Citation
Fate and Risk Assessment of Biosolids-Borne Ciprofloxacin (CIP) and Azithromycin (AZ)

Material Information

Title:
Fate and Risk Assessment of Biosolids-Borne Ciprofloxacin (CIP) and Azithromycin (AZ)
Creator:
Sidhu, Harmanpreet S
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (248 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Soil and Water Sciences
Committee Chair:
O'CONNOR,GEORGE A
Committee Co-Chair:
OGRAM,ANDREW V
Committee Members:
WILSON,PATRICK CHRISTOPHER
KRUSE,JASON KEITH
TOPP,EDWARD
Graduation Date:
5/4/2018

Subjects

Subjects / Keywords:
antibiotics -- azithromycin -- bioaccessibility -- bioaccumulation -- bioavailability -- biosolids -- ciprofloxacin -- risk -- toxicity
Soil and Water Sciences -- Dissertations, Academic -- UF
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Soil and Water Sciences thesis, Ph.D.

Notes

Abstract:
Ciprofloxacin (CIP) and azithromycin (AZ) are commonly prescribed antibiotics for various infections in humans and are, consequently, frequently detected in biosolids. Ecological and human health risks from biosolids-borne CIP and AZ are not well understood, but necessary for formulating policy on safe use and management of biosolids. A project funded by Water Environment & Reuse Foundation (WE&RF) was designed to identify and fill various data gaps in the fate of biosolids borne CIP and AZ and to facilitate a scientifically sound ecological and human health risk assessment. Data from a batch equilibration retention/release study formulated our central hypothesis that the limited bioaccessibility of strongly sorbed biosolids-borne CIP and AZ minimizes human and environmental health risks. Bioavailabilities of biosolids-borne CIP and AZ were assessed in subsequent organism response (plant, earthworm, and microbial systems) studies and (where applicable) correlated with chemical bioaccessibilities. The organism response data revealed limited bioavailability (plant bioaccumulation factor (BAF) values 0.01 (CIP) and 0.1 (AZ), depurated earthworm BAF values ~4 (CIP) and ~ 7 (AZ), minimal impacts on overall microbiota) of the biosolids-borne antibiotics under environmentally relevant scenarios. The data generated herein, and environmentally relevant data collected from pertinent literature, were utilized in a scientifically sound integrated risk assessment (IRA), following the World Health Organization framework. Human and ecological exposure hazards were identified and potential adverse effects from CIP and AZ in land-applied biosolids were assessed using a tier approach. The IRA, consistent with our hypothesis, estimated negligible risks from biosolids-borne CIP and AZ under real-world based biosolids management practices. Even unrealistically high exposures from land application of biosolids pose minimal human and ecological health risks. Preliminary pollutant limits, calculated based on the most sensitive organisms, suggest that long term real-world based land application of biosolids is without appreciable human and ecological health risks. Chemical load tracking is not needed for the majority of USA biosolids, but may be necessary for some biosolids that contain greater than 12 mg CIP/kg and 2.2 mg AZ/kg. The IRA needs refining by including more data, especially on biosolids-borne antibiotic resistance, before suggesting modifications to current land-application regulations. ( en )
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.
Thesis:
Thesis (Ph.D.)--University of Florida, 2018.
Local:
Adviser: O'CONNOR,GEORGE A.
Local:
Co-adviser: OGRAM,ANDREW V.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2020-05-31
Statement of Responsibility:
by Harmanpreet S Sidhu.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Embargo Date:
5/31/2020
Classification:
LD1780 2018 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

FATE AND RISK ASSESSMENT OF BIOSOLIDS BORNE CIPROFLOXACIN (CIP) AND AZITHROMYCIN (AZ) By HARMANPREET SINGH SIDHU A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2018

PAGE 2

2018 Harmanpreet Singh Sidhu

PAGE 3

To my parent, Varinder Singh and Narinder pal Kaur

PAGE 4

4 ACKNOWLEDGMENTS A fter more than 4 years of intensive work to d ay is an exuberant d ay for me as I give the finishing touches to my dissertation with this note of thanks to all the people responsible for my success. It has been an intense period of exponential growth at a scientific as well as a personal level The biggest thanks go to my mentor and major not only in scientific and academic but also individual and beyond to support me in my ups an d downs, both academic and personal, and made completion of this dissertation possible. I have learned a great deal about highest regards and have immense respect and gratitu de towards him. I would like to express gratitude to my committee member s Drs. Jason Kruse, Andrew Ogram, Edward Topp, and Chris Wilson. Their help throughout my research made my success possible. I have learned a lot from this exceptional group of peopl e, may it be valuable techniques, scientific and research methods or important life lessons. An extra thank you to Dr. Andrew Ogram for helping me with microbial studies and data analysis. I was fortunate to receive his wisdom and knowledge, which unargua bly made this dissertation possible. Thanks go to Drs. Liabin Huang, Hee Sung Bae, and Abid Al Agely for they, especially Liabin, assisted me and help ed me get acclimated to the world of microbial research techniques I would also like to thank Dr. Drew Mc Avoy for partially supporting my research and assisting in completion of this dissertation.

PAGE 5

5 A big thank you to Dr. Kuldip Kumar for procuring biosolids, soils, and earthworms that made this dissertation possible. I would also like to thank Dr. Ramesh Redd y for supporting me during periods of doubt His support helped me finish this dissertation. I would also like to thank Dr. Jonathan Judy and Caleb Gravesen for their help and support with my research Special thanks go to the Soil and Water Sciences D epar tment for funding my degree, the Water Environment & Re use Foundation (WE & RF ) for funding my research, and Mike Sisk for being an excellent departmental coordinator and helping me throughout my Ph.D. A final thank s to the two most important individuals in my life, my late parents for it is their teachings that always encourage me to be a better person and achieve success I will always be grateful to them at every moment of my life.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 11 LIST OF FIGURES ................................ ................................ ................................ ........ 13 LIST OF ABBR EVIATIONS ................................ ................................ ........................... 15 ABSTRACT ................................ ................................ ................................ ................... 17 CHAPTER 1 INTRODUCTION, BACKGROUND, AND PROJECT OBJECTIVES ...................... 19 Introduction ................................ ................................ ................................ ............. 19 Structure, Properties, Use, and Mechanism of Action ................................ ...... 20 Typical Expected Concentrations in Biosolids and Amended Soils .................. 26 Objectives ................................ ................................ ................................ ............... 29 Intermediate Objec tives ................................ ................................ .................... 30 Intermediate objective 1. To assess retention release behavior of CIP and AZ ................................ ................................ ................................ .... 30 Intermediate objective 2. To assess plant responses to biosolids borne CIP and AZ. ................................ ................................ ............................ 31 Intermediate objective 3. To assess microbial responses to biosolids borne CIP and AZ. ................................ ................................ .................. 32 Intermediate objective 4. To assess earthworm responses to biosolids borne CIP and AZ. ................................ ................................ .................. 32 Ultimate Objective: To Conduct Human and Ecological Health Risk Assessment of Biosolids borne CIP and AZ. ................................ ................. 33 2 RETENTION/RELEASE BEHAVIOR OF CIPROFLOXACIN AND AZITHROMYCIN ................................ ................................ ................................ .... 34 Synopsis ................................ ................................ ................................ ................. 34 Introduction ................................ ................................ ................................ ............. 35 Retention/Release Mechanisms ................................ ................................ ....... 35 Extent of Retention/Release ................................ ................................ ............. 37 Materials ................................ ................................ ................................ ................. 40 Methods ................................ ................................ ................................ .................. 41 Confirmation of No I sotope Exchange with the Surroundings .......................... 41 ................................ ....... 42 Ciprofloxacin Sorption onto Centrifuge Tubes ................................ .................. 46 Equilibration of Biosolids with CIP and AZ ................................ ........................ 46 Desorption from Biosolids Previously Equilibrated with CIP or AZ ................... 47

PAGE 7

7 Soil effects on desorption of biosolids borne CI P or AZ ............................. 47 TOrC competition effects on desorption ................................ ..................... 48 Specifically sorbed metal effects ................................ ................................ 48 Detection Limits and Statistical Analysis ................................ .......................... 49 Results and Discussion ................................ ................................ ........................... 50 CIP Sorption onto Centrifuge Tubes ................................ ................................ 50 ................................ ....... 51 Sorption ................................ ................................ ................................ ...... 51 Desorption ................................ ................................ ................................ .. 52 Isotherms ..................... 54 Desorption from Biosolids Previously Equilibrated with CIP or AZ ................... 61 Bioaccessibility versus Bioavailability ................................ ................................ ..... 63 Conclusions ................................ ................................ ................................ ............ 64 3 BIOSOLIDS BORNE CIPROFLOXACIN (CIP) AND AZITHROMYCIN (AZ): PLANT SYSTEM ................................ ................................ ................................ .... 66 Synopsis ................................ ................................ ................................ ................. 66 Introduction ................................ ................................ ................................ ............. 67 Materials ................................ ................................ ................................ ................. 73 Methods ................................ ................................ ................................ .................. 74 Plant Uptake and Phytotoxicity Study (Biosolids borne TOrCs) ....................... 74 Phytotoxicity Study in Soils (No Biosolids) ................................ ....................... 81 Definitive Phytotoxicity Tests in Soils (No Biosolids) ................................ ........ 82 Statistical Analysis ................................ ................................ ............................ 83 Results and Discussion ................................ ................................ ........................... 83 Phytotoxicity in Soil (no Biosolids) Definitive Tests for CIP .............................. 83 Plant Uptake Study (Biosolids borne TOrCs) ................................ ................... 84 Dry weight yields from the plant uptake study ................................ ............ 84 Uptake of biosolids borne CIP and AZ ................................ ....................... 89 Conclusions ................................ ................................ ................................ ............ 95 4 BIOSOLIDS BORNE CIPROFLOXACIN (CIP) AND AZITHROMYCIN (AZ): MICROBIAL SYSTEM ................................ ................................ ............................ 97 Synopsis ................................ ................................ ................................ ................. 97 Introduction ................................ ................................ ................................ ............. 97 Materials ................................ ................................ ................................ ............... 102 Methods ................................ ................................ ................................ ................ 103 Respiration ................................ ................................ ................................ ..... 107 RNA Extraction a nd Pre qPCR Treatment ................................ ..................... 107 Target Gene Selection for Real Time PCR (qPCR) ................................ ........ 108 Nitrogen cycle genes ( amoA and nirK/nirS) ................................ ............. 108 Phosphatase genes ( phoN and phoD ) ................................ ..................... 109 Ciprofloxacin ( qnrA qnrB qnrS ) and AZ ( ermB mefE ) resistance genes 109 Real Time qPCR ................................ ................................ ............................ 110 Standard Curve Development and qPCR Efficiency ................................ ...... 110

PAGE 8

8 Assessment of CIP and AZ Extractability and Stability ................................ ... 111 Chemical extraction methods ................................ ................................ ... 111 TLC analysis ................................ ................................ ............................ 113 Determination of compound mineralization ................................ .............. 114 Recoveries from methanol:water and ASE extraction methods ............... 116 Efficiencies and recoveries from TLC analysis ................................ ......... 117 Description of Treatments ................................ ................................ .............. 117 Detection Limits and Statistical Analysis ................................ ........................ 118 Results and Discussion ................................ ................................ ......................... 118 Respiration ................................ ................................ ................................ ..... 118 Microbial Re sponse to Biosolids borne CIP and AZ ................................ ....... 120 Ammonia oxidizing bacteria (AOB) gene (bacterial amoA ) ...................... 120 Ammonia oxidizing archaea (AOA) gene (archaeal amo A) ...................... 124 Nitrite reductase genes ( nirK and nirS ) ................................ .................... 124 Phosphatase genes ( phoN and phoD ) ................................ ..................... 125 Ciprofloxacin resistance genes ( qnrA qnrB and qnrS ) ........................... 125 Azithromycin resistance genes ( ermB and mefE ) ................................ .... 126 16S rRNA analysis ................................ ................................ ................... 128 Antibiotic Resistance Development and Spread: Possibilit ies and Unknowns 136 TLC Analysis ................................ ................................ ................................ .. 138 Extraction of CIP and AZ from The Solid Matrices ................................ ......... 140 Calcium chloride extraction ................................ ................................ ...... 140 Methanol:water extraction ................................ ................................ ........ 141 ASE extraction ................................ ................................ ......................... 141 Combustion ................................ ................................ .............................. 142 Chemical Extractability versus Bioavailability ................................ ................. 144 Microbial Response Study Limitations ................................ ................................ .. 145 Conclusions ................................ ................................ ................................ .......... 146 5 BIOSOLIDS BORNE CIPROFLOXACIN (CIP) AND AZITHROMYCIN (AZ): EARTHWORM SYSTEM ................................ ................................ ...................... 148 Synopsis ................................ ................................ ................................ ............... 148 Introduction ................................ ................................ ................................ ........... 149 Materials ................................ ................................ ................................ ............... 151 Methods ................................ ................................ ................................ ................ 152 Laboratory Study ................................ ................................ ............................ 152 Determination of bioaccumulation factors (BAF) ................................ ...... 155 3 H detection limits ................................ ................................ ................... 156 Field Study ................................ ................................ ................................ ..... 156 Statistical Anal ysis ................................ ................................ .......................... 157 Results and Discussion ................................ ................................ ......................... 158 Laboratory Study ................................ ................................ ............................ 158 TOrC toxicity to earthworms ................................ ................................ ..... 158 Bioaccumulation of CIP and AZ by earthworms ................................ ....... 158 Extraction and analysis of CIP and AZ from the solid matrices ................ 163 Field Study ................................ ................................ ................................ ..... 165

PAGE 9

9 Conclusions ................................ ................................ ................................ .......... 166 6 RISK ASSESSMEN T OF BIOSOLIDS BORNE CIPROFLOXACIN (CIP) AND AZITHROMYCIN (AZ) ................................ ................................ ........................... 168 Synopsis ................................ ................................ ................................ ............... 168 Introduction ................................ ................................ ................................ ........... 169 Integrated Risk Assessment Approach ................................ ................................ 171 Pathways Considered ................................ ................................ .................... 172 Selection of HEIs ................................ ................................ ............................ 174 Risk Characterization ................................ ................................ ..................... 175 Reference Dose Calculations ................................ ................................ ......... 175 Human toxicity benchmark (RfD) calculations ................................ .......... 175 Ecological toxicity benchmark (RfD) calculations ................................ ..... 176 Tiered Risk Assessment Approach ................................ ................................ 177 First tier (screening level) assessment ................................ ..................... 177 Second tier assessment ................................ ................................ ........... 178 Thir d tier assessment ................................ ................................ ............... 179 Surface Water Chemical Concentration Calculations ................................ ..... 183 Data Used in Risk Assessment ................................ ................................ ...... 186 Results ................................ ................................ ................................ .................. 186 First Tier (Screening Level) Risk Assessment ................................ ................ 186 Pathways of concern for AZ ................................ ................................ ..... 186 Pathways of concern for CIP ................................ ................................ .... 186 Second Tier Assessment ................................ ................................ ................ 199 Third Tier Assessment ................................ ................................ .................... 199 Puttin g the HQ Values into Perspective ................................ ......................... 201 Calculation of Biosolids borne CIP and AZ Pollutant Limits ................................ .. 204 Cumulative Pollutant Loading Rate (CPLR) ................................ ................... 204 Annual Pollutant Loading Rate (APLR) ................................ .......................... 205 Ceiling Concentration Limit ................................ ................................ ............. 205 Pollutant Concentration Limit ................................ ................................ .......... 206 Sources of Uncertainty and IRA Limitations ................................ .......................... 206 Conclusions ................................ ................................ ................................ .......... 210 7 SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS ................................ 212 Introduction ................................ ................................ ................................ ........... 212 Intermediate Objectives ................................ ................................ .................. 213 Intermediate objective 1. To assess retention release behavior of CIP and AZ ................................ ................................ ................................ .. 213 Intermediate objective 2. To assess plant responses to biosolids borne CIP and AZ. ................................ ................................ .......................... 213 Intermediate objective 3. To assess microbial responses to biosolids borne CIP and AZ. ................................ ................................ ................ 214 Intermediate objective 4. To assess earthworm responses to biosolids borne CIP and AZ. ................................ ................................ ................ 215

PAGE 10

10 Ultimate Objective: To Conduct Human and Ecological Health Risk Assessment of Biosolids borne CIP and AZ. ................................ ............... 215 Future Research Priorities ................................ ................................ .................... 216 APPENDIX A AXYS METHOD MLA 075: ANALYSIS OF AZITHROMYCIN AND CIPROFLOXACIN IN SOLID, AQUEOUS, AND TISSUE BY LC MS/MS ............. 219 Extraction ................................ ................................ ................................ .............. 219 Analysis ................................ ................................ ................................ ................ 219 Calibration ................................ ................................ ................................ ............. 219 Analyte Identification ................................ ................................ ........................... 219 Quantitation ................................ ................................ ................................ .......... 220 Reporting limits ................................ ................................ ................................ ..... 220 Quality Assurance/Quality Control ................................ ................................ ........ 220 Limitations to Performance Soil Samples ................................ ........................... 221 B MICROBIAL DNA QPCR DATA ................................ ................................ ............ 222 LIST OF REFERENCES ................................ ................................ ............................. 226 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 248

PAGE 11

11 LIST OF TABLES Table page 1 1 Select literature on, and properties of, CIP and AZ. ................................ ........... 24 1 2 Ciprofloxacin and AZ concentrations (mg/kg) in the 7 biosolids matrices and respective reporting limit (mg/kg) for each biosolids for each compound. .......... 28 2 1 Select properties of soils and biosolids used in retention/release studies (measured average values from duplicate samples). ................................ ......... 41 2 2 Concentrations (mg chemical per kg solid matrix) of CIP and AZ added to different solid matrices. ................................ ................................ ....................... 43 2 3 Sorption (Kds) and desorption (Kdd) partitioning coefficients predicted by linear model standard deviation (SD), Koc and Kcec values, average mass balance (%) SD, and hysteresis index (H), for CIP and AZ. ............................ 55 3 1 Properties and nutrient contents of soils and biosolids involv ed in the study. .... 74 3 2 Nutrient supplementation to plants from various sources. ................................ .. 80 3 3 Average dry weight yields (g) standard deviations (SDs) for radish, lettuce, and fescue grass grown in sand directly spiked with different concentrations of CIP (in the a bsence of biosolids). ................................ ................................ ... 83 3 4 Average dry weight yields (g) in the controls (no chemical or biosolids) and the biosolids amended treatments standard deviations (SDs) for radish, lett uce, and fescue grass. ................................ ................................ ................... 87 3 5 Average reporting limit (RL) standard deviation (SD) and corresponding average CIP and AZ concentr ations SDs in control and biosolids amended sand (mg/kg dry weight) for various treatments. ................................ ................. 90 3 6 Reporting limit (RL) stand ard deviation (SD) and corresponding chemical concentrations of the target TOrCs (mg/kg) in each repli cate of the three plants grown in the controls and the biosolids amended treatments. ................. 92 4 1 Properties of biosolids and manured sand used in the incubation study (measured average values from duplicate samples). ................................ ....... 103 4 2 Experimental design ................................ ................................ ......................... 107 4 3 Average CIP and AZ percent recoveries standard deviations (SDs), and % relative standard deviation (RSD) from methanol:water and ASE extraction methods. ................................ ................................ ................................ ........... 116

PAGE 12

12 4 4 Average CIP and AZ percent recoveries standard deviations (SDs), % relative standard deviation (RSD), and retardation factors (RFs) from TLC analysis ................................ ................................ ................................ ............ 117 5 1 Select properties of soils and biosolids used in the lab earthworm study (measured, average values from duplicate samples). ................................ ...... 153 5 2 Bioaccumulation factors (BAFs) standard deviations (SDs) for depurated and un depurated worms (on dry weight basis), and the average chemical (%) excret ed by the earthworms during depuration. ................................ ......... 161 6 1 Pathways of exposure and corresponding highly exposed individuals (HEI)s considere d in the risk assessment. ................................ ................................ ... 173 6 2 Azithromycin toxicity data ................................ ................................ ................. 180 6 3 Ciprofloxacin toxicity data ................................ ................................ ................. 181 6 4 Selected CIP and AZ references dose (RfD) (mg/kg d) for various HEI groups. ................................ ................................ ................................ ............. 183 6 5 Parameters and assumptions for calculating screening level hazard quotient (HQ) values. ................................ ................................ ................................ ..... 187 6 6 Azithromycin hazard quotient (HQ) values, and equations and assumptions to calculate HQ values. ................................ ................................ ..................... 193 6 7 Ciprofloxacin hazard quotient (HQ) values, and equations and assumptions to calculate HQ values ................................ ................................ ...................... 196 6 8 Second tier risk assessment hazard quotient (HQ) values for plants (pathway 8) and earthworms (pathway 9) ................................ ................................ ........ 200 6 9 Second tier risk assessment hazard quotient (HQ) values for predators (pathway 10) ................................ ................................ ................................ ..... 200 6 10 Third tier risk assessment hazard quotient (HQ) values ................................ ... 201

PAGE 13

13 LIST OF FIGURES Figure page 1 1 Structures of ciprofloxacin (CIP) and azithromycin (AZ) along with pKa values. ................................ ................................ ................................ ................ 22 1 2 Species distribution diagrams at environmentally relevant pH .......................... 23 2 1 Representative sorption/desorption isotherms for CIP and AZ. .......................... 60 3 1 Radish plants exposed to biosolids borne CIP. ................................ .................. 85 3 2 Fescue grass exposed to biosolids borne AZ ................................ .................... 86 3 3 Lettuce plants exposed to biosolids borne CIP. ................................ .................. 86 4 1 Incubation study set up (Constant air flow apparatus). ................................ ..... 106 4 2 A representative figure, with standard error bars, showing no effects of CIP or AZ treatments on microbial respiration in the biosolids or the amended manured sand. ................................ ................................ ................................ .. 119 4 3 Expression of bacterial amoA in biosolids. ................................ ....................... 128 4 4 Expression of bacterial amoA in manured sand. ................................ .............. 129 4 5 Expression of archaeal amoA in biosolids. ................................ ....................... 129 4 6 Expression of archaeal amoA in manured sand. ................................ .............. 130 4 7 Expression of nirS in biosolids. ................................ ................................ ......... 130 4 8 Expression of phoN in biosolids. ................................ ................................ ....... 131 4 9 Expression of phoN in manured sand. ................................ .............................. 131 4 10 Expression of phoD in biosolids. ................................ ................................ ....... 132 4 11 Expression of phoD in manured sand. ................................ .............................. 132 4 12 Expression of qnrS in biosolids. ................................ ................................ ........ 133 4 13 Expression of qnrS in manured sand. ................................ ............................... 133 4 14 Expression of mefE in biosolids. ................................ ................................ ....... 134 4 15 Expression of mefE in manured sand. ................................ .............................. 134

PAGE 14

14 4 16 Expression of ermB in biosolids. ................................ ................................ ....... 135 4 17 Expression of ermB in manured sand. ................................ .............................. 135 4 18 First order AZ degradation kinetics. ................................ ................................ .. 140 4 19 Percent chemical recoveries with standard error bars from biosolids using various fractionation schemes. ................................ ................................ ......... 143 4 20 Percent chemical recoveries with standard error bars from biosolids amended manured sand using various fractionation schemes. ........................ 143 5 1 CIP bioaccumulation factors (depurated worms; dry weight basis). ................. 162 5 2 AZ bioaccumulation factors (depurated worms; dry weight basis) ................... 162 5 3 Percent CIP recoveries from various chemical treatments using various fractionation schemes. ................................ ................................ ...................... 164 5 4 Percent AZ recoveries from various chemical treatments using various fractionation schemes. ................................ ................................ ...................... 164 B 1 AOB DNA quantification from controls and 95 th percentile TOrC treatments on day 0 and day 90. ................................ ................................ ........................ 222 B 2 qnrS DNA quantification from c ontrols and 95 th percentile TOrC treatments on day 0 and day 90. ................................ ................................ ........................ 222 B 3 ermB DNA quantification from controls and 95 th percentil e TOrC treatments on day 0 and day 90 ................................ ................................ ........................ 223 B 4 mefE DNA quantification from controls and 95 th percentile TOrC treatments on day 0 and day 90. ................................ ................................ ........................ 223 B 5 16S rRNA quantification from controls and 95 th percentile TOrC treatments on day 0 and day 90. ................................ ................................ ........................ 224 B 6 Representative figure showing KCl extractable NH 4 N and NOx, and water extractable P values over time for various CIP and AZ treatments in the biosolids. ................................ ................................ ................................ .......... 224 B 7 Representative figure showing KCl extractable NH 4 N and NOx, and water extractable P values over time for various CIP and AZ treatments in the manured san d. ................................ ................................ ................................ .. 225 B 8 Representative figure showing KCl extractable NH 4 N and NOx, and water extractable P values over time for various CIP and AZ treatments in the biosolids amended manured sand. ................................ ................................ ... 225

PAGE 15

15 LIST OF ABBREVIATIONS AOA Ammonia oxidizing archaea AOB Ammonia oxidizing bacteria ASE Accelerated solvent extraction AZ Azithromycin BAF Bioaccumulation factor CEC Cation exchange capacity CIP Ciprofloxacin Dow Distribution coefficient EPI Suite Estimation Programs Interface Suite IRA Integrated risk assessment Kd Soil water partitioning coefficient Koc Soil organic carbon water partitioning coefficient Kow n Octanol water partitioning coefficient LC MS/MS Liquid chromatography tandem mass spectrometry MDTA Minimum detectable true activity MIC Minimum inhibitory concentration MWRDGC Metropolitan Water Reclamation District of Greater Chicago NOAEC No observ ed adverse effect concentration NOEC No observed effect concentration P W HC Pot water holding capacity RF Retardation factor RfD Reference dose TOrC Trace organic chemical TLC Thin layer chromatography

PAGE 16

16 USEPA United States Environmental Protection Agency WE & RF Water Environment & Reuse Foundation WHO World Health Organization WWTP Wastewater treatment plant

PAGE 17

17 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FATE AND RISK ASSESSMENT OF BIOSOLIDS BORNE CIPROFLOXACIN (CIP) AND AZITHROMYCIN (AZ) By Harmanpreet Singh Sidhu M ay 2018 Chair: Major: Soil and Water Sciences Ciprofloxacin (CIP) and azithromycin (AZ) are commonly prescribed antibiotics for various infections in humans and are, c onsequently frequently detected in biosolids E cological and human health risks from biosolids borne CIP and AZ are not well understood but necessary for formulating policy on safe use and management of biosolids. A project funded by Water Environment & Reuse Foundation (WE & RF) was designed to identif y and fill various data gaps in the fate of biosolids borne CIP and AZ and to facilitate a scientifically sound ecological and human health risk assessment. Data from a batch equilibration retention/release study form ulated our central hypothesis that the limited bioaccessibility of strongly sorbed biosolids borne CIP and AZ minimizes human and environmental health risks Bioavailabilit ies of biosolids borne CIP and AZ w ere assessed in subsequent organism response ( plant, earthworm, and microbial systems ) s tudies and (w here applicable ) correlated with chemical bioaccessibili ties The organism response data revealed limited bioavailability (plant bioaccumulation factor ( BAF ) values 0.01 (CIP) and 0.1 (AZ) depurated earthworm BAF values ~4 (CIP) and ~ 7 (AZ), minimal impacts on overall microbiota) of the biosolids borne antibiotics under environmentally relevant scenarios. The data

PAGE 18

18 generated herein and environmentally relevant data collected from pertinent literature were utiliz ed in a scientifically sound integrated risk assessment (IRA), following the World Health Organization framework Human and ecological exposure hazards were identified and potential adverse effects from CIP and AZ in land applied biosolids were assessed us ing a tier approach The IRA consistent with our hypothesis estimated negligible risks from biosolids borne CIP and AZ under real world based biosolids management practices Even unrealistically high exposures from land application of biosolids pose mini mal human and ecological health risks. Preliminary pollutant limits calculated based on the most sensitive organism s suggest that long term real world based land application of biosolids is without appreciable human and ecological health risks. Chemical l oad tracking is not neede d for the majority of USA biosolids but may be necessary for some biosolids that contain greater than 12 mg CIP/kg and 2.2 mg AZ/kg The IRA needs refining by including m ore data, especially on biosolids borne antibiotic resistance before suggest ing modifications to current land application regulations

PAGE 19

19 CHAPTER 1 INTRODUCTION BACKGROUND, AND PROJECT OBJECTIVES Introduction For ages, biosolids have been applied to agricultural lands for their beneficial effects on soils and plant growth. However, with advancements in human civilization and incorporation of various compounds including pharmaceuticals and personal care products in our daily activities, numerous chemicals find their way into the biosolids. Concerns particularly due to incomplete knowledge about human and ecological risks, about biosolids borne trace organic co mpounds (TOrCs) threaten the viability and sustainability of land application of biosolids. H igh detection frequencies and numerous gaps in fate and toxicity data make the human and ecological health risks posed by two TOrCs ciprofloxacin ( CIP ) and azit hromycin ( AZ ) uncertain Data on the e nvironmental fate of CIP, and to a lesser extent AZ, exist, but little pertains directly to biosolids amended soil systems. Numerous data gaps prompted t he USEPA to identif y CIP and AZ as high priority compounds for the risk assessment of biosolids borne TOrCs. In a risk assessment, the general assumption is that if exposure of a contaminant to a population of concern is minimal, the associated risks are also minimal. Therefore, knowledge of the potential fate of a c ontaminant is of utmost importance in assessing the associated risks. The first step in determining the exposure potential is assessing the environmental concentrations of a contaminant. Up to 60% of CIP (Girardi et al., 2011) and 75% of AZ (Zuckerman, 200 0) is excreted from the human body as unchanged parent compound The m ajority of the excreted compounds ends up in primary and secondary sludge in wastewater treatment plants ( WWTPs ; Golet et al.,

PAGE 20

20 2003; Carmosini and Lee, 2009 ; Senta et al., 2013 ). The slu dge accumulated within WWTPs is often processed to produce biosolids intended for land application. The biosolids generated in the US A contain both CIP and AZ at concentrations ranging from several hundred g/kg to several mg/kg (Table 1 1 ) About half of US A biosolids are land applied (NRC, 2002), resulting in a systematic release of biosolids borne CIP and AZ into the terrestrial environment. Once land applied, biosolids borne CIP and AZ are subject to various transformations and translocations in the re ceiving soils. P rocesses like irreversible sorption onto the receiving soils or biotic or abiotic degradation can decrease environmental and/or human exposure of the target TOrCs. On the other hand, uptake by soil biota (including plants, earthworms, micro organisms), volatilization, leaching etc. can increase the exposure to higher organisms including humans. Also, potential antibiotic resistance development and spread and associated risks are a cause for concern with land applied antibiotics (Girardi et al ., 2011; Munir and Xagoraraki, 2011 ; Bengtsson Palme and Larsson, 2016 ). Assessing human and ecological health risks associated with biosolids borne CIP and AZ are, therefore, crucial for devising science based regulatory policies for safe use and manageme nt of biosolids. Structure, P roperties, U se, and M echanism of A ction Ciprofloxacin and AZ (Figure 1 1) are broad spectrum antibiotics used to treat a number of bacterial infections in humans (Girardi et al., 2011; Parnham et al., 2014). Ciprofloxacin is t he most widely used second generation quinolone antibiotic, and acts by inhibiting one or more of a group of enzymes called topoisomerases in bacterial cells. The topoisomerases are needed for supercoiling, replication and separation of bacterial DNA (Hoop er, 1999). At low exposure concentrations, CIP i s bacteriostatic

PAGE 21

21 inhibiting topoisomerase and DNA replication. At high exposure concentrations, CIP is bactericidal causing the release of free DNA ends from the DNA topoisomerase CIP complex, and leading t o chromosomal DNA fragmentation (Silva et al., 2011). A zithromycin is a second generation macrolide antibiotic that inhibits bacterial protein synthesis by binding with the 50S ribosomal subunit and inhibiting translation of mRNA (Parnham at al., 2014). A zithromycin, like other macrolides, is bacteriostatic in nature (Dorfman et al., 2008). The commonly prescribed HCl formulation of CIP is readily water soluble and practically insoluble in organic solvents. On the other hand, AZ is sparingly soluble in wa ter but readily soluble in organic solvents (Table 1 1) Low water solubility of AZ suggests limited mobility in (water based) soil systems, but water solubility alone is insufficient to assess fate and transport of contaminants in the environment. Other factors like chemical nature (e.g., organic vs. inorganic; presence of ionizable moi eties, pKa values size, biodegradability), soil/biosolids properties (e.g., CEC, pH, OM, presence of chemical degrading microbes), and interaction s between the environment and chemical (e.g., soil solution partitioning coefficient ( Kd ) values ) are critica l to the fate, transport, and potential risks associated with an environmental contaminant. For instance, although CIP is readily soluble in water, its cationic nature results in adsorption of more than 95% of the dissolved chemical onto negatively charged soil particles (Nowara et al., 1997), potentially limiting its bioaccessibility. Ciprofloxacin exists as a zwitterion and AZ is a cation in the environment (Figures 1 1 and 1 2). The predominant CIP and AZ species have positively charge moieties (NH + fun ctional group; Figure 1 1) throughout the environmentally relevant pH

PAGE 22

22 range of 4 9, so cation exchange is more thermodynamically favorable than the hydrophobic interactions typically assigned to TOrCs (Horvath et al., 1976). Ciprofloxacin also has a negati vely charged moiety (COO function group; Figure 1 1) through most of the environmentally relevant pH (5 9; Figure 1 2) and anion exchange, cation bridging, and surface metal complexation can also contribute to CIP retention at pH values >6. U nlike many no n polar/neutral TOrCs, the fate and transport of (and, thus, potential environmental and human health risks from) CIP and AZ are expected to be strongly influenced by the pH and CEC of the biosolids and receiving soils. Understanding the sorption mechanism s involved in CIP and AZ retention is important because sorption influences CIP bioaccessibility, which in turn influences bioavailability. Bioaccessible, herein, refers to fraction of the total chemical that can become available to be taken up by an organ ism B ioavailable refers to the fraction of the bioaccessible chemical that is actually taken up by an organism. Figure. 1 1. Structures of ciprofloxacin (CIP) and azithromycin (AZ) along with pKa values.

