index Comments/assumptions 16. Biosolids soil surface water aquatic organism SWC / RfD CF Lepomis macrochirus : 0.14 Salmos gairdneri : 0.14 Ceriodaphnia sp.: 3.6 Pseudokirchneriella subcapitata : 0.68 Microcystis aeruginosa : 0.68 Pimephales promelas : 1.4 Mysidopsis bahia : 110 Lifetime exposure to maximum concentration of contaminated surface water 199

PAGE 200

Table 9-8. Parameters and assumptions for cal culating adjusted hazard indices (HI values) Abbreviation Parameter definition Value/ Equation Assumptions/Explanation Reference Asms Area affected by land application of biosolids 1074 ha Part 503 risk assessment parameter USEPA, 1995 Aws Area of the watershed (ha) 440,300 Part 503 risk assessment parameter USEPA, 1995 Csed TCC concentration in sediment entering surface water Csed = Csma DF USEPA, 1995 Csma TCC concentration in soil eroding from biosolidsamended land Csma = (Pa fero) / MEsms CF USEPA, 1995 Csw Surface water concentration associated with the Csed Csw = Csed / [KDsw + (Pl / Ps) (1/ w)] USEPA, 1995 DF Eroded soil dilution factor 0.0066 Part 503 risk assessment parameter USEPA, 1995 fero Fraction of total TCC loss by erosion USEPA, 1995 KDsw Partition coefficient between solids and liquid within the surface water 2512 L kg-1 Indigenous biosolids-borne TCC partitioning coefficient Chapter 3 MEsma Estimated rate of soil loss for the SMA 8,500 kg ha-1 y-1 Part 503 risk assessment parameter USEPA, 1995 Pa Annual application rate of TCC (mg ha-1 y-1) Worst-case scenario: 6.55 x 106 131 mg TCC kg biosolids1; 50 Mg ha-1 application rate USEPA, 1995 TNSSS, 2009b 200

PAGE 201

Table 9-8. Continued Abbreviation Parameter definition Value/ Equation Assumptions/Explanation Reference Pl / Ps Percent liquid in the water column divided by percent solids in the water column 62,500 (unitless) Part 503 risk assessment parameter USEPA, 1995 Ssma Sediment delivery ratio for the SMA 0.46 Part 503 risk assessment parameter USEPA, 1995 Sws Sediment delivery ratio for the watershed 0.17 Part 503 risk assessment parameter USEPA, 1995 201

PAGE 202

202 Table 9-9. Adjusted hazard indices (HI values) for exposure pathway 10 Pathway Hazard Index Equation Adjusted worst-case hazard index Adjusted 100-year hazard index Adjusted agronomic hazard index Assumptions 10. Biosolids soil soil organism predator (ECd FI / BW) / RfD CF / 0.5 Shrew: 4.4 American woodcock: 5.5 Shrew: 6.5 American woodcock: 8.0 Shrew: 0.440.88 American woodcock: 0.55-0.88 50% bioavailability (USEPA, 1995) TCC t1/2 = 20y (Chapter 8) Agronomic biosolids application rate: 510 Mg ha-1 Shrew specific: 59% of diet is earthworms (USEPA, 1993c)

PAGE 203

Table 9-10. Adjusted hazard indices (HI values) for exposure pathway 16 Pathway Hazard Index Equation Adjusted Hazard Indices Using Adjusted Surface Water Concentration (ng L-1) Assumptions 190 (TCC Consortium, 2002a) 3.4 (Topp, personal communication) 13 (sedimentation) 16. Biosolids soil surface water aquatic organism SWC / RfD CF Lepomis macrochirus : 0.0039 Salmos gairdneri: 0.0039 Ceriodaphnia sp.: 0.10 Pseudokirchneriella subcapitata : 0.019 Microcystis aeruginosa : 0.019 Pimephales promelas : 0.038 Mysidopsis bahia : 3.17 Lepomis macrochirus : 6.9 10-5 Salmos gairdneri: 6.9 10-5 Ceriodaphnia sp.: 0.0018 Pseudokirchneriella subcapitata : 0.00034 Microcystis aeruginosa : 0.00034 Pimephales promelas : 0.00068 Mysidopsis bahia : 0.057 Lepomis macrochirus : 0.00027 Salmos gairdneri : 0.00027 Ceriodaphnia sp.: 0.0068 Pseudokirchneriella subcapitata : 0.0013 Microcystis aeruginosa : 0.0013 Pimephales promelas : 0.0026 Mysidopsis bahia : 0.22 203

PAGE 204

Figure 9-1. Depiction of the integrated health risk assessment framework (WHO, 2001) 204

PAGE 205

Biosolids intended for land-application soil home garden soil human plant animal air-borne dust soil organism surface water & aquatic organisms groundwater plant application method and rate transformation Biosolids intended for land-application soil home garden soil human plant animal air-borne dust soil organism surface water & aquatic organisms groundwater plant application method and rate transformation volatilization Figure 9-2. Conceptual model of human and ecological exposures to biosolids-borne triclocarban (TCC) 205

PAGE 206

CHAPTER 10 SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS Introduction The ultimate objective of the project was to pe rform an integrated human/ecological health risk assessment that could support one of two hypotheses: 1) the risk associated with the environmenta l presence of TCC in land-applied biosolids is negligible, or 2) the risk is sufficient to warrant further regulatory attention a nd research to reduce environmental contamination and potential adverse effects. Meeting intermediate objectives (i.e. de termine TCC physicochemical properties and partitioning, biosolids-borne TCC concentrations, environmental transport, persistence, and soil organism impacts) provided the data nece ssary to meet the ultimate objective. Risk was assessed in a tiered process by cal culating screening-leve l hazard indices (HI values) for relevant human and ecological recep tors under the Part 503 Biosolids Rule worstcase and 100-year application scenarios. Hazard indices >1 s uggested the predicted environmental concentrations were greater than concentrations determined to induce adverse effects in a given receptor. H azard indices <1 suggested that a given receptor would not be exposed to problematic concentrations of biosolids-borne TCC over a lifetime. The screeninglevel HI values for the majority of exposure pathways (including all human exposure pathways) under both land-application scenarios were below the critical level of one. The two pathways for which the screening-level HI values were >1 included the ecological receptors selected to represent terrestrial pr edators feeding on earthworms grown in biosolids206

PAGE 207

amended soil (Pathway 10) and aquatic organisms (Pathway 16), and were selected for a more detailed risk assessment. Parameters and assump tions used to calculate the screening-level HI values were reevaluated, and modi fied where appropriate (e.g. in corporation of TCC degradation rates into estimates of environmental concentrat ions). The resulting adjusted HI values for Pathway 10 remained >1, however, and supported selection of Hypothesis 2. The adjusted HI values for aquatic organism s and earthworm predators were subsequently used to calculate preliminary TCC pollutant lim its for land-applied biosolids that could support further discussion of possible environmental regu lation of TCC for the protection of human and ecological health. Four TCC limits (i.e. the cumulative pollutant rate, annual pollutant loading rate, ceiling concentration limit, and pollutant concentration lim it) applicable to land-applied biosolids were calculated based on the guideline s provided in the Part 503 Rule risk assessment (1995) (Figure 10-1). The limits were expressed in varying units to provide applicators and regulators with guidelines addressing multiple appr oaches to land-application of biosolids. The preliminary TCC pollutant limits, particularly th e Pollutant concentration limit, could have important implications for current biosolids land-a pplication practices, but need to be evaluated in light of available soil concentration data and remaining environmental eff ects data gaps before modifications to current land-appl ication regulations are suggested. The following sections summarize the results of the intermediate objectives, explore the implications and limitations of the calculated preliminary TCC pollutant limits, and propose future research to address remaining and newly identified TCC data gaps. Relevance of the data developed under the project to the USEPA Propo sed Rule for Antimicrobial Data Requirements (Federal Register, 2008) is also discussed. 207

PAGE 208

Summaries of Intermediate Objective Results and Future Research Needs Intermediate Objective 1: Confirm Physicochemical Properties of TCC The measured solubility and Kow of TCC is 0.045 mg L-1 and 3.5, respectively. Solubility and Kow are indicators of hydrophobicity and lipophilicity, and are genera lly considered good predictors of a compounds tendency to move between various environmental compartments, bioaccumulate, and degrade. The sparing solubility and moderate Kow of TCC was reflected in the preferential partitioning of TCC to bi osolids, limited environmental mobility (< 0.18% of measured biosolids-borne TCC leached from amended soil columns), and minimal biodegradation (biosolids-borne TCC half-life of 8-20 y). Intermediate Objective 2: Characterize TCC Concentrations in Multiple Biosolids and Partitioning of Indigenous TCC in the Biosolids Matrix The mean TCC concentration in the 23 biosolids analyzed herein (20 mg kg-1) was less than the mean TCC concentration in the sewage sludge products (39 mg kg-1; n = 84) analyzed in the TNSSS (2009a). The 95th percentile value from the sewage sludge products analyzed in the TNSSS was 131 mg TCC kg-1, and was used to calculate all relevant HI values. Triclocarban concentrations in the 23 bioso lids analyzed suggest aerobic sludge treatment methods such as composting and aerobic digestion can reduce biosolids-borne TCC concentrations. Analysis of additional bios olids, however, is necessary to confirm the hypothesis. The publicly available TNSSS data make no distinctions between biosolids treatment methods and resultant TCC concentrations, and it is unclear whether materials listed as final products are all processed biosolids. Studies quantifying TCC in fluent concentrations and corresponding biosolids-born e TCC concentrations as a f unction of biosolids processing methods at multiple wastewater treatment plants could help identify the conditions under which 208