PAGE 23

23 Figure.1 2. Species distribution diagrams at e nvironmentally relevant pH values for A). CIP [Created using pKa values of 6.1 and 9.2; Aristilde and Sposito, 2013] and B). AZ [Created using pKa values of 8.74 and 9.45; McFarland et al., 1997]. The log Dow (pH dependent octanol water partitioning coefficient ) of AZ (Table 1 1 ) varies from ~ 0 (low pH) to strongly positive (high pH). This variation in log Dow is explained by the predominant AZ species at a particular pH (Figure 1 2 ). The Dow values and speciation diagram suggest that cation adsorption govern s AZ retention/release at pH values less than ~8.5 A t pH values greater than ~ 8 .5 the neutral AZ species become s important and hydrophobic interactions are expected to contribute to AZ retention. In contrast, the negative log Dow values for CIP (Table 1 1 ) throughout the environmentally relevant pH range suggest that hydrophobic interactions with soil and organic matter have little influence on CIP adsorption. Estimated vapor pressures greater than 10 12 mm Hg (Table 1 1 ) suggest that vo latilization of both CIP and AZ is negligible and not a concern in determining fate of (or risks from) the two TOrCs in the environment. The low to moderate water solubility, along with positive charge and high log Koc (Table 1 1 ), suggest that CIP and AZ have limited mobility in biosolids and soils at environmentally relevant pH values Limited mobility (and thus limited leaching potential) of CIP and AZ has been confirmed in

PAGE 24

24 column leaching studies (von Hellens, 2015) and in actual field scale studies ( US National Library of Medicine, 2002 ; Golet et al., 2003 ; Gottschall et al., 2012 ) D egradation and persistence data suggest that CIP can accumulate in the soil for years especially after repeated application of highly contaminated biosolids ( Al Ahm a d et al., 1999 ; Chenxi et al., 2008; Girardi et al., 2011; Gottschall et al., 2012 ; Garcia Santiago et al., 2016 ) D ata on AZ degradation and persistence are however, scarce. Gottschall et al. (2012) estimated a disappearance half life (DT 50 ) of ~70 d in land applied biosolids aggregates in a yearlong field scale study Azithromycin concentrations i n biosolids amended soil samples were, however, un detectable Maier and Tjeerdema (2018) suggested AZ DT 50 of ~ 80 d in natural s ediments under aerobic conditions but the regression coefficient (r 2 ) explaining AZ disappearance with time was only ~0.3 and s tandard error around the data was high (>6 5 %) B oth CIP and AZ reported ly pose risk s to aquatic environments but studies conducted in biosolids amended soil sys tems are scarce (Table 1 1 ) Table 1 1 Select literature on, and properties of CIP and AZ. Property CIP AZ Source (Superscript 1=CIP; 2=AZ) Molecular mass (g/mol) 331.3 749 1,2 US National Library of Medicine, 2002. Typical concentrations in US A biosolids (median; mean; 95 th percentile) (mg/kg) 5.1; 10.5; 36.1 0.26; 0.83; 3.2 1,2 USEPA, 2009 Water s olubility (mg/L) (20 0 C) 30,000 (at 20 0 C) 5.43 (at 25 0 C) 1 Varanda et al. 2006 2 Ericson, 2007 pKa1, pKa2 6.10.2, 9.20.5 8.74, 9.45 1 Aristilde and Sposito, 2013 2 McFarland et al., 1997 log D ow (pH=7.4) 1.11 0.53 1 Takacs Novak et al., 1992 2 Ericson, 2007

PAGE 25

25 T abl e 1 1. Continued Property CIP AZ Source (Superscript 1=CIP; 2=AZ) log D ow (pH=5;7;9) 1.05; 0.98; 1.46 0; 0.22; 3.28 1 Ross et al., 1992 2 Calculated using appropriate equations from McFarland et al., 1992 and principles from Horvath et al., 1977 Kd (L/kg) 2500 20000 (biosolids) 150 5 00 00 (soils) 386 (biosolids) 150 (soils) 1 1 Nowara et al., 1997 1 Thiele Bruhn 2003 2 Gobel et al., 2005 1 Vasudevan et al., 2009 2 Maier and Tjeerdema, 2018 Koc (L/kg) 61000 (at pH=5) 180 00 600 00 1 Nowara et al., 1997 2 Ericson, 2007 2 Maier and Tjeerdema, 2018 Vapor pressure 2.85 X10 13 mm Hg at 25 0 C (est.) 2.65 X10 24 mm Hg at 25 0 C (est.) 1,2 USEPA EPI Suite, 2011 t in water (Photolysis shallow rivers; Hydrolysis; Biodegradation) 0.5 1.5 h; not readily hydrolyzed; not readily biodegraded 3.7 20 h; likely readily hydrolyzed; not readily biodegraded 1 Al Ahmad et al., 1999 1 Burhenne et al., 1997 a,b 1 Golet et al. 2002 a,b 1 Girardi et al., 2011 2 US National Library of Medicine, 2002 2 Ericson, 2007 2 Tong et al., 2011 2 Maier and Tjeerdema, 2018 t in soil/biosolids (Photolysis, Hydrolysis, Biodegradation) DT 50 of ~5 y; not readily photolyzed; not readily hydrolyzed; not readily biodegradable DT 50 ranging from 70 d to 3 y; no data available on photolysis, hydrolysis, and biodegradation 1 Girardi et al., 2011 1 Al Ahmed et al., 1999 1 Chenxi et al., 2008 1,2 Walters et al., 2010 2 Gottschall et al., 2012 2 Maier and Tjeerdema, 2018 MIC90 0.008 2 mg/L for enterobacteriaceae, P. aeruginosa, H. influenza, streptococci, Staphylococcus aureus, Salmonella 0.03 2 mg/L for enterobacteriaceae P. aeruginosa, H. influenza, streptococci, Staphylococcus aureus, Salmonella 1 Wise et al., 1983 1 Eltahawy, 1993 2 Gordillo et al., 1993 2 LeBel, 1993 MIC for 1% of variates (MIC1) 0.004 mg/L (for 6 bacterial genera) 0.001 mg/L (for 29 bacterial genera) 1,2 Bengtsson Palme and Larsson, 2016

PAGE 26

26 Table 1 1. Continued. Property CIP AZ Source (Superscript 1=CIP; 2=AZ) Toxicity to aquatic organisms High potential for acute and/or chronic toxicity Some potential for toxicity 1 Halling Sorensen et al., 2000 1 Ortiz de Garcia et al., 2014 1 Papageorgiou et al., 2016 2 Harada et al., 2008 2 Iatrou et al., 2014 Toxicity to terrestrial organisms (bacteria; earthworms; rats) Toxic to soil bacteria; data not available; Rats (LD 50 >5000 mg/ Kg) Likely not toxic to soil microbes; data not available; Rats (LD 50 >3000 mg/kg) 1 Girardi et al., 2011 1, 2 US National Library of Medicine, 2002. Human toxicity May cause liver toxicity in at risk individuals at clinically prescribed doses Calculated RfD = 0.00 4 m g/kg d* May cause liver toxicity in at risk individuals at clinically prescribed doses Calculated RfD = 0. 067 m g/kg d* 1, 2 Andrade and Tulkens, 2011 1, 2 Maggioli et al., 2011 Dr. Andrew Maier, personal communication Reference doses (RfD) were determined by Dr. Andrew Maier, who is a Co PI on the WE & RF funded project. RfD values were calculated by applying appropriate safety factors to high quality data extracted from a thorough review of pertinent literature. Typical Expected C oncentrations in Biosolids and Amended Soils The concentrations of contaminants in WWTP influents and in the processed biosolids can vary widely with the geographical region and population served, as well as with season, sewage treatment processes, etc. To account for such variations, the USEPA periodically performs nationwide surveys to assess representative concentrations of various pharmaceuticals and personal care products in US A biosolids. T argeted national sewage sludge surveys (TNSSS) are the broades t and most thorough nationwide surveys of biosolids in the USA and use USEPA approved analytical methods. The latest TNSSS was conducted more than a decade ago and involved 74 treatment plants scattered throughout US T he average ( mean ) median (50 th perce ntile) and 95 th percentile concentrations were 10.5 17.7, 5.4, and 36.1 mg, respectively, for CIP per kg biosolids (USEPA, 2009) The average median, and 95 th

PAGE 27

27 percentile concentrations for AZ were 0.83 2.3, 0.25, and 3.2 mg, respectively, per kg bios olids. Additionally published analyses suggest that many individual biosolids contain CIP and AZ at concentrations close to the median concentrations previously reported by USEPA (Gottschall et al., 2012; Sabourin et al., 2012; Youngquist et al., 2014; Angelo and Starnes, 2016 ; Garca Santiago et al., 2016). A recent, but much smaller and regionally focused, survey of biosolids from 11 treatment plants in the north western region of US, found CIP concentrations of 3.06 2 (mean), 2.78 (median), and 7.6 (95 th percentile) mg per kg biosolids; and AZ concentrations of 0.41 0.47 (mean), 0.25 (median), and 1.74 (95 th percentile) mg per kg biosolids (Kennedy/Jenks Consultants, 2015). Eight biosolids samples obtained from the American Mid west and S outheast were analyzed to assess CIP and AZ concentrations, as a part of this research. The biosolids included one commercially available biosolids and seven biosolids directly obtained from wastewater treatment plants. T he biosolids represented Class A an d Class B materials synthesized using different methods including air drying, anaerobic digestion, aerobic digestion, composting, and heat drying. The samples were stored at 0 0 C prior to analysis by AXYS (BC, Canada) using AXYS method MLA 075, which is ba sed on USEPA Method 1694 (USEPA, 2007). Detail s o f AXYS method MLA 075 are provided in Appendix A. The a nalysis results (Table 1 2) are consistent with the recent literature in that the chemical concentrations in all eight biosolids were near the median co ncentration of ~5 (CIP) and ~0.25 (AZ) mg/kg reported in the TNSSS (USEPA, 2009). The data suggest that the concentrations

PAGE 28

28 of CIP and AZ in modern USA biosolids have decreased since the TNSSS samplings and that accurate assessment of the national average me dian and 95 th percentile concentrations of biosolids borne CIP and AZ require revisions to better reflect modern products Biosolids are typically land applied at or below an agronomic (N based) rate of ~ 2 0 Mg/ ha and subsequently dispersed in top 15 cm of soil. A dilution factor of around 100 is, therefore, expected in most top soils resulting in biosolids amended soil concentrations of <1 mg/kg for CIP and << 1 mg/kg for AZ. Nevertheless, the environmental and human health risks from even such low environmentally relevant concentrations of CIP and AZ in biosolids amended soils are incompletely known. Table 1 2. Ciprofloxacin and AZ concentrations (mg/kg) in the 8 biosolids products and respective reporting limit (mg/kg) for each biosolids for each compound. Chemical Biosolids Reporting limit (mg/kg) Concentration found (mg/kg) CIP 1 Class A 0.018 3.57 2 Class A 0.066 3.28 3 Class A 0.019 0.006 0.954 0.04 4 Class B 0.016 4.64 5 Class B 0.014 5.90 6 Class B 0.023 4.81 6 Heat dried 0.025 0.815 6 Compost 0.212 Below reporting limit AZ 1 Class A 0.017 0.435 2 Class A 0.009 0.694 3 Class A 0.004 0.009 0.06 0.04 4 Class B 0.006 0.205 5 Class B 0.005 0.268 6 Class B 0.004 0.217 6 Heat dried 0.008 0.012 6 Compost 0.004 Below reporting limit

PAGE 29

29 Objectives The ultimate objective of this research was to perform a scientifically sound integrated risk assessment of biosolids borne CIP and AZ to guide regulatory agencies on the safe use and management of biosolids. T he goal was to attain a better understanding of fate and exposure of biosolids borne CIP and AZ and their potential effects on human and ecological health, to inform decision makers, waste water management agencies, regulators, product manufacturers, and cons umers. The central hypothesis was that the limited bioaccessibility of the strongly sorbed CIP and AZ minimizes risks to huma n and environmental health. The ultimate objective was achieved through a number of intermediate objectives that focused on genera ting/gathering high quality data on the fate, exposure, and toxicity of the target TOrCs. The intermediate objectives were accomplished by either obtaining data from the literature by conducting greenhouse/lab oratory studies and /or from analyses of sampl es from biosolids amended fields An emphasis was placed on data generated in environmentally relevant biosolids soil matrices, with realistic loading rates, and under real world scenarios. A detailed literature review (summarized in Table 1 1 ) identified major data gaps in the bioavailability ( retention / release terrestrial bioaccumulation factors, bio uptake), degradation (AZ), and toxicity (human and terrestrial systems) benchmarks of the target TOrCs. Human toxicity benchmarks were generated by Co PI s on the WE & RF project funding this work and were directly used in the risk assessment I ntermediate objectives were designed to generate data needed to fulfill the main objective and included characterizing adsorption/desorption of, and plant, earthworm, and microbial response s to biosolids borne CIP and AZ. Using data from literature, from

PAGE 30

30 peers (human toxicity), and that generated through this work a tiered integrated risk assessment was accomplished using frameworks developed by WHO (2001) and the USEPA Part 503 Biosolids Rule risk assessment (USEPA, 1995). Intermediate Objectives Intermediate o bjective 1. To a ssess r etention r elease b ehavior of CIP and AZ The first step in predicting the fate of, and risks from, a chemical is understanding its rete ntion release behavior. A thorough retention/release study involves several soils and biosolids var ying in physicochemical properties expected to influence the retention/release behavior of a chemical. The factors expected to influence retention/release of CIP and AZ in soils/biosolids include: pH, CEC, OM, Fe/Al oxides, clay minerology, chemical concentrations, and solution ionic strength ( Nowara et al., 1997; Hari et al., 2005; Carmosini and Lee, 2009; Vasudevan et al., 2009; Aristilde and Sposito, 2013; Jiang et al., 2013; Wu et al., 2013 ; Goulas et al., 2016 ) A detailed mechanistic retention/release behavior study can require months of effort, and was beyond our scope Rather our focus was to generate Kd values f or CIP and AZ in biosolids and soils per tinent to s ubsequent plant, microbial, and earthworm exposure studies We conducted a non mechanistic retention/release study using batch equilibration methods similar to that used by Agyin Birikorang et al. (2010) The study involv ed CIP or AZ adsorption/desorption in 3 soils, 3 biosolids amended soils (at 1% rate, w/w), and one biosolids Also, b iosolids pre equilibrated with spiked CIP and AZ at varying concentrations w ere amended to the 3 soils (at 1% rate, w/w) to assess the release of b iosolids borne TOrCs after land application Release from pre equilibrated biosolids was assessed using various approaches to understand potential desorption

PAGE 31

31 (and bioaccessibility) of the target TOrCs and associated environmental and human health consequen ces Anaerobically digested air dried Class A biosolids ( 3 Class A in Table 1 2) was selected for sorption/desorption and subsequent bioavailability studies because it fulfilled our two criteria of selection: i) a biosolids containing low concentrations of the target TOrCs, and ii) a biosolids from a large, diverse, metropolitan area serving a large population and representative of typical WWTP practices and land application programs in the USA. T he study generated environmentally relevant data (especially Kd values) on fate and transport (thus bioaccessibility) of the two TOrCs. Utilization of 3 H labeled compounds minimize d the need for the extensive sample clean up required in other analytical methods provided q uick analysis, and resulted in very low detection limits of the target analytes The detailed methods, results, and implications are presented in Chapter 2. Intermediate o bjective 2 To a ssess plant responses to b iosolids borne CIP and A Z Accumulation by plants grown in biosolids amended soils represent an important pathway of potential CIP and AZ introduction into human and ecological foo d chains. D ata on plant uptake of, and toxicities from, environmentally relevant concentrations of biosolids borne CIP and AZ are limited and warrant determination To that end, g reenhouse studies were conducted to assess phyto a ccumulation and phytotoxicity of environmentally relevant concentrations of biosolids borne CIP and AZ in 3 crops grown in two biosolids amended so ils. The crops represent a range of plants with different morphologies, physiologies, and exposure scenarios to biosolids borne chemicals When possible, plant response curves were generated and utilized to determine

PAGE 32

32 bioaccumulation factors. The study meth odology, results, and discussion a re detailed in Chapter 3 Intermediate o bjective 3 To a ssess m icrobial r esponse s to b iosolids borne CIP and AZ. Microorganisms are highly sensitive to changes in the soil/biosolids environment and represent a crucial but often ignored, part of a scientifically sound risk assessment Data on microbial response s to biosolids borne CIP and AZ are scarce and the limited data suggest potential adverse effects and development of antibiotic resistance (Girardi et al., 2011; Cui et al., 2014 ). The reported effects of CIP and AZ on soil microorganisms warrant studies on microbial response to biosolids borne CIP and AZ under environmentally realistic conditions using molecular biology approaches A microbial response incubation study (for up to 120 d ) in biosolids and biosolids amended soils was conducted using varying concentrations of the two 3 H labeled compounds. The study assessed the impacts of biosolids borne CIP and AZ on soil microorganisms through changes in soil respir ation, expression of genes involved in N and P cycles and antibiotic resistance gene expressions. The study also assessed changes in CIP and AZ extractability and stability over time and correlated bioaccessibility (based on chemical extractability) with bioavailability (microbial response). The generated data and their implications are detailed in Chapter 4. Intermediate o bjective 4 To a ssess earthworm responses to b iosolids borne CIP and AZ Earthworms are excellent organisms for bio monitoring potentia l bioavailability of biosolids borne CIP and AZ and consequent environmental and human health risks. Further, earthworms are a significant fraction of the diet of some terrestrial vertebrates (Suter et al., 2000) and a pathway of CIP and AZ introduction i nto ecological food

PAGE 33

33 chains Data on earthworm response s to biosolids borne CIP and AZ are essentially absent, but crucial for a scientifically sound risk assessment. Earthworm uptake of, and toxicity from, biosolids borne 3 H CIP and 3 H AZ was studied in three biosolids amended soils under laboratory conditions. The study generated critical information necessary for ecotoxicological risk assessment of biosolids borne CIP and AZ. Analysis of field samples supplemented the laboratory studies. The study detai ls are provided in Chapter 5. Ultimate Objective : To C onduct H uman and E cological H ealth R isk A ssessment of B iosolids b orne CIP and AZ. The principle hypothesis was tested by conducting an integrated risk assessment (IRA) of biosolids borne CIP and AZ. A t iered IRA was conducted based on the principles and framework published by WHO (2001) and USEPA Part 503 Biosolids Rule making risk assessment. (USEPA, 1995). Human and ecological exposure hazards were identifie d using the hazard quotient (HQ) approach A t ier I (screening level) risk assessment assumed conservative to unrealistic exposure scenarios and assessed risks to highly exposed organisms. S ubsequent tier risk assessment s refined assumed biosolids and chemical characteristics and application scenario s (towards less conservative and more realistic values) while still focused on pathways routes, and receptors of potential concern (Chapter 6) Implications of the study and future environmental regulation of biosolids containing CIP and AZ for the protection of human and environmental health are provided in Chapter 7.

PAGE 34

34 CHAPTER 2 RETENTION/RELEASE BEHAVIOR OF CIPROFLOXACIN AND AZITHROMYCIN Synopsis Lim i ted data are available on the retention/release of ciprofloxacin ( CIP ) and azithromycin ( AZ ) for predicting potential chemical bioaccessibilit ies in biosolids and biosolids amended soils The p resent work focused on environmentally relevant aspects of retention/release behavior of CIP and AZ via non mechanistic sorption/desorption studies. A b atch equilibration metho d was employed to generate sorption/desorption isotherms and to determine partitioning coefficient (Kd) values of CIP and AZ in seven solid matrices, including 3 soils, 3 biosolids amended soils, and a class A biosolids In a separ ate study, the same biosolids was pre equilibrated with the target compounds for a week before CIP and AZ desorption was assessed using 4 approaches desorption batch experiments (using calcium chloride ), (ii) addition of lead chloride ( l ead ( Pb 2+ ) being a specifically sorbing cation), (iii) competitive desorption of CIP in presence of AZ, and vice versa, and (iv) application of the contaminated biosolids to three soil media. The batch experiments yielded hysteretic sorption /desorption an d soil Kd values rang ing from ~10 L/kg to over 200 L/kg Sorption of CIP and AZ by the biosolids produced moderately high Kd values of 357 L/kg for CIP and 428 L/kg for AZ and extremely small hysteresis coefficients ( less than 0.003 ) Desorption of biosoli ds borne CIP and AZ was negligible using any of the 4 desorption approaches. The results strongly suggest that the bioaccessibility of biosolids borne CIP and AZ is minimal and that biosolids (not soils) control desorption when CIP and AZ are biosolids bor ne.

PAGE 35

35 Introduction Predicting the fate of trace organic compounds (TOrCs) in the environment requires understanding the extent and mechanisms of retention/release. Most reported retention/release studies of the targeted TOrCs have been limited to the sorpti on behavior of CIP in soils, and studies on biosolids borne CIP or AZ are scarce. Further, data on desorption behavior (which is equally or perhaps more important) are rare for both compounds. Retention/ R elease Mechanisms As CIP and AZ are catio nic at env ironmentally relevant pH values (Figure s 1 1, 1 2 ), soil water partitioning coefficients ( K d values) are expected to depend on the pH and CEC of biosolids and/or biosolids amended soil. Both CIP and AZ cationic moieties can interact with negatively charged soil surfaces via a host of interactions in volving non specific columbic interactions and/or specific binding. Indeed, the literature suggests that cation exchange, cation bridging, and complexation with surface bound metal oxides, along with potential co ntribution s from p p interactions and H bondin g are the major mechanisms of CIP retention (Nowara et al., 1997; Hari et al., 2005; Carmosini and Lee, 2009; Vasudevan et al., 2009; Aristilde and Sposito, 2013; Jiang et al., 2013 ; Wu et al., 2013). Consequen tly, pH and CEC are the major factors expected to influenc e retention/release behavior of cationic trace organics such as CIP and AZ whereas hydrophobic interactions are of minimal significan ce (Carmosini and Lee, 2009; Gong et al., 2012; Aristilde and Sp osito, 2013, Goulas et al., 2016). Information pertaining directly to mechanisms involved in the retention/release behavior of AZ in soils and biosolids is largely absent. Ericson (2007) reported that more than 90% of AZ exists as bound (sorbed) residues of limited bioaccessibility but

PAGE 36

36 offered no mechanistic explanation Azithromycin is predominantly a divalent cation between the environmentally relevant pH values of 4 to ~ 8.5 (Figure 1 2 ); thus like CIP, AZ retention/release to/from soil particles is lik ely influenced by soil /biosolids cation exchange capacity. S oil Koc values ranging from 1 8 0 00 to 600 00 L/kg are reported for AZ (Table 1 1) but a log D ow (at pH 7) of 0.53 (Ericson, 2007) suggests that hydrophobic interactions are not dominant at pH value s up to ~8. Hydrophobic interactions may become important at high pH in high organic matter solids because of t molecular structure (mostly made up of nonpolar moieties) and a log D ow (at pH 9) of 3.28 (McFarland et al., 1997). Li ke CIP, AZ retention via H bonding is also possible C ation bridging and complexation with surface bound aluminum, iron, and manganese oxides and are likely not mechanism s of AZ retention because of the lack of carboxylic (or other negatively charged) functional gro ups in the molecular structure. A typical 1% (w/w) application rate of a biosolids containing the 95 th percentile concentration s of CIP (36.1 mg/kg) and AZ (3.2 mg/kg) (USEPA, 2009) adds only ~ 10 4 cmolc CIP and ~10 5 cmolc AZ per kg of soil. Even 100% sa nd has thousands of fo ld more CEC (e.g., 1 3 cmolc/kg) than needed to adsorb virtually all of the CIP and AZ added in the biosolids. Such miniscule concentrations of CIP and AZ added to soils suggest that sorbent sorbate interactions, with minimal influenc e from competing cations, should influence the adsorption of the CIP and AZ onto charged soil particles, especially if sorption is specific and/or strong. However, many studies have observed or suggested competition for exchange sites between organic catio ns (including CIP) and/or between organic and inorganic cations (Hendricks et al., 1941; Weed and Weber, 1968; Margulies et al., 1988; Fabrega et al., 2001; Sassman and Lee, 2005;

PAGE 37

37 Carmosini and Lee, 2009; Vasudevan et al., 2009 ; Conkle et al., 2010; Aristi lde and Sposito, 2013; Wu et al., 2013). The apparent competition for exchange sites is attributed to: 1) the preferential adsorption on exchange sites (structural charge on soil particles can be expressed as discrete adsorption sites rather than a smear o f surface charge) (Weed and Weber, 1968; Margulies et al., 1988; Droge and Goss, 2013), and/or 2) separations of charges on the chemicals or size exclusion of large cations (Weed and Weber, 1968; Fabrega et al., 2001 ; Carrasquillo et al., 2008). However, o nce an organic cation accesses an adsorption site, the retention should be stronger than for common inorganic cations because of the additional involvement of van der Waals forces, H bonding, and/or p p bonding (Weed and Weber, 1968; Aristilde and Sposito, 2013). Extent of Retention/ R elease Ciprofloxacin strongly and extensively binds to many soil and biosolids particles (Nowara et al., 1997; Castela Papin et al., 1999; Hari et al., 2005; Carmosini and Lee, 2009; Vasudevan et al., 2009; Aristilde and Sposi to, 2013; Jiang et al., 2013; and Starnes, 2016 ; Goulas et al., 2016). Ciprofloxacin p artitioning coefficient (Kd) values range from ~100 to >45,000 L/kg in soils (Nowara et al., 1997; Vasudevan et al., 2009) and from ~2700 to >19000 L/kg in biosolids and activated sludge (Nowara et al., 1997; Wu et al., 2009 ; T hus, literature suggests relatively high CIP Kd values ( implying limited compound bioaccessibility ) under environmentally relevant conditions but only few studies address AZ retention/release in soils and biosolids. Maier and Tjeerdema (2018) reported a K d value of ~150 L/kg in loamy sand soil sediments (~1% OM). Gobel et al. (2005) reported a Kd value of 376 86 L/kg for AZ in activated sludge and suggest that hydrophobic interactions, rather than ionic

PAGE 38

38 interactions, dominate AZ sorption onto activated s ludge. Gobel et al. (2005), however, provid es no characterization data for the activated sludge (including pH value), experienced analytical interferences, and acknowledges possible non equilibrium state s of the collected samples. Literature pertaining di rectly to CIP and, especially, AZ desorption is limited and t he desorption behavior of biosolids borne CIP/AZ following biosolids application to soils has not been studied. Several studies suggest that desorption of organic TOrCs that strongly adsorb to so il/biosolids particles (like CIP) should not be significant under environmentally relevant conditions (Nowara et al., 1997; Wu et al., 2013; Chen et al., 2015; Berhane et al., 2016 ; Carrillo et al., 2016 ). However, such studies were conducted under conditi ons not directly pertinent to biosolids borne chemicals and/or were limited by analytical errors, poor recoveries, unrealistically high TOrC biosolids borne CIP and suggeste d that a considerable fraction of sorbed CIP (16%) becomes bioaccessible through desorption. Accurate assessment of desorption dynamics of environmentally relevant concentrations of biosolids borne CIP and AZ is, therefore, critical for a sound risk assess ment and deserves additional stud ies The present study focused on the environmentally relevant aspects of retention/release behavior of biosolids borne CIP and AZ and conducted non mechanistic sorption/desorption studies. E nvironmentally relevant concent rations of CIP and AZ can vary widely for different sources (or sinks). Herein, environmentally relevant concentrations are discussed in context of biosolids borne chemicals. The concentrations of the target TOrCs in USA biosolids typically range between median

PAGE 39

39 and average concentrations (i.e., 5 to 11 mg (CIP) and 0.25 to 0.83 mg (AZ) per kg biosolids ) reported in the targeted national sewage sludge survey (USEPA, 2009). The uppermost end (95 th percentile ) concentrations of environmental relevance per kg biosolids are ~36 mg (CIP) and ~ 3.2 mg (AZ) (USEPA, 2009). Based on typical (median to average) biosolids concentrations and the typical 1% (dw/dw) land application rate, most biosolids amended soils nominally contain 0.05 to 0.11 mg CIP and 0.003 to 0.008 mg AZ per kg. Using the 95 th percentile concentrations, the uppermost end of environmental relevance is 0.36 mg CIP and 0.032 mg AZ per kg amended soil. The latter concentrations also represent soil concentrations (without attenuation) from ~7 (C IP) and ~10 (AZ) years of repeated application of biosolids (at 1% (dw/dw) application rate) contaminated with median chemical concentrations. Based on the literature and c ompound c hemical properties we hypothesized that both CIP and AZ sorb extensively a nd strongly to biosolids and soils and that desorption of biosolids borne CIP and AZ is minimal. The primary objective was to generate Kd values for biosolids soil systems that can be used as first approximations of biosolids borne CIP and AZ bioaccessibil ity for subsequent studies involving biological responses (plant earthworm, and microbial response). The secondary objective was to assess potential cumulative release of the target TOrCs using various desorption encouraging approaches. We also used envir onmentally relevant and/or high background solution ionic strengths (using CaCl 2 ) to minimize the influence of media CEC on chemical sorption to qualitatively assess sorption mechanisms other than CEC dependent cation exchange.

PAGE 40

40 Preliminary studies generate d information on sorption/desorption equilibrium and kinetics, and were followed by definitive triplicated sorption/desorption experiments that: 1. generated sorption desorption isotherms, 2. determined Kd, Kcec (Kd normalized to media CEC) and Koc (Kd normaliz ed to media organic carbon ) values 3. assessed sorption/desorption hysteresis and 4. assessed desorption of biosolids borne target chemical Materials 3 H Ciprofloxacin (CAS No. 85721 33 1; 97.4% radiochemical purity) and 3 H azithromycin (CAS No. 117772 70 0; 98.4% radiochemical purity) were custom synthesized by Moravek Biochemicals (Brea, CA). 2 14 C Ciprofloxacin (CAS No. 85721 33 1; >97% radiochemical purity) was also obtained from Moravek Biochemicals (Brea, CA). Pharmaceutical secondary standard s (>99% pur e) of CIP and AZ, analytical grade calcium chloride (CaCl 2 ), and double deionized water were purchased from Sigma Aldrich (St. Louis, MO). The biosolids (3 Class A, Table 1 2) used in the study w as anaerobically digested air dried Class A biosolids from Me tropolitan Water Reclamation District of Greater Chicago (MWRDGC) and contained low CIP (1 mg/kg) and AZ (0.06 mg/kg) concentrations (analyzed by AXYS, BC, Canada) Cattle manure ( obtained from University of Florida Dairy Research Unit, located in Hague, Florida ) contain ed undetectable (<0.035 mg/kg) CIP and 0.01 mg AZ /kg (analyzed by AXYS, BC, Canada) Materials Company (Birmingham, AL). A s ilt loam (fine loamy, mixed, superact ive, mesic Fluventic Eutrudepts) was obtained from Wisconsin. The sand amended with the served as one of

PAGE 41

41 the three soils Select properties of the soils and the biosolids involved in the study are listed in Table 2 1. Table 2 1 Select p roperties of soils and biosolids used in retention/release studies (measured average values from duplicate samples). Media pH OM ( g/kg ) CEC (cmolc/kg) Mehlich 3 Fe (mg/kg) Mehlich 3 Al (mg/kg) Biosolids 6.5 43 0 185 3 60 600 Sand 6 1. 6 2.7 0.67 2.8 Silt loam 6.25 1 0 19.2 50 19 0 Manured sand 6.8 2 4 9.1 5. 7 15 Biosolids amended sand 6 9. 1 4.8 0.83 20 Biosolids amended silt loam 6.5 1 4 21.6 6 6 22 0 Biosolids amended manure d sand 6.6 30 10.9 1 6 3 7 [Solid matrix characterization was conducted by the Analytical Research Laboratories (ARL) at the University of Florida, Gainesville, FL. Chemical extraction and analysis was performed using the following methods: EPA 200.7 (Fe and Al), EPA 150.1 (pH), EPA 351.2 (TKN), Loss on Ignition (OM), and BaCl 2 compulsive exchange method, Gillman and Sumpter, 1986 (CEC) The data represent average values from duplicate samples ]. Methods Confirmation of No Isotope Exchange with the Surroundings The tritium atoms were substi tuted inside the rings of CIP and AZ molecules so isotope exchange with the surroundings was not expected The lack of isotope exchange and the stability of the 3 Colucci and Topp (2002). Briefly, 6 samp les of 20 mL scintillation vials were prepared by adding 5 mL water to ~ 8 00 becquerel (Bq) of 3 H CIP or 3 H AZ. Immediately thereafter and following 7 d of incubation at 30 0 C in the dark, triplicate vials were removed and evaporated to dryness under a strea m of nitrogen. 3 H activity was counted in a liquid scintillation counter ( LS 6500, Beckman Coulter, Brea, CA ) at the beginning, and at the end of the incubation to confirm that there was no evaporative or any other loss of 3 H, and therefore no significant tritium exchange with the surroundings.

PAGE 42

42 S orption/ D esorption B atch E xperiments B atch equilibration method s similar to those described by Agyin Birikorang et al. (2010) w ere used to obtain CIP and AZ adsorption / desorption isotherms in 7 solid matrices consisting of soils, biosolids, and biosolids amended soils. Briefly, 1 g of each so il media (the three soils (sand, silt loam, and manured sand), and the three soils amended with the biosolids (1% dw/dw)), was placed in a 15 mL polypropylene cent rifuge tube and was reacted with CIP or AZ dissolved in 5 mL of 0.01 M CaCl 2 Similarly, 0.5 g of the biosolids was placed in a 15 mL polypropylene centrifuge tube and reacted with CIP or AZ dissolved in 2.5 mL of 0.2 M CaCl 2 The latter salt was used beca use Ca 2+ and Cl are typically the most abundant ions making up soil solutions U sing 0.01 M (soils) or 0.2 M (biosolids) CaCl 2 likely provided adsorption competition similar to or much greater than typically encountered by non specifically sorbing CIP and AZ in soil solutions. The 0.01 M CaCl 2 solution provided cmolc (from Ca 2+ ) of ~5 times (sand) equal to (manured sand), or 0.5 times (silt loam) the CEC of the solid matrices The 0.2 CaCl 2 solution provided cmolc of ~ 2 times the CEC of the biosolids. Using CaCl 2 as the background solution s essentially saturated all/most of the cation exchange sites, thus, minimizing the influence of CEC on CIP/AZ sorption. The influence of media CEC on CIP and AZ sorption was intentionally abated to qualitatively asse ss retention mechanisms other than cation exchange (e.g., specific sorption, H bonding, p p bonding, etc.). Based on USEPA (2009) biosolids survey data, f our initial CIP or AZ concentrations were used in the sorption/desorption studies (Table 2 2) T o ensure detectable 3 H activity in the solution phase t he lowest chemical concentrations ( 3 H compound only) were based on the expected Kd values in each solid matri x For

PAGE 43

43 instance, because biosolids were expected to retain significantly more CIP/AZ than s and, ~ 17000 Bq of 3 H activity was added to the biosolids but 10 fold less 3 H activity was added to the sand The other spiked concentrations of CIP and AZ ( 3 H plus un labeled compound) mimicked the mean (10.5 mg/kg for CIP and 0.83 mg/kg for AZ), the 95 th percentile (36.1 mg/kg for CIP and 3.2 mg/kg for AZ), and 5 times the 95 th percentile (180.5 mg/kg for CIP and 16 mg/kg for AZ) concentrations present in U SA biosolids (USEPA, 2009). The 5 x 95 th percentile concentration simulated accumulation of the targ et TOrCs (without attenuation) in the solid matrices following: 1) 5 years of land application of severely contaminated biosolids (annually at 1% rate, w/w) or 2) 30 years of land application of biosolids containing typical (median) CIP/AZ concentrations Assuming the typical 1% (w/w) biosolids application rate, soils (both amended and un amended) were spiked with 100 times less CIP and AZ concentrations than used for the biosolids. The lowest possible concentrations of 3 H AZ in soil matrices (Table 2 2) we re greater than the uppermost limit of environmental relevance, but d etection limitations for the chemical in the solution phase a nd low specific activity ( 29.6 GBq /mmol) of the radiolabel, prompted the use of high AZ concentrations. Table 2 2 Concentrat ions (mg chemical per kg solid matrix) of CIP and AZ added to different solid matrices. Media Lowest a to greatest CIP concentration s spiked ( mg/kg ) Lowest to greatest AZ concentration s spiked ( mg/kg ) Biosolids 0.27 to 180.5 0.84 to 16 Sand 0.012 to 1.8 0.042 b to 0.4 Manured sand 0.027 to 1.8 0.084 b to 0.4 Silt loam 0.068 to 1.8 0.21 b to 0.6 A mended sand 0.027 to 1.8 0.084 b to 0.4 A mended silt loam 0.082 to 1.8 0.25 b to 0.6 A mended manured sand 0.082 to 1.8 0.20 b to 0.5 a Lowest CIP and AZ concentrations for each media were adjusted based on expected Kd values to obtain quantifiable activity in the solution phase. b Lowest AZ concentration exceeds the 95 th percentile concentration found in the US A biosolids (USEPA, 2009 ).

PAGE 44

44 The total number of samples was 168 (2 chemicals x 3 rep licates x 7 media x 4 concentrations). There were 36 controls (2 chemicals x 3 rep licates x (solution s without solid matrices). The lowest and greate st chemical concentrations spiked to solid free 0.2 M and 0.01 M CaCl 2 (solutions without solid matrices) served as measures of initia l chemical activity and accounted for loss of chemicals to process es others than sorption by the solids matrices (e.g., chemical retention onto the polypropylene centrifuge tubes and tube caps). The samples and controls were agitated in the dark on an end over end shaker held at 150 rpm for 36 h. [A preliminary kinetics study showed that 36 h was sufficient for CIP and AZ to reach sorption equilibrium.] Following reaction a ll samples were centrifuged at 3220 x g for 10 min at constant temperature (~25 0 C). An aliquot (1 mL) of the supernatant was mixed with 10 mL of scintillation fluid (Ecoscint A; National Diagnostics, Atlanta, GA) in a 20 mL HDPE scintillation vial. The supernatant remaining in each centrifuge tube was decanted and a subsample of solid matrix (~0.2 g wet weight) was collected and air dried. About 0.1 g of the dried solids was oxidized for independent determinations of the CIP and AZ activities in the sorbed phase. The 3 H activity in the supernatant from each sample was measured for 10 min using a liquid scintillation counter (LS 6500, Beckman Coulter Brea, CA ) with background correction against blank samples. To account for quenching, each sample wa s recounted after addition of an internal standard ( 3 H water (PerkinElmer, Waltham, MA)). The scintillation counter automatically corrected for the radioactive decay since the manufacturing of the radiolabeled chemicals. To minimize chemiluminescence, the scintillation vials were stored for 24 h in the dark before analysis. [The time to minimize

PAGE 45

45 chemiluminescence was determined in a preliminary study where 3 replicates of 3 H CIP or 3 H AZ in 0.2 M CaCl 2 were equilibrated with biosolids for 24 h, centrifuged at 3220 x g, collected, and analyzed for significant decrease in 3 H activity over 36 h] Desorption was initiated immediately following the sorption step. Five mL of fresh 0.01 M or 0.2 M CaCl 2 solution was added to each centrifuge tube to encourage desorpt ion. Samples were agitated on an end over end shaker held at 150 rpm for 36 h, centrifuged for 10 min at 3220 x g, and the 3 H activity in each supernatant sample determined. Desorption kinetics w ere studied by carrying out multiple desorption steps where s ample agitation was repeated in 24 h intervals with additions of fresh volumes of salt solutions for up to several d ays (until the percentage of desorbed chemical was <0.1% of the initial 3 H chemical concentration added). Following completion of the exper iments, solid samples were dried and 0.5 g of the soil samples (amended and un amended) and 0.1 0.2 g of the biosolids samples were oxidized. Combustion occurred in a sample oxidizer (OX 500; RJ Harvey Corp., NY), and the 3 H 2 O produced was collected in 10 mL of a scintillation fluid (Scinti Safe 50; Fisher Scientific, Hampton, NH). The solutions were analyzed on the liquid scintillation counter as described earlier for determination of the total remaining (sorbed) 3 H CIP or AZ for mass balance purposes. T he pH of each sample was measured every 36 h and remained constant (less than 0. 2 change) during the course of the experiments. Centrifuge tubes, biosolids, amended and un amended soils, salt solutions, etc. were weighed at every step to account for 3 H i n solutions remaining in the samples after sample drying/decantation.

PAGE 46

46 Ciprofloxacin S orption onto C entrifuge T ubes Sorption of CIP onto the centrifuge tubes was a concern. Labware, including centrifuge tubes, can re move significant ( up to 7 0%) CIP from the solution depending upon the tube material (e.g., glass, Teflon, polypropylene) and the background solution used (Goulas et al., 2016). We conducted preliminary studies to assess CIP retention by centrifuge tubes for our experimental conditions. Briefly, ~ 1700 Bq of 3 H CIP or ~ 250 Bq of 14 C CIP was added to polypropylene centrifuge tubes containing 5 mL of double deionized (DDI) water, 0.01 M CaCl 2 or 0.2 M CaCl 2 An aliquot (1 mL) from each centrifuge tube was drawn immediately after CIP addition and ana lyzed on the liquid scintillation counter. Following 36 h agitation in the dark on an end over end shaker held at 150 rpm, another 1 mL aliquot was drawn and analyzed on the scintillation counter. Equilibration of B iosolids with CIP and AZ Rather than spi king soils already amended with biosolids (the typical approach), amendment of pre equilibrated biosolids to soils represents a more environmentally realistic scenario where CIP and AZ are truly biosolids borne. In addition to the traditional sorption/de sorption batch experiments, an alternate procedure involved equilibration of the biosolids with added CIP or AZ before assessing release potential of biosolids borne CIP and AZ Biosolids were equilibrated with 3 H CIP or 3 H AZ at the same four ( 3 H or 3 H + un labeled compound) concentrations used in the traditional batch experiments. The biosolids and spikes were equilibrated for a week in the dark at 23 25 0 C on an end over end shaker held at 150 rpm. Preliminary studies suggested spike (and isotopic) equil ibrium after only 36 h of incubation.

PAGE 47

47 After equilibration, biosolids were rapidly flushed with water to remove CIP/AZ in the solution phase while minimizing equilibration of the added water and the biosolids. The set up consisted of a filter paper (0.45 m, 5 cm diameter) placed inside a V shaped glass funnel. The glass funnel was attached to a 12 fold vacuum manifold attached to a vacuum held at 12 bars. Approx imately 500 mg of each biosolids sample was placed on the filter paper. One mL of double deioniz ed water was added to the biosolids and the vacuum was turned on for 1 min. The solution passing through the filter paper was discarded. The process of water addition was repeated twice more (1 mL each time) at successive intervals of 45 sec. Therefore, a total of 3 mL water was added to each of the biosolids samples. The vacuum was run for a total of 3 min. The biosolids were removed from the filter paper by scraping and spread onto a plastic sheet inside a fume hood and covered by another plastic sheet. T he biosolids were then allowed to air dry for 1 d in the dark. The final product was assumed to represent a scenario where the target compounds were truly biosolids borne, i.e., sorbed (reversibly and/or irreversibly) onto the biosolids particles. Desorpti on from B iosolids P reviously E quilibrated with CIP or AZ Desorption of biosolids borne CIP and AZ from the pre equilibrated biosolids was assessed using the following three approaches: S oil effects on desorption of biosolids borne CIP or AZ borne CIP and AZ after land application better simulates environmental scenarios and may be different than that assessed in borne CIP and AZ, s ubsamples of the biosolids previously equilibrated with CIP or AZ were added to each of the 3 soil media (sand, manur e d sand, and silt loam) at a 1%

PAGE 48

48 application rate (w/w). One gram of each biosolids amended media was placed in a 15 mL polypropylene centri fuge tube followed by addition of 5 mL of 0.01 M CaCl 2 solution to encourage desorption. The samples were agitated in the dark on an end over end shaker for 36 h and desorption was studied for up to 3 desorption steps similar to as described earlier. The s tudy consisted of triplicates and the total number of initial samples was 72. TOrC competition effects on desorption Because CIP and AZ could compete for similar sorption sites, we assessed the effect of CIP on AZ desorption (and vice versa). A 0.2 g subsample of biosolids pre equilibrated with AZ was reacted with un labeled CIP at concentrations at least 5 times the initial AZ concentration. Similar studies of CIP effects on AZ desorption were conducted. Desorption of biosolids borne CIP and AZ was as sessed for up to 2 desorption steps. The study was conducted in triplicate and the total number of initial samples was 24. S pecifically sorbed metal effect s Lead strongly and specifically adsorb s to soils/biosolids matrices (Appel et al., 2008), and may co mpete with specifically bound TOrCs. To assess CIP and AZ desorption behavior in presence of ions like Pb 2+ 1 mL of 0.01 M PbCl 2 was added to 0.2 g subsamples of the pre equilibrated biosolids. The 0.01 M PbCl 2 provided at least 100 times more cmolc Pb 2+ than the TOrCs added to the biosolids. We then assessed desorption of CIP and AZ from the biosolids for up to 2 desorption steps. The study was conducted in triplicate and consisted of 24 samples.

PAGE 49

49 Following completion of the desorption studies, 500 mg amen ded soils and 100 200 mg of the biosolids samples were combusted and analyzed as described earlier to assess mass balance. Detection Limits and Statistical Analysis Tritium counting efficiencies of 42 58% were obtained on the liquid scintillation counter i n all samples. An operationally defined cut off quantification limit of 1.7 Bq was used such that an activity below 1.7 Bq was considered insignificant for the study purposes. Although, the 3 H activities in all the solid matrices were greater than the mini mum detectable true activity (MDTA) value of 0.46 Bq radioactivity was present in the systems), the activities less than 1.7 Bq were regarded The MDTA is defined as the smallest amount of activi ty required to be in a sample in order that a measurement can be expected to correctly imply the presence, and correctly quantitatively assay the activity with a predetermined degree of confidence ( Altshuler and Pasternack, 1963 ). The total chemical in the sorbed and in the solution phase was calculated based on the underlying assumption that both un labeled and 3 H labeled fractions of each chemical behave identically in the sorption/desorption system. The same 3 H CIP/AZ concentrati 5 ) in the sorbed phase ( see Desorption from B iosolids P reviously E quilibrated with CIP or AZ section in results ) from 36 h to a week suggest attainment of isotope equilibrium. The data were analyzed using R for normality (Shapiro Wilk test) and homogeneity of variance (Levene test). The ANOVA was used for analysis of normal and homogeneous data whereas the Kruskal Wallis test was used for non normal data. All data analysis w as

PAGE 50

50 Results and Discussion CIP S orption on to C ent rifuge T ubes As much as 7 0% of initially added CIP can sorb to various types of labware (including polypropylene centrifuge tubes used herein) when CIP is added in deionized water ( Goulas et al. 2016) However in the presence of background salts cation competition for exchange sites increases and CIP retention on the labware decreases The centrifuge tubes containing CIP in DDI water retained 74 4 % 3 H CIP or 47 32% 14 C CIP. Ciprofloxacin retention was also significant in 0.01 M CaCl 2 solutions where 27 5% and 22 2% of 3 H and 14 C CIP, respectively, was retained on the centrifuge tubes. There was no CIP retention by the tubes in presence of 0.2 M CaCl 2 background. These results confirm that CIP sorption to labware depends upon background sa lt solution but that h competitively impede CIP sorption to the labware. M ost sorption/desorption studies (including this one) typically utilize 0.01 M (or lower) ionic strength of common ions su ch as Ca 2+ so CIP sorption to the labware is likely Some published CIP sorption literature fails to mention CIP sorption by centrifuge tubes even in the soil/biosolids less controls which confounds interpretation of the data. The mass balance analysis from the sample oxidizer, after correction for percent recoveries (93 98%), accounted for 94 107% of the initial 3 H added to the samples. Thus, the mass balance analys e s (for both sorption an d desorption steps by solids/biosolids) show no important loss of the chemicals by retention onto the centrifuge tubes. Likely, the greater affinity of the solid matrices for CIP than the centrifuge tubes minimized extraneous losses of added chemical. Othe rs (e.g.,

PAGE 51

51 Vasudevan et al., 2009; Goulas et al., 2016) reported similarly high (> 90%) mass balances in studies of solid matrices as in our study. The results suggest that: 1. even if CIP sorption to the labware is significant from solutions (without sorbent s), sorption to labware is minimal when appropriate sorbents (e.g., biosolids or soils) are present in the system, and 2. CIP sorption to labware may be problematic in the sample preparation steps (e.g., solid phase extraction and compound purification) prior to TOrC analysis by instruments like LC MS/MS Such retention could result in significant chemical losses that can be inappropriately attributed to sorption of the chemical by solid matrix. S orption/ D esorption B atch E xperiments Sorption At e quilibrium, biosolids sorbed essentially all (~99%) of the added CIP and AZ. The percent sorption trend among soils was: sand (~90% CIP and ~70% AZ) manu re d sand (~93% CIP and ~73% AZ) < silt loam (~98% CIP and ~97% AZ). The biosolids amended soils followed a similar sorption trend: sand (~94% CIP and ~73% AZ) manured sand (~94% and ~75% AZ) < silt loam (~98.5% CIP and ~97.5% AZ). The addition of biosolids to the soils provided an estimated additional CEC of around 1.85 cmolc per g of soil ; CEC increases of ~70%, ~20%, and ~10% in sand, manured sand, and silt loam soils, respectively. Increases in CIP and AZ retention were, however, only marginal (~1 3% inc rease in sorption) suggesting limited influence of media CEC on the chemical retention Retention of CIP or AZ increased with increasing soil CEC values, but even the sand (CEC = 2.7 cmolc/kg) retained ~90% CIP and ~70% AZ. Silica, with a point of zero ch arge of 2.9, is negatively charged at environmentally relevant pH values (Peterson et al., 2009). Organic cations can sorb onto silica by electrostatic attraction, van der Waals attraction, cation bridging, and/or irreversible

PAGE 52

52 binding, and zwitter ions (li ke CIP) can also sorb by COO bridge complexes ( Turku et al., 2007; Peterson et al., 2009; Chen et al., 2013). A typical 1% (w/w) application rate of a biosolids containing the 5 x 95 th percentile concentration of CIP (180.5 mg/kg) and AZ (16 mg/kg) adds o nly ~ 5 x 10 4 cmolc CIP and ~10 4 cmolc AZ per kg of soil. The amounts of the target TOrCs were, therefore, thousands to tens of thousands fold smaller than media CEC, and could partly explain extensive retention even by the sand (CEC = 2.7 cmolc/kg). Sec ondary solid phases (including Fe and Al oxides) can also contribute to extensive sorption of TOrCs by sand. Chen et al. (2013) compared CIP sorption onto sand coated with small amounts (few hundred mg/kg) of total Al and Fe with sorption by acid (coated sand). Thus, although CIP sorption was 2 0 times greater and stronger on the Herein, the small amoun ts of Melich 3 extractable Al (~0.03 cmolc) and Fe (~0.003 cmolc) per kg of the sand were, nevertheless, at least ~50 times more than the cmolc of added target TOrCs. Thus, TOrC complexation with the oxides, especially for CIP, may have contributed to exte nsive retention by the sand. Sand sorbed ~70% AZ, likely because of weak electrostatic attractions as evident from ~50% cumulative desorption over 7 desorption steps (see next section) The true nature and mechanisms of chemical adsorption by silica are st ill unknown and require additional studies to fully understand but we re beyond the scope of this work Desorption Both CIP and AZ desorbed continuously but slowly (Figure 2 1 ). Cumulatively, considerably more AZ desorbed from each of the solid matrices than CIP, suggesting

PAGE 53

53 AZ sorption is weaker than that of CIP. Ciprofloxacin can form surface complexes with metal oxides because it possesses a COO moiety, whereas AZ does not ( Figure 1 1 ). Surface complexation with metals (e.g., Fe and Al oxides) as an additional retention mechanism may explain the greater sorption and less desorption of CIP than AZ. Cumulative desorption of both compounds from biosolids was minimal (~0.5% CIP and ~1% AZ) over the course of 7 desorption steps. The percent desorption trend for the un amended soils was: sand (~26% CIP and ~50% AZ) > manured sand (~4.5% CIP and ~32% AZ) > silt loam (~3.7% CIP and ~6% AZ). Biosolids a mendment of the soils reduced c ompound desorption, but the trend among soils remained unchanged : sand (~13.5% CIP and ~36% AZ) > manured sand (~4% CIP and ~28% AZ) > silt loam (~1.5% CIP and ~4% AZ). The results are in accordance with literature reports of minimal desorption of many bi osolids borne TOrCs. Goulas et al. (2016) spiked CIP to silty loam soils amended with animal manure or compost of sewage sludge and green wastes (3% w/w) to study CIP desorption using various extractants over 28 d Ciprofloxacin desorption was minimal (<1 5%) in soil/compost systems, but significantly greater (even >10% for some extractions) in soil/manure systems, suggesting retention/release of CIP depends upon the type and nature of organic sorbent. Similarly, some literature suggest s variation in the re tention/release behavior of the progenitor compound of AZ by different biosolids (Radjenovic et al., 2009), which could affect desorption behavior. Future studies should include different biosolids materials to more fully characterize biosolids type influences on compound sorption/desorption behavior from biosolids.

PAGE 54

54 Contrary to the results obtained herein and those suggested by literature, orted that ~16% CIP desorbed from biosolids over 2 weeks, as measured with the diffusion gradient in thin films (DGT) technique. The researchers suggested that many batch experiments show little CIP desorption because ,000 x g) used. Sample centrifugation at such high g force likely removes colloidally bound TOrCs from the suspensions generated in the batch experiments, whereas the DGT technique detected colloidally bound TOrCs and estimated considerably greater CIP des orption. Colloidally bound TOrCs may be bioaccessible, so bioaccessible fractions of CIP may be underestimated in some batch experiments. However, we used significantly less centrifugal force (3220 x g) and still found CIP desorption to be minimal. Bioacce ssibility/availability of a chemical is ultimately decided by the exposed organism and, thus, it is important to relate bioaccessibility with bioavailability by studying target organism responses. Subsequent organism response data (discussed in later Chapt ers) identified the true bioavailability of biosolids borne CIP and AZ. S orption/ D esorption B atch E xperiments Isotherms Sorption of CIP and AZ by the solid matrices is well described by linear models (r 2 >0.99) over the entire concentration range (Figure 2 1 ). Partitioning coefficients (Kd) standard deviations (SDs) for each solid matrix derived from the linear models are provided in Table 2 3 The Kds values refer to sorption and Kdd refers to desorption for the 1st desorption step (36 h after sorption step). Both sets of Kds and Kdd values were determined using least squares linear regression.

PAGE 55

55 Table 2 3. Sorption (Kds) and desorption (Kdd) partitioning coefficients predicted by linear model standard deviation (SD), Koc and Kcec values average mass balance (%) SD, and hysteresis index (H), for CIP and AZ in the 7 solid matrices. Chemical Matrix K ds SD (L/kg) K dd SD (L/kg) Kcec a (L/ cmol c ) Koc b (L/kg) Mass balance SD (%) H CIP Biosolids 357 5.3 1035 89 1.93 1661 94.8 0.7 0.003 Sand 37.6 1.9 49 1.6 13.91 57846 103 2.6 0.52 Silt loam 362 7.5 434 19 18.84 69615 97.4 1.8 0.003 Manured sand 61.9 2.9 309 17 6.8 5115 99.7 3.5 0.34 Amended Sand 72.7 2.4 108 3.1 15.13 16523 104 1.9 0.33 Amended Silt loam 334 8.8 1660 84 15.48 49541 98.6 1.2 0.003 Amended manured sand 75.7 3.3 442 22 6.94 5081 98.9 4.1 0.25 AZ Biosolids 428 23 619 94 2.32 1993 96.3 1.3 0.002 Sand 11.2 0.61 12.4 1.1 4.13 17231 101.4 0.9 0.65 Silt loam 188 4.2 250 8.5 9.81 36231 95.7 1.6 0.007 Manured sand 14.2 0.52 21.1 3.1 1.55 1174 100.5 2.4 0.53 Amended Sand 12.5 0.46 21.9 2.4 2.6 2841 98.1 2.9 0.51 Amended Silt loam 202 3.7 356 3.1 9.36 29956 97.7 1.6 0.006 Amended manured sand 15.3 0.71 22 1.6 1.41 1027 95.8 1.4 0.29 a Kcec= Kds normalized to media CEC. b Koc= Kds normalized to media OC (assuming OC to be 50% of the OM). The Kds values for AZ ranged from ~10 to 200 L/kg for the soils in our study, and averaged ~430 L/kg for the biosolids (Table 2 3). The Kds values for AZ in the biosolids (at background ionic strength of 0.2 M) are similar to the value of 376 86 L/kg obt ained by Gobel et al. (2005) in activated sludge (background ionic strength not provided). Literature reports a soil Kds value of ~150 L/kg (in a silty loam soil) (Maier and Tjeerdema, 2018), similar to the Kds values determined herein. Soil Koc values ran ging from 22800 to 59600 L/kg reported by Ericson (2007) yield soil Kds values for AZ similar to results obtained herein. For instance, a Koc value of 22800 L/kg yields a Kds of ~15 L/kg for AZ in the sand (OM= 0.1 6 %) used in our study. Literature suggests 19000 L/kg (Nowara et al., 1997; Wu et al., 2009 ;

PAGE 56

56 using CIP concentrations in biosolids similar to those used by others (e.g., Nowara et al., 1997; Vasudevan e t al., 2009; Wu et al., 2009 ; the Kds values obtained herein are much smaller. For instance, Kds values for biosolids sorbed CIP are 34 times smaller than the Kds value (~12,000 L/kg) reported by a similar concentration range. The lower Kds values c ould be due to the use of 0.2 M CaCl 2 instead of the typical 0.01 M CaCl 2 background salt concentration used by others. Differences in ionic strength can affect Kds values to some extent (Sassman and Lee, 2005; Carmosini and Lee, 2009 ; Vasudevan et al., 2009), but cannot explain the order of magnitude difference found herein. W e used lower centrifugal force (3220 x g for 10 min) compared to others (10,000 x g for 30 min) and did not filter the supernatants. Thus, colloidally bound CIP and AZ may have been present in our supernatants resulting in smaller Kds values than measured by others To that end, we assessed effects of centrifugal force and supernatant filtration on the Kds values of CIP and AZ for the biosolids. In one experiment, the samples (after sorption equilibration) were centrifuged at 3220 x g for 10 min or at 10,00 0 x g for 30 min, followed by analysis of an aliquot (1 mL) of the supernatants via liquid scintillation as usual. The experiment involved 3 H and 14 C CIP and 3 H AZ, the biosolids, 3 rep licates and 2 chemical concentrations (low and high used in the batch experiments; total samples = 18). We included 14 C CIP in the study to confirm the accuracy of the results obtained from the 3 H labeled compounds. In the other experiment, a fraction of the supernatants after centrifugation at 3220 x g for 10 min were analy zed directly via liquid scintillation, whereas another fraction was passed through a 0.2 m filter before analysis.

PAGE 57

57 Sorption partitioning coefficient ( Kds ) values for centrifugation at 3220 x g for 10 min were 402 35 L/kg ( 14 C CIP), 376 16 L/kg ( 3 H CI P), and 399 27 L/kg ( 3 H AZ). The Kds values are the same as those obtained from initial sorption/desorption experiment and the Kds values for 14 C and 3 H CIP were statistically the same confirming the results obtained from the 3 H labeled compound. Us ing greater centrifugal force increased the Kds values to 487 29 L/kg ( 14 C CIP), 457 33 L/kg ( 3 H CIP), and 511 22 L/kg ( 3 H AZ). The Kds values obtained from the filtered supernatants were ~150 L/kg (CIP) and ~100 L/kg (AZ) greater than those obtained from unfiltered supernatants. The considerable increases (42% for CIP and 23% for AZ) in Kds values indicate significant influence of colloidally bound CIP and AZ, but still do not explain the huge differences between Kds values for CIP obtained herein and those reported in the literature. Colloids bind metals and organic phosphorus in Wang and Guo, 2000; Chen and W ang, 2001 ; v an Moorleghem et al., 2013). Literature is currently missing on bioavailability of colloidally bound TOrCs, and apparently warrants investigation. Some colloidally bound CIP and AZ is likely bioaccessible (perhaps to earthworms if not plants an d/or microorganisms) and should be accounted for in risk assessments. However, force and supernatant filtration effects are not the only reasons for the smaller Kds values obtained herein Perhaps the single biosolids used herein had unusual ly low CIP sorption capacity. Previously reported CIP/AZ sorption/desorption studies used loss of chemical from the solution phase to estimate sorbed chemical concentrations, and some studies experienced inadequate mass balances due to poor extraction reco veries. The analytes

PAGE 58

58 can be lost by processes other than soil sorption such as by adsorption to labware before sample analysis (Goulas et al., 2016). The sorption/desorption isotherms generated herein were generated with adequate mass balance, recoveries, and independent measurements of the 3 H TOrCs in the sorbed and solution phases. Th e same results utilizing 14 C and 3 H labeled CIP indicate that the Kds values obtained herein are reliable. Dividing Kds values by media CEC and OC contents (Table 2 3) fa iled to normalize Kds values to constant values across the different media. The vastly different Kcec and Koc values obtained for the 7 media suggest that CEC and OM had limited influence on compound sorption. The limited influence of media CEC can be expl ained by the addition of sufficient Ca 2+ to saturate (out compete cationic CIP or AZ species) all biosolids, sand, and manured sand and most silt loam cation exchange sites. The Ca 2+ addition thus essentially minimized the influence of media CEC on the che mical sorption. 0.01 M CaCl 2 (soils) and 0.2 M CaCl 2 (biosolids) were used as background salts, but the ratios of cmolc added (via Ca 2+ ) and the solid matrix CEC were different for different media. The different ratios likely resulted in different exchange site competition in different media and could explain the failed Kds normalization to media CEC. However, other explanations like greater colloidally bound CIP and AZ in the biosolids than in the silt loam that was not removed from the solution by centrif ugation, more interferences from other chemicals and ions in the biosolids than soils, etc. cannot be ruled out. Importantly, all solid matrices strongly sorbed the target compounds and showed significant desorption hysteresis (Figure 2 1 ). The Kdd values of CIP and AZ were

PAGE 59

59 greater than Kds values (Table 2 3 ). The Kdd values indicate that CIP sorbed more strongly to the solid matrices than AZ. Greater steric hindrances (due to larger size and greater hydrophobicity) may have limited AZ access to some adsorption sites available to CIP. Also as CIP possesses a COO moiety, some CIP likely formed surface complexes with positively charged metal oxides, but AZ could not. Nonetheless, CIP and AZ strongly bound to the solid matrices. Desorption was less in t he 2nd desorption step and continued to decrease through the last desorption step (Figure 2 1 ), suggesting formation of some irreversibly bound residues and/or bound residues with severely limited reversibility. The ratio of slopes for desorption and adsorption is a measure of desorption hysteresis (Agyin Birikorang et al., 2010). The hysteresis coefficient, H, was calculated for the highest concentration of the two compounds in the desorption experiments using the following equation: The H values of the target TOrCs for all seven matrices were <1 (Table 2 3) and decreased with increasing media CEC. Values of 0.002 to 0.003 for the biosoli ds, 0.003 to 0.007 for the silt loam (amended and un amended), 0.25 to 0.53 for the manured sand (amended and un amended), and 0.52 to 0.65 for the sand (amended and un amended) (Table 2 3) clearly demonstrate significant hysteresis (Agyin Birikorang et al ., 2010). The especially small hysteresis coefficients for the biosolids suggest that desorption of the biosolids borne CIP and AZ is negligible.

PAGE 60

60 Figure 2 1. Representative sorption/desorption isotherms for CIP (top) and AZ (bottom). The solid black lines (sorption) and the solid red lines (desorption) represent the fit of the data. The dashed blue lines (sorption) indicate upper and lower 95% confidence limits

PAGE 61

61 Desorption from B iosolids P reviously E quilibrated with CIP or AZ Sorption coefficients (Kds values) for the pre equilibrated biosolids were 397 44 L/kg (CIP) and 414 18 L/kg (AZ), which are the same as values obtained in the 36 h of sorption study (Table 2 3). Thus, sorption/desorption behavior of the two compounds did not change, up to at least a week, after attainment of sorption equilibrium (at 36 h in the sorption study), suggesting attainment of isotope equilibrium. Desorption coefficients (Kdd values) of the compounds from the pre equilibrated biosolids were 987 17 L /kg (CIP) and 714 41 L/kg (AZ) Following application of the pre equilibrated biosolids to the soils, the Kdd values were 947 59 L/kg (CIP) and 672 50 L/kg (AZ). Th us, the Kdd values for the biosolids for each compound were statistically the same wi th or without application to the soils. The results indicate that Kdd values do not change after land application of contaminated biosolids and that borne CIP and AZ is minimal. No significant competition occurred between CI P and AZ for sorption sites. The cumulative desorption of AZ from biosolids in the presence of CIP (and vice versa) was <2.5%. Addition of 0.01 M PbCl 2 provided 10 cmolc of Pb 2+ per kg biosolids, but had negligible effect (<1.5% desor ption ) on CIP and AZ d esorption. The biosolids had been previously equilibrated with CIP or AZ for a week and negligible desorption in presence of (CIP versus AZ) and Pb 2+ suggests one or more of several possibilities: 1. most of the sorption sites for CIP o n the biosolids surfaces are different than sorption sites for AZ; which is possible especially considering the differences in sizes of the two compounds, zwitter ionic versus cationic nature, and separation of charges on the molecules (Figure 1 1) ;

PAGE 62

62 2. some or most of the sorption sites on the biosolids are similar for CIP and AZ, but the biosolids has sufficient sorption sites to negate competition between the compounds; 3. the biosolids strongly retain the initially sorbed compound (most likely explanation in lieu of cumulative CIP and AZ desorption data); 4. CIP and AZ sorption is non and sorption/desorption data); and/or 5. specific sorption sites for CIP/AZ and Pb 2+ are different. We did not separately analyze sorption of the CIP, AZ, or Pb 2+ in the competition studies and therefore, are limited in our discussion of the actual mechanisms operating in the biosolids systems. Cations in soils/biosolids could either facilitate or imped e CIP and (likely) AZ sorption depending on if the cations are present on the sorbent or are in solution phase (Chen et al., 2013). A thorough assessment of the ongoing mechanisms, requires a detailed competition study involving varying chemical concentrat ions (and possibly other environmentally relevant extractants) and assessing sorption / desorption of each chemical in presence of the other. Such experiments we re, however, beyond the scope of current study. Considerable colloidally bound CIP and AZ may hav e remained on the 0.45 m filter paper used for the preparation of pre equilibrated biosolids. Therefore, we likely excluded some of the more labile (i.e., colloidally bound) CIP and AZ from the desorption studies. Desorption of CIP and AZ from colloidal f ractions may be of interest because some researchers identify colloidally bound chemical as bioavailable (e.g., T he biosolids were equilibrated with the target TOrCs using 0.2 M CaCl 2 as the background salt that represented ~ 2 times greater cmolc than CEC of the biosolids. I onic strengths of the typical background solutions in biosolids are less than assessed

PAGE 63

63 herein. Therefore, we likely underestimated CIP and AZ sorption to biosolids and thus, overestimated percent desorptio n. Minimal d esorption of biosolids borne CIP and AZ using 0.2 M CaCl 2 as background salt and the 4 approaches i.e., Sorption competition between CIP and AZ addition of PbCl 2 and application of the contaminated biosolids to the three soils; strongly suggest that the bioaccessibility of biosolids borne CIP and AZ is minimal. Bioaccessibility v ersus B ioavailability Chemical b ioavailability ultimately depends upon the exposed organism but assessing bioaccessibility for example using extractants or determining desorption isotherms, provides valuable information on the potential bioavailability Our desorption data suggest minimal bioaccessibility of the biosolids borne CIP and AZ. B ioaccessibility mu st however, be correlated with bioavailability to be meaningful. To that end plant, earthworm, and microbial exposure studies were conducted using the same biosolids. The organism exposure studies are detailed in subsequent chapters. The present study inv olved only one class A biosolids. Retention/release behavior (and thus bioaccessibilities) of chemicals can differ among different biosolids because of differen ce s in physico chemical properties. Literature data reasonably support minimal bioaccessibility potential of CIP, but scarce data on AZ require additional studies using different biosolids.