PAGE 209

TCC concentrations in biosolids are minimized. The information could be useful in minimizing the concentration of TCC and other similar organics in final products. Partitioning behaviors of indigenous (logKd/logKoc = 3.4/3.9) and spiked (Kd/Koc = 3.3/3.9) TCC in biosolids are similar, which suggests the use of spiked compound in TCC fate and transport studies adequately mimics the beha vior of indigenous co mpound. Two sequential water extractions at the begi nning of the biosolids-borne 14C-TCC biodegradation study recovered a total of 14 and 3% of spiked compound from an amended fine sand and silty clay loam, respectively. After 7.5 months, water extractability was reduced to 3 and 0.7%, respectively. Percent recovery in the second biotic fine sand wa ter extract was ~30% less than recovery in the first biotic fine sand water ex tract for the first 6 wks of the experiment. Differences between recoveries in the two fine sand water extracts were greatly reduced for the remaining 24 wks (absolute percent recovery in e ach water extract ranged from ~0.5-1.5%). A nearly opposite trend was observed in the biotic s ilty clay loam samples. Recoveries in the two sequential silty clay loam extracts remained simila r for the first 3 wks, and then recovery in the second water extract decreased to ~40% less than recove ry in the first extract. Given minimal 14C-TCC degradation in both soils and an eff ective 7.5 month spike eq uilibration period, the differences between sequential wate r extractability within each treatment at the conclusion of the experiment could be assumed to represent the behavior of indigenous TCC. Additional water extractability work with indi genous TCC could confirm the a ppropriateness of using spiked compound in studies of biosolids-borne TCC fate and transport, or whet her current and future data should be adjusted to acc ount for use of spiked TCC. Equilibrium partitioning of spiked TCC to biosol ids is faster and more extensive (~99% at 24 h) than partitioning to sandy (~ 60% at 24 h) and loamy (~75% at 24 h) soils. Conversely, 209

PAGE 210

equilibrium desorption of spiked TCC from biosolids was slower and less extensive (~50% at 10 d) than desorption from sandy (~80% at 5 d) an d loamy (~70% at 6 d) soils. The sorption and desorption behaviors in the (unamended) spiked so ils were similar in spiked biosolids-amended soils (22 Mg biosolids ha-1 rate). The TCC solubility and Kow data, the extensive sorption of TCC to biosolids, moderate sorption to sandy an d loamy soils, and incomplete desorption from the three solids were all consistent with the limited environmental mobility and minimal biodegradability of bioso lids-borne TCC measured. If future models of biosolids-borne TCC e nvironmental transport, or explanations of observed environmental behaviors, are to incorp orate measured data characterizing TCC sorption and desorption kinetics, the ad sorption/desorption experiments should be expanded to include soils amended with biosolids spiked with TCC prior to mixing. The TCC sorption and desorption kinetics of biosolids differed greatly from the kinetics of soils and spiked biosolidsamended soils. The data suggest that the delivery of TCC to soil as a component of biosolids could significantly retard the rates and extent of TCC transf er between environmental matrices. The inclusion of soils with a wi de range of physicochemical propert ies, biosolids produced from multiple processing methods, and replicates containing biosolids equilibrated with spiked TCC for >>24 h could improve the generalizability of the results, identify the most influential parameters, and characterize change s in desorption kinetics of aged materials. Determining the desorption and bioavailability of biosolids-borne TCC in shortand long-term field soils is of particular interest. Intermediate Objective 3: Characterize Leac hability of Biosolids-Borne TCC in Amended Soils Less than 0.2% of indigenous biosolids-borne TCC leached (~4 pore volumes of leachate) from replicated soil columns (sand) amended with one of four biosolids of known TCC content. 210

PAGE 211

Leachate data were also collected for repli cated columns receiving one of seven additional biosolids for which TCC content was not know n. The mean TCC concentration in the wastewater treatment products analyzed in the TNSSS (2009a) was used to estimate the percent applied TCC leached from soils amended with biosolids of unknown content. The estimated minimum and maximum percents of applied T CC leached from all columns were 0.01% and 0.69%, respectively. The greatest TCC concentration recorded was 3.3 ng mL-1, but no TCC was detected in any leachates after the fifth leaching event. None of the TCC concentrations in the amended soil column leachates posed a risk to human health through the ingestion of contam inated groundwater pathway. Most quantified TCC concentrations from early leaching events were greater than the lowest reported chronic NOEC (0.06 ng mL-1) and acute NOEC (2 ng mL-1) concentrations for aquatic invertebrates, and one TCC concentration was also greate r than the lowest reported acute EC50 concentration (3 ng mL-1) for aquatic invertebrates. However, TCC concentrations in biosolids-amended soil leachates are expected to be diluted upon interception by groundwater or surface water, acutely toxic levels of TCC in bios olids-amended soil leachates we re rare, and both acutely and chronically toxic levels were short-lived following a singl e application of biosolids. An additional mechanism that could further reduce the potential toxicity of TCC in biosolids-amended soil leachates is sorption to DOM and/or DOC, as DOC-bound contaminants are reportedly unavailable for uptake by some aq uatic organisms. Characterization of TCC sorption to DOM/DOC, development of analyti cal methods that distinguish between free (unassociated) and DOM/DOC-associated T CC, and studies of DOM-associated TCC bioavailability could improve the biosolid s-borne TCC risk assessment for pathways characterizing exposures vi a surface and ground waters. 211

PAGE 212

Other areas of continued study mi ght include leachability of TCC in shortand long-term field application sites; impacts of soils, bios olids, application me thods, and ground cover on leachability; and TCC contamination in ground and drinking water. The contribution of leachate-borne TCC to ground and drinking water contamination is expected to be minimal, however, given the preferential partitioning of TCC to the solid fraction of the soil. Of greater importance is characterizing the extent of TCC tr ansport in runoff from biosolids-amended soils. To date, TCC concentrations in amended soil ru noff have been characterized in only one field study (Topp, personal communication). Intermediate Objective 4: Characterize Biodegradation of Biosolids-Borne TCC in Amended Soils During a 7.5-month 14C-TCC biodegradation study in biosolids-amended soils, <5% of the spiked radiolabel was mineralized. Estimated T CC half-lives in fine sa nd and silty clay loam soils were 20 y and 8 y, respectively. The am ended soil samples were sequentially extracted with water, MeOH, and NaOH at regular interv als, and subsequently combusted to assess compound lability and changes in ex tractability with time. Extractability decreased with time in both soils, as evidenced by reductions in water and MeOH (i.e. predicted labile, and potentially bioavailable, fractions) recoveries and increase d recovery in the combusted fraction. Water extractability tended to be a better predictor of 14C-TCC mineralization in the silty clay loam than in the fine sand. Most of the 14C-TCC remained MeOH extractable in both soils, yet ultimate biodegradation of the compound was mi nor, suggesting MeOH is a poor predictor of TCC mineralization. Appr oximately half of the 14C-TCC spike quickly converted to the combusted (i.e. bound residue) fraction in the silty clay loam, as compared to ~20% conversion in the fine sand at the conclusion of the study. Radio-thin-layer-chromatography (RAD-TLC) analysis of the MeOH extracts detected no intermediate degradation products. 212