PAGE 64

64 Conclusions Both CIP and AZ extensively and strongly sorb to biosolids and soils at environmentally relevant as well as unrealistically high chemical concentra tions Desorption of biosolids borne CIP and AZ is minor and, accordingly, bioaccessibility and (likely) bioavailability are negligible. The results are consistent with our hypothesis of strong sorption and minimal desorption of biosolids borne CIP and AZ. Strong linearity of both sorption and desorption isotherms over the tested concentration range suggests minimal bioaccessibility of target TOrCs even when severely contaminated biosolids are land applied over several years. However, the study involved onl y one class A biosolids and retention/release behavior of different biosolids can logically be expected to vary L iterature confirms strong retention and minimal release of CIP from biosolids and soils Based on 10s to 1000s fold greater Kd values of CIP reported in the literature, the bioaccessibility o f CIP is likely equal to or less than measured here However, data on AZ retention/release are scarce and warrant further investigations utilizing various biosolids of different characteristics Accounting for the lability of colloidally bound CIP and AZ is desirable for conservatively estimat ing Kd values ( bioaccessibilit ies) of biosolids borne CIP and AZ. S ome colloidally bound CIP and AZ may be bioavailable to certain organisms. studies suggest that excess concentrations of CIP, AZ, or Pb 2+ have minimal effect on the desorption of the target TOrCs and that both of the target TOrCs likely specifically bind to the biosolids matrices. However, a full sorption / desorption competition study involving varying concentrations of similar compounds, ionic strengths, background salts, etc., is necessary to better assess the

PAGE 65

65 mechanisms involved in retention/release behavior of the two compounds in biosolids systems.

PAGE 66

66 CHAPTER 3 BIOSOLIDS BORNE CIPROFLOXACIN (CIP) AND AZITHROMYCIN (AZ): PLANT SYSTEM Synopsis Plant uptake of trace organic chemicals ( TOrCs ) can result in phytotoxicities and/ or serve as a pathway of human/animal exposure Information on phytotoxicity and phytoaccumulation of environmentally relevant concentrations of ciprofloxacin (CIP) and azithromycin (AZ) from biosolids amended soils is limited To that end, a greenhouse stud y was conducted to assess the phytoavailability of biosolids borne CIP and AZ. The study involved tw o biosolids amended soil matrices (including sand as a worst case scenario), environmentally relevant chemical concentration range s and three crops The plants ( lettuce Lactuca sativa var. buttercrunch ; radish Raphanus sativus var. crunchy royale ; and t all fescue grass Festuca arundinacea var. Kentucky 31 ) represen t different morphologies, physiologies, and chemical exposure scenarios. Phytotoxicity and phytoaccumulation w ere negligible even at the uppermost end of environmentally relevant target TOrC c oncentrations even in sand Separate p hytotoxicity experiments (sand directly spiked with target TOrCs without biosolids) revealed a n o observed adverse effect concentrations (NOAEC) of 3.2 mg/kg for AZ for the three crops ; and of 0.36 mg/kg (lettuce) and 1.1 mg/kg (radish and fescue) for CIP. The NOAEC value s are greater (100 fold in case of AZ) t han t he environmentally relevant c oncentrations of the two TOrCs in biosolids L and application of biosolids borne CIP and AZ appears to pose De minimis risks to plants and the impacts of the target TOrCs o n human and/or animal food chains are likely insignificant

PAGE 67

67 Introduction Plants grown in biosolids amended soils can potentially take up various biosolids borne trace organic chemicals ( TOrCs ) and in turn, serve as sources for human and/or animal exposure (Grote et al., 2007). Land application of biosolids is increasing with increasing population and urbanization. A bout 50 60% of biosolids generated in the USA are applied to crop and range lands as soil amendment s (McLain et al., 2017) thus increasing the potential exposure of plants to biosolids borne TOrCs. Quantifying the transfer of biosolids borne TOrCs from the environment to the ecological and/or human food web is, therefore, crucial in a proper risk assessment of potential adverse effects on environmental and human health. Ciprofloxacin (CIP) and azithromycin (AZ) are two TOrCs of interest for plant uptake studies because of limited (CIP) or scar c e (AZ) environmentally relevant data in biosolids amended soils. Plant uptake depends upon properties of the chemical, the matrix containing the chemical, exposure time, plant species, soil properties, humidity, and temperature (Miller et al., 2016). Bioaccessibility is the first step in determ ining plant uptake of a chemical. Strong adsorption of CIP and AZ to soil/biosolids surfaces and tendency to form non extractable (non plant available) residues over time ( Sidhu, Chapter 2; Nowara et al., 1997; Castela Papin et al., 1999; Ericson, 2007; Va sudevan et al., 2009; Girardi et al., 2011; Chenxi et al., 2008; Maier and Tjeerdema, 2018 ) should limit plant uptake of the two ionic TOrCs. However, in a dynamic plant soil continuum various factors can release sorbed chemicals increasing chemical availa bility for plant uptake. For instance, ion exchangeable (reversibly adsorbed) TOrCs may become plant available through changes in ionic strength of soil solution (during fertilization, irrigation, etc.). Also, rhizosphere interactions can affect uptake of many TOrCs. Rhizospheres

PAGE 68

68 typically support significantly lower pH values (up to 2 units in certain cases), which can alter the speciation of CIP from that in the bulk soil (Miller et al., 2016). Transpiration is considered the main mechanism responsible f or uptake of plant available pharmaceuticals and can facilitate chemical uptake by passive (e.g., diffusion) or active (e.g., proton pumps) transport ( Collins et al., 2011 ; Dodgen et al., 2015). Lipophilicity (as measured using Dow (pH dependent Kow)) and acidity (pKa) are the two main chemical properties usually associated with plant uptake of TOrCs. Lipophilicity can predict plant uptake of neutral organic compounds (Trapp et al., 2000), but is often insufficient to predict uptake of ionic chemicals. Addi tional mechanisms such as electrical attraction (or repulsion), uptake and transport through water, facilitated diffusion through protein channels, protein mediated energy dependent uptake, and ion trapping require consideration (Trapp, 2000; Trapp 2009; M iller et al., 2016). The Dow is a weak indicator of partitioning potential of ionic compounds also because plant membranes (lipid bilayers) can better accommodate charged organic compounds than suggested by solubility in n tion coefficients are better predictors of interactions between polar and ionizable compounds with phospholipid based bio membranes. (Miller et al., 2016). Goldstein et al. (2014) suggested that because ions cross bio membranes at a slower rate than neutra l molecules, uptake and accumulation of ionic chemicals is less than that of neutral chemicals. Uptake of cationic compounds with low lipophilicity (like CIP and AZ) likely occurs by electrical attraction of the cation to the negative charge on the plasmal emma and/or accumulation into the vacuole by ion trap (Wu et al., 2015). Chemicals with pKa values between 5 7 (e.g., CIP) can accumulate through ion trapping because of changes in chemical speciation

PAGE 69

69 when the chemical moves from cell wall (pH~4 5) into ce ll cytoplasm (pH~7.5) (Miller et al., 2016). Chemicals can move through the plant root and enter the vascular system by apoplastic (i.e., between cells along cell walls), symplastic (i.e., through cells via plasmodesmata), or transmembrane (i.e., through c ells via cell membranes) pathways (Miller et al., 2016). Translocation of organic cations into the plant vascular system is likely limited by interactions with negatively charged cell membrane (Miller et al., 2016). Also, the C asparian strip can block apop lastic transport of both CIP and AZ. L arge molecular size (e.g., AZ) and low lipophilicity (e.g., CIP and AZ) can limit symplastic and transmembrane transport of a chemical (Miller et al., 2016). Differences in the composition of cytoplasmic membranes amon g plant species and tissues can result in an order of magnitude differences in uptake potential of a chemical (Miller et al., 2016). Ideal plant accumulation studies should, therefore, involve a variety of plants representing groups with different morpholo gies and physiologies. Ciprofloxacin has negative log Dow values at environmentally relevant pH values (Ross et al., 1992), and log Dow values of AZ are less than 1 at pH less than ~ 8.5 (McFarland et al., 199 7 ). In traditional plant uptake models, negative log Dow values suggest minimal tendency to partition to organic phases like plant tissues. However, several studies suggest a significant potential for plant uptake of ionic organic contaminants (Redshaw et al., 2008; Herklotz et al., 2010; Jones Lepp et al., 2010; Wu et al., 2010; Eggen et al., 2011; Shenker et al., 2011; Hyland et al., 2015; Franklin et al., 2016). The studies involved hydroponic conditions or soils where wastewater effluents or irrigation waters spiked with TOrCs were used instead of us ing biosolids borne

PAGE 70

70 TOrCs. Nonetheless, the studies collectively establish that plants can take up ionic compounds like CIP and AZ, if readily available, in the growth medium. Kumar and Gupta (2016) reasoned that 4 chemical properties -Dow, molecular we ight, number of H bond donors (i.e., OH and NH groups), and the number of H bond acceptors (i.e., O and N atoms) in the structure of a chemical -can predict the potential for uptake of a chemical into cell membranes. Based on the 4 properties, Kumar an d Gupta (2016) devised The negative Dow values, the molecular weight of ~331 g/mole, the presence of two H bond donors, and seven H bond acceptors, suggest accumulation of CIP in plants (if bioacces sible), and potential translocation to leaves and fruits through the plant vascular system (Kumar and Gupta, 2016 ; Miller et al., 2016 ). The large molecular weight (~749 g/mole), five H bond donors, and fourteen H bond acceptors, suggest limited accumulati on or translocation of AZ ( Kumar and Gupta, 2016; Miller et al., 2016). consequently posing an environmental and human health risk (Kipper et al., 2010; Lillenberg et al., 2010 ; Eg gen et al., 2011 ). However, the studies were conducted at unrealistically high TOrC concentrations directly spiked to the soil (i.e., not biosolids borne). Further, the calculated bioaccumulation factors (BAFs) for roots, shoots, and leaves from the studie s by Kipper et al. (2010) and Eggen et al. (2011) suggesting limited plant uptake of CIP. Lillenberg et al. (2010) reported BAF values greater than 2 in lettuce, but CIP concentrations ranged from 10 to 500 mg per kg soil (i.e., at least 25 times greater than the uppermost end of environmentally relevant CIP concentrations of biosolids borne CIP in biosolids amended soils).

PAGE 71

71 Two field scale studies using biosolids amended soils suggest minimal potential for plant uptake of biosolids borne CIP and/or AZ ( Gottschall et al., 2012 ; Sabourin et al., 2012). The biosolids used were relatively low in the target TOrCs (CIP and AZ concentrations were similar to median concentrations found in USA biosolids, USEPA, 2009) and the biosolids application r ate was the same as (Gottschall et al., 2012), or one half, the typical 1% rate (Sabourin et al., 2012) used in USA. Therefore, the biosolids amended soils used for crop growth contained relatively low concentrations of the two TOrCs. Further, crops were p lanted 7 months (Gottschall et al., 2012) or 13 months (Sabourin et al., 2012) after biosolids application. Target TOrCs were detected in only a few plant samples. Sabourin et al. (2012) attributed the minimal plant uptake of TOrCs to natural attenuation p rocesses during the one year delay mandated in Canada between biosolids application and crop harvesting. However, a similar time offset is not required in the USA and crops planted soon after biosolids application could take up significantly more TOrCs th an found by Gottschall et al. (2012) and Sabourin et al. (2012) Further, growing plants in systems where biosolids are contaminated with varying (and greater) concentrations of the chemicals is important to fully assess uptake potential. Data on phytotox icities of the target TOrCs are also scarce (CIP) or absent (AZ). A few studies have been conducted on the toxicity of CIP to aquatic plants, but terrestrial phytotoxicity studies are limited. Some studies suggest that CIP can be toxic to terrestrial plant s ( Migliore et al., 2003 ; Kipper et al., 2010; Eggen et al., 2011), but CIP concentrations used were unrealistically high. Aristilde et al. (2010) reported CIP toxicity to spinach chloroplasts, roots, and shoots in a hydroponics system where the

PAGE 72

72 majority o f the CIP was fully bioaccessible, and CIP concentrations were ~10 times greater than uppermost end of environmentally realistic concentrations. Environmentally relevant concentrations can vary widely for different sources (or sinks) of a chemical. Herein environmentally relevant concentrations in growth media (i.e., biosolids amended soils) are defined in the context of biosolids borne chemicals. The typical concentrations of the target TOrCs in biosolids range between median and average concentrations r eported in the targeted national sewage sludge survey (USEPA, 2009 ) Based on the typical (median to average) chemical concentrations in biosolids and the typical 1% (dw/dw) land application rate, most biosolids amended soils (i.e., typical plant growth me dium) nominally contain 0.05 to 0.11 mg CIP and 0.003 to 0.008 mg AZ per kg. Based on the 95 th percentile concentrations found in USA biosolids, the uppermost end of environmental relevan ce is 0.36 mg CIP and 0.032 mg AZ per kg amended soil. The latter con centrations also represent soil concentrations (without attenuation) from ~7 (CIP) and ~10 (AZ) years of repeated application of biosolids (at 1% (dw/dw) application rate) contaminated with median chemical concentrations. The present work evaluate d the CIP and AZ uptake and toxicity in several crops grown in soils amended with variously contaminated biosolids. Our hypothesis was that biosolids borne CIP and AZ are minimal ly phytotoxic and phytoaccumulat ive because of limited bioaccessibility caused by strong sorption onto biosolids/soils. Although we expected negligible phytotoxicities and phytoaccumulation, the dearth of literature on environmentally relevant concentrations of the biosolids borne target TOrCs rationalized the study. We utilized greenh ouse studies to assess CIP and AZ phytotoxicities to 3

PAGE 73

73 crops (representing a variety of plant types) grown in two biosolids amended soils and an un amended sand Uptake of the target TOrCs by the 3 crops was assessed in a biosolids amended sand as a worst case scenario. Rather than using point (single concentration) estimates of uptake and toxicity potentials found in the already scarce literature, we focused on develop ing response curves (slopes) over varying environmentally relevan t chemical concentration s. Plant response curves are more transferrable and more representative of a multitude of soil conditions than point developing the plant response aspect of ecological implica tions of biosolids borne TOrCs. Materials C IP (CAS No. 85721 33 1; >99.9% purity), AZ (CAS No. 117772 70 0; >99.9% purity) standards, and double deionized water were purchased from Sigma Aldrich (St. Louis, MO). The biosolids (3 Class A, Table 1 2) used w a s anaerobically digested air dried Class A biosolids, from Metropolitan Water Reclamation District of Greater Chicago (MWRDGC) and contained low CIP (1 mg/kg) and AZ (0.06 mg/kg) concentrations (analyzed by AXYS, BC, Canada) Cattle manure ( obtained from University of Florida Dairy Research Unit, located in Hague, Florida ) contain ed undetectable (<0.035 mg/kg) CIP and 0.01 mg AZ /kg (analyzed by AXYS, BC, Canada ) Materials Compa ny (Birmingham, AL). A silt loam soil (fine loamy, mixed, superactive, mesic Fluventic Eutrudepts) was obtained from Wisconsin. Select properties of the soils and the biosolids involved in the study are listed in Table 3 1.

PAGE 74

74 Table 3 1. Properties and nutrie nt contents of soils and biosolids involved in the study. Media pH OM ( g/kg ) CEC (cmolc /kg) TKN (mg/kg) KCl extract. NO x (mg/kg) KCl extract. NH 4 N (mg/kg) Water extract. P (mg/kg) Melich 3 extract. K (mg/kg) Biosolids 6.5 43 0 180 29000 37 3700 7100 4200 Sand 6 1. 6 2.7 4.4 0.11 1.0 0.37 10 Silt loam 6.3 1 0 19 1600 210 250 11 150 Manured sand 6.8 2 4 9.1 1000 0.54 3.4 53 260 Amended sand 6 9. 1 4.8 290 1.0 16 64 52 Amended manured sand 6.6 3 0 11 1300 1.2 35 96 300 [Solid matrix characterization was conducted by the Analytical Research Laboratories (ARL) at the University of Florida, Gainesville, FL. Chemical extraction and analysis was performed using following methods: EPA 350.1 (NH4 N), EPA 353.2 (NOx N); EPA 200.7 (P and K), EPA 150.1 (pH), EPA 35 1.2 (TKN), Loss on Ignition (OM), and BaCl 2 compulsive exchange method, Gillman and Sumpter, 1986 (CEC) The data represent average values f rom duplicate samples] Methods Plant U ptake and P hytotoxicity S tudy ( B iosolids borne TOrCs) A study was conducted under greenhouse conditions (controlled temperature between 20 25 0 C and average daily sunlight of ~12 h) at the University of Florida Turfgrass Envirotron in Gainesville, FL (29.64N, 82.36W) in the spring of 2016. The triplicated study involved: 1. two che micals CIP and AZ, 2. t wo soils the sand and the sand amended with 4% (dw) cattle manure (hereafter referred to as manured sand), subsequently amended with the biosolids containing varying chemical concentrations (described below), 3. three crops lettuce (La ctuca sativa), radish (Raphanus sativus), and tall fescue grass (Festuca arundinacea); and 4. sand and manured sand controls without biosolids or TOrCs. The total number of samples/pots w as (2 chemicals x 4 treatments (i.e., 3 concentrations + control) x 2 soils x 3 crops x 3 replicates) = 144. The greenhouse study was conducted with the intent of generating a plant uptake response curve to the applied compounds (CIP and AZ) at three concentrations, in addition to a control (no

PAGE 75

75 added TOrC). The chemical conc entrations were selected based on the latest targeted national sewage sludge survey (USEPA, 2009) of biosolids produced in the USA. Targeted national sewage sludge surveys are the broadest and most thorough nationwide surveys of biosolids in the USA conduc ted by USEPA and use USEPA approved analytical methods. Compound concentrations in the biosolids were: 1. the inherent CIP/AZ concentration of the biosolids (unspiked), 2. the biosolids spiked with 10.5 mg CIP/kg (average, USEPA, 2009) or 36.1 mg CIP/kg (95 th p ercentile, USEPA, 2009), and 3. the biosolids spiked with 0.83 mg AZ/kg (average, USEPA, 2009) or 3.2 mg azithromycin/kg (95 th percentile, USEPA, 2009). The biosolids were spiked with CIP or AZ using double deionized water as the carrier solvent. The biosoli ds spiking was performed in the greenhouse by spreading a thin layer of the biosolids in fiber glass trays and uniformly spraying appropriate concentrations of CIP or AZ. The double deionized water was allowed to evaporate as the biosolids equilibrated wit h the spikes for a week. During equilibration, the trays were loosely covered with dark sheets to avoid direct exposure to light (some TOrCs, especially CIP, are photosensitive (Lam et al., 2003; Cardoza et al., 2005)). Following equilibration, biosolids (spiked and unspiked) were mixed with the soils for 30 min, at the typical 1% agronomic application rate (dw/dw) in a concrete mixer. The nominal total (i.e., spiked plus inherent) concentrations of CIP per kg of each biosolids amended soil w ere : 0.01 mg, 0.115 mg, and 0.371 mg. The nominal total concentrations of AZ per kg of the biosolids amended sand were 0.0006 mg, 0.0089 mg, and 0.033 mg ; and were 0.001 mg, 0.0093 mg, and 0.033 mg per kg of the biosolids amended manured sand. The manure contained 0.01 mg AZ/kg, so the AZ

PAGE 76

76 concentration in unspiked un amended manured sand (4% manure, dw basis) was 0.0004 mg AZ per kg manured soil. Inclusion of the three chemical concentrations provided an adequate range of environmentally relevant concentrations for deve loping the plant response curves. The sand used herein represents a worst case scenario of minimal CIP and AZ retention (Sidhu, Chapter 2) and, likely, maximum compound phytoavailability. Addition of 4% (dw) manure to the sand provided more retention capa city by adding OM and CEC (Table 3 1 ) and was expected to strongly affect CIP and AZ uptake. Four kg of each biosolids amended soil was then packed into individual 12.7 cm x 24.1 cm (dia x ht) plastic pots (resulting bulk density = ~1.3 Mg/m 3 ) and brought to pot water holding capacity (PWHC) The bottom of each plastic pot (equipped with drainage holes) was lined with water permeable landscape cover to retain the soil in pot while allowing water drainage. In a preliminary study, the PWHC was determined by saturating a pot of dry soil with excess water and allowing free drainage for 24 h. Water retained by the soil in the pot after free drainage stopped was considered the PWHC A soil moisture content of 80% the P W HC was targeted because p ast experiences suggest that a soil water content ~80% of the P W HC equates to the field capacity of the soil (Pannu et al., 2012) The three crops lettuce ( Lactuca sativa var. buttercrunch ), radish ( Raphanus sativus var. crunchy royale ), and tall fescue grass ( Festuca arundinacea var. Kentucky 31 ) were grown in the controls and biosolids amended soils. The crops represent a monocot (fescue grass), dicots (lettuce, radish), and aboveground (lettuce, radish leaves, and fescue grass) and belowground (radis h) biomasses. The three plants were

PAGE 77

77 chosen because of different morphologies, different growth and rooting behaviors and different exposure mechanisms to the contaminated soils that can affect uptake of and/or toxicity to various chemicals (Schroll and Sch eunert, 1992 ; Duarte Davidson and Jones, 1996; Miller et al., 2016 ) including CIP and AZ. Fescue and lettuce/radish represent two major groups of flowering plants. Also, lettuce and radish crops represent a pathway of exposure to humans whereas fescue repr esents a pathway of exposure to grazing animals. The three crops represent several mechanisms of chemical uptake that are common in the majority of food and feedstock crops and, thus, represent chemical uptake potential for a variety of crops. For instance exposure of a biosolids borne chemical is significantly greater to roots than other plant parts, and the transport to above ground biomass is governed by the nature and make up of plant vascular system (different in monocots versus dicots). The location and thickness of Casparian strips that can affect apoplastic transport of chemicals into the plant vascular system is different in monocots (fescue grass) versus dicots (radish and lettuce) All the crops were planted from seed at rates of 5 seeds per pot for lettuce, 20 seeds for radish, and 300 500 seeds for fescue grass (for uniform coverage of pot surface) at a depth of ~ 0.64 cm (fescue grass) and ~ 1.3 cm (lettuce and radish). The excess seeding rates for each pot of lettuce and radish were to ensure g ermination of adequate number of plants per pot. The upper 2.5 5 cm soil inside the pot was kept moist by daily misting with deionized water until seedling emergence. After seedling emergence, two thinnings were performed for radish and lettuce. The first was performed when plant height was at least 2.5 cm where the number of seedlings was reduced to about half of the seeding rate. The second thinning occurred one week later,

PAGE 78

78 resulting in recommended planting rates of 2 lettuce plants and 3 radish plants per pot ( B rown et al., 2015). No thinning was necessary for fescue grass. The thinned plants were discarded. The watering regime after seedling emergence involved irrigating when the water content in the pot was ~10% less than 80% of the P W HC (Pannu et al. 2012). The pots needed watering every 2 3 d and were watered by weighing and adding the appropriate amount of water to each pot to maintain the water content near 80% of PWHC. The fertilization regime involved application of a commercially available wate r (N: P: K 24: 3.5: 13.3, plus micronutrients) and potassium nitrate (N: K 22.6: 100). The fertilizer solutions for lettuce and radish were prepared, separately, by dissolving 60 g of solid Miracle Gro and 15 g KNO 3 in 12 L of water. The fertilizer solution for fescue grass was prepared by mixing 24 g of Miracle Gro and 15 g KNO 3 in 12 L water. Each pot was fertilized with a total of 250 mL fertilizer solution split into 4 5 weekly applications over the entire growing seas on. About 50 mL of the fertilizer solution was applied to each pot after the pot was filled with soil (before planting seeds). Starting ten d ays after germination, pots were provided with the liquid fertilizer (50 62.5 mL per pot) from the remaining fertil izer solution at weekly intervals until harvest. The total fertilizer dose per pot was equivalent to ~185 kg/ha N, ~22 kg/ ha P, and ~239 kg/ha K (radish and lettuce) and ~95 kg/ha N, ~9 kg/ha P, and ~190 kg/ha K (fescue grass). The fertilizer scheme was based on a preliminary study conducted in sand (no biosolids or manure) and recommendations provided by local vegetable and grass experts (Personal Communications, G. Hochmuth and J. Kruse, 2016). The sand

PAGE 79

79 treatments (i.e., un amende d sand controls) warranted use of fertilizer because the sand was low in all essential nutrients and OM (Table 3 1). We opted for fertilizing all the treatments (including biosolids amended sand and manured sand) to maximize plant nutrition and to minimize nutrition deficiency differences associated with the treatments. Table 3 2 shows the amounts of inorganic N, P, and K provided through fertilization ; i norganic N (NH 4 + and NO x ), water extractable P, and Melich 3 extractable K provided through organic ame ndment additions (biosolids and manure) ; and r ecommended fertilizer doses for the three crops. The nutrition needs of the plants, even grown in the un amended sand (except P for lettuce), were met via fertilization and likely minimized effects of treatment induced nutritional differences on crop yields. The amounts of plant available P from the manure and the biosolids far exceeded that provided through fertilizer application. Plant responses to nutrient supply differences could confound comparisons of the yields of the three crops in the different soils. The effects, if any, would likely be greatest when comparing yields in the controls and biosolids amended treatments, but a re likely negligible across biosolids amended treatments for a particular crop, soi l, and chemical. Pots were rotated periodically to minimize greenhouse variations of temperature (johnnyseeds.com), Fairfield, Maine) and local experts, the radishes were harve sted at 32 d after planting and the lettuce plants were harvested at 46 d after planting. The fescue grass was harvested at 32 d after planting by clipping the grass ~0.5 cm above the soil using scissors. After harvesting, radish roots and shoots were sepa rated physically and treated as separate samples. The harvested radish roots, radish shoots,

PAGE 80

80 lettuce, and fescue grass were thoroughly washed in buckets containing deionized water followed by a stream of deionized water. The samples were then patted dry an d weighed to obtain wet weight yields. The wet weight yields (data not shown) were determined to assess consumer exposure because crops are typically consumed fresh ( wet weight basis ) The samples were dried at 40 0 C in paper bags to a constant weight to ob tain dry weight yields. A 25 30 g soil sample from each pot was collected to a depth of 15 cm using a small hoe, homogenized, and dried at 40 0 C to a constant weight. Table 3 2. Nutrient supplementation to plants from various sources. Source Crop(s) N (NH 4 + and NO x ) (kg/ha) P (kg/ha) K (kg/ha) Reference Fertilization Radish, Lettuce 185 22 239 Personal communication, G. Hochmuth, 2016 Fescue 95 9 190 Personal communication, J. Kruse, 2016 Manure (at 4% rate) Radish, Lettuce, Fescue 6.87 105* 500* Measured values (Table 3 1) Biosolids (at 1% rate) Radish, Lettuce, Fescue 74.8 142* 84* Measured values (Table 3 1 ) Recommendations for nutrient poor soils Radish 100 17 95 Hochmuth et al., 1996 Lettuce 102 99 190 Hochmuth et al., 1996 Fescue (per month) 75 11 42 Personal communication, J. Kruse, 2016 *Values are based on water extractable P and Melich 3 extractable K (Table 3 1). The dried plant and soil samples were packed in paper bags and then enclosed in individual Ziploc bags. Bagged samples were stored in a freezer ( 16 0 C) until delivery for analysis. Samples were packaged in ice bags and shipped via overnight delivery to AXYS Analytical Services (BC, Canada) for analysis of CIP and AZ in the plant tissue and in the amen ded soils. The target TOrCs were analyzed by LC MS/MS

PAGE 81

81 following AXYS Method MLA 075 (based on EPA method 1694 for analysis of pharmaceuticals (USEPA, 2007)), details of which are provided in Appendix A Plant uptake of CIP and AZ was assessed and, where po ssible, BAF values were calculated from the slope of plant uptake response curves. The data w ere used in scientifically sound risk assessment of biosolids borne CIP and AZ (Chapter 6) Phytotoxicity Study in Soils ( N o B iosolids) A preliminary study, to as sess the adverse effects of extraordinarily high CIP or AZ concentrations on seed germination, crop growth, etc., was performed in two soils: a sand (for both CIP and AZ) and a silt loam (CIP) or a manured sand (AZ). The sand was low in CEC and OM whereas the silt loam or manured sand had greater CEC and OM values (Table 3 1). The silt loam soil (fine loamy, mixed, superactive, mesic Fluventic Eutrudepts) was used in similar work on an other biosolids borne TOrC (Pannu et al., 2012). The manured sand was us ed for AZ because the supply of the silt loam was limited, but manure (4%, dw) amendment to the sand significantly increased media CEC and OM (Table 3 1). The two TOrCs were directly spiked into the un amended soils instead of being biosolids borne becaus e of expected greater bioaccessibilities of the soil borne chemicals ( Sidhu, Chapter 2 ). The sand and the silt loam were directly spiked with 36.1 mg CIP per kg soil. The sand and the manured sand were directly spiked with 3.2 mg AZ per kg soil. The sand r epresented the worst case exposure scenario (i.e., chemical bioaccessibility considerably greater than in a typical agricultural soil). The 36.1 mg CIP and 3.2 mg AZ per kg soil concentrations are 100 times greater than uppermost end of environmentally rel evant CIP/AZ concentrations. In fact, the concentrations mimic

PAGE 82

82 chemical accumulation (i.e., no attenuation) from 100 years of land application (at the typical 1% rate) of biosolids containing 95 th percentile concentrations of CIP or AZ. Adverse effects of CIP and AZ on growth of the three crops (radish, lettuce, and fescue) were assessed by visual inspection and by measuring the plant biomass (dw) yield. Based on the various vegetative measures (e.g., emergence, height, fresh weight, biomass yield), there was no AZ phytotoxicity in any crop in any soil at the extraordinarily high concentration used. No CIP phytotoxicity was observed in the silt loam soil. However, all the crops grown in the sand spiked with 36.1 mg CIP/kg exhibited toxicity (stunted growth and height, and reduced yields) leading to additional, definitive tests. The differences in CIP phytotoxicity between the sand and the silt loam soil l ikely reflect the large differences in soil retention capacities, as reflected in CEC values (2.7 cmolc/k g for sand versus 19.2 cmolc/kg for silt loam (Table 3 1)). The greater CEC in the silt loam soil likely resulted in greater adsorption of CIP to soil particles, and less bioavailability, than in the sand. Definitive Phytotoxicity Tests in Soils (No Biosolids) Three CIP concentrations 0.36 mg/kg (corresponding to the uppermost end of environmentally relevant concentration), 1.1 mg/kg (3 times greater than uppermost environmentally relevant concentration), and 3.61 (10 times greater than uppermost envi ronmentally relevant concentration) were selected for the definitive tests. Based on results from the preliminary study, only un amended (no biosolids or animal manure) sand was spiked with the three CIP concentrations. The study was triplicated and seedin g and growth conditions (including water regime, fertilizer regime, management practices, etc.) were similar to those used in the plant uptake study. The CIP was not

PAGE 83

83 biosolids borne and represented a worst case scenario of direct exposure of CIP to plants growing in a low CEC soil. Statistical Analysis The data were analyzed using R for normality (Shapiro Wilk test) and homogeneity of variance (Levene test). The ANOVA (and subsequent Tukey HSD) was used for analysis of normal and homogeneous data, wherea s the Kruskal Wallis test was used for non Results and Discussion Phytotoxicity in Soil (no Biosolids) Definitive Tests for CIP Germination rates in all the CIP treatments for the three crops exceeded 75%. No adverse effects on plant growth were observed immediately after germination or during early stages of crop growth. However, a clear difference in the crop growth and above ground biomass, especially between the control and 3.61 mg CIP/kg sand treat ments, was observed for all crops beginning 2 weeks after germination. Table 3 3 shows the average radish, lettuce, and fescue dry weight yields from sand spiked with different concentrations of CIP. Table 3 3 Average dry weight yields (g) s tandard dev iations (SDs) for radish, lettuce, and fescue grass grown in sand directly spiked with different concentrations of CIP (in the absence of biosolids). Concentration (mg CIP/kg Sand) Radish yield (dw) (g)SD Lettuce yield (dw) (g)SD Fescue yield (dw) (g)SD 0 2.21 a* 3.60.46 a 3.10.7 a 0.361 2.030.35 a 3.670.55 a 3.770.71 a 1.10 1.670.2 ab 2.430.21 b 2.970.46 a 3.61 0.60.1 b 1.070.38 c 1.430.21 b *Letters (a, b, c) mean dry weight yields among treatments for a particular crop.

PAGE 84

84 The uppermost end of environmentally relevant CIP concentrations (i.e., 0.36 mg/kg soil) did not affect yields of any of the crops. The yields of radishes (roots+shoots) and fescue grass were unaffected by CIP concentrations up to 1.1 mg/kg sand (~3 times greater than uppermost end of environmentally relevant concentrations). The results suggest that environmentally relevant concentrations of CIP do not adversely affect growth and yields of ra dishes, lettuce, and fescue grass. Thus, 0.36 mg CIP/kg sand is the no observed adverse effect concentration (NOAEC) of CIP for lettuce, and 1.1 mg CIP/kg is the NOAEC for radish and fescue yields. The phytotoxicity study was conducted by directly spiking CIP/AZ to the soils. The amendment of biosolids borne CIP and AZ is likely to further increase the NOAEC values because of strong adsorption of CIP/AZ to the biosolids. Environmentally relevant concentrations of biosolids borne CIP and AZ pose negligible t oxicity to the three crops. Plant Uptake Study (Biosolids borne TOrCs) Dry weight yields from the plant uptake study Germination rates for radish and lettuce were greater than 75% and for fescue on plant growth were observed immediately after, or at early stages of, germination. The shoots of the three crops were observably similar in all the treatments except in the un amended sand controls, where growth was slower than in the biosolids amended a nd/or manured soils. However, at harvest, the growth in un amended sand controls was adequate and visually similar to almost all of the other treatments. Inexplicably, growth of lettuce grown in sand amended with unspiked biosolids was clearly better than in the other treatments (Table 3 4). Figures 3 1 to 3 3 are representative pictures of lettuce, radish

PAGE 85

85 and fescue grass, respectively, and show no obvious differences (except for lettuce) in the health of plants exposed to the various concentrations of bio solids borne CIP and AZ. Due to some unknown factor, only 2 radish plants (per pot) survived in 2 of the 3 manured sand replicates amended with unspiked biosolids. For statistical analysis, the average weights of individual radish plants were used to esti mate the yields from the two pots. Briefly, the average dry weight of radish roots in the treatment was ~0.49 g and the yields from the 2 replicates, each with 2 radishes, were 0.7 and 1 g, respectively. The total yields from the 2 replicates were estimate d to be ~1.2 g (i.e., 0.7 + 0.49) and ~1.5 g (i.e., 1 + 0.49). Figure 3 1 Radish plants exposed to biosolids borne CIP (photo courtsey of author)

PAGE 86

86 Figure 3 2 Fescue gras s exposed to biosolids borne AZ (photo courtsey of author) Figure 3 3 Lettuce plants exposed to biosolids borne CIP (photo courtsey of author).

PAGE 87

87 Table 3 4. Average dry weight yields (g) in the controls (no chemical or biosolids) and the biosolids amended treatments standard deviations (SDs) for radish, lettuce, and fescue g rass. Chemical Growth medium Treatment (expected chemical concentration in the growth medium) Radish Root Radish Shoot Lettuce Fescue Average yield (dw) (g) SD Average yield (dw) (g) SD Average yield (dw) (g) SD Average yield (dw) (g) SD CIP Sand Control 1.80.95 a* 0.730.32 a 3.601.30 a 3.430.58 a Unspiked biosolids (0.01 mg/kg) 1.300.30 a 1.570.84 a 8.930.57 b 4.070.61 a Biosolids spiked with 10.5 mg CIP/kg (0.115 mg/kg) 1.770.72 a 1.120.26 a 3.770.61 a 3.600.50 a Biosolids spiked with 36.1 mg CIP/kg (0.371 mg/kg) 1.400.27 a 1.130.15 a 4.500.66 a 3.600.87 a Manured Sand Control 1.530.42 a 1.070.15 a 3.930.49 a 2.830.42 a Unspiked biosolids (0.01 mg/kg) 1.40 @ 0.2 a 1.10 @ 0.10 a 3.700.80 a 3.100.20 a Biosolids spiked with 10.5 mg CIP/kg (0.115 mg/kg) 2.570.86 a 1.130.21 a 5.770.91 a 4.771.39 a Biosolids spiked with 36.1 mg CIP/kg (0.371 mg/kg) 2.070.91 a 1.100.30 a 5.900.61 a 4.430.65 a AZ Sand Control 1.000.27 a 0.830.25 a 3.770.61 a 3.100.5 a Unspiked biosolids (0.0006 mg/kg) 1.630.84 a 1.260.25 a 8.270.70 b 4.501.21 a Biosolids spiked with 0.83 mg AZ/kg (0.0089 mg/kg) 1.830.87 a 1.600.50 a 4.301.13 a 3.771.15 a Biosolids spiked with 3.2 mg AZ/kg (0.033 mg/kg) 1.630.35 a 1.600.10 a 5.901.47 ab 4.430.76 a Manured Sand Control (0.0004 mg/kg) 1.430.61 a 0.700.20 a 3.970.32 a 2.970.32 a Unspiked biosolids (0.001 mg/kg) 1.971.03 a 1.030.21 a 4.230.61 a 2.930.58 a Biosolids spiked with 0.83 mg AZ/kg (0.0093 mg/kg) 1.601.32 a 1.010.27 a 5.331.10 a 3.570.25 a Biosolids spiked with 3.2 mg AZ/kg (0.033 mg/kg) 1.070.32 a 0.930.29 a 4.500.53 a 4.601.32 a *Letters (a, b) rop, soil, and chemical. @Results adjusted to account for the dry weights of 2 missing radish plants.

PAGE 88

88 The crops had ~6 5% (radish), ~75% (lettuce), and ~80% (fescue) moisture on average, and the wet weight yield (data not shown) trends were the same as those in dry weight yields. The yields of radishes (root and shoots) and fescue were unaffected by trace organic additions at any concentration in any soil. Lettuce yields were greater in sand amended with unspiked biosolids than in other treatments. The yields of lettuce in biosolids amended manured sand were unaffected by trace organic additions. Moreover, lettuce yields f rom the sand and the manured sand controls were similar. Adverse effects on lettuce yields from CIP and AZ are not likely given that no phytotoxicity was observed when high CIP and AZ concentrations were directly spiked to un amended sand. Thus, the greate r yields of lettuce in the unspiked biosolids amended sand appear unique and difficult to explain. One possible explanation is the hormetic effects of CIP and AZ on the lettuce. Hormesis is a phenomenon where low dose of a stressor (e.g., CIP) simulates a positive response (e.g., lettuce yield) (Calabrese and Baldwin, 1997; Poschenrieder et al., 2013). Hormetic effects of various contaminants (including antibiotics) on plant growth have been reported (Calabrese and Blain, 2009). Some nutrient concentrations especially P, supplied by biosolids and/or manure additions were nominally greater than those added through fertilization, and greater than needed for adequate plant growth (Table 3 2). However, no significant differences were observed between the dry we ight yields of radishes (roots and shoots) and fescue grass grown in controls and amended soils. Further, the dry weight yields were generally similar (except for lettuce grown in unspiked biosolids amendments) in the

PAGE 89

89 sand and the manured sand (Table 3 4), suggesting that plant responses to nutrient differences among treatments were negligible. Uptake of biosolids borne CIP and AZ The nominal concentrations of the target TOrCs in soils were based on the inherent and inherent + spiked concentrations in the b iosolids. The expected concentrations considered a 100 fold dilution of biosolids TOrCs when biosolids are applied at a 1% (dw/dw) rate to uncontaminated soil. The actual TOrC concentrations in the soil matrices were measured to determine spiking success ( Table 3 5). Reporting limits (RL values ) were determined for each sample based on the LC MS/MS detected signal to noise ratios and peak areas of specific internal standards or surrogates. Generally, the spiking was successful with reported TOrC concentrat ions ~90% of the spiked concentrations after accounting for background (soil) TOrC concentrations (Table 3 5). Some results were highly variable. Reported concentration was only ~20% of the spiked concentration in sand amended with 1% biosolids that was sp iked with average (i.e., 0.83 mg/kg) AZ concentration (Table 3 5). Variability can result because 1) 13 C 3 15 N Ciprofloxacin used by AXYS as internal standard can sometimes interact with soil samples such that analysis results do not meet method criteria, a nd 2) ~0.5 g soil containing only 5 mg biosolids (source of CIP or AZ) was analyzed per sample. Difficulty in taking a representative sample for analysis is common when dealing with small amounts and can explain large variabilities between replicates and t he much lower than expected AZ concentration in one of the treatments. Concern about the low reported versus expected AZ concentration is, however, mute because plant uptake of the target TOrCs was negligible even in growth media containing uppermost end of environmentally relevant concentrations (Table 3 6).

PAGE 90

90 The expected AZ concentration in the sand amended with unspiked biosolids the AXYS protocol require s analyz ing for both CIP and AZ in all samples. Concentrations of CIP and AZ in the controls (un amended and unspiked sand) were essentially the same as in the sand amended with unspiked biosolids (Table 3 5), but approached non detection in both media. In a dose response study, unspiked biosolids (especially those v ery low in chemical of interest) are better amended treatments. Comparisons between treatments in such cases are more robust because only one major variable (i.e., compound dose) dictates the response. Table 3 5. Average reporting l imit (RL) standard deviation (SD) and corresponding average CIP and AZ concentrations SDs in control and biosolids amended sand (mg/kg dry weight) for various treatments. Chemical Treatment, i.e., sand amended with: RL SD (mg/kg) Concentration expected (mg/kg) Concentration reported (mg/kg) SD % of total that was reported CIP No biosolids (control) 0.023 0.009 0 0.015 0.009* N/A Unspiked biosolids 0.025 0.008 0.01 0.024 0.013 95.2 Biosolids spiked with 10.5 mg CIP/kg 0.031 0.001 0.11 0.117 0.009 92.7 Biosolids spiked with 36.1 mg CIP/kg 0.015 0.003 .371 0.328 0.066 84.4 AZ No biosolids (control) 0.003 0.0001 0 0.0015 0.00006* N/A Unspiked biosolids 0.003 0.0002 0.0006 0.0015 0.00009* N/A Biosolids spiked with 0.83 mg AZ/kg 0.004 0.0004 0.0089 0.002 0.0002* 22 Biosolids spiked with 3.2mg AZ/kg 0.004 0.0002 0.032 0.030 0.08 89.1 *Concentrations below RL were taken as one half of the corresponding RL values for statistical purposes (USEPA, 1991). N/A = not applicable.

PAGE 91

91 The target TOrCs were found in less than half of the plant samples and frequently in only 1 or 2 replicates of a treatment (Table 3 6). Plant concentrations were negligible compared to the CIP and AZ concentra tions in the growth medium (biosolids amended sand). According to the criteria that chemical must be detected in all the replicated treatments and not in the controls (Sabourin et al., 2012), only the AZ uptake by lettuce from the soil containing environme ntally relevant AZ concentrations is significant (Table 3 6). However, lettuce accumulated minimal (~3 g kg) AZ compared to the soil concentration, even at the uppermost end of environmentally relevant concentrations. Because CIP and AZ concentrations for many replicates were below reporting limits, generating plant response curves was not possible. P oint estimates of BAF value s generated from the samples where all replicates were above reporting limits were ~ 0.01 (CIP) and 0.1 (AZ) T he chemical concentr ations in plants are based on dry weight basis. The crops are typically consumed on a wet weight basis, so exposure based on the actual plant BAF values are likely significantly less than 0.01 (CIP) and 0.1 (AZ) Overall, the data suggest minimal uptake an d accumulation potential of environmentally relevant concentrations of biosolids borne CIP and AZ by the 3 crops. Consequently, entry of the target chemicals into plant based human and ecological food chains is miniscule. The results are in accordance with the literature reports ( Gottschall et al., 2012; Sabourin et al., 2012; Prosser and Sibley, 2015 ; Wu et al., 2015; ) and with our hypothesis of minimal uptake potential of the biosolids borne target compounds from biosolids amended soils.

PAGE 92

92 Table 3 6. Reporting limit (RL) standard deviation (SD) and corresponding chemical concentrations of the target TOrCs (mg/kg) in each replicate of the three plants grown (dry weight basis) in the controls and the biosolids amended treatments. Chemical Treatment, i.e., sand amended with: Radish Lettuce Fescue grass RL SD (mg/kg) Concentrati on found (mg/kg) RL SD (mg/kg) Concentration found (mg/kg) RL SD (mg/kg) Concentration found (mg/kg) CIP No biosolids (control) 0.0084 0.0007
PAGE 93

93 The plants bioaccumulated ~10 fold more AZ than CIP. Azithromycin is a larger molecule and has more aromaticity and a greater log Dow (i.e., pH dependent Kow) value than CIP at environmentally relevant pH values. Octanol water partitioning coefficients (Do w values) of organic compounds are often well correlated with uptake into organic tissues; thus, the greater the Dow values, the greater the expected bioaccumulation potential (ECETOC, 2013). The partitioning coefficients (Kd values) of ~360 (CIP) and ~430 (AZ) L/kg in the biosolids (Table 2 3) are similar, so the greater Dow value of AZ is consistent with greater bioaccumulation (on a per unit basis) of AZ than CIP. Results suggest that, like other TOrCs, CIP and AZ accumulated via partitioning into plant lipids. Sabourin et al. (2012) and Gottschall et al. (2012) are the only major field studies that analyzed plant uptake of several biosolids borne pharmaceuticals (including CIP and AZ), and the studies occurred in silty loam soils. The authors concluded that the potential for CIP and AZ to enter edible parts of food crops was negligible under normal farming conditions. However, the studies used only single concentrations of each TOrC and followed Canadian best management practices that require a waiting p eriod of several months to a year before harvesting crops grown on a biosolids amended land. Further, the biosolids amended soils contained the lower end of environmentally not case scenarios and minimal chemical exposure and accumulation was expected. Prosser and Sibley (2015) and Wu et al. (2015) conducted literature reviews of uptake potential of several TOrCs by terrestrial plants. The reviews included laborator y, greenhouse, and field studies all of which suggested minimal plant uptake of many

PAGE 94

94 TOrCs. However as noted by the authors, the studies either used single, unrealistic (high or low) chemical concentrations and/or followed Canadian best management practic es. Herein, although a greenhouse study was conducted, plant uptake under a worst case scenario in biosolids borne systems was assessed. Over a range of realistic biosolids borne CIP and AZ concentrations, CIP and AZ uptake by 3 crops, was negligible. The three crops studied represented plants with different morphologies and physiologies. Ciprofloxacin and AZ uptake was negligible in monocot (fescue grass), dicots (lettuce, radish), aboveground (lettuce, fescue grass), and belowground (radish roots) biomass ; suggesting minimal uptake of environmentally relevant concentrations of the target TOrCs by terrestrial crops grown in biosolids amended soils. Plants can metabolize TOrCs after uptake (Coleman et al., 1997; Huber et al., 2009; Wu et al., 2015 ; Miller et al., 2016). Conjugation with certain plant compounds is a common mechanism by which plants detoxify xenobiotics (Coleman et al., 1997 ; Huber et al.; 2009, 2012; Bartha et al., 2010 ). Internal compartmentation (storage in vacuoles) or apoplastic deposition may result in the presence of the conjugates in plant tissues for enzymes cleave many of the conjugates and release the parent compounds or their activated metabolites (Colem an, 1997), thus posing potential risk to consumers (Wu et al., 2015). Typical analysis techniques do not search for conjugated TOrCs and can under estimat e true plant uptake of the TOrC and possible human exposure. Data on conjugated TOrCs and total compou nd uptake by plants are scarce. Dodgen et al. (2013) found that total accumulation (analyzed by 14 C labeled compound) of a few other TOrCs was low and similar to that of the parent TOrCs; dietary uptake of the total

PAGE 95

95 TOrCs was predicted to be negligible. Limited CIP and AZ bioaccessibilities and the small estimated BAF values (0.01 for CIP and 0.1 for AZ ) reported here suggest minimal risks from conjugates. Confirmation of critical chemical accumulation often requires use of radio labeled compounds, but limited resources and unavailability of radio labeled compounds often thwart such investigations. Conclusions Greenhouse studies were conducted to assess p lant uptake and phytotoxicity potentials of CIP and AZ. The plants studied represent a range of terrestrial crops with different morphologies and physiologies and different exposure scenarios to the target TOrCs; all of which could affect chemical uptake and to xicity. Phytotoxicity experiments (without biosolids) found no observed adverse effect concentrations (NOAEC) of 3.2 mg/kg for AZ, and 0.36 mg/kg (lettuce) and 1.1 mg/kg (radish and fescue) for CIP. The NOAEC value for AZ is 100 fold greater than the uppe rmost end of environmentally relevant AZ concentrations and represent accumulation of AZ from land application of a severely contaminated biosolids over 100 years. The NOAEC values for CIP are also greater than environmentally relevant concentrations in bi osolids amended soils. The b iosolids borne status of the chemicals and the presence of adequate inherent organic matter in soils (common in typical agricultural soils) should further increase the NOAEC values. Plant uptake of, and toxicity from, biosolids borne CIP and AZ were negligible even in a worst case scenario (i.e., sand as a growth medium) and in the presence of uppermost end of environmentally relevant chemical concentrations, consistent with our hypothesis of limited bioavailability The study em ployed only one biosolids, and different biosolids can exhibit different retention/release (bioaccessibility) behaviors.

PAGE 96

96 However, a variety of plants (representing food chain crops and pasture grass ) showed negligible response to a range of concentrations of the biosolids borne chemicals. Thus, present data suggest that several years of land application of biosolids containing typical (~median concentrations, USEPA, 2009) concentrations of the target TOrCs, at 1% or greater agronomic rates, pose De minimis risks to plants; even when assuming no chemical attenuation. Even severely AZ or CIP contaminated biosolids pose insignificant risks to plants. The estimated dry weight bioaccumulation factor (BAF) values of 0.01 (CIP) and 0.1 (AZ) suggest that plants gro wn in soils amended with even severely contaminated biosolids pose minimal risks to consumers (i.e., humans and animals).

PAGE 97

97 CHAPTER 4 BIOSOLIDS BORNE CIPROFLOXACIN (CIP) AND AZITHROMYCIN (AZ): MICROBIAL SYSTEM Synopsis Impacts of biosolids borne c iproflox acin (CIP) and azithromycin (AZ) on biosolids and biosolids amended soil microbes are scarcely known To that end, microbial response s to varying environmentally relevant concentrations of biosolids borne CIP and AZ were assessed in an incubation study We studied changes in microbial respiration and employed molecular techniques (RNA analysis) to study various microbial responses involved in the nitrogen and phosphorus cycles and antibiotic resistance development over time. Using 3 H labeled compounds, we assessed CIP and AZ bioavailability to microorganisms and correlated chemical extractability (potential bioaccessibility) and stability to microbial responses in biosolids and amended soil media. Microbial responses to environmentally relevant concentra tions of biosolids borne CIP and AZ were small and showed moderate to weak correlation with bioaccessible fractions of the target chemicals. Positive effects of biosolids amendment appear to outweigh adverse TOrC effects from an agronomic viewpoint. Howeve r, inhibition of some gene expressions ( at least initially) and expression of antibiotic resistance genes warrant further stud ies L ong term field studies are needed to fully assess the potential for b iosolids borne TOrC impacts on microb ial physiology co mmunity structure, and pertinent resistance genes. Introduction Microorganisms are an integral part of biosolids and soils and can affect biosolids/soil health by influencing, for example, nutrient cycling and availability, and enhanc ing chemical bioaccess ibility, biodegradation, and /or exposure Assessing

PAGE 98

98 potential impacts of biosolids borne trace organic contaminants (TOrCs) on microbial activity is therefore, critical to a thorough assessment of ecological and human health risks from land application of biosolids. T he positive impacts of land application of biosolids (e.g., improved nutrient availability, soil physico chemical properties, overall microbial activity, etc.) usually overshadow the adverse effects (Young et al., 2011 ; Park et al., 2013 ). Nevertheless, the ecological risks posed by the introduction of TOrCs into the environment through land application of biosolids could be profound. For example, biosolids containing environmentally relevant antibiotic concentrations can reportedly serve as a potential source of antibiotic resistance in the environment (Munir and Xagoraraki, 2011) E nvironmentally relevant concentrations of CIP and AZ can reported ly c hange long term soil microbial community structure and function (Girardi et al., 2011). The fate s and microbial bioavailabilit ies of biosolids borne TOrCs, particularly those that are ionic, are not well understood. Ciprofloxacin (CIP) and azithromycin (AZ), two such ionic TOrCs common in biosolids, are broad spectrum antibiotics used to treat a number of bacterial infections in humans (Girardi et al., 2011; Parnham et al., 2014). Because of widespread use in consumer products, data gaps in ecotoxicological and/or human health benchmarks, and high detection frequencies reported in literature, the risk posed to environmental and/or human health by CIP and AZ is unclear. The USEPA has identified azithromycin (AZ) and ciprofloxacin (CIP) amongst the high priority compounds for the risk assessment of biosolids borne TOrCs. Ciprofloxacin is the most wi dely used second generation quinolone antibiotic, and acts by inhibiting one or more of a group of enzymes called topoisomerases in

PAGE 99

99 bacterial cells (Silva et al., 2011). Topoisomerases are needed for supercoiling, replication and separation of bacterial DN A (Hooper, 1999). At low concentrations, CIP can be bacteriostatic and inhibit the replication of DNA by inhibiting topoisomerase. At higher concentrations, CIP can also induce the release of free DNA ends from the DNA topoisomerase CIP complex, leading to chromosomal DNA fragmentation (Silva et al., 2011). Azithromycin is a second generation macrolide antibiotic that inhibits bacterial protein synthesis by binding with the 50S ribosomal subunit and inhibiting translation of mRNA (Parnham at al., 2014). Azi thromycin, like other macrolides, is bacteriostatic in nature (Dorfman et al., 2008). Shifts in soil microbial community composition in response to antibiotics application (including CIP) suggest that although overall microbial health c an recover over tim e, microbial community structure can change quickly after antibiotic exposure and remain changed long term (Demoling et al., 2009; Girardi et al., 2011 ; Cui et al., 2014; Ding et al., 2014 ). A cclimation of soil microbial activity over time after exposure t o antibiotics ( such as AZ) results in shorter recovery times for microbial populations after subsequent antibiotic application s (Girardi et al., 2011; Fang et al., 2014 ; Topp et al., 2016). Land application of biosolids borne antibiotics may increase the abundance of antibiotic resistance genes and mobile genetic elements and aid in the spread of antibiotic resistance for at least a few months after biosolids application ( Munir and Xagoraraki, 2011 ; Chen et al., 2016; Rahube et al., 2016 ; Singer et al., 20 16 ). Enrichment of antibiotic resistance genes and resistance bacteria in soils after biosolids application however, depends upon the background soil antibiotic resistance gene

PAGE 100

100 reservoir, soil microbial diversity, and soil microbial numbers ( Brooks et al., 2007; Munir and Xagoraraki, 2011). Some antibiotics, including CIP, reportedly retain bioactivity even when sorbed onto soil matrices ( Chander et al., 2005; Subbiah et al., 2011; Chen et al., 2015). Notably, the evidence was accumulated usi ng exceptionally high initial antibiotic concentrations, involved artificial growth media, and fail ed to represent environmentally realistic exposure scenarios. Microbial exposure to TOrCs from the slow degradation of biosolids aggregates may cause contin uous stress on microbes (Gottschall et al., 2012). S lowed and delayed exposure of antibiotics to microbes degrading the contaminated biosolids theoretically s elect s for resistance and aid s in the spread of antibiotic resistance in the environment. Extended studies (at least a few months) are needed to fully assess microbial response to biosolids borne antibiotics. F ew studies address microbial response s to CIP and AZ and only one utilized biosolids borne chemicals. D ata directly pertaining to the influence of AZ on soil microorganisms are absent, but studies show that CIP can adverse ly a ffect soil respiration, N cycling, and soil microbial community structure and function (Girardi et al., 2011; Cui et al., 2014). Girardi et al. (2011) also reported emergenc e of antibiotic resistance genes over 4 months in soils spiked with environmentally relevant CIP concentrations. Environmentally relevant concentrations can vary widely for different sources (or sinks) of a chemical. Herein, environmentally relevant conce ntrations in growth media (i.e., biosolids or biosolids amended soils) are discussed in the context of biosolids

PAGE 101

101 borne chemicals. The typical concentrations of the target TOrCs in USA biosolids typically range between median and average concentrations (i.e ., 5 to 11 mg CIP/kg and 0.25 to 0.83 mg AZ / kg ) reported in the target national sewage sludge survey ( USEPA, 2009). The uppermost end (95 th percentile ) concentrations of environmental relevance per kg biosolids are ~36 mg (CIP) and ~ 3.2 mg (AZ) (USEPA, 2009). Based on typical (median to average) biosolids concentrations and the typical 1% (dw/dw) land application rate, most biosolids amended soils nominally contain about 0.05 to 0.11 mg CIP /kg and 0.003 to 0.008 mg AZ / kg. Using the 95 th percentile concentrations in biosolids the uppermost end of environmental relevan ce is 0.36 mg CIP and 0.032 mg AZ per kg amended soil. The latter concentrations also represent soil concentrations (without attenuation) from ~7 (CIP) and ~10 (AZ) years of repeated application of biosolids (at 1% (dw/dw) application rate) contaminated with median chemical concentrations. Youngquist et al. (2014) reported that the CIP residues in composted biosolids had no observable adverse effects on soil microbes. The biosolids were spiked with CIP at greater than environmentally relevant concentrations before composting and the composting process was postulated to minimize CIP activity. Unfortunately, missing data on extraction procedures and some un replicated samples in the experimental design limit the rigorous interpretation of the study. Further, the culture based techniques used in Youngquist et al. (2014) can fail to identify the full extent of CIP effects on microorganisms. Ciprofloxacin can a lso change microbial gene expression, microbial community structure, and microbial respiration (e.g., Girardi et al., 2011 ; Cui et al., 2014), expressions not adequately assessed with culture based techniques. The

PAGE 102

102 reported effects of CIP and AZ on soil mic roorganisms warrant studies on microbial response to biosolids borne CIP and AZ under environmentally realistic conditions using molecular biology approaches. Sorption/desorption results suggest that the bioaccessibilit ies of biosolids borne CIP and AZ ar e negligible due to moderately high partitioning coefficient values > 350 L/kg and extremely small (< 0.003) hysteresis coefficients. Thus we hypothesized minimal microbial response to environmentally relevant concentrations of biosolids borne CIP or AZ. To test the hypothesis, an incubation study (for up to 120 days) was conducted to determine microbial response s to varying concentrations of biosolids borne CIP and AZ. The study assessed the following important parameters pertaining to impacts of biosolid s borne CIP and AZ on microbes: 1. Bioavailability to the microorganisms as indicated by effects on various microbial responses (respiration, expression of genes involved in N and P cycles) and antibiotic resistance development over time. 2. C hanges in CIP and AZ extractability (potential bioaccessibility) and degradation (including mineralization) over time. Extractability, degradation, and bioaccessibilit ies of CIP and AZ w ere correlated with bioavailabilit ies defined by sequential extraction of biosolids and an amended soil. Materials 3 H C IP (CAS No. 85721 33 1; 97.4% radiochemical purity) and 3 H AZ (CAS No. 117772 70 0; 98.4% radiochemical purity) were custom synthesized by Moravek Biochemicals (Brea, CA). Pharmaceu tical secondary standards (>99% pure) of CIP and AZ, analytical grade calcium chloride (CaCl 2 ), and double deionized water were purchased from Sigma Aldrich (St. Louis, MO). The biosolids (3 Class A, Table 1 2)

PAGE 103

103 used in the study w as anaerobically digested air dried Class A biosolids from MWRDGC, and contained low CIP (1 mg/kg) and AZ (0.06 mg/kg) concentrations (analyzed by AXYS, BC, Canada) Cattle manure (obtained from University of Florida Dairy Research Unit, located in Hague, Florida) contai ned undetec table (<0.035 mg/kg) CIP and 0.01 mg AZ /kg (analyzed by AXYS, BC, Canada ). A medium textured sand (Birmingham, AL). The sand amended with the manure (at 4% rate, dw/dw) (hereafter re served as one of the incubation med ia Select properties of the soils and the biosolids involved in the study are listed in Table 4 1. Table 4 1. Properties of biosolids and manured sand used in the incubation study (measured average values from duplicate samples). Media pH OM ( g /kg ) CEC (cmolc /kg) TKN (mg/kg) KCl extract. NO x (mg/kg) KCl extract. NH 4 N (mg/kg) Water extract. P (mg/kg) Melich 3 extract. K (mg/kg) Biosolids 6.5 43 0 180 29000 37 3700 7100 4200 Manured sand 6.8 2 4 9.1 1000 0.54 3.4 53 260 Biosolids amended manured sand 6.6 3 0 11 1300 1.2 35 96 300 [Solid matrix characterization was conducted by the Analytical Research Laboratories (ARL) at the University of Florida, Gainesville, FL. Chemical extraction and analysis was performed using the following methods: EPA 350.1 (NH4 N), EPA 353.2 (NOx N); EPA 200.7 (P and K), EPA 150.1 (pH), EPA 351.2 (TKN), Loss on Ignition (OM), and BaCl2 compulsive exchange method, Gillman and Sumpter, 1986 (CEC) The data represent average values from duplicate samples ]. Methods Varying concentrations of CIP and AZ were added to 2 solid matrices (Table 4 2): manured sand amended with 1% (w/w) biosolids ; and 100% biosolids. T he biosolids were pre equilibrated with the antibiotics for a week in the dark at ambient conditions (~25 0 C) prior to manured sand amendment. The equilibration was performed using 1:1.5 solid:solution ratio on an end to end shaker at 150 rpm for 7 d and the

PAGE 104

104 equilibrated biosolids were air dried before amendmen t to the manured sand. Three chemical concentrations were used: biosolids spiked with either 0.15 mg CIP or 0.40 mg AZ ( 3 H compound only) ; 10.5 mg CIP or 0.83 mg AZ, or 36.1 mg CIP or 3.2 mg AZ per kg. The 10.5 mg CIP and 0.83 mg AZ per kg biosolids values represent average concentrations detected in US biosolids, and the 36.1 mg CIP and 3.2 mg AZ per kg biosolids values represent 95 th percentile concentrations detected in US biosolids (USEPA, 2009). Tritium labeled CIP and AZ were used for chemical quantif ication and analysis. The target concentrations of the other treatments were reached by adding non labeled compounds along with the 3 H compounds. After equilibration, 0.25 g of the spiked biosolids were added to 24.75 g of each manured sand (1% w/w), a rat e typical to that used in agriculture. The nominal (inherent + spike) concentrations of CIP and AZ per kg manured sand were: 0.012 mg, 0.12 mg, and 0.37 mg for CIP and 0.005 mg, 0.009 mg, and 0.033 mg for AZ. For the 100% biosolids treatments, 25 g of bio solids were spiked with the 3 concentrations of CIP or AZ ( 3 H compound only, the average, or the 95 th percentile concentrations) 4 5 h prior to incubation. Spiking was performed by spraying the target TOrC (in double deionized water) on a thin layer of the biosolids. The nominal (inherent + spike) concentrations of CIP and AZ per kg biosolids were: 1.01 mg, 11.5 mg, and 37.1 mg for CIP and 0.075 mg, 0.89 mg, and 3.26 mg for AZ. Initiating the incubation after only 4 5 h of spiking (less than the time requir ed to reach equilibrium) was intentional because we wanted to assess the effects of CIP or AZ on microbes present in the biosolids prior to microbial acclimation and/or compound attenuation. We attempted to contrive conditions similar to those affecting mi crobes when the biosolids

PAGE 105

105 are being formed (e.g., direct exposure of microbes to TOrCs prior to both becoming a part of biosolids). Each sample was moistened to field capacity (~10% manured sand (w/w), and ~35% biosolids (w/w)) in a 300 mL glass Mason jar Field capacity was estimated from pot water water holding capacity) determined in a greenhouse study. A constant air flow apparatus (Figure 4 1 ) equipped with an oil less pump aerated the samples at ~0.5 L/min of CO 2 free humidified air. Multiple pressure regulators ensured uniform gas pressure (flow) throughout the apparatus. The experimental design is summarized in Table 4 2. In coming air was stripped of CO 2 and humidified by pumping ambient air first through 2 M KOH, then through a column of soda lime chips (Ca (OH) 2 >80%, KOH< 3%, NaOH< 2%, Ethyl violet<1%), again through 2 M KOH, and then finally through more CO 2 free water (Snyder et al., 2010). The experiment was conducted at ~25 0 C. Samples were weighed periodically, and sterile CO 2 free water was added, as needed, to maintain a moisture content near field capacity. Each Mason jar containing biosolids or amended manured sand was connected, via Dow Corning Silastic laboratory tubing, to another 300 mL Ma son jar containing 50 mL 2 M KOH (to trap CO 2 evolved). Each Mason jar with 2 M KOH was connected to two 30 mL scintillation vials (in a series), via S i lastic tubing, each containing 20 g of 4 molecular sieves (Figure 4 1). The molecular sieve absorbs m oisture up to 23% of its 0 C) (MSDS, Sigma Aldrich, St. Louis, MO). Thus, the molecular sieve trapped 3 H 2 O released as a result of compound mineralization. The vials used to trap 3 H 2 O were rep laced every 4 5 d and

PAGE 106

106 composite molecular sieve samples were combusted to quantify CIP and AZ mineralization each month. We assessed the following parameters over the course of the study: Microbial respiration: D ays 6, 21, 42, 63, and 90. RT qPCR (mRNA) a nalysis of genes involved in N and P cycles and emergence of resistance to CIP and AZ: D ays 0, 5, 45, and 90. Analysis of 3 H 2 O evolved during the incubation study to assess CIP and AZ mineralization: D ays 0, 30, 60, 90 and 120 Analysis of extractable CIP and AZ: D ays 0, 45, and 90. Where time dependent changes in gene expression were expected beyond 90 days, 120 day samples were analyzed. The assessed parameters represent a range of microbial health endpoint indicators often analyzed in microbial health s tudies Figure 4 1. Incubation study set up (Constant air flow apparatus ; c ourtesy of author).