PAGE 213

Multiple questions regarding biosolids-borne TCC biodegradation remain. Biodegradation products have been identified in bench-top ac tivated sludge systems, but not in biosolidsamended soils. Extensive TCC degradation was reported in spiked non-amended soils (Ying et al., 2007; Xia et al., 2008), but the degradation products were not isolated. The lab-based biodegradation study described herein assessed compound loss in two soils amended with one biosolids at one rate. Possible impacts of biosolids loading rate biosolids characteristics, TCC concentration, increased time following biosolid s addition(s), and changing environmental conditions are unknown. Data gaps also remain pertaining to the degrading microorganisms, TCC impacts on soil microbial community structur e, TCC bioavailability versus degradability, and potential development of antimicrobial resistance in amended soils. If TCC contributes to antibacterial resistance development in biosol ids-amended soils, the risk assessment conducted herein could require significant amendment. Intermediate Objective 5: Characterize Biosolids-Borne TCC Toxicity to Terrestrial Organisms Biosolids-borne TCC effects on earthworm mortality, and the microbially mediated processes of ammonification, nitrif ication, and respiration were mini mal. There was no effect on mortality in earthworms incubated in silty clay loam and artificial soils amended with biosolids (22 Mg ha-1 rate) containing 10,000 mg TCC kg-1 (soil concentration of ~100 mg TCC kg soil-1 ). Earthworm mortality was only affected in th e fine sand, with a resultant estimated LC50 of 40 mg TCC kg amended soil-1. Results suggest biosolids-borne T CC does not pose a significant risk to adult earthworm survival, but the impacts on sublethal health outcomes (e.g. growth, reproduction, and cocoon survival) are not known. An earthworm mortality study, modified to include assessments of sublethal effects asso ciated with biosolids-borne TCC, should be conducted. 213

PAGE 214

Biosolids-borne TCC had no apparent effect on microbial respiration or ammonification up to the maximum concentration test ed (i.e. 717 mg TCC kg biosolids-1 or ~0.717 mg TCC kg amended soil-1). Similarly, there was no TCC effect on nitrification up to 73 mg TCC kg biosolids-1, and the effects at th e 717 mg TCC kg biosolids-1 were minimal and statistically inconclusive. The lack of signi ficant TCC treatment effects on respiration, ammonification and nitrification was not unexpected, as the processe s in soil are performed mostly or entirely by gram negative microorganisms, and TCC affects primarily gram positive microorganisms. An examination of biosolids-borne TCC impacts on the processes of respiration, ammonification, and nitrification were warranted, however, given their importance to hea lthy ecosystem function. Future studies addressing impacts of biosolids-borne TCC on gram positive soil microorganisms are needed. No data on the toxicity of TCC to terrestrial plants are available. The limited TCC uptake by plants, and the lack of an obs erved adverse effect on plants gr own in biosolids-amended soils under current land-application pr actices suggest the impact of biosolids-borne TCC on plant health is negligible. Intermediate Objective 6: Characterize Biosolids-Borne TCC Bioaccumulation Bioaccumulation factors were calculated for earthworms and Bahia grass tissue harvested from biosolids amended soils. The calculated earthworm BAFs following 30 d of exposure were 18 (fine sand), 20 (silty clay loam ), and 5.2 (artificial so il), and appeared to be inversely related to the OM fraction of the soil. The relationship was partially attributed to the known preferential consumption of OM by Eisenia fetida worms. The OM content of the artificial soil was 10x and 2.5x greater than the OM contents of the fine sand and the silty clay, respectively, and thus provided the worms with a larger uncontamin ated food source. The results suggest 214

PAGE 215

bioaccumulation in earthworms will be less pronounced in soils with elevated OM contents (prior to equilibration with the indigenous soil organic matter). Bioaccumulation of biosolids-borne TCC wa s much less in Bahia grass than in earthworms. Tissue samples of Bahia grass grow n in soil columns amended with biosolids at varying rates during an unrelated study of biosolids-borne phosphorus loss were analyzed for TCC content and the resulting calculated BAFs ranged from 0.00041 to 0.008. The Bahia grass BAFs indicate minimal plant uptake of indigenou s biosolids-borne TCC. Similar levels of accumulation were also documented in corn stov er from biosolids-amended field soil. Implications of the Preliminary TCC Pollutant Limits and Recommended TCC Research Priorities The preliminary TCC pollutant limits calcul ated in Chapter 9 are not intended to be suggestions for immediate changes to current biosolids land-applic ation regulations; rather, they should be used to guide future discussions and TCC research. The preliminary limits highlight the importance of assessing the risk of emerging contaminants of concern in biosolids and comparing calculated pollutant limits to concentr ations in the field under current land-application practices. One of the biggest challenges associated w ith conducting the bios olids-borne TCC risk assessment, and calculating the preliminary limits, was identifying the most appropriate receptor and health outcome endpoints. The risk asse ssment presented herein identified the most important biosolids-borne TCC exposure pathways and health outcomes for the most sensitive species endpoints using the best da ta available. The risk of bi osolids-borne TCC was found to be much greater for wildlife than for humans. Ef forts to reduce biosolids-borne TCC risks to the most sensitive ecological species can be optimized with an improved understanding of the parameters in the relevant exposure pathways and by confirming that the selected health 215

PAGE 216

outcome endpoints are based on complete toxicity da ta. Thus, future research efforts should be prioritized to address Exposure Pathways 10 and 16, and to fill remaining TCC toxicity gaps. Research Priority #1: Improve Characterizati on of the Most Sensitive Exposure Pathways The two most critical pathways influencing the preliminary TCC pollutant limit calculations were Pathways 10 (i.e. biosolids soil soil organism predator) and 16 (i.e. biosolids soil surface water aquatic organism). The multiple parameters used to calculate the corresponding adjusted HI values for each pathway (Equations 10-1 and 10-2) were reviewed to identify research needed to improve characte rizations of exposures an d health outcomes for the most sensitive ecological endpoints. Pathway #10: HI = (ECd FI / BW) / RfD CF / 0.5 (10-1) where: ECd = TCC concentration in earthworm (d.w.) FI = earthworm predator food intake rate BW = bodyweight RfD = reference dose CF = unit conversion factor 0.5 accounts for assumed 50% bioavailability Pathway #16: HI = SWC / RfD CF (10-2) where: SWC = surface water concentration Pathway 10-specific research priorities The ECd calculations for Pathway 10 (Equation 10 -1) relied on benc h-top study results (Chapter 6) derived from the analysis of TCC bioaccumulation in worms incubated in jars containing soils amended with spiked biosolids. In the field, the impacts of environmental processes, the effects of TCC aging, and the ability of worms to migrate to and from 216

PAGE 217

contaminated surface soil are expected to re duce bioaccumulation of biosolids-borne TCC. Future studies of bioaccumulati on in earthworms living in fields receiving shortand long-term applications of biosolids (paired with appropriate control sites) w ould be particularly helpful to confirming whether earthworm uptak e in the laboratory accurately models earthworm uptake in the real world, and to quan tifying typical TCC concentrati ons in earthworms consumed by wildlife. The RfD used to calculate the HI values for Pathway 10 (25 mg kg b.w.-1 d-1) was derived from a study in which rats received TCC in diet ary feed. Observed eff ects at doses >25 mg kg b.w.-1 d-1 in the study included anemia, reduced body weight, and increased liver and spleen weights. Studies surveying for similar health out comes in predators of ea rthworms in biosolidsamended soils would be useful for determining wh ether adverse health effects documented in the laboratory occur at the field level. If advers e effects were identified, however, laboratory-based feeding studies would be require d to determine if biosolids-bo rne TCC (and not other biosolids contaminants and/or environmental factors) was the causative agent. The effects of earthworm-borne TCC on pred ators will partially be a function of bioavailability. The adjusted HI values assu med 50% of TCC consumed in earthworm tissue was bioavailable, and, in effect, assumed that bioavailability of earthworm-borne TCC to predators was half the bioavailability of dietar y feed TCC fed to rats in the study used to calculate the RfD. Such an assumption may or may not be correct. Work to characterize the bioavailability of TCC in earthworm tissue is needed to evaluate the accuracy of the HI values. If TCC bioavailability in earthworm tissue is greater than estimated herein, the adjusted HI values will increase. Further, if TCC bioavailab ility in earthworm tissue is equal to that in dietary feed, the adjusted HI values will incr ease to the screening-level HI values, but if 217

PAGE 218

earthworm-borne TCC bioavailability is greater than that in dietar y feed, the true HI values will be greater than any estimates ca lculated herein. Preliminary TCC pollutant limits based on the RfD would also need to be adjusted if di fferences in bioavailability were identified. Bioavailability of chemicals in food is quant ified according to the assimilation efficiency (i.e. the amount sorbed per amount ingested in food) (Newman, 2001). The assimilation efficiency can be calculated by rearranging an equa tion used to calculate ti ssue concentrations in an organism following food and water intake: Ct = (kuC1 + RC2) / ke (1 e-k e t) (10-3) Or = [((Ctke) / (1 e-k e t)) kuC1] / RC2 (10-4) where: Ct = concentration in the organism (ug g-1) ke = elimination rate constant (h-1) t = time ku = uptake clearance (mL g-1 h-1) C1 = concentration in water (ug mL-1) C2 = concentration in food (ug g-1) R = specific ration, or mass of food consumed per mass of organism (g g-1 h-1) = assimilation efficiency (ug ug-1) Although C1 and C2 are known, and R, ke, and ku can be estimated from the literature, the biosolids-borne TCC assimilation efficiency from earthworms cannot be calculated due to the lack of data for TCC concentrations in earthwo rm predators. Work to characterize TCC blood concentrations, or bioaccumulation in the tissu e, of earthworm predators is required to characterize the bioavailability of earthworm-borne TCC. 218