PAGE 107

107 Table 4 2. Experimental design System Number of samples Total Manured sand biosolids systems (TOrCs are biosolids borne) 2 x 3 x 1 x 5 (chemicals x concentrations x solid matrices x replicates) 30 Biosolids alone 2 x 3 x 1 x 5 (chemicals x concentrations x solid matrices x replicates) 30 Controls: 1. empty jar to confirm efficacy of the CO 2 scrubber system. 2. 5 rep licates of the 1 un amended manured sand spiked with 3 H compounds only ( 3 H Control) 3. 5 rep licates of the unspiked manured sand 4. 5 rep licates of unspiked biosolids 1 2 x 1 x 1 x 5 (chemicals x concentrations x solid matrices x replicates) 1 x 1 x 5 (concentratio ns x solid matrices x replicates) 1 x 1 x 5 (concentrations x solid matrices x replicates) 1 10 5 5 Total initial samples 81 Respiration Evolved CO 2 from each treatment jar was considered a measure of microbial respiration, and was collected in a base trap containing 50 mL of 2 M KOH. The base traps were removed and replaced at 6, 21, 42, 63, and 90 d into the study, and analyzed for carbonates (representing trapped CO 2 ) (Anderson, 1982). Briefly, a solution of 0.1 M BaCl 2 was u sed to first precipitate the carbonates in a known volume of base trap and the base trap was subsequently centrifuged at 3220 x g for 10 minutes. One mL of the supernatant was then transferred to a glass 20 mL scintillation vial, treated with phenolphthale in indicator, and titrated to neutrality with 0.2 M HCl. The unused KOH remaining in the base trap was used to calculate the moles of KOH neutralized by H 2 CO 3 (representing evolved CO 2 ). RNA E xtraction and Pr e qPCR T reatment At each sampling time, RNA was extracted from a 2 g sub sample using RNAeasy kits (Qiagen, Hilden, Germany) according to manufacturer instructions. After

PAGE 108

108 extraction, genomic DNA was removed from the RNA using DNAase Max Kit (Qiagen, Hilden, Germany) accord ing to manufacturer instructions. The RNA concentration and purity were determined using Biophotometer Plus (Eppendorf, Hamburg, Germany) and 1.5% agarose gel with ethidium bromide No significant DNA contamination was confirmed using PCR followed by gel e lectrophoresis using 1.5% agarose gel with ethidium bromide. The study employed use of reverse transcriptase quantitative PCR (RT qPCR). The RNA extracted from samples at various times was converted into cDNA The cDNA then served as the template for qPCR reaction. Approximately 500 ng RNA was reverse transcribed into complementary DNA (cDNA) in 20 L volume using RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, MA) according to manufacturer instructions. All samples were stored at 80 0 C prior to extraction or analysis. Target G ene Select ion for Real T ime PCR (qPCR) mRNA transcribed from the f ollowing genes were converted to cDNA and then targeted for quantification by qPCR using StepOne Plus (Applied Biosystems, Foster City, CA ) : Nitrogen cycle genes ( a moA and nirK/nirS ) Changes in expression of enzymes involved in the nitrogen cycle in the presence of biosolids borne TOrCs such as CIP and AZ can suggest impacts of biosolids borne constituents (including TOrCs) on soil microbial populations. Ammonia monooxygenase is a key enzyme in nitrification and nitr i te reductase is a key enzyme in denitrification. Ammonia monooxygenase ( amoA ) gene ( archaeal amoA for ammonia oxidizing archaea (AOA) an d bacterial amoA for ammonia oxidizing bacteria (AOB) ) and nirS and nirK genes (for dissimilatory nitrite reductase) are widely distributed in the

PAGE 109

109 environment (Tourna et al., 2008; Theerachat et al., 2011), and were selected for analysis. Comparing amoA and nirK/nirS exp ressions from unspiked controls with CIP or AZ spiked samples reflect ed the effects of increased concentrations of biosolids borne chemicals on ammonia oxidizing and denitrifying bacteria. The RT qPCR analysis was performed using primers and conditions det ailed in Rotthauwe et al. (1997) for bacterial amoA in Tourna et al. (2008) for archaeal amoA in Braker et al. (1998) for nirK and in Chenier et al. (2003) for nirS Phosphatase genes ( phoN and phoD ) All solid matrices involved in the study had pH value s slightly less than 7, where both the acid phosphatase gene phoN and alkaline phosphatase gene phoD function (Nannipieri et al., 2011). A nalyses of phoN and phoD were performed using primers and conditions detailed in Bergkemper et al. (2016) and Sakurai et al. (2008), respectively. Ciprofloxacin ( qnrA qnrB qnrS ) and AZ ( ermB mefE ) resistance genes The presence and expression of CIP and AZ resistance genes in the soil and biosolids were measured 1 ) to assess selection pressure on soil/biosolids microbes from CIP/AZ, and 2 ) to assess the adaptation of soil microbiota to CIP/AZ. Genes erm B and mef E are associated with AZ resistance in microorganisms (Marchandin et al., 2001; Ambrose et al., 2005). Gene erm B encodes for target site modification and can be transposon and/or plasmid mediated (Nguyen et al., 2009). Gene mef E encodes for efflux pumps and is mostly transposon mediated (Ambrose et al., 2005). erm B media ted macrolide resistance is the most common mechanism in many areas of the world, whereas mef E dominates in the United States (Del Grosso et al., 2007). Genes qnrA qnrB and qnrS are typically plasmid encoded quinolone resistance mechanisms The qnr genes produce Qnr proteins that bind with and protect

PAGE 110

110 DNA gyrase and topoisomerase IV from inhibition by CIP (Jacoby, 2005). The RT qPCR analys es w ere performed using primers and conditions detailed in Cattoir et al. (2007) for qnrA qnr B and qnr S and in Sutc liffe et al. (1996) for ermB and mefE. Real Time qPCR Gene quantification was performed using Maxima SYBR Green/ROX qPCR Master Mix (2X) (Thermo Fisher Scientific, Waltham, MA). Briefly, 10 L SYBR Green, 4 L of 10 M primers (forward and reverse each), 2 L cDNA (~10 100 ng), and Molecular Grade water were mixed in each 20 L reaction in a 96 well qPCR plate. The cDNA in each well was amplified using qPCR methods customized for each target gene. Standard C urve D evelopment and qPCR E fficiency Based on the target gene, appropriate dilution factors were determi ned from a preliminary study. Briefly, triplicate samples assessed cDNA amplification efficiency by four serial dilutions of cDNA: no dilution, 1:1, 1:10, and 1:100 in molecular grade water. The qPCR amplification efficiencies of the cDNA were determined from the slopes of plots of log cycle threshold (Ct) values versus relative cDNA, and efficiencies between 90 105% were considered acceptable. External standard curves from cloned cDNA (with known copy numbers calculated from concentrations determined usi ng Biophotometer Plus (Eppendorf, Hamburg, Germany ) ) were made by serial dilution, 6 points in duplicate. The PCR efficiencies of amplification, determined from slope s of graph of log Ct values versus log copy number of the cDNA, were between 90 105 % and considered acceptable. The lowest standard curve calibration point, and hence the quantification limits, corresponded to ~5000 copies/g media for genes involved in N and P cycles and ~2000

PAGE 111

111 copies/g for antibiotic resistance genes. Upon completion of each r un, a melting curve analysis was performed to check specificity of the primers and confirm adequate amplification and quantification from larger amplicons (up to 500 bp in size). Assessment of CIP and AZ Extractability and Stability Chemical extraction me thods A f ractionation scheme allowed sequential extraction of CIP and AZ from solids samples at d ay 0, 45, and 90 Briefly, a t each sampling time, a 2 g solid sub sample from each jar was collected and placed in a 15 mL polypropylene centrifuge tube. Sampl es were sequentially fractionated using i) 0.01 M CaCl 2 ii) sonication with 1:1 (v/v) methanol:water, iii) accelerated solvent extraction (ASE), and iv) combustion; to determine concentrations of CIP and AZ associated with operationally defined fractions (ECETOC, 2013). The 0.01 M CaCl 2 th borne CIP and AZ. The specifics of the sequential fractionation are detailed below : Extraction with 0.01 M CaCl 2 : Solid CaCl 2 (10 mL) suspensions were equilibrated for 36 h, in the dark, on an end over end shaker at 150 rpm. The tubes were centrifuged at 10,000 x g for 30 min after equilibration. An aliquot (1 mL) of the supernatant was collected with a pipette and mixed with 10 mL of scintillation fluid (Ecoscint A; National Diagnostics, Atlanta, GA) in a 20 mL HDPE scintillation vial. The CaCl 2 remaining in the tubes was collected for thin layer chromatography (TLC) analysis ( see TLC analysis section ).

PAGE 112

112 Extraction with methanol: water: After the CaCl 2 extraction, the remaining solids were suspended by sonication for 1 h with 10 mL of methanol: water (1:1, v/v). The suspensions were centrifuged at 10,000 x g for 30 min, and 1 mL of the supernatant was mixed with 10 mL of scintillat ion fluid in a 20 mL HDPE scintillation vial. The m ethanol:water remaining in the tubes was collected for TLC analysis. Extraction with ASE: Following methanol:water extraction, the samples were air dried and CIP and AZ w ere extracted on a Dionex ASE 200 ( Sunnyvale, CA) using methods described in Golet et al. (2002) for CIP and Ding et al. (2011) for AZ. Briefly, dried samples (500 mg of the biosolids or amended soils) were transferred into 11 mL stainless steel extraction cells from Dionex, and mixed with 10 g of hydromatrix sand. Extraction solvents consisted of 50 mM phosphoric acid (pH 2.0) and acetonitrile mixture (1:1, v/v) for CIP and acetonitrile/water mixture (7:3, v/v) for AZ. The other operation parameters were: extraction temperature = 100 C; e xtraction pressure = 100 bar; preheating period = 5 min; static extraction period = 15 min; extraction cycles = 4 (CIP) or 3 (AZ); solvent flush = 150% (CIP) or 100% (AZ) of the cell volume; nitrogen purge = 300 s (CIP) or 120 s (AZ). A one mL aliquot of t he extract was mixed with 10 mL of scintillation fluid in a 20 mL HDPE scintillation vial; 10 mL of the remaining extract was collected for TLC analysis. Combustion: Solid samples remaining after ASE were air dried, thoroughly mixed, and sub sampled ( 0. 2 0.3 g ) for combustion Combustion was carried out in a sample oxidizer (OX 500; RJ Harvey Corp., NY), and the 3 H 2 O produced was collected in 10 mL of a scintillation fluid (Scinti Safe 50; Fisher Scientific, Hampton, NH).

PAGE 113

113 The samples were weighed at each e xtraction step to account for carry over radioactivity from previous fractionation steps. The solutions were analyzed by liquid scintillation as described in Chapter 2, for determination of the 3 H activity to assess bound chemical and for mass balance pur poses. TLC analysis At each extraction step, up to 10 mL of the supernatant/extract collected was concentrated to a known volume under nitrogen and spotted (5 L) onto Whatman thin layer chromatography plates (Partisil LK5D, Silica Gel 150 , 20x20 cm). The plates were developed in sealed glass chambers containing the mobile phases. The mobile phase for CIP was 100 mL of dichloromethane methanol 2 propanol 25% NH 3 (3:3:5:2) mixture (Girardi et al., 2011). For AZ, the mobile phase was 100 mL of chloroform ethanol Low chemical concentration ( and t he energy of 3 H emitted beta) prohibited visualization on the developed plates Therefore, along with the extracts, a 5 L of an un labeled p harmaceutical secondary standards stock solution (100 mg/L) of each target TOrC was spotted on each TLC plate to determine retardation factor (RF) for each chemical. In TLC, RF is the ratio of the distance traveled by the center of a spot to the distance t raveled by the solvent front. Ciprofloxacin was visualized under UV rays at a wavelength of 330 nm (Krzek et al., 2005) and AZ was visualized as brown spots after spraying with sulfuric acid ethanol (1:4 v/v) followed by heating the plates at 120 0 C for 5 min. The un labeled standard s w ere used to determine and visualize the RF values of 0.81 0.05 (CIP) and 0.36 0.04 (AZ). After the RF was determined, silica was scrapped off below, at, and above the RF location. Silica scrapping enabled semi

PAGE 114

114 quantitati ve identif ication of the target parent chemicals (silica scraps at the RF) and chemical complexes with other entities and/or degradation products (scrapped regions above and below the RF respectively ) ( Troxler et al., 1978 ; Chhalotiya et al., 2017). Kwiec mobile phase ha ve measured RF values to those reported by others using similar mobile phases confirmed the ta rget TOrCs and increased confidence in the analysis method. T he T LC analysis provided semi quantitative analysis of the nature of the extracted chemical. Liquid scintillation counting of scraped TLC silica was performed by suspending the silica powder in 1 0 mL of Hydrofluor (National Diagnostics, Atlanta, GA). Briefly, T LC silica was added to Hydrofluor and the contents were rigorously shaken until formation of a gel following addition of 3 mL distilled water. The gel was then counted on the liquid scintill ation counter ( LS 6500, Beckman Coulter, Brea, CA ) Determination of compound mineralization The mineralization of CIP and AZ (as determined from 3 H 2 O produced) was assessed as follows: 1. Combustion of molecular sieve traps and 2. L iquid scintillation analysis of KOH traps (KOH absorbs water) using Hydrofluor 1. Combustion of molecular sieve traps : The trapping efficiency and recovery of 3 H 2 O from the molecular sieves was assessed in triplicate as follows: A pproximately 1700 Bq of 3 H 2 O (PerkinElmer, Waltham, MA), diluted in 2 mL of double deionized water, w ere added to a 20 mL glass scintillation vial. The scintillation vial was connected to two other 20 mL scintillation vials in series via S ilastic tubing,

PAGE 115

115 each containing 5 g of 4 molecular sieves. Nitrog en gas was used to evaporate the H 2 O from the first scintillation vial to dryness, such that the gas (containing water vapors) passed through the molecular sieves before exiting the apparatus. The molecular sieves in each scintillation vial were mixed and a 500 mg sub sample from each vial was combusted in a sample oxidizer (OX 500; RJ Harvey Corp., NY). The 3 H 2 O produced was collected in 10 mL of a scintillation fluid (Scinti Safe 50; Fisher Scientific, Hampton, NH). Ten mL of scintillation fluid (Ecoscint A, National Diagnostics, Atlanta, GA) was added to the first scintillation vial to analyze remnant 3 H 2 O. The solutions were analyzed by liquid scintillation for determination of the 3 H activity to assess recoveries and a mass balance was calculated. Liqu id scintillation counter 3 H counting efficiencies were 52 55%. Corrected for percent recoveries (96.5%) from the sample oxidizer, 49.6 1.5% and 36.9 2.7% 3 H was recovered from the first and second molecular sieves, respectively. The scintillation vial used to evaporate H 2 O contained 2.5 0.45 % 3 H activity. A total mass balance of 83 5.5% was obtained for the 3 H trapping efficiency of the molecular sieves. The recoveries were deemed acceptable considering that : a) little to no mineralization was expe cted for the target compounds during the incubation ; and b) a mass balance, conducted for each sample, accounted for evaporat ive losses of the mineralized compound. 2. Liquid scintillation analysis of KOH traps using Hydrofluor: Aliquots ( 1 mL ) from KOH tr aps w ere removed weekly and composite samples were analyzed via liquid scintillation to quantify CIP and AZ mineralization for each month of incubation.

PAGE 116

116 Recoveries from methanol:water and ASE extraction methods Tritium counting efficiencies of 52 57% were obtained on the liquid scintillation counter for methanol:water extracts. The ASE extracts yielded poor tritium counting efficiencies (25%), and were diluted (1:1, v/v) in double deionized water prior to liquid scintillation counting to obtain tritium coun ting efficiencies within the range of ~40 60% recommended by scintillation cocktail manufacturer (National Diagnostics, Atlanta, GA). Ciprofloxacin and AZ extraction recoveries from the methanol:water and ASE extraction methods (Table 4 3) were determined, separately, by spiking 7 replicates (1 g each) of the two solid matrices with 250 Bq of CIP or AZ. The spiked matrices were equilibrated in the dark for a week and air dried prior to extraction either by methanol:water or ASE. More CIP and AZ were extract able from manured sand than from biosolids for both extraction methods (Table 4 3), consistent with the results from sorption/desorption study where biosolids sorbed CIP and AZ more strongly than manured sand media. The relative standard deviations (RSD) w ere less than 12%, establishing the two extraction methods as acceptable. Table 4 3. Average CIP and AZ percent recoveries s tandard d eviations (SDs), and % r elative s tandard d eviation (RSD) from methanol:water and ASE extraction methods. Extraction method CIP AZ Biosolids Manured sand Biosolids Manured sand % recovery SD % RSD % recovery SD % RSD % recovery SD % RSD % recovery SD % RSD Methanol: w ater 31.7 3.2 10.1 39.5 4.7 11.9 21.3 1.9 8.9 29.6 1.0 3.3 ASE 73.1 4.3 5.9 86.4 3.4 3.9 29.8 3.6 12.4 38.6 2.8 7.3

PAGE 117

117 Efficiencies and recoveries from TLC analysis Efficiencies of (and CIP and AZ recoveries from) TLC analyses were determined by spotting ~1700 Bq 3 H labeled CIP and AZ in 10 L double deionized water onto the TLC plates. The experiment involved 7 replicates and the TLC plates were developed and scraped, as described earlier. Hydrofluor (National Diagnostics, Atlanta, GA) generated 42 48% tritium efficiency on the LSC. Chemical recoveries from the TLC analyses wer e 96.7 1.6% (CIP) and 89.9 2.1% (AZ), and the RF values were close to expected RF values of 0.8 (for CIP) and 0.4 (for AZ) (Table 4 4) Collectively, the data indicate a reliable analysis method. Table 4 4. Average CIP and AZ percent recoveries s tand ard deviations (SDs), % relative standard deviation (RSD), and retardation factors (RFs) from TLC analysis Chemical % recovery SD %RSD RF CIP 96.7 1.6 1.7 0.81 AZ 89.9 2.1 2.3 0.36, 0.97* *AZ TLC yielded an additional spot near the mobile phase parent TLC analysis suggested 5.4 0.9% non parent compound entity (likely a degradation product) of 3 H AZ in the stock solution initially added to the biosolids However, the mass of the non parent compound entity added to the manured sand and the biosolids was miniscule (<0.001 mg/kg) and deemed insignificant to the experimental results. Further, the non parent com pound entity was accounted for in the chemical extraction and analysis. Description of T reatments 1. Unspiked control refer s to unspiked biosolids 2. Un amended control refers to un amended and unspiked manured sand

PAGE 118

118 3. Un amended 3 H control refer s to manur ed sand directly spiked only with 3 H labeled CIP or AZ (no biosolids) Because biosolids inherently contain ed both CIP and AZ, un amended 3 H controls were included to serve as the lowest concentrations (few g/kg) of CIP and AZ in the study 4. Treatments of b iosolids or biosolids amended manured sand spiked with 3 H CIP or 3 H AZ only (few g/ kg biosolids) are referred to as 3 H compound 5. Treatments of biosolids or biosolids amended manured sand spiked with average concentrations of CIP or AZ (based on USEPA, 2009 data) are referred to as Average 6. Treatments of biosolids or biosolids amended manured sand spiked with 95 th percentile concentrations of CIP and AZ (based on USEPA, 2009 data) are referred to as 95 th percentile Detection Limits and Statistical Analysis Tritium counting efficiencies of 42 58% were obtained on the liquid scintillation counter in all samples. An operationally defined cut off quantification limit was set at 1.7 Bq Although, the 3 H activities of unspiked manured sand and biosolids w ere greater than the minimum detectable true activity value of 0.46 Bq radioactivity was present in the systems), the activities less than 1.7 Bq were regarded The data were analyzed using R for normality (Shapiro Wilk test) and homogeneity of variance (Levene test). The ANOVA was used for analysis of normal and homogeneous data whereas the Kruskal Wallis test was used for non normal data. All data analysis was Results and Discussion Respiration Neither CIP nor AZ affected CO 2 evolution throughout the incubation at any concentration (Figure 4 2). Microbial respiration was significantly greater in the biosolids alone than in the amended manured sand treatments, reflecting the greater organic

PAGE 119

119 matter and nutrient contents of biosolids than of the manured sand. Amendment with biosolids (1%, w/w) increased the organic matter and nutrient content of the manured sand but did not significantly affect microbial respiration. More CO 2 evolved during d 6 21, 21 42, and 42 63 than d ay 0 6 of incubation, likely because of greater time interval between samplings. As observed in similar respiration studies (e.g., Snyder et al., 2011; Pannu et al., 2012), CO 2 evolution tapered off with time. Figure 4 2. A representative figure, with standard error bars, showing no effects of CIP or AZ treatments on microbial respiration in the biosolids or the amended manured sand. The t reatments consist of unspiked controls, solid matrices spiked with only 3 H CIP or 3 H AZ (few g/ kg biosolids) and matrices spiked with average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009) Literature on effects of AZ on mic robial respiration are absent, and conflicting data are reported for CIP. Girardi et al. (2011) suggests CIP negatively affects soil

PAGE 120

120 respiration at environmentally relevant concentrations, but that microbial respiration recovers over time. On the other han d, Cui et al. (2014) found no effects on soil respiration even at CIP concentrations well above environmental relevance. Ciprofloxacin was not biosolids borne in either study. Data in Figure 4 2 show that environmentally relevant concentrations of biosolid s borne CIP and AZ have negligible impacts on overall microbial respiration. Microbial R esponse to B iosolids borne CIP and AZ Ammonia oxidizing bacteria ( AOB ) gene (bacterial amoA ) For both TOrCs, bacterial amoA expression in the biosolids was ~2 orders of magnitude greater than in the amended manured sand (Figure 4 3) The impacts likely reflected significantly greater NH 4 + concentration (Table 4 1) and 100 fold dilution of biosolids in the amended manured sand On d ay 0, bacterial amoA expression for both compounds in the biosolids varied statistically with the TOrC concentration: 3 H compound > Average > 95 th percentile The likely reason for the effects on bacterial amoA expressions on d ay 0 is non equilibrium of AZ and CIP with the biosolids (sampling was conducted 10 h after spiking the biosolids, whereas equilibrium was determined to require 36 h). About 1% CIP and chemical extraction results ) on d ay 0, corresponding to CIP and AZ concentrations greater than the minimal inhibitory concentrations for 90% variates (MIC90) of some microbes (Wise et al., 1983; Eltahawy, 1993; Gordillo et al., 1993; LeBel, 1993). By d ay 5, bacterial amoA expression s in AZ incubations were similar across all treatments. The bacterial amoA expression s in CIP incubations on d ay 5 increased for the average and the 95 th percentile CIP treatments compared to d ay 0, but remained

PAGE 121

121 less than that in the controls and 3 H compound treatments. The bacterial amoA expression s in all CIP treatments were similar by d ay 45, suggesting one or more of the following scenarios: 1) AOB recovery from adverse effects induced by the TOrC, 2) TOrC attenuation, and/or 3) antibiotic resistance development/acquisit ion in AOB The three scenarios can overlap. C iprofloxacin and AZ reversibly bind to target sites in microbes. The binding affinities vary depending on species of microbes, and environmental conditions such as pH, salt concentrations, moisture content, etc ( Shen and Pernet, 1985; Noble et al., 2003; Petropoulos et al., 2009; Jelic and Antolovic, 2016). Ammonia oxidizing bacteria can recover from adverse effects induced by CIP after d issociation of initially bound CIP from the target sites in static bacteri a. Often the d issociation is a result of resistance mechanisms such as antibiotic efflux and/or decreased ambient concentrations of the target TOrCs (such as by attenuation). Ciprofloxacin attenuation (see chemical extraction results ) appears to be a facto r affecting bacterial amoA expression in the average and 95 th percentile CIP treatments, along with some contribution from increases in antibiotic resistance (see results for resistance gene expression analysis ). The recovery of bacterial amoA expression s in AZ treated biosolids (Figure 4 3B ) by d ay 5 can be explained by attainment of TOrC retention equilibrium with the biosolids. The increased bacterial amoA expression s in the average and 95 th percentile CIP treatments from d ay 0 to d ay 5 (Figure 4 3 A ) is also likely due to attainment of TOrC equilibrium with the biosolids. However, the bacterial amoA expression remained less than that in the controls and 3 H compound treatments, suggesting that the microbes did not fully recover within 5 d Greater con centrations (10.5 and 36.1 mg/kg

PAGE 122

122 biosolids), greater toxicity to microbes (US National Library of Medicine, 2002 ; Girardi et al., 2011 ), and smaller MIC values of CIP than AZ may explain the longer recovery time of bacterial amoA expression in CIP treatmen ts than in AZ treatments. Different microbial target sites and different binding affinities of CIP and AZ for different organisms (Barnard and Maxwell, 2001; Garza Ramos et al., 2001 ; Petropoulos et al., 2009; Mustaev et al., 2014) could also have affected bacterial amoA expression. A nother (possibly supplementary) explanation for the observed bacterial amoA expression results could be the sorption partitioning coefficients (Kd) values of CIP and AZ in the biosolids. Sorption studies revealed Kd values of 360 L/kg (CIP) and 430 L/kg (AZ) in the same biosolids. The Kd values yield solution concentrations of ~ 0.03 mg CIP and ~ 0.007 mg AZ per liter in the average CIP and in 95 th percentile AZ treatments, respectively. Such concentrations of CIP and AZ exceed MIC90 (CIP) and MIC1 (minimal inhibitory concentrations for 1% variates of some microbes) (AZ) values for some organisms (Wise et al., 1983; Eltahawy, 1993 ; Bengtsson Palme and Larsson, 2016 ). Therefore, even if equilibrium was reached by d ay 5, sufficient CIP or AZ was present in the solution phase to adversely affect susceptible microbes in some TOrC tre atments. Biosolids amendment increased bacterial amoA expression in the manured sand media (Figure 4 4 ), likely in response to ~ 1 0 fold increase in NH 4 + concentrations (Table 4 1). All biosolids amended manured sand treatments had greater bacterial amoA e xpression s than the controls (Figure s 4 4 ), suggesting that biosolids amendment significantly increased activity of at least some microbes. Initially ( day 0 ) bacterial amoA expression s in the average (CIP) and 95 th percentile (both CIP and AZ) treatments w ere

PAGE 123

123 less than those in the control and 3 H compound treatments. By day 5 the expression s recovered in AZ treatments Th e e xpression s increased in CIP treatments by day 5 but remained lower in the average and the 95 th percentile treatments than in other CIP treatments. The observed lower bacterial amoA expressions for both compounds are likely explained by the fact that the AOB in biosolids (representing >90% of AOB in amended manured sand media, Figure 4 4 ) were already adversely affected prior to biosolids amendment to manured sand. The bacterial amoA expression s in amended manured sand on day 0 mimicked the trends of bacterial amoA expression s in 100% biosolids treatments on day 5 Apparently, sufficient CIP (in the average and 95 th percentile tre atments) was bioavailable to adversely affect bacterial amoA expression even after a week. Data suggest that biosolids dictate bioaccessibility of biosolids borne CIP and AZ and the activity of at least some microbes. The recovery of bacterial amoA express ion s over time suggests that the high concentrations of CIP/AZ affected bacterial amoA expression s but (likely) not the number of AOB To confirm this hypothesis, we quantified the number of bacterial amoA from DNA extracted from 95 th percentile treatments on d ay 0 and 90. R esults ( Appendix B Figure B 1 ) show negligible change in relative copy numbers of bacterial amoA consistent with the hypothesis that CIP/AZ inhibited bacterial amoA expression s but not the total number of bacteria l amoA The DNA results suggest that effects of biosolids borne CIP and AZ were bacteriostatic (i.e., inhibited bacterial growth), rather than bactericidal, and explain the recovery of bacterial amoA expression s over time. The b acteriostatic nature of CIP under environmentally relevant scenarios is well documented (e.g., Thiele Bruhn and Beck, 2005 ; Girardi et al., 2011)

PAGE 124

124 Ammonia oxidizing archaea ( AOA ) gene (archaeal amo A) Expressions of archaeal amoA (Figures 4 5, 4 6 ) were an order of magnitude greater than bacterial amoA expression s in both media for both compounds, and were largely unaffected by TOrC treatments. As with bacterial amoA determinations, biosolids amendment significantly increased archaeal amo A expressi on in the manured sand media. The archaeal amo A expression s in 95 th percentile CIP treatments w ere slightly inhibited on d ay 0, but recovered by d ay 5 and w ere similar across treatments thereafter. None of the other treatments for either compound adversely affected archaeal amo A expression. Ammonia oxidizing archaea abundance usually exceeds AOB in terrestrial ecosystems ( Taylor et al., 2012 ; Banning et al., 2015; Sterngren et al., 2015) and most archaea (including AOA ) are resistant to common antibiotics (Hilpert et al., 1981; Khelaifia and Drancourt, 2012), including CIP and AZ. Therefore, environmentally relevant concentrations of biosolids borne CIP and AZ are unlikely to affect overall ammonia oxidation rates in biosolids and amended soils. Indeed, NH 4 + concentrations were the same across all treatments for both chemicals (Appendix B Figures B 6 to B 8). Other TOrCs (Triclocarban and Triclosan) had similar negligible adverse effects on ammonia oxidation even at concentrations well above environmental relevance (Snyder et al., 2010; Pannu et al., 2012). Nitrite reductase genes ( nirK and nirS ) Expressions of nirK were not detected throughout the incubation, but low levels (~10 3 ) of nirS expression were detected on d ay 45 and 90 in all CIP and AZ treatmen ts (Figure 4 7 ). Negligible NOx (Table 4 1) and oxic experimental conditions likely favored negligible dissimilatory nitrite reduction initially. R eduction increased as more NOx

PAGE 125

125 became available by ammonia oxidation over time (Appendix B Figures B 6 to B 8). The nirS expression was unaffected by TOrC treatments for either compound, indicating negligible effects of biosolids borne CIP and AZ. Phosphatase genes ( phoN and phoD ) Both phoN and phoD expressions were unaffected by TOrC treatments. The expressions increased on d ay 5, likely in response to moisture addition (samples were dry prior to d ay 0), but remained the same thereafter (Figures 4 8 to 4 11 ). The expressions were low (~10 3 ) in the biosolids, likely due to high P content (Table 4 1) but were an order of magnitude greater in manured sand because of relatively low P content. Unlike genes involved in the N cycle, expressions of phoN / phoD were the same in amended and un amended manured sand because of similar P concentrations (Table 4 1; Appendix B Figures B 6 to B 8). Nitrogen and phosphorus chemical data : Potassium chloride extractable NH 4 + and NOx, and water extractable P concentrations (Appendix B Figures B 6 to B 8) were similar for a particular media at a particular time irrespec tive of TOrC treatments for either compound Data s uggest negligible effects of biosolids borne CIP and AZ on N and P cycling in the environment. Th e expression data of genes involved in N and P cycling are in accordance with the chemical (extractable NH 4 + NOx, and P) data. Ciprofloxacin resistance genes ( qnrA qnrB and qnrS ) Expressions of CIP resistance genes qnrA and qnrB were not detected during the incubation. qnrS was express ed (~10 3 10 4 copies/g ), but only in the average and 95 th percentile CIP treatments (Figure s 4 12 4 13 ) The qnrS expression indicat es influence of biosolids borne CIP on microbes The gene expression was detectable from d ay 5 through the end of incubation in the biosolids treatments (Figure 4 12) whereas

PAGE 126

126 qnrS expression was below detectable levels by d ay 45 in the manured sand (Figure 4 1 3 ). The differences in qnrS expression in different matrices likely reflect greater stress on microbes (because of greater CIP concentrations) in the biosolids than the biosol ids amended manured sand Apparently, borne CIP concentrations can modestly induce expression of qnr resistance genes, but expression diminishes with time possibly in response to diminishing CIP bioa ccessibility (Figures 4 1 9 4 20 ). The question remains whether this low level (few to several thousand copies per gram media) expression is clinically and/or environmentally relevant Jacoby (2005) suggests that the presence of qnr facilitate s selection o f additional resistance mutations by 10 fold by inducing and/or enriching mutant microbes exposed to CIP. Thus, even low level qnr expression could be potentially important, but more research on a longer time scale (encompassing several months to years) an d under field conditions is necessary to adequately test the hypothesis. The results of qPCR on DNA extracted from the 95 th percentile treatments on d 0 and 90 (Appendix B Figure B 2 ) show that the relative copy numbers of qnrS marginally increased over t ime in the biosolids but not in the amended manured sand Results suggest possible ( although marginal ) enrichment of CIP resistance bacteria. Azithromycin resistance genes ( ermB and mefE ) Low levels (~10 4 copies/g) of ermB and mefE expression were expressed in the biosolids throughout the incubation, even in the controls (Figures 4 1 4 4 1 6 ). The mefE expression increased relative to controls only in the 95 th percentile AZ treatments (Figure 4 1 4 ). The low levels of mefE expression in other treatments are likely not the impacts of added AZ because increases in expression were not proportional to increases in AZ concentration and an increase was only observed in the greatest AZ

PAGE 127

127 treatment. The enhanced mefE expression decreased over ti me and was the same as control and other treatments by d ay 90 (Figure 4 1 4 ). The ermB expression increased with increasing AZ concentrations (Figure 4 1 6 ), suggesting increased AZ stress on the microbes. U nlike with mefE the enhanced ermB expression was p roportional to AZ concentrations throughout incubation (even by 90 d ). Microbes expressing ermB may have increased in the biosolids and stabilized thereafter at least through 90 d Expressions of mefE and ermB were quantified only in the average and 95 th p ercentile treatments in manured sand (Figures 4 1 5 4 1 7 ). Significantly fewer microbes and less microbial activity were observed in manured sand compared to biosolids, and mefE and ermB expressions were below detection limits in some manured sand treatments. The expressions were about an order of magnitude less than those in the biosolids. mefE expression s in the amended manured sand were not quantifi able by d ay 90 (Figure 4 1 5 ), whereas ermB expression r emained quantifiable and treatment dependent (Figure 4 1 7 ). The ermB data prompted a nalysis of select sub samples from 120 d incubation s for ermB expression The ermB expressions a t 120 d incubation remained treatment dependent in the biosolids and the ame nded manured sand samples and were similar to d ay 90 (data not shown). Th us, the expression of at least some antibiotic resistant genes can increase and/or stabilize in soil (at least for a few months) after contaminated biosolids are land applied. qPCR analyses of DNA extracted from the 95 th percentile treatments on d ay 0 and 90 (Appendix B Figure B 3) reveal no changes in ermB copy number and (likely) microbes carrying ermB gene. Similarly, t he c opy numbers of mefE (Appendix B Figure

PAGE 128

128 B 4) did not change overtime, perhaps because the mefE expression responded to some unknown factor, rather than AZ treatment 16S rRNA analysis E ffects of biosolids borne CIP and AZ on the overall microbial populations were assessed via RT q PCR on extracted 16S rRNA Extractions were from the control and the 95 th percentile CIP and AZ treatments of both incubation media on d ay s 0 and 90. Primers and conditions detailed in Harms et al. (2003) were used and the results appear in Appendix B (Figure B 5). The n e gligible impacts of biosolids borne CIP and AZ on microbial populations are consistent with the reported bacteriostatic nature of the target TOrCs ( Dorfman et al., 2008 ; Silva et al., 2011). Figure 4 3. Expression of bacterial amoA in biosolids for A) CIP treatments B) AZ treatments The treatments consist of unspiked controls, solid matrices spiked with only 3 H CIP or 3 H AZ, and matrices spiked with average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letters test) among the cDNA copies per gram between treatments for a particular media, chemical, and time point.

PAGE 129

129 Figure 4 4. Expression of bacterial amoA in manured sand for A) CIP treatments B) AZ treatments The treatments consist of un amended manured sand controls, u n amended manured sand spiked only with 3 H CIP or 3 H AZ (few g/kg soil) and manured sand amended with biosolids (1% w/w rate) spiked with only 3 H CIP or 3 H AZ, or average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letters (a, b, c d ) represent gram between treatments for a particular media, chemical, and tim e point. Figure 4 5. Expression of archaeal amoA in biosolids for A) CIP treatments. B) AZ treatments The treatments consist of unspiked controls, solid matrices spiked with only 3 H CIP or 3 H AZ, and matrices spiked with average or 95 th percentile co ncentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letter s (a b test) among the cDNA copies per gram between treatments for a particular media, chemical, and time point.

PAGE 130

130 Figure 4 6. Expression o f archaeal amoA in manured sand for A) CIP treatments. B) AZ treatments The treatments consist of un amended manured sand controls, un amended manured sand spiked only with 3 H CIP or 3 H AZ (few g/kg soil), and manured sand amended with b iosolids (1% w/w rate) spiked with only 3 H CIP or 3 H AZ, or average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letters (a, b) represent significant gram between treatments for a particular media, chemical, and time point. Figure 4 7 Expression of nirS in biosolids for A) CIP treatments. B) AZ treatments The treatments consist of unspiked controls, solid matrices spiked with only 3 H CIP or 3 H AZ, and matrices spiked with average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letter (a) represent s the cDNA copies per gram between treatments for a particular media, chemical, and time point.