PAGE 219

Pathway 16-specific research priorities One of the most critical parameters (i.e. surface water concentrati on) in the HI value calculations for Pathway 16 (Equation 10-2) wa s based on a single, unpublished, field-based biosolids-borne TCC runoff study (3.4 ng L-1). Additional studies of the varying impacts of soil type, biosolids characteristics, loading rate, and a pplication method on TCC concentrations in field runoff are needed before it is reasonable to generalize results and ca lculate final pollutant limits. The estimated surface water concentration attr ibutable to partitioning of sediment-bound TCC into the water column was ~4x greater than the measured TCC concentration in biosolidsamended soil runoff. The relatively large estimat ed TCC concentration was primarily due to the assumption that 100% of the sediment was biosolids contai ning 131 mg TCC kg-1. Another likely source of error, and an area warranting additional study, was the distribution coefficient used to estimate equilibrium partitioning of biosolids-borne TCC in the water column. The partition coefficient was based on the mean measured indigenous TCC Kd (Chapter 3), but assumed reversible desorption. Studies of 14C-TCC spiked to biosolid s and biosolids-amended soils (Chapter 4), however, suggest desorption is incomplete. Futu re studies of sediment-borne TCC partitioning in the water column would be usef ul for characterizing the relative contribution of TCC in land-applied biosolids to surface water contamination. Research needs relevant to both critical exposure pathways One of the most practical approaches to a ssessing the reasonableness of the preliminary pollutant limits and the estimated environmental con centrations used to calculate the HI values is characterization of TCC concentrations in soils that have received multiple applications of biosolids over time. The results of such a study could be used to assess whether TCC concentrations in biosolids-amended soils unde r current typical and worst-case application 219

PAGE 220

scenarios reach critical levels in the field. For example, the soil co ncentrations at two field sites in Illinois receiving three annual applications of biosolids at 20-37 Mg ha-1 rates (i.e. much greater than typical agronomic application rate s) were 0.043-0.51 mg TCC kg-1, and below the CPLR (0.53 mg TCC kg soil-1 when incorporated into th e top 15 cm, or 1.2 kg TCC ha-1). Amended soil should be collected fr om fields differing in appli cation rates and methods, number of applications, soil type, use (e.g. pasture, crop production), a nd geographical distribution to obtain samples representative of land-applicati on scenarios across the United States. Research Priority #2: Fill Remaining TCC Toxicity Gaps The toxicity data used to estimate the ri sks of biosolids-borne TCC to humans and terrestrial vertebrates describe d such adverse effects as an emia, reduced body weight, and increased liver and spleen weights. If new TCC toxicity data iden tify adverse health effects at lower concentrations than previously documen ted, the biosolids-borne TCC risk assessment would have to be updated using the most recent information. Two additional areas of concern with regard to possible adverse health outcomes associated with TCC include endocrine effects and antimicrobial resistance development, but the information currently available is insufficient to estimate potential associated human or ecological health risks. Triclocarban enhanced estradiolor testoste rone-dependent activation of ER (estrogen receptor)and AR (androgen receptor)-respons ive gene expression in recombinant cell bioassays, but induced little or no activity al one (Ahn et al., 2008). Chen et al. (2008) documented increased accessory sex organ and repr oductive tract tissue weights in castrated rats injected with testosterone propionate and fed TC C-spiked chow (0.25% wt/wt) for 10 d. Results suggest TCC might act as an endoc rine enhancer rather than an endocrine disruptor. The study authors suggest future research on TCC-induced endocrine enhancement should address impacts on normal physiological functions and/or reproduction in both men and women, potential 220

PAGE 221

harmful effects to women with ER-positive breast cancer, and interaction of TCC with oral contraceptives, hormone replacement therapy, sy nthetic androgens, and glucocorticoid therapy. Studies addressing potential TCC resistance development are also limited. Agar minimum inhibitory concentration (MIC) tests compari ng the susceptibilities of MRSA (methicillin resistant staphylococcus aureus) with MSSA (met hicillin susceptible st aphylococcus aureus) and VRE (vancomycin-resistant enterococci) with VSE (vancomycin-sensitive enterococci) to triclocarban found no difference be tween the respective resistant a nd susceptible st rains (Suller and Russell, 1999). Subsequent efforts to i nduce TCC tolerance (as measured by increased MICs) in E. coli and Pseudomonas aeruginosa through such techniques as paper disk diffusion, cup plate diffusion, and gradient agar plates were unsuccessful (Walsh et al., 2003). Future research utilizing the pollutioninduced-community-tolerance (PICT) approach could be used to characterize potential TCC resistance development in complex microbial communities of biosolids-amended soils. The PICT concept is founded on the assumption that communities chronically exposed to a contaminant will eventu ally develop an increased tolerance to the contaminant (Newman and Clements, 2007). The increased tolerance might be caused by death of the most sensitive community members, and/or physiological and genetic adaptations. The resulting shift in community structure could be assessed with a va riety of short-term ecotoxicological tests, which might include, for example, monitoring changes in substrate utilization patterns or leucine incorporation. Pollution-i nduced-community-tolerance is confirmed if communities previously exposed to the contaminant of intere st are less susceptible to increased concentrations than communities from a contaminant-free environment. 221

PAGE 222

Applications to the USEPA Data Requirements for the Antimicrobial Pesticides Proposed Rule The importance of filling data gaps for TCC (and other antimicrobial compounds), and the necessity of considering the unique effects of a biosolids matrix, is highlighted in the USEPA 40 CFR Parts 158 and 161 Data Requirements for Antimicrobial Pesticides Proposed Rule published for public comment in October, 2008. An expressed goal of the updated data requirements is to address the issue of down-th e-drain antimicrobials that reach WWTPs. The document thoroughly outlines data requirements for compounds in the aquatic environment and food intended for human consumption, but fails to explicitly require measurement of antimicrobial physicochemical properties, concentrat ions in biosolids, and environmental fate in biosolids-amended soil systems. Results of th e project herein demonstrate the necessity of considering risks associated with biosolids-borne antimicrobials, and supports the inclusion of similar efforts in the proposed rule requireme nts. The work measuring TCC solubility, TCC Kow, and biosolids-borne TCC toxicity demonstrates that an understanding of down-the-drain antimicrobial fate, and a second expressed goal of the proposed rule (i.e. facilitate well-informed regulatory decisions protective of human and envi ronmental health), can be enhanced by the use of physical measurements obtained according to the standardized OPPTS test methods appropriately adapted to the unique situa tions of biosolids impacted systems. 222

PAGE 223

Submission of project proposal, identification of assessment endpoints, and development of conceptual model (Figure 9-2) Biosolids-borne TCC concentration: 39 mg kg-1(95thpercentile in TNSSS, 2009a) 16 human and ecological exposure pathways (Table 9-2) Biosolids application scenarios Worst-case 1-time 50 Mg ha-1100-year 5 Mg ha-1annually Data generation and collection, including: physicochemical properties partitioning biosolids-borne TCC concentrations leachability plant uptake earthworm toxicity earthworm accumulation microbial toxicity half-life Calculation of screening-level hazard indices (HI values) HI values < 1 all human exposure pathways 4 ecological exposure pathways HI values > 1 Biosolids soil soil organismpredator Biosolids soil surface water aquaticorganism Reevaluation using measured data Calculation of Preliminary Biosolids-Borne TCC Pollutant Limits CPLR: 1.2 kg TCC ha-1APLR: 0.6 kg TCC ha-1y-1Ceiling concentration limit: 277 mg TCC kg biosolids-1Pollutant concentration limit: 1.2 mg TCC kg biosolids-1 Adjusted HI values > 1 Biosolids soil soil organismpredator Problem formulation Analysis Risk characterization Risk management Submission of project proposal, identification of assessment endpoints, and development of conceptual model (Figure 9-2) Biosolids-borne TCC concentration: 39 mg kg-1(95thpercentile in TNSSS, 2009a) 16 human and ecological exposure pathways (Table 9-2) Biosolids application scenarios Worst-case 1-time 50 Mg ha-1100-year 5 Mg ha-1annually Data generation and collection, including: physicochemical properties partitioning biosolids-borne TCC concentrations leachability plant uptake earthworm toxicity earthworm accumulation microbial toxicity half-life Calculation of screening-level hazard indices (HI values) HI values < 1 all human exposure pathways 4 ecological exposure pathways HI values > 1 Biosolids soil soil organismpredator Biosolids soil surface water aquaticorganism Reevaluation using measured data Calculation of Preliminary Biosolids-Borne TCC Pollutant Limits CPLR: 1.2 kg TCC ha-1APLR: 0.6 kg TCC ha-1y-1Ceiling concentration limit: 277 mg TCC kg biosolids-1Pollutant concentration limit: 1.2 mg TCC kg biosolids-1 Adjusted HI values > 1 Biosolids soil soil organismpredator Submission of project proposal, identification of assessment endpoints, and development of conceptual model (Figure 9-2) Biosolids-borne TCC concentration: 39 mg kg-1(95thpercentile in TNSSS, 2009a) 16 human and ecological exposure pathways (Table 9-2) Biosolids application scenarios Worst-case 1-time 50 Mg ha-1100-year 5 Mg ha-1annually Data generation and collection, including: physicochemical properties partitioning biosolids-borne TCC concentrations leachability plant uptake earthworm toxicity earthworm accumulation microbial toxicity half-life Calculation of screening-level hazard indices (HI values) HI values < 1 all human exposure pathways 4 ecological exposure pathways HI values > 1 Biosolids soil soil organismpredator Biosolids soil surface water aquaticorganism Reevaluation using measured data Calculation of Preliminary Biosolids-Borne TCC Pollutant Limits CPLR: 1.2 kg TCC ha-1APLR: 0.6 kg TCC ha-1y-1Ceiling concentration limit: 277 mg TCC kg biosolids-1Pollutant concentration limit: 1.2 mg TCC kg biosolids-1 Adjusted HI values > 1 Biosolids soil soil organismpredator Problem formulation Analysis Risk characterization Risk management 131 mg kg-1Figure 10-1. Summary of the biosolids-borne triclocarban (TCC) risk assessment process 223

PAGE 224

WORKS CITED Ahn, K.C., Zhao, B., Chen, J., Cherednichenko, G., Sanmarti, E., Denison, M., Lasley, B., Pessah, I., Kultz, D., Chang, D., Gee, S.J., Hammock, B.D., 2008. In vitro biologic activities of the antimicrobials triclocarban, its analogs, and triclosan in bioassay screens: receptor-based biassay screens. Environmental Health Perspectives 116, 1203-1208. Alexander, M., 1999. Biodegradation and Bior emediation. Academic Press, San Diego, California. Anderson, J.P., 1982. Soil respiration, in Page, A.L., Miller, R.H., Keeney, D.R. (Eds.), Methods of Soil Analysis, Part 2 Chemical and Microbiological Properties, Second Edition, pp.842-842. Aronson, D., Boethling, R., Howard, P., and Stitele r, W., 2006. Estimating biodegradation halflives for use in chemical screening. Chemosphere 63, 1953-1960. Banks, M.K., Schwab, A.P., Cofield, N., Alleman, J.E., Switzenbaum, M., Shalabi, J., Williams, P., 2006. Biosolids-amended soils: Part I. Eff ect of biosolids app lication on soil quality and ecotoxicity. Water Environment Research 78, 2217-2230. Barns, S.M., Takala, S.L., Kuske, C.R., 1999. Wide distribution and diversity of members of the bacterial kingdom Acidobacterium in the e nvironment. Applied and Environmental Microbiology 65, 1731-1737. Bastian, F., Bouziri, L., Nicolardot, B., Ranjard, L., 2009. Impacts of wheat straw decomposition on successional patterns of soil microbial community structure. Soil Biology and Biochemistry 41, 262-275. Beaver, D., Roman, D., and Stoffel, P. 1956. Th e preparation and bacteriostatic activity of substituted ureas. Journal of the American Chemical Society 79, 1236-1245. Beyer, W.N.; Connor, E.E.; Gerould, S., 1994. Estim ates of soil ingestion by wildlife. Journal of Wildlife Management 58, 375-382. Boese, B.L., Ozretich, R.J., Lamberson, J.O., Swar tz, R.C., Cole, F.A., Pelletier, J., Jones, J., 1999. Toxicity and phototoxicity of mixtur es of highly lipophilic PAH compounds in marine sediments: Can the PAH model be extra polated? Environmental Contamination and Toxicology 36, 270-280. Bonin, A.M.; Arkauskas, A.P.; Angus, D.S.; Baker, R.S.U.; Gallagher, C.H.; Greenoak, G.; Lane Brown, M.M.; Meher-Homji, K.M.; Reeve, V., 1982. UV-absorbing and other sunprotecting substances: genot oxicity of 2-ethylhexyl Pmethoxycinnamate. Mutation Research 105, 303-308. Boyle, M., Paul, E.A., 1989. Nitrogen transformati ons in soils previously amended with sewage sludge. Soil Science Society of America Journal 53, 740-744. 224

PAGE 225

Bromilow, R.H., Chamberlain, K., 1989. Designing molecules for systemicity, in Atkin, R.K., Clifford, D.R. (Eds.), Mechanisms and Regulation of Transport Processes. Monograph 18, British Plant Growth Regulator Group, pp.113-128. Calabrese, E., 1992. ATSDR Public Health Risk Assessment Guidance Manual. CRC Press, Boca Raton, Florida. Chalew, T.E.A.; Halden, R.U., 2009. Environmental exposure of aquatic a nd terrestrial biota to triclosan and triclocarban. JAWRA Journal of the American Water Resources Association 45, 4-13. Chambliss, C., Miller, P., Lord, E., 2001. Florida Cow-Calf Management, second ed. UF/IFAS, Gainesville, Florida. Chen, J., Ahn, K., Gee, N., Mohamed, M., Duleba A., Zhao, L., Gee, S., Hammock, B., Lasley, B., 2008. Triclocarban enhances testosterone action: A new type of endocrine disruptor? Endocrinology 149, 173. Chilom, G., Bestetti, G., Sello, G., Rice, J.A., 2004. Formation of bound residues by naphthalene and cis-naphthalene1,2-dihydrodiol. Chemosphere 56, 853-860. Chinault, S., 2007. Agronomic and environmental characterization of phosphorus in biosolids produced and/or marketed in Florida. (Masters thesis, Univers ity of Florida, 2007). Retrieved January, 2009 from UF Online Dissertations. Chinault, S., OConnor, G.A., 2008. Phosphorus release from a biosolids-amended sandy spodosol. Journal of Environmental Quality 37, 937-943. Chiou, C.T., 2002. Partition and Adsorption of Organic Contaminants in Environmental Systems. John Wiley & Sons, Inc., Hoboken, New Jersey. Chu, S., Metcalfe, C., 2007. Simultaneous determ ination of triclocarb an and triclosan in municipal biosolids by liquid chromatogra phy tandem mass spectrometry. Journal of Chromatography 1164, 212-218. Chung, N.; Alexander, M., 1999. Effect of sequest ration and bioavailabili ty of two polycyclic aromatic hydrocarbons. Environmental Science and Technology 33, 3605-3608. Cofield, N., Banks, M.K., Schwab, A.P., 2008. Lability of polycyclic aromatic hydrocarbons in the rhizosphere. Chemosphere 70, 1644-1652. Collins, C., Fryer, M., Grosso, A., 2006. Plan t uptake of non-ionic organic chemicals. Environmental Science and Technology 40, 45-52. Connell, D.W., Markwell, R.D., 1990. Bioaccumulation in the soil to earthworm system. Chemosphere 20, 91-100. 225

PAGE 226

Coogan, M.A., Edziyie, R.E., La Point, T.W., Venables, B.J., 2007. Algal bioaccumulation of triclocarban, triclosan, and met hyl-triclosan in a North Texa s wastewater treatment plant receiving stream. Chemosphere 67, 1911-1918. Coogan, M.A., La Point, T.W., 2008. Snail bi oaccumulation of triclo carban, triclosan, and methyltriclosan in a North Texas, USA, str eam affected by wastewater treatment plant runoff. Environmental Toxicology and Chemistry 27, 1788-1793. Cornelis, P., 2008. Pseudomonas: genomics and molecular biology. Caister Academic Press, Norfolk, United Kingdom. Cousins, I.T., Mackay, D., 2001. Strategies for including vegetation compartments in multimedia models. Chemosphere 44, 643-654. Darbre PD., 2006. Environmental oestrogens, cosm etics and breast cancer. Best Practice and Research Clinical Endocri nology and Metabolism 20, 121-143. Daughton, C., 2004. PPCPs in the environment: Future research beginning with the end always in mind, in: Kummerer, K. (Ed.), Phar maceuticals in the Environment, second ed. Springer, Heidelberg, Germany, pp. 463-495. Daughton CG, Ternes TA., 1999. Pharmaceuticals and personal care products in the environment: agents of subtle change? Environmental Health Perspectives 107 (Supplement 6), 907-938. Di Toro, D. M., 1985. A particle interaction model of rever s ible organic chemical sorption. Chemosphere 14, 1503-1538. Di Toro, D.M., Zarba, C.S., Hansen, D.J., Be rry, W.J., Schwartz, R.C., Cowan, C.E., Pavlou, S.P., Allen, H.E., Thomas, N.A., Paquin, P. R., 1991. Technical basis for establishing sediment quality criteria for nonionic organic-chemicals usin g equilibrium partitioning. Environmental Toxicology and Chemistry 10, 1541-1583. Dokianakis, S.N., Kornaros, M.E., Lyberatos, G., 2004. On the effect of pharmaceuticals on bacterial nitrate oxidation. Water Science a nd Technology 50, 341-346. DuarteDavidson, R., Jones, K.C., 1996. Screen ing the environmental fate of organic contaminants in sewage sludge applied to agricu ltural soils. 2. The potential for transfers to plants and grazing animals. Scienc e of the Total Environment 185, 59-70. Dytczak, M.A., Londry, K.L., Cleszkiewicz, J.A. 2008. Nitrifying genera in activated sludge may influence nitrification rates. Water Environment Research 80, 388-396. Epstein, E., 2003. Land Application of Sewage Sludge and Biosolids. CRC Press, Boca Raton, Florida. 226