PAGE 131

131 Figure 4 8 Expression of phoN in biosolids for A) CIP treatments. B) AZ treatments The treatments consist of unspiked controls, solid matrices spiked with only 3 H CIP or 3 H AZ, and matrices spiked with av erage or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letter the cDNA copies per gram between treatments for a particular media, chemical, and time point. Figure 4 9 Expression of phoN in manured sand for A) CIP treatments. B) AZ treatments The treatments consist of un amended manured sand controls, un amended manured sand spiked only with 3 H CIP or 3 H AZ (few g/kg soil), and manured sand amended with biosolids (1% w/w rate) spiked with only 3 H CIP or 3 H AZ, or average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letter (a) represent s no significant DNA copies per gram between treatments for a particular media, chemical, and time point.

PAGE 132

132 Figure 4 10 Expression of phoD in biosolids for A) CIP treatments. B) AZ treatments The treatments consist of unspiked controls, solid matrices spiked with only 3 H CIP or 3 H AZ, and matrices spiked with average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letter the cDNA copies per gram between trea tments for a particular media, chemical, and time point. Figure 4 11 Expression of phoD in manured sand for A) CIP treatments. B) AZ treatments The treatments consist of un amended manured sand controls, un amended manured sand spiked only with 3 H CIP or 3 H AZ (few g/kg soil), and manured sand amended with biosolids (1% w/w rate) spiked with only 3 H CIP or 3 H AZ, or average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letter (a) represents no significant di between treatments for a particular media, chemical, and time point.

PAGE 133

133 Figure 4 1 2 Expression of qnrS in biosolids The treatments consist of unspiked controls solid matrices spiked wi th only 3 H CIP or 3 H AZ, and matrices spiked with average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letters (a, b) represent significant differences between treatments for a particular media, chemical, and time point. Symbol (*) indicates that mean expression (cDNA copies/g) was below quantification limit. Expression was detected only in some of the five replicates, but the figure shows average of all 5 replicates (expression below quantification levels was considered equal to half of the quantification limit, USEPA, 1991). Figure 4 13. Expression of qnrS in manured sand. The treatments consist of un amended manured sand controls, un amended manured sand spiked only with 3 H CIP or 3 H AZ (few g/kg soil), and manured sand amended with biosolids (1% w/w rate) spiked with only 3 H CIP or 3 H AZ, or average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Symbol (*) indi cates that mean expression (cDNA copies/g) was below quantification limit. Expression was detected only in some of the five replicates, but the figure shows average of all 5 replicates (expression below quantification levels was considered equal to half of the quantification limit, USEPA, 1991).

PAGE 134

134 Figure 4 1 4 Expression of mefE in biosolids The treatments consist of unspiked controls solid matrices spiked with only 3 H CIP or 3 H AZ, and matrices spiked with average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letters (a, b) represent significant differences treatments for a particular media, chemical, and time point. Figure 4 15. Expression of mefE in manured sand. The treatments consist of un amended manured sand controls, un amended manured sand spiked only with 3 H CIP or 3 H AZ (few g/kg soil), and manured sand amended with biosolids (1% w/w rate) spiked with only 3 H CIP or 3 H AZ, or average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letter (a) represent no test) among the cDNA copies per gram between treatments for a particula r media, chemical, and time point. Symbol (*) indicates that mean expression (cDNA copies/g) was below quantification limit. Expression was detected only in some of the five replicates, but the figure shows average of all 5 replicates (expression below qua ntification levels was considered equal to half of the quantification limit, USEPA, 1991).

PAGE 135

135 Figure 4 1 6 Expression of ermB in biosolids. The treatments consist of unspiked controls solid matrices spiked with only 3 H CIP or 3 H AZ, and matrices spiked with average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letters (a, b c ) represent significant differences treatments for a particular media, che mical, and time point. Figure 4 17. Expression of ermB in manured sand. The treatments consist of un amended manured sand controls, un amended manured sand spiked only with 3 H CIP or 3 H AZ (few g/kg soil), and manured sand amended with biosolids (1% w /w rate) spiked with only 3 H CIP or 3 H AZ, or average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Letter s (a b test) among the cDNA copies per gram between treatments for a particular media, chemical, and time point. Symbol (*) indicates that mean expression (cDNA copies/g) was below quantification limit. Expression was detected only in some of the five replicates, but the figure shows average of all 5 replicates (expression below quantification levels was considered equal to half of the quantification limit, USEPA, 1991).

PAGE 136

136 Antibiotic Resistance Development and Spread: Possibilit ies and Unknowns Antibiotic minimum effect concentrations (MEC) are about 10 fold less than MIC values (Bengtsson Palme and Larsson, 2016) and represent environmentally relevant concentrations of many biosolids borne antibiotics, including CIP and AZ S ub inhibitory concen trations especially MEC, of antibiotics may facilitate development and spread of antibiotic resistance in the environment ( Clarke and Smith, 2011 ; Larrson, 2014; Mao et al., 2015; Bengtsson Palme and Larsson, 2016 ; Chen et al., 2016; Rahube et al., 2016; Singer et al., 2016; Vikesland et al., 2017 ). Biosolids reportedly contain many antibiotic resistance genes and bacteria, and can serve as a source of antibiotic resistance in the environment ( Larrson, 2014; Mao et al., 2015; Chen et al., 2016; Singer et a l., 2016; Vikesland et al., 2017 ). I ncreased antibiotic resistan ce activity in the biosolids amended soils often depends upon: a) microbial diversity and abundance and resistance gene pool of the soils receiving contaminated biosolids ; and b) soil characteristics /processes that limit antibiotic bioavailability (Brooks et al., 2007; Eriksen et al., 2009; Munir and Xagoraraki, 2011). For example, large soil CEC values might result in extensive sorption of antibiotics initially present in the solution phase of land applied biosolids Thus, biosolids borne antibiotic resistance determinants (i.e., resistant bacteria resistance genes and mobile genetic elements ) may or may not increase the resistance in biosolids amended soils ( Brooks et al., 2007; Erik sen et al., 2009; Munir and Xagoraraki, 2011 ; Marti et al., 2013; Rahube et al., 2016; Singer et al. 2016; Vikesland et al., 2017 ). Also, increase s in soil antibiotic resistance determinants usually dissipate to background levels over time ( several month s) after biosolids application (Rahube et al., 2016; Singer et al., 2016). Marti et al. (2013) and Rahube et al. (2016) found th at

PAGE 137

137 biosolids amended soils and plants grown in biosolids amended soils shortly after biosolids application had increased abundan ce of some resistance determinants compared to unamended soils Exposure to the antibiotic resistance determinants from soils shortly after biosolids amendment therefore, can be greater than from unamended soils and from soils s everal months after biosoli ds application (Rahube et al., 2016) Unfortunately, risks from the increased exposure to biosolids borne antibiotic resistance determinants are unknown. Horizontal gene transfer (HGT) of biosolids borne resistance genes to microbes in amended soils can accelerate spread of antibiotic resistance (Chen et al., 2016), but additional (field) studies are needed to assess HGT rates in biosolids and biosolids amended soils. Pathogen regulation in land applied biosolids suggests that the adverse impacts of should be clinically minimal Resistance spread is, however, often not localized, and ecological impacts from antibiotic spread (localized or not) are unknown. E vidence for biosolid s as a source of antibiotic resistance determinants is mostly based on studies utilizing Class B biosolids. Class A biosolids contain significantly less pathogens, microbial populations, and resistance gene determinants than Class B biosolids. Thus, antibi otic resistance development and spread in Class A biosolids is likely limited compared to Class B biosolids, but confirmations are necessary. S ite restrictions ranging from 30 days (animal grazing) to greater than 14 months (food crops) after land applicat ion of Class B biosolids, may reduce potential impacts (especially on humans) of biosolids borne antibiotic resistance. However, a ssessments

PAGE 138

138 of biosolids borne antibiotic resistance exposures especially shortly after biosolids applications (e.g., for graz ing animals) are crucial for risk determination. Based on measured Kd values (Sidhu, Chapter 2), equilibrium solution phase concentrations of CIP and AZ in the biosolids were 0.02 mg CIP/L and 0.002 mg AZ/L in the average concentration treatments, and ~0. 1 mg CIP/L and 0.007 mg AZ/L in the 95 th percentile treatments. These concentrations, especially in the average treatments, approach MEC values for some microbes (Table 1 1) Antibiotic resistance gene expression and possible enrichment of resistant bacter ia observed in this study were however, minor (~10 4 copies/g or less). Thus, our data are insufficient to definitively s upport, but qualitatively back biosolids facilitated antibiotic resistance development and spread concerns suggested by the literature S everal data gaps need to be addressed to adequately assess consequences of antibiotic resistance determinants in biosolids. M ore and l ong term field scale investigations are critical, as is a quantitative framework to address risks from biosolids born e antibiotic resistance TLC Analysis T 3 H recovery was attributed to the parent compounds O 3 H entities (degradation products and/or chemical complexes) were negligible. The majority of the non parent CI P entities were present in parent CIP entities likely represent CIP complexes because of negligible expected degradation ( Al Ahm a d et al., 1999; Thiele Bruhn, 2003; Chenxi et al., 2008; Girardi et al., 2011; Subbiah et al., 2011; Gottschall et al., 2012) and reported CIP complexation with cations (like Ca 2+ ), common in biosolids and soils (Aristilde and Sposito, 2008; Uivarosi, 2013). Similarly, excluding the non parent AZ entity present in the stock solution (RF = 0 .97), non parent AZ entities were

PAGE 139

139 4.1 1% on d ay 45 and 5.5 0.8% on d ay 90 in the two incubation media Almost all of the non parent AZ entities we re degradation products S is expected (Lyman et al., 1990; Gottschall et al., 2012; Topp et al., 2016) and no significant AZ complexes are reported in literature. No 3 H activity was observed in KOH traps or molecular sieves throughout the incubation (and up to 120 d ), suggesting n egligible mineralization of either compound. The results are consistent with literature reports of persistence of CIP and AZ in the biosolids and amended soils ( Ahm a d et al., 1999; Thiele Bruhn, 2003; Girardi et al., 2011; Subbiah et al., 2011; Gottschall et al., 2012; Topp et al., 2016). T he stock 3 H AZ degraded at the rate of ~2% per month, so the detected non parent AZ entities during the incubation likely reflect radiolabel in stability and are unrelated to microbial action and/or incubation conditions In either case, AZ disappearance was first order and was described by following equation, (r 2 > 0.95): where [A] t = chemical concentration at time t, [A] 0 = initial chemical concentration, k 1 = constant for first order reaction, and t = time. A plot of ln [AZ] t /[AZ] 0 versus time ( d ) for the biosolids (Figure 4 1 8 A) yields a slope of 0.00071 or a k 1 value of 0.00071 (k 1 = slope). Similarly, plot of ln [AZ] t /[AZ] 0 versus time ( d ) for the biosolids amended manured sand (Figure 4 1 8 B) yields a k 1 value of 0.00066.

PAGE 140

140 Figure 4 1 8 First order AZ degradation (actually, parent compound disappearance) kinetics in A) biosolids. B) biosolids amended manured sand The solid black lines represent the fit of the data. The dashed blue lines indicate upper and lower 95% confidence limits. T he resulting first order half lives of AZ in the biosolids and in the biosolids amended manured sand were the same 976 100 d and 1050 1 38 d respectively. However, as the incubation study did not reach even the first AZ half life, predicting AZ half life in the biosolids and the amended soil is problematic. The estimated half lives should be regarded as only first approximations, but are similar to those suggested by others (e.g., Walters et al., 2010 ). Extraction of CIP and AZ from The Solid Matrices Calcium chloride extraction and the amended manured sand. On d ay 0, ~1% CIP and ~2% AZ were extracted from the 100% biosolids and u n amended 3 H control respectively (Figures 4 1 9 4 20 ). Non equilibrium conditions in biosolids (only 10 h equilibration) and ~10 fold smaller Kd values in the manured sand than in the biosolids, likely explain detectable CaCl 2 extractable TOrCs in these systems on d ay 0.

PAGE 141

141 and AZ in the solid matrices. However, we used low amounts ( 600 Bq /g in biosolids and 250 Bq /g in manured sand) of 3 H labeled compound, restricting our limit of detection. Sufficient CIP (but, likely, not AZ) could still be present in the CaCl 2 extractabl e phase to affect microorganisms. For instance, a CaCl 2 extractable CIP concentration representing 0.2% (below detection limit) of added CIP in the 95 th percentile biosolids treatment represents a solution phase concentration of ~0.07 mg/L; which is within the MIC90 range for some microbes (Wise et al., 1983; Eltahawy, 1993; Gordillo et al., 1993; LeBel, 1993; Bengtsson Palme and Larsson, 2016). Methanol: w ater extraction Methanol:water extracted ~ 30 35% CIP and ~ 13 20% AZ from both biosolids and amended manured sand on d ay 0 (Figures 4 1 9 4 20 ). Methanol:water extractable CIP decreased to ~ 17% by d ay 45 and to ~10% by d ay 90 (Figures 4 1 9 4 20 ), suggesting stronger binding of CIP with solid matrices over time. Methanol:water generally extracted similar amounts (~18%) of AZ from the solid matrices (Figures 4 1 9 4 20 ) at d ay 0 and 45, and extractability decreased only marginally to ~ 15% by d ay 90 suggesting: limited chemical interactions between AZ and solid matrices after initial sorption equilibrium i s reached (36 h in this system). ASE extraction ASE extracted ~ 44 48% CIP and ~ 27 30% AZ on d ay 0. Like methanol:water extraction, ASE extractable CIP decreased with time to ~30% by d ay 45 and to ~25% by d ay 90, whereas ASE extractable AZ changed more sl owly to ~25% by d ay 90 (Figures 4 1 9 4 20 ).

PAGE 142

142 Similar time dependent decreases in extractable chemicals (including CIP and AZ) have been reported by others (e.g., Reid et al., 2000; Ericson, 2007 ; Girardi et al., 2011 ; Ma et al., 2015 ). Girardi et al. (2011), and several others (e.g., Sabourin et al., 2012; Aristilde and Sposito, 2013 ; Cui et al., 2014), attributed the decreases in ASE extractable CIP to bound residue formation. Methanol:water and ASE extractable CIP and AZ values in b iosolids amended manured sand and biosolids were similar, but less than values in un amended 3 H control The data are consistent with results from the sorption/desorption study that for biosolids borne CIP and AZ, the biosolids and not the soil dictates contaminant retention/release behavior. Combustion The mass balance analysis from the sample oxidizer on d ay 0, 45, and 90 after correction for percent recoveries (89 98%), accounted for 86 110% of the initial 3 H added to the samples. Generally, the rec overies were similar across treatments for a particular chemical, media, and fractionation method (Figures 4 1 9 4 20 ). Given compound recoveries from various fractionation methods and total mass balance, a cumulative error of about 10% is associated wit h the extraction data. Decreases in CIP extractability were statistically significant and trends in CIP extractability changes over time are valid even after considering the error associated with extractability measurements (Figures 4 1 9 A, 4 20 A ). The tren ds in AZ extractability changes over time (Figures 4 19 B, 4 20 B) on the other hand, are likely within the natural variation of biological systems. The changes in AZ extractability and stability over time, thus, should be considered only 1st order approxim ations of the likely extraction trends. Longer term extractability studies ( up to years) can provide more quantitative results, but no such studies have been conducted to our knowledge.

PAGE 143

143 Figure 4 1 9 Percent chemical recoveries with standard error bars from biosolids on days 0, 45, and 90 using various fractionation schemes. A) CIP treatments (with concentrations) B) AZ treatments (with concentrations) The treatments consist of solid matrices spike d with only 3 H CIP or 3 H AZ, and matrices spiked with average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Figure 4 20 Percent chemical recoveries with standard error bars from biosolids amended manured sand on days 0, 45, and 90 using various fractionation schemes. A) CIP treatments (with concentrations) B) AZ treatments (with concentrations) The treatments consist of un amended manured sand spiked only with 3 H CIP or 3 H AZ (few g/kg soil), and manured sand amended with biosolids (1% w/w rate) spiked with only 3 H CIP or 3 H AZ, or average or 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009).

PAGE 144

144 Chemical Extractability versus B ioavailability Accessing TOrC stability and extractability can be valuable in estimating microbial bioavailability, but values must be correlated with microbial response to be meaningful. Generally, bioaccessibility, rather than the total concentrations, determines chemi cal bioavailability in the environment (Rosendahl et al., 2012). We sequentially extracted incubation samples at 3 time intervals : 0, 45, and 90 d The 0.01 M CaCl 2 cons and AZ w ere actions of the two TOrCs were present in both media at concentrations exceeding MIC90 and/or MIC1 for some microbes (Wise et al., 1983; Eltahawy, 1993; Gordillo et al., 1993; LeBel, 1993; Bengtsson Palme and Larsson, 2016). Expressions of bacterial amoA q nr S ermB and mefE confirm that CIP and AZ were bioaccessible and, more importantly, bioavailable (at least to some microbes) throughout the incubation. Correlating bioaccessibility to bioavailability was, however, problematic because: 1) of limited samp ling data points (3 for chemical extractability and 4 for gene expression), 2) some gene expressions were independent of TOrC treatments (e.g., phoN and phoD ), and 3) AZ extractability was independent of incubation time. Da ta reveal no definitive relations hips between AZ extractability and bioavailability, but suggest that low concentrations, high sorption, high MIC values, and slow degradation fraction of CIP signific antly correlated with bacterial amoA (r = 0.52, p <0.05) and qnrS

PAGE 145

145 expression s often associated with environmental samples. Nonetheless, CIP correlation data indicate certain trends i.e., the decreased CIP stress (due to decreased bioaccessibility) on some microbes (e.g., AOB ) over time was likely reflected by increased microbial activities and/or expression of antibiotic resistance genes. Longer term extractability and gene expression data are needed to confirm the trends. The extractants used in the fractionation scheme (ECETOC, 2013) may offer promise as indicators of bioaccessible pool for the t wo TOrCs, but more research is needed for conclusiveness Microbial Response Study L imitations The study involved only one Class A biosolids. Retention/release (i.e., bioaccessibility) behavior of CIP and AZ, and microbial populations are expected to vary with physico chemical properties of different biosolids. However, adverse effects on overall microbial health were minimal over a range of concentrations of the biosolids borne target TOrCs, including the uppermost end of environmental relevance. Also, th e biosolids used in the present study had much smaller Kd values of CIP than reported for several other biosolids (see Chapter 2 ) so bioavailability of CIP in typical biosolids is likely equal to or less than assessed herein. The reverse transcriptase qPCR data show relative (not absolute) copy numbers. For reliability, we only compared data obtained from the samples subjected to RNA extraction, reverse transcription, and/or qPCR at the same time as all reaction conditions were constant between the sam ples.

PAGE 146

146 E xpressions of antibiotic resistance genes for some CIP/AZ treatments (at a particular time) are significantly different based on replicates (Figures 4 12 to 4 1 7 ). However, the differences are small (less than half an order of magnitude) and may no t reflect true differences in microbial populations. Nevertheless, data show consistent trends. Conclusions Environmentally relevant concentrations of biosolids borne CIP and AZ are bioavailable and adversely affect (at least) a f ew microbes ( but only init ially) in biosolids and (to much less extent) in biosolids receiving soils. The adverse effects are muted from an agronomic viewpoint (e.g., minimal effects on microbial respiration and on genes involved in N and P cycling) and largely overshadowed by bene fits from land application of biosolids. A dverse effects of CIP on microbes (although minimal) exceed those of AZ, likely because of greater CIP concentrations in biosolids, and greater potency in the environment. Even several years of land application (w ithout chemical attenuation) of biosolids containing typical (~median or average) concentrations of the target TOrCs, at 1% or greater agronomic rates, appears to pose minimal risks to overall microbial activity from an agronomic viewpoint (i.e., N and P c ycling) Expected chemical attenuation between biosolids applications should further abate the microbial health risks. M inimal effects of TOrCs on the overall soil/biosolids microbial activity is a welcomed news However, i nhibition of bacterial amoA expre ssion indicates that the two TOrCs can stress microbes (at least initially). Also chemical concentration induced 4 copies/g) in antibiotic resistance gene expression s ( qnrS, mefE, ermB ), possible maintenance ( ermB ) of resist ance genes and possible

PAGE 147

147 antibiotic resistant bacterial enrichment occurred Present data are insufficient to fully document but qualitatively support the notion of stress induced antibiotic resistance development and spread facilitated by biosolids. Relating biosolids borne antibiotic resistance determinants to increase and spread of antibiotic resistance in biosolids amended soils and associated ecological and human health risks is, however, problematic. L onger term (especially field) studies using v arious Class A and Class B biosolids and developing quantitative frameworks are needed to fully assess potential biosolids borne TOrC impacts on microbial resistance genes.

PAGE 148

148 CHAPTER 5 BIOSOLIDS BORNE CIPROFLOXACIN (CIP) AND AZITHROMYCIN (AZ): EARTHWORM SY STEM Synopsis Earthworms can accumulate trace organic chemicals ( TOrCs ) from biosolids amended soils and their c onsumption by terrestrial vertebrates is a pathway of TOrC introduction into ecological food web. Due to their u nique exposure s, earthworms mak e excellent organisms for monitoring contaminant bioavailability and consequent environmental and human health risks from biosolids borne TOrCs. Earthworm response data on two such biosolids borne TOrCs ciprofloxacin (CIP) and azithromycin (AZ) are largely absent, but necessary for comprehensive risk assessment. Laboratory studies assess ed earthworm response s to biosolids borne CIP and AZ over a range of environmentally relevant to unrealistically high target TOrC concentrations Slopes of response curves yielded bioaccumulation factors (BAF values) useful for risk assessment E nvironmentally relevant (and much greater) concentrations of biosolids borne CIP and AZ are not toxic to earthworms but result in compound accumulation Earthworm responses w ere linear (r 2 >0.97) even up to unrealistically high TOrC concentrations. O nly a fraction (~20% for CIP and ~40% for AZ) of the total ingested CIP and AZ accumulated in earthworm tissue yielding BAF values of ~4 (CIP) and ~7 (AZ) in depurated worms and ~ 20 (CIP and AZ) in un depurated worms. Analyses of field samples of soils and worms from control and biosolids amended plots complemented the laboratory study, but w ere too limited to support definitive conclusions. The laboratory generated BAF values sugg est potential for biosolids borne TOrCs entry into the ecological food web but field validation is essential.

PAGE 149

149 Introduction Earthworms, by virtue of eating and mixing soils, organic matter, minerals in soil etc., can be exposed to many trace organic chemic als ( TOrCs ) present in biosolids amended soils Proximity to the contaminated amended soil, thin and permeable cuticle s and consumption of large amounts of soil, make earthworms excellent organisms for monitoring potential TOrC bioavailability and consequ ent environmental and human health risks ( Jager et al., 2005 ; ECETOC, 2013). E arthworms can accumulate some trace organic chemicals (TOrCs) from biosolids amended soils (e.g., Kinney et al., 2008 ; Higgins et al., 2010; Snyder et al., 2011; Pannu et al., 2012), but data on the bioavailability of biosolids borne ciprofloxacin (CIP) and azithromycin (AZ) to earthworms are essentially absent. The only data available are for CIP uptake and accumulation in earthworms f rom studies addressing inorga nic metals (e.g., copper and cadmium) uptake by earthworm in the presence of CIP (Huang et al., 2009; Wen et al., 2011). Authors suggested that CIP accumulates in earthworm tissues and increases accumulation of metal ions like copper (Huang et al., 2009). However, the experimental de tails (freely bioavailable CIP at exceptionally high, g/kg, concentrations) limit the environ mental relevance of the two studies. E arthworms are a significant fraction of the diet of many terrestrial vertebrates (Suter et al., 2 000), and represent a potential pathway for TOrC s (including CIP and AZ ) into the ecological food web. Earthworm exposure to biosolids borne CIP and AZ is especially relevant given the immobility and persistence of the compounds in the surface layer of ame nded soil ( Ericson et al., 2007 ; Girardi et al., 2011; Gottschall et al., 2012 ). Lab oratory based bioassays and field based data on earthworm toxicity and

PAGE 150

150 bioaccumulation under environmentally relevant scenarios are crucial for comprehensive risk assessmen t of biosolids borne CIP and AZ. We examined the bioavailability and toxicity of biosolids borne CIP and AZ to earthworms ( Eisenia fetida ) at concentrations ranging from environmentally relevant to unrealistically high. Environmentally relevant concentrati ons can vary widely for different sources (or sinks) of a chemical. Herein, environmentally relevant concentrations in growth media (i.e., biosolids amended soils) are discussed in context of biosolids borne chemicals. The typical concentrations of the target TOrCs in biosolids range between median and average concentrations found in the latest targeted national sewage sludge survey (USEPA, 2009). Based on the typical concentrations and the typical 1% (dw/dw) land applic ation rate, most biosolids amended soils nominally contain about 0.05 to 0.11 mg CIP /kg and 0.003 to 0.008 mg AZ / kg. Based on the 95 th percentile concentrations found in USA biosolids, the uppermost end of environmental relevance is 0.36 mg CIP and 0.032 m g AZ per kg amended soil. The latter concentrations also represent soil concentrations (without attenuation) from ~7 (CIP) and ~10 (AZ) years of repeated application of biosolids (at 1% (dw/dw) application rate) contaminated with median chemical concentrat ions. B iosolids were spiked with varying concentrations of CIP or AZ and amended to soils in a laboratory setting to assess earthworm ( Eisenia fetida ) bioaccumulation and toxicities. Eisenia fetida are easily cultured in the laboratory and form the basis o f an extensive database on the effects of numerous classes of chemicals on earthworms (Lee et al., 2008). A limited number of field samples (biosolids amended and un amended soils) and corresponding earthworm samples were also analyzed. F ield

PAGE 151

151 samples: 1) a void issue associated with spik ing the biosolids, 2) allow validation of data from lab oratory studies, and 3) enhance the reliability of risk assessment. The unique nature of the field samples used (single biosolids application in 2008 at exceptionally hig h rate of 228 Mg/ha) also allowed assessment of residual and compound persistence effects on earthworm uptake and toxicity. We hypothesized that 1) environmentally relevant concentrations of biosolids borne CIP and AZ are too low to adversely affect earth worms, and 2) strong sorption to biosolids minimizes CIP and AZ bioaccumulation The experiment generated critical information necessary for a science based risk assessment of biosolids borne CIP and AZ. Materials 3 H C IP (CAS No. 85721 33 1; 97.4% radioch emical purity) and 3 H AZ (CAS No. 117772 70 Biochemicals (Brea, CA). Pharmaceutical secondary standards (>99% pure) of CIP and AZ and double deionized water were purchased from Sigma Aldrich (St. Louis, MO). The biosolids ( 3 Class A Table 1 2) used in the study w as an anaerobically digested air dried Class A biosolids from MWRDGC, and contained low CIP (1 mg/kg) and AZ (0.06 mg/kg) concentrations (analyzed by AXYS, BC, Canada) A medium textured (Birmingham, AL). A silty clay loam (fine, mixed, superactive, mesic Typic Endoaquolls) from an adjacent location, amended with an exceptionally high biosolids application and corresponding earthworm samples were collected from a field s ite in Illinois

PAGE 152

152 Earthworms ( Eisenia fetida ) used in the lab oratory studies were purchased from Carolina Biological (NC). Prior to use in laboratory experiments the worms were grown in moist peat moss (growing medium) and fed worm food (Magic Worm Food; Magic products Inc., WI) each d ay for at least 21 d Methods Laboratory S tud y Toxicity of biosolids borne CIP and AZ to earthworms was evaluated in the laboratory using modifications of Earthworm s ub chronic t oxicity t est (OCSPP 850.3100) (USEPA, 2012). Bioaccumulation of the target biosolids borne chemicals was also assessed under the same conditions. Chemical toxicity and bioaccumulation response curves were generated from s biosolids borne CIP and AZ, r anging from environmentally relevant (i.e., ~0.015 mg CIP and ~0.0089 mg AZ per kg amended soil) to unrealistically high (i.e., ~1.8 mg CIP and ~0.16 mg AZ per kg amended soil) (USEPA, 2009). The unrealistically high concentrations correspond to 5 times th e soil concentrations expected from application of biosolids contaminated with 95 th percentile CIP or AZ concentrations (USEPA, 2009) and represent a worst case scenario (i.e., chemical accumulation, without attenuation, from land application of severely contaminated biosolids over 5 years at 1% (dw/dw) application rate). The chemical concentrations were selected based on the USEPA targeted national sewage sludge survey (USEPA, 2009) of biosolids produced in the USA, which is the broadest and most thorough nationwide survey, and use USEPA approved analytical methods. The triplicated study involved: 1. two antibiotics CIP and AZ,

PAGE 153

153 2. four biosolids borne chemical concentrations 3. two solid matrices a biosolids amended (1% w/w) sand and the control field soil that was amended with the biosolids (1%, w/w) prior to lab experiments, 4. sand and control field soils without biosolids or TOrCs, and 5. the heavily amended field soil spiked with a single (very low few g/kg) concentration of CIP or AZ. Select characteristics of the soils are listed in Table 5 1. The total number of samples were (2 chemicals x 4 treatments x 2 soils x 3 replicates + 2 ch emicals x 1 concentration ( 3 H compound only ) spiked to the heavily amended field soil x 3 replicates) = 48 + 6 = 54. The number of controls was 12 (2 chemicals x 2 un amended soils x 3 replicates). Table 5 1. Select properties of soils and biosolids used in the lab earthworm study (measured, average values from duplicate samples). Media pH OM ( g/kg ) Biosolids amended sand 6 9. 1 Biosolids amended control field soil 8 4 8 Heavily amended field soil 7.8 7 7 [Solid matrix characterization was conducted by the Analytical Research Laboratories (ARL) at the University of Florida, Gainesville, FL. Methods used: EPA 150.1 (pH) and Loss on Ignition (OM) The data represent average values from duplicate samples ]. Prior to soil amendment, the biosolids used in the lab oratory studies was pre equilibrated with the 4 concentrations of the 3 H compounds or 3 H plus unlabeled compounds for 7 d in the dark at ~25 0 C, a time shown in preliminary studies to promote chemical and isotopic exchange equilibria. Using 3 H labeled CIP and AZ for chemical quantification and analysis minimized sample treatment and chemical extraction steps and provided very low (~ng/ kg) detection limits The equilibration utilized 1:2 solid:solution ratio on an end to end shaker at 150 rpm for 7 d T h e equilibrated biosolids were air dried before amendment to the two soils. Two g rams of the spiked

PAGE 154

154 biosolids were amended, separately, to 198 g of the sand or the control field soil at the 1% (d w/ d w) rate typically used in agriculture. The nominal concentr ations of CIP and AZ (mg ) per kg of the two amended soil s were: 0.015, 0.12, 0.37, and 1.8 for CIP and 0.0089, 0.016, 0.032, and 0.16 for AZ. Only one CIP or AZ spiking concentration was used for the heavily amended field soil, where the purpose was to co mpare lab oratory generated earthworm bioaccumulation factors (BAF) with the in situ field BAF values from the same soil. The comparison allowed assessing effects of CIP and AZ persistence on earthworm uptake and toxicities. Additionally, including the two field soils (heavy amended and control) in the laboratory study made comparing the results with those from the field study more direct, as the same solid matrix was used in both studies. Like in the pre equilibration of biosolids, 10 g of the heavily amend ed field soil was spiked with 3 H compound only and equilibrated in dark for 7 d Two g rams of the equilibrated heavily amended field soil w ere mixed with 198 g unspiked heavily amended field soil. The nominal 3 H CIP/AZ concentrations per kilogram of the heavily amended field soil were 0.005 mg CIP or 0.0067 mg AZ. Two hundred g ram sub samples of each solid matrix were placed in 1 L glass Mason jars and brought to field capacity (~6% for the sand and ~20% for the two field soils (w/w)). Earthworms (un iform in shape, size, and each weighing between 400 600 mg) were handpicked from the growing medium and washed with deionized water before use in the study. Ten earthworms were added to each sample and the jars were covered with cotton cloth using rubber b ands to reduce moisture loss and to prevent earthworm escape. The samples were incubated at room temperature (~25 0 C) for 28 d

PAGE 155

155 The earthworms were exposed to a 12 h light and dark cycle (using a white light lamp in a dark room) to better simulate environme ntal scenarios. The samples were weighed periodically, and water was added whenever moisture content was ~10% less than the field capacity. Dead earthworms (if any) at the soil surface were screened for, counted, and removed as necessary each d ay and the number of living earthworms were tallied every 7 d for 28 d Following the 28 d incubation, earthworms were removed from the soils, gently rinsed with double deionized water, placed on a paper towel to remove excess moisture and weighed. The earthworms wer e then transferred to petri dishes lined with moistened filter paper, and allowed to depurate for 48 h. The filter papers were replaced in the morning, mid d ay and at the end of d ay during depuration. Depurated earthworms were dried to constant weight, grou nd, and a subsample was oxidized using a Harvey OX 500 oxidizer (RJ Harvey Corp., NY). The filter papers with earthworm excreta were also dried, composited, ground, and combusted to assess the total amount of CIP and AZ ingested by the earthworms and to co mpare that to bioaccumulated CIP and AZ. Sub samples of the soils (not including excreta), taken at the beginning and end of the study were sequentially extracted and analyzed for mass balance purposes and to: 1) confirm radio labeled CIP and AZ initially added, 2) assess CIP/AZ extractability and degradation, and 3) correlate CIP and AZ bioaccessibility with bioavailability to earthworms. The chemical extraction, fractionation, and thin layer chromatography (TLC) analysis schemes were the same as describe d in Chapter 4. Determination of bioaccumulation factors (BAF) The process of bioaccumulation considers all pathways and routes of exposure and is valuable in assessing bioavailability of a chemical (Alexander, 1999). Chemical