PAGE 227

Er, F., Ogut, M., Mikaylov, F.D., Mermut, A.R ., 2004. Important factor s affecting biosolids nitrogen minerlization in soils. Communica tions in Soil Science and Plant Analysis 35, 2327-2343. Federal Register, 2008. Data Re quirements for Antimicrobial Pe sticides; Proposed Rule 73, 196. http://www.regulations.gov/fdmspublic/c omponent/main?main=DocketDetail&d=EPAHQ-OPP-2008-0110 Accessed April 15, 2009. Fierer, N., Jackson, J.A., Vilgalys, R., Jackson, R.B., Assessment of soil microbial community structure by use of taxon-specific quantitative PCR assays. Applied and Environmental Microbiology 71, 4117-4120. Franco-Hernandez, O., Dendooven, L., 2006. Dynamics of C, N and P in soil amended with biosolids from a pharmaceutical industry producing cephalosporines or third generation antibiotics: A laboratory st udy. Bioresource Technology 97, 1563-1571. Fuhr, F., Ophoff, H., 1998. Pesticide bound residues in soi l: Workshop, 3-4 September 1996. Garten Jr., C.T.; Trabalka, J.R., 1983. Evaluation of models for predicting terrestrial food chain behavior of xenobiotics. Envir onmental Science and Technology 17, 590-595. Gevao, B., Semple, K.T., Jones, K.C., 2000. Bound pesticide residues in soil: a review. Environmental Pollution 108, 3-14. Gledhill, W.E., 1975. Biodegradation of 3,4,4`-trichlorocarbanilide, TCC, in sewage and activated sludge. Wa ter Resources 9, 649-654. Gobas, F., Clark, K.E., Shiu, W.Y., Mackay, D., 1989. Bioconcentration of polybrominated benzenes and biphenyls and related super hydrophobic chemicals in fish role of bioavailability and elimination in the feces. Environmental Toxicology and Chemistry 8, 231-245. Gouin, T., Cousins, I., Mackay, D., 2004. Compar ison of two methods for obtaining degradation half-lives. Chemosphere 56, 531-535. Guy, R., Hadgraft, J., 2002. Transdermal Drug Delivery. Informa Health Care, New York, New York. Halden, R.U., Paull, D.H., 2004. Analysis of triclocarban in aquatic samples by liquid chromatography electrospray ionization mass spectrometry. Environmental Science and Technology 38, 4849-4855. Halden, R.U., Paull, D.H., 2005. Co-occurrence of triclocarban and triclosan in U.W. water resources. Environmental Science and Technology 39, 1420-1426. 227

PAGE 228

Halden, R., 2007. Emerging knowledge on emerging contaminants. Northeast Water Science Forum, August 8-9, Portland, Maine. http://www.neiwpcc.org/ppcpconference/ppcpP resentations07.asp. Accessed March 1 2009. Hansch, C., Quinlan, J., Lawrence, G., 1968. The linear free-energy relationships between partition coefficients and the aqueous solubil ity of organic liquids. Journal of Organic Chemistry 33, 47-350. He, Z., Alva, A., Yan, P., Li, Y., Calvert, D., Stoffella, P., Banks, D., 2000. Nitrogen mineralization and transformation from composts and biosolids during field incubation in a sandy soil. Soil Science 165, 161-169. Heidler, J., Sapkota, A., Halden, R., 2006. Partitioning, persistence, and accumulation in digested sludge of the topical antiseptic triclocarban during wastewater treatment. Environmental Science and Technology 40, 3634-3639. Hickman, Z.A.; Reid, B.J., 2005. Towards a more appropriate water based extraction for the assessment of organic contaminant availa bility. Environmental Pollution 138, 299-306. Hulster, A., Muller, J.F., Marschner, H., 1994. So il-plant transfer of pol ychlorinated dibenzo-pdioxins and dibenzofurans to vegeta bles of the cucumber family ( Cucurbitaceae). Environmental Science and Technology 28, 1110-1115. Integrated Risk Information System (IRIS), 2009. IRIS Glossary. www.http://www.epa.gov/IRIS/help_gloss.htm#c. Accessed April 2, 2009. Jetten, M.S.M., 2008. Special issue: The microbial nitrogen cycle. E nvironmental Microbiology 10, 2903-2909. Johnson, R.R., Navone, R., Larson, B.S., 1963. An unusal epidemic of methemoglobinemia. Pediatrics 31, 222-225. Johnson, D.L., Jones, K.C., Langdon, C.J., Piearce, T.G., Semple, K.T., 2002. Temporal changes in earthworm availability and extrac tability of polycyclic aromatic hydrocarbons in soil. Soil Biology 34, 1363-1370. Jussara, R.; Koskinen, W.; Sadowsky, M., 2006. In fluence of soil aging on sorption-desorption and bioavailability of simazine. Journal of Agricultural and Food Chemistry 54, 13731379. Kah, M., Brown, C.D., 2008. Log D: lipophilicit y for ionisable compounds. Chemosphere 72, 1401-1408. Kelsey, J.W., Kottler, B.D., Alexander, M., 1997. Selective chemical extractants to predict bioavailability of soil-aged organic chemicals. Environmental Science and Technology 31, 214-217. 228

PAGE 229

Kelsey J. W., Kottler B. D., Alexander M., 1997. Selective chemical extractants to predict bioavailability of soil-aged organic chemicals. Environmental Science and Technology 31, 214-217. Kemmitt, S., Wright, D., Goulding, K., Jones, D., 2006. pH regulation of carbon and nitrogen dynamics in two agricultural soils. Soil Biology and Biochemistry 38, 898-911. Kerr, K.M., McAvoy, D.C., 1999. Determination of so rption coefficients for triclocarban (TCC). Experiment Summary (E98-001). Corporat e Professional and Regulatory Services Division, Environmental Sciences Department, Procter & Gamble. Khalil, M., Hossain, M., Schmidhalter, U., 2005. Carbon and nitrogen mineralization in different upland soils of the subtropics treate d with organic materials. Soil Biology and Biochemistry 37, 1507-1518. Kielak, A., Pijl, A.S., van Veen, J.A., Kowalchuk, G.A., Phylogenetic diversity of Acidobacteria in a former agricultural soil. International So ciety of Microbial Ecology Journal 3, 378382. Klaus, U., Pfeifer, T., Spiteller, M., 2000. APC I-MS/MS: A Powerful tool for the analysis of bound residues resulting from the interaction of pesticides with DOM and humic substances. Environmental Science and Technology 34, 3514-3520. Lapen, D.R., Topp, E., Metcalfe, C.D., Li, H., Ed wards, M., Gottschall, N., Bolten, P., Curnoe, W., Payne, M., Beck, A., 2008. Pharmaceutical a nd personal care products in tile drainage following land application of municipal bioso lids. Science of the Total Environment 399, 50-65. Laturnus, F., von Arnold, K., Gron, C., 2007. Organic contaminants from sewage sludge applied to agricultural soils False alarm rega rding possible problems for food safety? Environmental Science and Po llution Research 14, 53-60. Lee, S., Gan, J., Liu, W.P., Anderson, M.A., 2003. Evaluation of K-d underestimation using solid phase microextraction. Enviro nmental Science and Technology 37, 5597-5602. Lichstein, H., Soule, M., 1944. Studies of the effect of sodium azide on microbic growth and respiration. Journal of Bacteriology 47, 221-230. Liste, H.-H., Alexander, M., 2002. Butanol extracti on to predict bioavailability of PAHs in soil. Chemosphere 46, 1011-1017. Luby, S., Agboatwalla, M., Feinkin, D.R., Painter, J., Billhimer, W., Altaf, A., Hoeskstra, R.M., 2005. Effect of handwashing on child health: a randomized controlled trial. Lancet 366, 225-233. Luo, Y., Zhou, X., 2006. Soil Respiration and the Environment. Academic Press, Burlington, Massachusetts. 229

PAGE 230

Lyman, W., Reehl, W., Rosenblatt, D., 1982. Handbook of Chemical Property Estimation Methods: Environmental Behavior of Organic Compounds. McGraw-Hill, New York, New York. Macleod, C.J.A.; Semple, J.T., 2003. Sequential extr actions of low concentrations of pyrene and formation of non-extractable residues in ster ile and non-sterile soils. Soil Biology and Biochemistry 35, 1443-1450. McDonnell, G.E., 2007. Antisepsis, Disinfection, and Sterilization. Wiley-Blackwell, Hoboken, New Jersey. Meylan, W.M., Howard, P.H., 1994a. Upgrade of PCGEMS Water Solubility Estimation Method (May 1994 Draft). prepared for Robert S. Boethling, U.S. Environmental Protection Agency, Office of Pollution Prevention a nd Toxics, Washington, DC; prepared by Syracuse Research Corporation, Environm ental Science Center, Syracuse, NY 13210. Meylan, W.M., Howard, P.H., 1994b. Validation of Water Solubility Estimation Methods Using Log Kow for Application in PCGEMS & EP I (Sept 1994, Final Report). prepared for Robert S. Boethling, U.S. Environmenta l Protection Agency, Office of Pollution Prevention and Toxics, Washington, DC; pr epared by Syracuse Research Corporation, Environmental Science Center, Syracuse, NY 13210. Meylan, W.M., Howard, P.H., Boethling, R.S., 1996. Improved method for estimating water solubility from octanol/water partition co efficient. Environmental Toxicology and Chemistry 15, 100-106. Miller, T., Heidler, J., Chillrud, S., DeLaquil, A., Ritchie, J., Mihalic, J ., Bopp, R., Halden, R.U., 2008. Fate of triclosan and triclo carban in estuarine sediment Environmental Science and Technology 42, 4570-4576. Mintel International Group, Ltd., 2004. Soap, bath, and shower products -US. http://www.marketresearch.com/product/display.asp?Pr oductID=1073723&SID=91040044 -329058991-384836483&kw=anti%2Dbacterial Accessed March 1, 2009. Nachabe, M., N. Shah, M. Ross, J. Vomacka., 2005. Evapotranspiration of two vegetation covers in a shallow water table environment. Soil Science Society of America Journal 69, 492499. Nagy, K.A., 1987. Field metabolic rate and food re quirement scaling in mammals and birds. Ecological Monographs 57, 111-128. National Research Council (NRC), 1983. Risk Assessment in the Federal Government: Managing the Process. National Academy Press, Washington, DC. National Research Council (NRC) 2002. Biosolids Applied to Land: Advancing Standards and Practices. United States Environmental Pr otection Agency, Committ ee on Toxicants and Pathogens in Biosolids Applied to Land. Na tional Academies Press, Washington, D.C. 230

PAGE 231

Nelson, D.W., Somers, L.E., 1996. Total car bon, organic carbon, and organic matter. Methods of soil analysis part 3chemical methods. Pages 995-1001. Madison, Wisconsin: Soil Science Society of America. Newman, M.C., 2001. Fundamentals of Ecotoxicology. Lewis Publishers, Boca Raton, Florida. Newman, M.C., Unger, M.A., 2003. Fundamentals of Ecotoxicology. CRC Press, Boca Raton, Florida. Newman, M.C., Clements, W.H., 2007. Ecot oxicology, A Comprehensive Treatment. CRC Press, Boca Raton, Florida. Obeysekera, J., J. Browder, L. Hornung, M.A. Ha rwell., 2004. The natural south Florida system I: Climate, geology, and hydrol ogy. Urban Ecosystems 3, 223-244. OConnor, G.A., 1996. Organic compounds in sl udge-amended soils and their potential for uptake by crop plants. Science of the Total Environment 185, 71-81. OConnor, G., Granato, T., Basta, N., 2001. Bi oavailability of biosolids molybdenum to soybean grain. Journal of Environmental Quality 30, 1653-1658. OConnor, G.A., Sarkar, D., Brinton, S.R., Elliot, H.A., Martin, F.G., 2004. Phytoavailability of biosolids phosphorus. Journal of Environmental Quality 33, 703-712. Opperhuizen, A., Schrap, S.M., 1988. Uptake effi ciencies of two polychlorobiphenyls in fish after dietary exposure to fi ve different concentrations. Chemosphere 17, 253-262. Perencevich, E., Wong, M., Harris, A., 2001. National and regional assessment of the antibacterial soap market: A step toward dete rmining the impact of prevalent antibacterial soaps. American Journal of Infection Control 29, 281-283. Rehm, B., 2008. Pseudomonas: Model Organism, Pathogen, Cell Factory. Wiley-VHM, Germany. Reid, B.J., Stokes, J.D., Jones, K.C., Semple K.T., 2000. Nonexhausti ve cyclodextrin-based extraction technique for the ev aluation of PAH bioavailability Environmental Science and Technology 34, 3174-3179. Roberts, T.R., 1984. Non-extractab le pesticide residues in soil and plants. IUPAC Reports on pesticides (17). Pure a nd Applied Chemistry 56, 945-956. Roman, D., Barnett, E., Balske, R., 1957. Cutaneous antiseptic activity of 3,4,4`trichlorocarbanilide. The Toilet Good Association 28, 12-13. Sapkota, A., Heidler, J., Halden, R., 2007. Dete ction of triclocarban a nd two co-contaminating chlorocarbanilides in US aquatic environments using isotope dilution liquid chromatography tandem mass spectrometry. Environmental Research 103, 21-29. 231

PAGE 232

Statistical Analysis System (SAS), 1989. SAS/ST AT users guide. Version 6. 4th ed. SAS Inst., Cary, NC. SCCP European Commission, 2004. Opinion on Triclocarban for Other Uses than as a Preservative. SCCP/0851/04. Scientific Committee on Consumer Products. http://ec.europa.eu/health/ph_risk/committees/04_sccp/docs/sccp_o_016.pdf Accessed March 6, 2009. Schnitzer, M., 1982. Organic matter characterizati on, in Page, A.L., Miller, R.H., Keeney, D.R. (Eds.), Methods of Soil Analysis, Part 2 Chemical and Microbiological Properties, Second Edition, pp.581-594. Semple, K., Morriss, A., Paton, G., 2003. Bioava ilability of hydrophobic or ganic contaminants in soils: Fundamental concepts and technique s for analysis. Europ ean Journal of Soil Science 54, 809-818. Singh, A., Ward, O.P., 2004. Biodegradation and Bioremediation. Springer, Berlin, Germany. Soil Science Society of America (SSSA), 2009. Glossary of Soil Science Terms. https://www.soils.org/pub lications/so ils-glossary Accessed March 11, 2009. Suller, M.T.E., Russell, A.D., 1999. Antibiotic a nd biocide resistance in methicillin-resistant Staphylococcus aureus and vancomycin-resis tant enterococcus. Journal of Hospital Infection 43, 281-291. Suter, G.W., Efroymson, R.A., Sample, B.E., Jones, D.S., 2000. Ecological Risk Assessment for Contaminated Sites. CRC Press, Boca Raton, Florida. Suter, G.W., 2007. Ecological Risk Assessm ent. CRC Press, Boca Raton, Florida. Talmage, S.S.; Walton, B.T., 1993. Food chain transf er and potential renal toxicity of mercury to small mammals at a contaminated terr estrial field site. Ecotoxicology 2, 243-256. Tang, J., Liste, H.-H., Alexander, M., 2002. Chemical assays of availabili ty to earthworms of polycyclic aromatic hydrocarbons in soil. Chemosphere 48, 35-42. TCC Consortium, 2002a. IUCLID Data Set, Report 201-14186B. http://www.epa.gov/chemrtk/tricloca/cl14186rs.pdf Accessed March 1, 2009. TCC Consortium, 2002b. High Production Volume (HPV) Chemical Challenge Program Data Availability and Screening Level Assessment for Triclocarban, CAS#: 101-20-2; Report 201-14186A. http://www.epa.gov/chemrtk/tricloca/c14186tp.pdf Accessed March 1, 2009. Trapp, S., McFarlane, J.C., 1995. Plant Contam ination, Modeling and Simulation of Organic Chemical Processes. CRC Press, Boca Raton, Florida. 232

PAGE 233

Trent University, 2009. Biosolids Amended Soil Level IV (BASL-4), v 1.10. Trent University, Peterborough, Ontario. United States Environmental Protection Agency (USEPA), 1989. Needs Survey: Report to CongressAssessment of Needed Publicly Owned Wastewater Treatment Facilities in the United States; EPA 430/9-89-001. USEPA, Washington, DC. United States Environmental Protection Agency (USEPA), 1993a. Determination of nitratenitrite by automated colorimetry, EPA Me thod 353.2 (Revision 2.0). Methods for the determination of inorganic substances in environmental samples. EPA/600/R-93/100. USEPA, Cincinnati, Ohio. United States Environmental Protection Agency (USEPA), 1993b. Determination of ammonia nitrogen by semi-automated colorimetry, EPA Method 350.1 (Revision 2.0). Methods for the determination of inorganic substances in environmental samples. EPA/600/R-93/100. USEPA, Cincinnati, Ohio. United States Environmental Protection Agency (USEPA), 1993d. Reference dose (RfD): Description and use in health risk assessments, Background document 1A. http://www.epa.gov/IRIS/rfd.htm Accessed April 20, 2009. United States Environmental Protection Agency (USEPA), 1993c. Wildlife Exposure Factors Handbook. EPA/600/R-93/187. USEPA, Washington, DC. United States Environmental Protection Agency (U SEPA), 1995. A Guide to the Biosolids Risk Assessments for the EPA Part 503 Rule. EPA/832-B-93-005. USEPA, Washington, DC. United States Environmental Protection Agency (USEPA), 1996a. Pr oduct Properties Test Guidelines, OPPTS 850.6200, Earthworm Subchroni c Toxicity Test. OPPTS, Washington, D.C. United States Environmental Protection Agen cy (USEPA), 1996b. Product Properties Test Guidelines, OPPTS 850.5100, Soil Microbial Co mmunity Toxicity Test. OPPTS, Washington, D.C. United States Environmental Protection Agency (USEPA), 1996c. Pr oduct Properties Test Guidelines, OPPTS 830.7840, Water Solubility: Column Elution Method; Shake Flask Method. OPPTS, Washington, D.C. United States Environmental Protection Agen cy (USEPA), 1996d. Product Properties Test Guidelines, OPPTS 830.7550, Part itioning Coefficient (n-octanol/water); Shake Flask Method. OPPTS, Washington, D.C. United States Environmental Protection Agency (USEPA), 1997. Exposure Factors Handbook. National Center for Environmental Assessmen t, Office of Research and Development, Washington, DC. 233

PAGE 234

United States Environmental Protection Agency (USEPA), 1998a. Pr oduct Properties Test Guidelines, OPPTS 835.3300, Soil Biodegradation. OPPTS, Washington, D.C. United States Environmental Protection Agency (USEPA), 1998b. Guidelines for Ecological Risk Assessment. EPA/630/R-95/002F. Ri sk Assessment Forum, US Environmental Protection Agency, Washington, DC. United States Environmental Protection Agen cy (USEPA), 2002. Exposure Analysis for Dioxins, Dibenzofurans, and Co-planar Polych lorinated Biphenyls in Sewage Sludge. The Office of Wastewater, USEPA, Washington, DC. United States Environmental Protection Agen cy (USEPA), 2006. Terms of Environment: Glossary, Abbreviations and Acronyms. http://www.epa.gov/OCEPAterms/bterms.html Accessed March 11, 2009. United States Environmental Protection Agency (USEPA), 2008. Screening Level Hazard Characterization of High Production Volu me Chemicals, Sponsored Chemical: Triclocarban. High Production Volume Chem icals Branch, Risk Assessment Division, Office of Pollution, Prev ention, and Toxics. http://www.epa.gov/hpvis/hazchar/101202_T riclocarban_HC_INTERIM_March%202008. pdf Accessed March 1, 2009. United States Environmental Protection Agency (USEPA), 2009a. Targeted National Sewage Sludge Survey Sampling and Analysis Technical Report ("Technical Report", EPA-822-R08-016). USEPA, Washington, D.C. United States Environmental Protection Agency (USEPA), 2009b. Targeted National Sewage Sludge Survey Statistical Analys is Report ("Statistical Report", EPA-822-R-08-018). USEPA, Washington, D.C. United States Environmental Protection Agency (USEPA), 2009c. Estimation Programs Interface Suite for Microsoft Window s, v 3.20. United States Environmental Protection Agency, Washington, DC, USA. United States Environmental Protection Agency (USEPA), 2009d. Persistent Bioaccumulative Toxic (PBT) Profiler, v 1.203. United States Environmental Protection Agency, Washington, DC, USA United States Environmental Protection Agency (USEPA), 2009e. EPA PROBIT Analysis Program, v 1.5. Ecological Monitoring Resear ch Division, Environmental Monitoring Systems Laboratory, United States Environm ental Protection Agency, Cincinatti, OH, USA. Walkley, A. and I. A. Black.,1934. An examination of Degtjareff method for determining soil organic matter and a proposed modification of the chromic acid titration method. Soil Science 37, 29-37. 234

PAGE 235

Waller, N.J., Kookana, R.S., 2009. Effect of triclo san on microbial activity in Australian soils. Environmental Toxicology and Chemistry 28, 65-70. Walsh, S.E., Maillard, J.Y., Russell, A.D., 2003a. Activity and mechanisms of action of selected biocidal agents on Gram-positive and -negativ e bacteria. Journal of Applied Microbiology 94, 240-247. Walsh, S.E., Maillard, J.Y., Russell, A.D., Ca trenich, C.E., Charbonneau, D.L., Bartolo, R.G., 2003b. Development of bacterial resistance to several biocides and effects on antibiotic susceptibility. Journal of Hospital Infection 55, 98-107. White, J.C., Hunter, M., Nam, K.; Pignatello, J.J., Alexander, M., 1999. Correlation between biological and physical availabi lities of phenanthrene in so ils and soil humin in aging experiments. Environmental Toxicology and Chemistry 18, 1720-1727. Wild, A, 1993. Soils and the Environment: An Introduction. Cambridge University Press, New York, NY. Wild, S.R., Jones, K.C., 1992. Organic chemicals entering agricultural soil s in sewage sludges screening for their potential to transfer to cr op plants and livestock. Science of the Total Environment 119, 85-119. Wolna-Maruwka, A., Klama, J., Niewiadomsk a, A., 2007. Effect of fertilization using communal sewage sludge on respiration activity and counts of selected microorganisms in the grey brown podzolic soil. Polish J ournal of Environmen tal Studies 16, 899-907. World Health Organization (WHO), 2001. Report on Integrated Risk Assessment. WHO/IPCS/IRA/01/12. World Health Organization, Geneva, Switzerland. Xia, K., Kwon, J-W., Armbrust, K., 2008. Transf ormation and extractab ility of antibacterial agents triclosan and triclocarban in soils a nd biosolids-amended soils. Proceedings from the 2008 Joint Annual Meeting, October 5-9, 2008, Houston, Texas. http://a-cs.confex.com/crops/2008am/webprogram/Paper50897.html Accessed January 5, 2009. Yaron, B., Calvet, R., Prost, R., 1996. Soil Pollution, Processes and Dynamics. Springer, Heidelberg, Germany. Youssef, N.H., Elshahed, M.S., 2009. Diversit y rankings among bacterial lineages in soil. International Society of Micr obial Ecology Journal 3, 305-313. Ying, G., Yu, X., Kookana, R., 2007. Biological degr adation of triclocarban and triclosan in a soil under aerobic and anaerobi c conditions and comparison with environmental fate modeling. Environmen tal Pollution 150, 300-305. Yuan, F., Ran, W., Shen, Q., Wang, D., 2005. Characterization of nitrifying bacteria communities of soil from different ecologi cal regions of China by molecular and conventional methods. Biology and Fertility of Soils 41, 22-27. 235

PAGE 236

236 Zhao, F., Subramanya, N.T., Yu, C.P., Chu, K. H., 2006. Biodegradation of pharmaceuticals and personal care products (PPCPs) by Nitrosomonas europaea. Abstracts of the General Meeting of the American Society for Microbiology 106, 472.

PAGE 237

BIOGRAPHICAL SKETCH Liz is daughter to the adventurous Mark and Lisa Hodges, sister to the quick-witted musician/builder Mark Hodges, Jr., and the lucky wife of the fly-fishing religionist Samuel Snyder. She spent most of her first 28 years in the beautiful state of Florida, but had the good fortune to also live for two years at the foothill s of the Pakistan Himalayas, one year in the capital of Taiwan, and two years in the city of Atlanta, Georgia. Li z is, or has been, many things: gymnast, scientist, coach, runner, teacher and in-law to a fabulous family of Texans. She looks forward to soon adding mother to the list, and, in the company of her husband and future family, in search of work and advent ure. (Do not worry, Mom and Pops, you can park your Airstream anywhere.) With her hard-earned degrees in hand, Liz is hell-bent on making a living that makes the world a better place for cu rrent and future residents, allows her to continually grow as a scientist and an educator, and takes her to pl aces even her Dad hasnt been to yet.