PAGE 156

156 uptake response curves were generated for chemical accumulation (depurated worms) and total chemical uptake (un depurated worms) by plotting chemical concentration in the earthworms as a function of chemical concentration in the biosolids amended soils. Bioaccumulation factors (BAFs) were calculated from the slopes of response curves instead of point estimates suggested by standard tests (e.g., OECD 2010 ) As only one concentration was used for the heavily amended field soil, a point BAF value was calculated by dividing 3 H chemical c oncentration found in the worms (mg/kg, dw) by 3 H chemical concentration found in the soil (mg/kg, dw) 3 H detection limits Tritium counting efficiencies of 42 55% were obtained on the liquid scintillation counter in all samples. An operationally defined cut off quantification limit of 1.7 Bq was used such that an activity below 1.7 Bq was considered insignificant for the study purposes. Although, the 3 H activities in all of the solid matrices were greater than the minimum detectable true activity value of 0.46 Bq was present in the systems), the activities were less than 1.7 Bq and were regarded as Field Stud y Longer term bioavailability of biosolids borne CIP and AZ was estimated by analyzing earthworms and corresponding soil samples collected at a field site in Illinois (Table 5 1). The biosolids amended site received a single application of an anaerobically digested cake biosolids (application rate of 228 Mg/ha) in 2008. Earthworms and the soil were collected from the site in 2017. Earthworms and soil were also collected from an adjacent control site (18 m away; silty clay loam soil with pH =8,

PAGE 157

157 OM = 4.8 %) with no known history of biosolids application. There were no known additional sources of CIP or AZ into the field site over the intervening 9 y Earthworms were collected from both the control and the heavily amended field soils using a mustard extraction method (Lawrence and Bowers, 2002). Briefly, allyl isothiocyanate solution was applied to the soil surface to encourage earthworm emergence. Approximately 20 earthworms were collected from each l ocation. The collected worms were kept in wet peat moss in aerated bags and shipped overnight to Cincinnati, Ohio under ice packs. The earthworms were cleaned with double deionized water but not depurated The collected (un depurated) worms and sub sample s of the two field soils (control and heavily amended) were frozen at 10 to 20 0 C until analysis. The chemical concentrations in the worms (4 replicates each) and the two field soils (3 replicates each) were analyzed by an independent laboratory (AXYS, BC Canada). The target TOrCs were determined by LC MS/MS following AXYS Method MLA 075 (based on EPA method 1694 for analysis of pharmaceuticals (USEPA, 2007)). Earthworms collected from the field sites were analyzed un depurated and on a wet weight basis whereas laboratory samples were analyzed depurated and on a dry weight basis. C omparisons between the field samples and laboratory studies required information on the amount of excreta presence in worms and the moisture contents of un depurated worms, depu rated worms, and worm excreta ; which were determined on sub samples of the collected worms Statistical A nalysis The data were analyzed using R for normality (Shapiro Wilk test) and homogeneity of variance (Levene test). The

PAGE 158

158 was used for analysis of normal and homogeneous data, whereas Kruskal Wallis test was used for non Results and Discussion Laboratory Stud y TOrC toxicity to earthworms Even at concentrations 5 times greater than the uppermost end of environment al relevance earthworm mortality in the after 28 d and statistically the same across all CIP and AZ treatments. Worms lost weight (~20%) in all treatment s due to limited food supply, but final earthworm weights were statistically the same across treatments for a particular media (data not shown). For both compounds, all earthworms in the sand control died in the first 7 d of the experiment. Mortality was attributed to insufficient food supply in the sand. The OCSPP 850.3100 guidelines specify that earthworm mortality in controls should not be more than 20%, but the guidelines are designed for artificial soils with ample food source. Results from artificial soils ( prepared according to the OCSPP method) are rarely transferable to real soils (Snyder et al., 2011; Pannu et al., 2012; Brami et al., 2017). Here real soils were used, and sand served as the worst case scenario. As the biosolids borne TOrC spiked treatments showed no acute or chronic toxicities to earthworms, we accept the toxicity results from all 3 soil media. Bioaccumulation of CIP and AZ by earthworms Bioaccumulation of CIP and AZ by (depurated) earthworms was well described by linear models (r 2 >0.97) over the entire concentration range (Figures 5 1 5 2 ). The data suggest that CIP and AZ bioaccumulation is linear even at CIP and AZ concentrations much greater than the environmentally relevant chemical concentrations

PAGE 159

159 Bioaccumulation factors ( BAF) standard deviations (SDs) derived from the linear models are listed in Table 5 2. The laboratory point estimated BAF values SD for both chemicals in the heavily amended field sample are also listed in Table 5 2. All BAF values are expressed on a d ry weight basis. All BAF values were greater than one demonstrating bioaccumulation of biosolids borne CIP and AZ. The accumulated concentrations were ~4 (CIP) and ~7 (AZ) times greater than the amended soil concentrations (Table 5 2) independent of spike d chemical concentrations. Bioaccumulation factors for a particular TOrC were similar across the various biosolids amended soils despite different OM contents and CEC values of the soils suggest ing that bioaccumulation of biosolids borne CIP and AZ wa s la rgely controlled by the contaminated biosolids (rather than the soils being amended). The BAF values were generated using 3 H compounds and are valuable for conservative risk assessment as they represent total (potential) accumulation of radiolabel. Tritium could be associated with parent compound and metabolites taken up from biosolids amended soil as well as with chemical metabolites produced within earthworm tissue. A wide range of BAF values for depurated worms is reported (less than 0.1 to greater than 20) for other (including ionic) biosolids borne TOrCs (Kinney et al., 2008; Higgins et al., 201 1 ; Snyder et al., 2011; Pannu et al., 2012 ), but no BAF values are available for biosolids borne CIP and AZ. Ciprofloxacin and AZ BAF values calculated herein f all within the general range reported for other TOrCs. Interestingly, the majority of the CIP (~80%) and the AZ (~60%) taken up by earthworms was excreted (Table 5

PAGE 160

160 2), and only ~20% (CIP) and ~40% (AZ) bioaccumulated. The similarity of BAF values for CIP a nd AZ in un depurated worms (Table 5 2) suggest that the same proportions (relative to soil concentrations) of CIP and AZ were taken up by the worms. However, like with plants, the worm s bioaccumulated more AZ than CIP. Azithromycin is a larger molecule an d has more aromaticity and a greater log Dow (i.e., pH dependent Kow) value than CIP at environmentally relevant pH values. Octanol water partitioning coefficients (Dow values) of organic compounds are often well correlated with uptake into organic tissues ; thus, the greater the Dow values, the greater the expected bioaccumulation potential (ECETOC, 2013). T he partitioning coefficient s (Kd values ) of ~360 (CIP) and ~430 (AZ) L/kg in the biosolids ( Table 2 3 ) are similar, so the greater Dow value of AZ is co nsistent with greater bioaccumulation (on a per unit basis) of AZ than CIP Results suggest that, like other TOrCs, CIP and AZ accumulated via partitioning into lipids in earthworm tissue. Bioaccumulation means that earthworms represent a pathway for CIP a nd AZ into the ecological food web. Total chemical uptake (BAF values) for un depurated worms is perhaps as (or more) important as BAF values for depurated earthworms for risk assessment P redators consume entire earthworms, which may, or may not include e xcreta. The potential CIP and AZ biomagnification in earthworm predators can be greater than predicted by depurated BAF values. Risk assessments, generally, use BAF values for depurated worms because the values represent equilibrium conditions. Bioaccumula tion factors derived from un depurated earthworms yield conservative estimates of risks to earthworm predators Importantly, the laboratory data represents a worst case scenario where earthworms spen t 100% time in the biosolids amended soil

PAGE 161

161 Bioaccumulation factors u n der field conditions are generally smaller than under laboratory conditions due to more heterogeneity under field conditions and/or greater chemical availability under laboratory conditions ( Pannu et al., 2012; Higgins et al., 201 0; van den Brink et al., 2015; Hoke et al., 2015 ). Also, earthworms do not spend 100% time in biosolids amended parts of the field so e arthworm BAF values under field conditions are l ikely less than laboratory determined values Table 5 2. Bioaccumulatio n factors (BAFs) standard deviations (SDs) for depurated and un depurated worms (on dry weight basis), and the average chemical (%) excreted by the earthworms during depuration. Chemical Media BAF SD (depurated worms) BAF SD (un depurated worms) Aver age percentage of the total excreted (%) CIP Biosolids amended sand 3.4 0.22 16 1.6 78.8 Biosolids amended control field soil 3.7 0.19 18 1.1 81 Heavily amended field soil 3.9 0.47 20 3.2 82.3 AZ Biosolids amended sand 6.3 0.61 17 3.4 61.7 Biosolids amended control field soil 7.0 0.67 20 2.9 65.4 Heavily amended field soil 7.6 0.78 21 3.8 63.8

PAGE 162

162 Figure 5 1. CIP bioaccumulation factors (depurated worms ; dry weight basis ) for A) sand. B) control field soil. The solid black lines represent the fit of the data. The dashed blue lines indicate upper and lower 95% confidence limits. Figure 5 2 AZ bioaccumulation factors (depurated worms ; dry weight basis ) for A) sand. B) control field soil. The solid black lines repr esent the fit of the data. The dashed blue lines indicate upper and lower 95% confidence limits.

PAGE 163

163 Extraction and analysis of CIP and AZ from the solid matrices Little or no target TOrC was extracted from the three treated soils using CaCl 2 (Figures 5 3, 5 4 ). On d ay 0, methanol:water extracted ~25% CIP and ~20% AZ, whereas ASE extracted ~40% CIP and ~ 30% AZ from all CIP or AZ treatments across all soils. Ciprofloxacin extractability either marginally increased or was unchanged from d ay 0 to day 28, and AZ extractability did not change over time. The mass balance analysis from the sample oxidizer on d ay 0 and 28, after correction for percent 10 5 % of the initial 3 H added to the samples. Generally, the recoveries were similar across treatments for all media for a particular time point (Figures 5 3, 5 4 ). T 3 H recover ed existed as the parent compounds based on RF assignment. O ther 3 H entities (degradation products and/or chemical complexes ) were negligible. The extractability data (Figures 5 3, 5 4 ) suggest that ~20% of the total target TOrCs is bioaccessible (similar to that found in microbial incubation studies ; Sidhu, Chapter 4 ) U nlike microbial study results however the trends in ch emical extractability over time suggest that earthworms could increase the bioaccessibility of TOrCs long term. However, the study was only 28 d (i.e., less than the 45 d second sampling time for microbial incubation study). C ompound recoveries from variou s fractionation methods and total mass balance have a cumulative error of about 10% t he extraction data from different treatments are within this 10% variation. Thus, the trends in chemical bioaccessibility changes over time are within the natural varia tion of biological systems and are deemed unremarkable.

PAGE 164

164 Figure 5 3 Percent CIP recoveries ( with standard error bars ) on days 0, 45, and 90 from various chemical treatments (with chemical concentrations) using various fractionation schemes from A) amended sand. B) amended control field soil. C) heavily amended field soil. The treatments consist of solid matrices spiked with only 3 H CIP or 3 H AZ, and matrices spiked with average 95 th percentile or 5 x 95 th perc entile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009). Figure 5 4 Percent AZ recoveries ( with standard error bar s) on day 0, 45, and 90 from various chemical treatments (with chemical concentrations) using various fractionation sch emes from A) amended sand. B) amended control field soil. C) heavily amended field soil The treatments consist of solid matrices spiked with only 3 H CIP or 3 H AZ, and matrices spiked with average 95 th percentile or 5 x 95 th percentile concentrations of CIP and AZ found in USA biosolids (USEPA, 2009).

PAGE 165

165 Field Study The heavily amended field soil had 228 Mg/ha of a cake biosolids applied in 2008. The soil nominally contained ~ 0.1 mg CIP /kg and 0.006 mg AZ /kg assuming the same chemical concentr ations in the biosolids that were land applied as the 3 Class A biosolids and no chemical attenuation. The 3 Class A biosolids (without air drying) were applied to the field soil in 2008. These nominal CIP and AZ concentrations were above the detection limits (~0.01 mg CIP and 0.003 mg AZ per kg) listed by AXYS for the two chemicals but neither compound was detected in the soil samples. Greater detection limits for CIP in the actual samples (~0.3 mg/kg ; due to matrix interference, see Appendix A ) and/or attenuation of the two compounds since the 9 years of biosolids application, likely resulted in undetectable soil concentrations. Earthworms (un depurated) bioaccumulated negligible CIP (~0.04 mg/kg) and AZ (un detectable) despite significantly lower rep orting limits (~100 fold less in case of CIP) in the earthworm tissues than in the soil. Thus, despite the expected long half lives of the compounds, soil dilution and bound residue formation likely protect ed against problematic earthworm accumulation of t he compounds even in soils amended at exceptionally high rates. In contrast with laboratory study, the field study suggests negligible accumulation of the target TOrCs by earthworms. However, several instances of undetected chemical concentrations and non quantifiable samples limit the usefulness of the field study. Further, there was a 9 y difference between biosolids application and collection of field samples. Attenuation processes likely severely limited chemical bioaccessibility to field biota. Bioacc umulation factors from relatively recently biosolids amended fields could be significantly greater, but such data are lacking. Conservatively, we will emphasize

PAGE 166

166 laboratory studies, which clearly indicate CIP and AZ accumulation potential in earthworms (Tab le 5 2). Implications of accumulation in earthworms are assessed in a risk estimation of biosolids borne CIP and AZ (Chapter 6). Conclusions U ptake and toxicity potentials of biosolids borne CIP and AZ to earthworms were assessed under laboratory condition s. Our hypothes es were that 1) environmentally relevant concentrations of biosolids borne CIP and AZ are too low to adversely affect earthworms, and 2) strong sorption to biosolids minimizes CIP and AZ bioaccumulation. R esults are consistent with our first hypothesis that biosolids borne CIP and AZ are not toxic to earthworms exposed to environmentally relevant (and much greater) concentrations. However, the target TOrCs accumulate d in the earthworms, contradicting our second hypothesis. Compound concentrations in e arthworms were ~20 times greater than in the biosolids amended soil but only a fraction (~20% for CIP and ~40% for AZ) of the ingested CIP and AZ bioaccumulated. The laboratory generated earthworm BAF values from depurated (~4 for CIP a nd ~7 for AZ) worms suggest potential for biosolids borne TOrCs entry into the ecological food web. Unfortunately t he analyses of worms and soil from a field site were too limited to support definitive conclusions A dditional investigations under field co nditions are needed Predators consume whole worms, which can include excreta. Thus, using BAF values for un depurated worms, or an average of BAF values for depurated and un depurated worms, likely represents a conservative 1st Tier assessment of risk Bioaccumulation appears to be controlled by biosolids because BAF values were similar across different soil matrices, but the study involved only one (Class A) biosolids.

PAGE 167

167 Different biosolids can have different physico chemical characteristics and, thus, di fferent chemical retention/release (bioaccessibility) behaviors. Spiking the biosolids used herein to extraordinarily high target TOrC concentrations, however, did not affect strongly linear (r 2 >0.97) earthworm response curves Thus, even if chemical bioac cessibility differs among biosolids, earthworm responses similar to the ones reported herein are likely from land application of typical (or even severely contaminated) biosolids over several years.

PAGE 168

168 CHAPTER 6 RISK ASSESSMENT OF BIOSOLIDS BORNE CIPROFLOXACIN (CIP) AND AZITHROMYCIN (AZ) Synopsis Ciprofloxacin (CIP) and azithromycin (AZ) are frequently detected in biosolids and biosolids amended soils, but risks from biosolids borne CIP and AZ are incompletely characterized Human and e cological he alth risks associated with biosolids borne CIP and AZ were evaluated using a tiered approach based on the W orld Health Organization (WHO) integrated risk assessment (IRA) framework. Human and ecological exposure hazards were identified under three biosolid s application scenarios using the hazard quotient (HQ) approach. Risks associated with various p athways, routes, and receptors of concern were initially unrealistically conservative (Tier 1), but were subsequently refined using less conservative (i.e., more realistic) exposure assumptions (Tier 2 and 3) The screening level (Tier 1) assessment identified three pathways of potential concern: b iosolids soil plant; b iosolids soil soil organism; and b iosolids soil soil organism predator. Subsequent tier (re fined) assessments suggested negligible human and ecological health risks from biosolids borne CIP and AZ under real world application scenarios. Risks are expected to be small even under unrealistically high exposure scenarios. Preliminary pollutant limit s calculated based on Coopers Hawk (CIP) and American Woodcock (AZ) as the most sensitive organism s suggest that single heavy application s of severely contaminated and long term application of typical biosolids are with out appreciable human and ecologic al risks. The IRA suggests that p ollutant load tracking is not needed for the majority of USA biosolids, but may be necessary for some biosolids containing greater than 12 mg CIP/kg and 2.2 mg AZ/kg

PAGE 169

169 However, the IRA needs refining by including more data, especially on biosolids borne antibiotic resistance, before modifying current land application regulations. Introduction T he presence of trace organic chemicals (TOrCs) in biosolids affects public and regulatory perception of land application safety Prev ious risk assessments on some biosolids borne TOrCs (both polar and non polar) suggest minimal human and ecological health risks (Eriksen et al., 2009; Jensen et al., 2012; Snyder et al., 2013) T he scientific community however, remains split on human and /or ecological risks from various biosolids borne TOrCs especially ionic compounds and antibiotics (Prosser and Sibley, 2015; Verlicchi and Zambello, 2015; Mohapatra et al., 2016; Clarke et al., 2017). Adverse health risks are dictated by chemical exposu re to the target organism, which depends upon c hemical properties and interaction with the environment. For instance, TOrCs that poorly sorb to biosolids and soils (e.g., carbamazepine) have high potential for exposure whereas strongly sorbing TOrCs (e.g. triclocarban) have much less potential (Clarke and Smith, 2011). S orption decreases contaminant bioavailability, but increases chemical persistence by limiting degradation (Alexander, 2000). The expected persistence and lack of target response data, iden tifie d biosolids borne antibiotics such as ciprofloxacin (CIP) and azithromycin (AZ) as compounds of interest for risk assessment (Higgins et al., 2010; Verlicchi and Zambello, 2015). A reliable risk assessment is essential for formulating sound polic ies to safeguard human and environmental health from biosolids borne CIP and AZ. A science based risk assessment requires knowledge of various physico chemical properties and the environmental behavior of the target TOrCs. Physico chemical properties and envir onmental fate dictate chemical exposure to an organism, but exposure does not

PAGE 170

170 necessary equal risk. Other factors such as chemical tolerance and elimination rates in the target organism are also important. Relating chemical exposure to risk is often thwart ed by lack of target organism response data. Consideration of real world based application scenarios and inclusion of measured values for various parameters (e.g., avian toxicity ) involved is, thus, crucial for a sound risk assessment. The persistence of C IP and AZ (Sidhu, Chapter 4; Walters et al., 2010; Girardi et al., 2011; Gottschall et al., 2012) portends that both chemicals remain in biosolids and biosolids amended soils for long times Longevity i ncreas es the potential exposure to human and ecological receptors. However, strong and extensive retention and limited release can severely limit bioavailability of persistent biosolids borne CIP and AZ (Sidhu, Chapter 4; Nowara et al., 1997; Ericson, 2007; Wu et al., 2013; Che n et al., 2015; Berhane et al., 2016; Carrillo et al., 2016). Some studies suggest that TOrCs like CIP and AZ constitute moderate to high risk in the environment (Halling Srensen et al., 2000; Ebert et al., 2011; Iatrou et al., 2014; Verlicchi and Zambell o, 2015). Most of the studies, however, only approximated various fate, response, and toxicity parameters and/or were not applicable to biosolids systems. There are few studies of risks from biosolids borne CIP and AZ. Most focused on human health risks th rough specific subset s of exposure pathways and suggested negligible risks (Eriksen et al., 2009; Jensen et al., 2012; Matamoros et al., 2012; Prosser and Sibley, 2015). Multi pathway and multi receptor integrated risk assessment (IRA) is lacking. Human an d ecological health are inseparable compartments of the environmental dynamics and detrimental impacts of a single hazard on one compartment often resonates in the other (Suter, 2007). Also, e cological receptors are

PAGE 171

171 often exposed to significantly greater c hemical concentrations than humans (Suter, 2007). E cological risks from environmental concentrations of a chemical can be profound even if human risks are minimal (Suter, 2007). An integrated risk assessment is, therefore, crucial to fully assess human and ecological risks from biosolids borne CIP and AZ. This dissertation focused on filling some of the critical data gaps needed for a sound risk assessment and conducting a tiered IRA of biosolids borne CIP and AZ. The IRA, similar to the one conducted by Sn yder et al. (2013), followed a framework developed by WHO (2001) and included (where applicable) assumptions, scenarios, and parameters from USEPA Part 503 Biosolids Rule making risk assessment (1995). The IRA involved experimental data generated in this d issertation, data gleaned from published literature, and toxicity benchmarks generated by Co PIs in a WE & RF funded project (Dr. Andy Maier, personal communication) Emphasis was on data generated at environmentally relevant chemical concentrations under re al world biosolids management practices. The goal was to inform policy decisions on science based regulation of land application of biosolids. The IRA tested the central hypothesis that limited bioaccessibility of the strongly sorbed CIP and AZ minimizes r isks to human and environmental health. Integrated R isk A ssessment A pproach The risk assessment involved the following scenarios: BAR 1 : A one time cropland application ( 10 0 Mg/ha dw ) of biosolids contain ing the 95 th percentile concentrations of CIP and AZ in USA biosolids (USEPA, 2009): The 95 th percentile is a high end chemical concentration utilized by USEPA in screening level risk assessments (USEPA, 2003). T ypical land application rates of biosolids to

PAGE 172

172 croplands are 10 to 20 Mg/ha (USEPA, 2003). Therefore, th e BAR 1 scenario assesses risks from a one time application of a severely contaminated biosolids at an exceptionally high rate. BAR 2 : Annual land application of biosolids containing average chemical concentration s (USEPA, 2009) at a rate of 2 0 Mg/ha y continuously, for 40 y: This scenario represents land application of typical biosolids, but still considers high end cumulative application (i.e., 40 y without any assumed chemical attenuation; USEPA, 2003) The sce nario is intended to assess long t erm risks from typical biosolids application s BAR 3 : Annual land application of biosolids containing 95 th percentile chemicals concentrations (USEPA, 2009) at the rate of 2 0 Mg/ha y for 40 y: This scenario assess es human and ecological risks from long term land application of severely contaminated biosolids. Although the scenario is unlikely, it model s risks from biosolids borne CIP and AZ under the uppermost extreme end of environmental relevance. Pathways C onsider ed Sixteen exposure pathways (Table 6 1 ) encompass ing the major routes of exposure to a highly exposed target organism (i.e., highly exposed individual) a re considered in the IRA. The first 14 exposure pathways a re directly incorporated from the risk asses sment supporting the Part 503 Biosolids Rule development (1995). In line with risk assessment revisions since the promulgation of the Part 503 Biosolids Rule, two additional pathways a re added to address potential entry of biosolids borne chemicals into su rface waters (USEPA, 2003; Snyder et al., 2013). For instance, runoff and erosion of biosolids amended soil to waterbodies can introduce a chemical (via complete dissociation of the contaminant from the biosolids amended soil) into a water

PAGE 173

173 pond (or a surfa ce water reservoir) near a biosolids application area. The contaminated water then pose s risk s to aquatic animals (pathway 16). Similarly, drinking contaminated water and eating contaminated fish pose potential risk s to humans and animals (pathway 15). The highly exposed individual (HEI) is an organism within a population with greatest, but realistic, exposure to a contaminant either by virtue of the environment in which they live/work or by virtue of their habits, developmental stages etc. (Epstein, 2003). A minimal risk to an HEI is perceiv ed as a negligible risk to rest of the population. Table 6 1. Pathways of exposure and corresponding highly exposed individuals (HEI)s considered in the risk assessment. Pathway 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 biosolids amended soil 3. Biosolids soil human Human (child) ingesting biosolids amended soil 4. Biosolids soil plant animal human Human lifetime ingestion of animal products (animals raised on forage grown on biosolids amended soil) 5. Biosolids soil animal human 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 amended soil 8. Biosolids soil plant Plant toxicity due to taking up biosolids borne chemicals 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 CIP or AZ dissociated from biosolids and biosolids amended soils that reach waterbodies.

PAGE 174

174 Table 6 1. Continued. Pathway HEI 13. Biosolids soil air human Adult human lifetime inhalation of volatilized CIP or AZ from biosolids amended soil 14. Biosolids soil groundwater human Human lifetime drinking well water containing CIP or AZ from biosolids that leached from soil to ground water 15. Biosolids soil surface water animal Animal lifetime drinking surface water and ingesting fish containing CIP or AZ dissociated from biosolids and biosolids amended soils that reach waterbodies. 16. Biosolids soil surface water aquatic organism Aquatic organism exposed to water containing CIP or AZ dissociated from biosolids and biosolids amended soils that reach waterbodies. Selection of HEIs The HEIs for each pathway were selected based on living situation s and/or behavior that put the individuals at high exposure risk to biosolids borne contaminants (USEPA, 1995). Based on the USEPA exposure factors handbook (USEPA, 2011), an adult human (age 20 y) and a child (age 5 y) were considered HEIs for human pathwa ys (1 to 5 and 11 14). The HEIs for non human pathways were selected based on literature (USEPA, 1993; USEPA, 2003). Predator and animal pathways (4, 5, 7, and 10) included mammals (Cow, Bos taurus ; Deer Mouse, Peromyscus maniculatus ; Red Fox, V ulpes vulpe s ; Short tailed Shrew, Blarina brevicauda ) and birds (American woodcock, Scolopax minor ; Coopers Hawk, Accipiter cooperii ; Red tailed Hawk, Buteo jamaicensis ). Earthworms ( Eisenia fetida ) and soil microbes were HEIs in soil (pathway 9). Surface water pathw ays (15) included Osprey ( Pandion haliaetus ) and River otter ( Lontra canadensis ) as animal HEIs. A cyanobacteria ( Microcystis aeruginosa ) was considered as a HEI for aquatic community (pathway 16) because current toxicity data (Tables 6 2 and 6 3) suggest it to be the most sensitive aquatic species to CIP and AZ

PAGE 175

175 in water. The risk assessment of the most sensitive species is protective of less sensitive species under the same exposure scenarios Risk Characterization Risks to the HEIs in each pathway are ass essed using the hazard quotient (HQ) approach. Hazard quotient is the ratio of potential exposure concentration of a substance to a concentration at which no adverse effects are expected (reference dose; USEPA, 2003). A HQ value less than 1 indicates negli gible risk, whereas a HQ value above 1 does not necessarily means risk and only implies that an exposure concentration exceeds the reference dose (RfD) by a certain amount. Neither CIP or AZ are carcinogenic (Tables 6 2 and 6 3), so acceptable risk based c ancer risk assessments are not applicable. Reference Dose Calculations The first step in risk assessment is determining reference doses (RfD) for receptors (human and ecological) of interest. Reference doses for the HEIs are calculated from the available toxicity data (summarized in Tables 6 2 and 6 3 ) by applying appropriate uncertainty factors. Where available, c hronic studies are preferred over acute studies and no observable effect concentrations (NOEC) are preferred over values for RfD calculations Human toxicity benchmark (RfD) calculations Human reference dose (RfD) values (Table 6 4) were determined by Co PIs of a WE & RF project (Dr. Andy Maier, personal communica tion) Briefly, the RfD for CIP is derived from the lowest therapeutic dose in humans (250 mg/ d ) by utilizing uncertainty factors for inter individual variability (10x), sub chronic to chronic exposure (10x), and

PAGE 176

176 extrapolation to a no effect level (10x). T he resulting value (0.25 mg/ d ) is adjusted to per kg dose by dividing with the default human body weight of 70 kg to yield a RfD value of 0.004 mg/kg d The RfD for AZ is based on a non toxic dose of 20 mg/kg d in a six month oral study in rats (Mayne et a l., 1996). The RfD is derived by utilizing uncertainty factors for inter individual variability (10x), sub chronic to chronic exposure (3x), and animal to human extrapolation (10x), resulting in an RfD of 0.067 mg AZ /kg d. Ecological toxicity benchmark (Rf D) calculations The RfD values for birds and animals (Table 6 4) are calculated from bird and chronic rat studies and incorporate uncertainty factors specific for ecosystem risk assessment (Suter, 2007). An uncertainty factor of 10 is applied to convert a cute to chronic exposure. An uncertainty factor of 5 is applied if the test species and endpoint species of interest are in the same class but different orders. Differences in species sensitivity, laboratory field extrapolation, and intraspecific variabili ty, each, are accounted for by using uncertainty factors of 2 (Suter, 2007). The starting point for CIP RfD for raptors is a NOEC value of 50 mg/kg based on a study in Buteo jamaicensis (Red tailed hawks) (Isaza et al., 1993). The RfD (CIP) for raptor birds incorporates uncertainty factors of 10 (acute to chronic study), of 2 (laboratory field extrapolation), and of 2 (intraspecific variability). The resultant RfD is 1.25 mg/kg d Based on sub chronic studies on birds and fast elimination rates, a lowes t NOEC of 15 mg/kg d serve as a starting point for CIP RfD calculations for birds other than raptors (Frazier et al., 1995; Samanta and Samiran 2017). Applying uncertainty factors of 3 (sub chronic to chronic), of 2 (laboratory field extrapolation), and o f 2 (intraspecific variability), yield a RfD of 1.25 mg/kg d.

PAGE 177

177 The CIP RfD values for mammals are based on a dose of 2 0 mg/kg d in a chronic rat study (US National Library of Medicine, 2002). Applying uncertainty factors of 2 (species sensitivity), of 2 (la boratory field extrapolation), and of 2 (intraspecific variability), yield s a RfD of 2.5 mg/kg d Similarly, the RfD values for AZ are determined from a dose of 20 mg/kg d in a chronic rate study ( Mayne et al., 1996 ). Applying uncertainty factors of 5 (sam e class but different orders), of 2 (species sensitivity), of 2 (laboratory field extrapolation), and of 2 (intraspecific variability), yield s a RfD value of 0. 5 mg AZ/kg d for birds. Applying uncertainty factors of 2 (species sensitivity), of 2 (laborator y field extrapolation), and of 2 (intraspecific variability), yield s AZ RfD value of 2.5 mg/kg d in mammals Tiered Risk Assessment Approach The risk assessment follow ed a tier ed approach where each increasing tier refined the applicable parameters (e.g., chemical properties, chemical environmental interactions, etc.) towards less conservative (and more realistic) values The HQ values of less than 1 for the corresponding HEIs eliminated concerns about many of the 16 pathways and subsequent assessments only involved the pathways where one or more HEIs had HQ values greater than 1. Characteristics of each tier are described below: First tier (screening level) assessment The screening level risk assessment follows key c onservative assumptions: 1. 100% chemical bioavailability and complete reversibility of chemical sorbed onto biosolids and soils. 2. No chemical attenuation, i.e., chemical accumulates with each subsequent biosolids application (in 40 y scenarios) and the cumul ative concentration of the chemical is 100% bioavailable.

PAGE 178

178 3. Maximum exposure of a HEI. For instance, 100% of the plants grown on contaminated land are consumed and 100% of predator diet consists of highly exposed earthworms. Second tier assessment Second tie r assessment refin es the following key assumptions: 100% chemical bioavailability assumption reduced to 50%: The bioaccessibility of biosolids borne CIP and AZ is << 100% (Sidhu, Chapters 2 and 4; Nowara et al., 1997; Ericson et al., 2007 ; Girardi et al., 2011 ). However, the bioavailability i s reduced to a still conservative 50%, based on limited bioavailabilities of long lived biosolids borne chemicals suggested by the 1995 USEPA Part 503 Biosolids rule guidance Chemical half life in soil assumed to b e 10 y: Both CIP and AZ can persist in biosolids and soils for several years (Sidhu, Chapter 4; Walters et al., 2010; Girardi et al., 2011; Gottschall et al., 2012). However, studies of biosolids borne CIP and AZ persistence often estimate half lives based on chemical disappearance, rather than actual confirmation of metabolism. The disappearance is often due to formation of non extractable (non bioaccessible) residues (Walters et al., 2010 ; Girardi et al., 2011 ) rather than degradation. Literature, based o n chemical disappearance estimates maximum average half lives of ~ 5 y (CIP) and ~ 3 y (AZ) (Walters et al., 2010). Herein, CIP was persistent, but a half life of ~ 3 y was estimated for AZ in a n incubation study (Sidhu, Chapter 4). Based on the literatur e and USEPA guidelines (USEPA, 2003), chemical half lives of 10 y a re selected as high end conservative values. A diet correction factor was used: For certain HEIs (e.g. predators) a correction factor i biosolids impacted prey. Values for prey (i.e., worms) as a fraction of predator diet

PAGE 179

179 (DF w ) are listed in Table 6 5. High end conservative diet correction factors a re used based on the exposure data from USEPA (USEPA, 1993; USEPA, 2003). Third tier assessment Third tier assessment refin es the following assumption: Chemical bioavailability reduced to 30%: Literature and data generated herein strongly support minimal bioaccessibility (an d bioavailability) of biosolids borne CIP and AZ (Sidhu, Chapter s 2 3, 4, and 5 ; Nowara et al., 1997; Ericson, 2007; Girardi et al., 2011; Goulas et al., 2016). The greatest reported bioaccessibility of biosolids borne Starnes, 2016). The screening level risk assessment assume s perfectly reversible sorption, whereas a plethora of literature suggests CIP and AZ sorption is highly hysteric (e.g., Sidhu, Chapter 2; Nowara et al., 1997; Wu et al., 2013; Chen et al., 2015; B erhane et al., 2016). Inferring chemical bioavailability from bioaccessibility data can be difficult because bioavailability is ultimately defined by the exposed organism. Bioaccessibility and plant, microbial, and earthworm response data generated in thi s dissertation are consistent with the notion of minimal bioavailability of biosolids borne CIP and AZ. In the one instance where a statistically sound response was measured (microbial response), the response moderately correlated with methanol:water extra ction, i.e., slowly bioaccessible (ECETOC, 2013) fraction of biosolids borne chemicals (Sidhu, Chapter 4). B ased on the freely bioaccessible (CaCl 2 extractable) plus the slowly bioaccessible concentrations of biosolids borne CIP and AZ, 30% bioavailability of the target chemicals seems a better estimate than the arbitrary value of 50%.

PAGE 180

180 Table 6 2. Azithromycin toxicity data Study Species Description Endpoint Results Reference Chronic studies Rats/mice Developmental or Reproductive Toxicity, up to 63 d. NOEC 2 0 mg/kg d US National Library of Medicine, 2002; Mayne et al., 1996 A cute Daphnia magna (planktonic crustacean) Immobilization, 48 h EC50 100 mg/L Minguez et al., 2014 EC50 120 mg/L Vestel et al., 2016 Chronic Daphnia magna Chronic effects on reproduction, growth, and survival NOEC 0.0044 mg/L Vestel et al., 2016 Genotoxicity/ Mutagenicity/ Carcinogenicity/ Teratogenicity Ames Salmonella typhimurium strains TA1535, TA1537, TA98 and TA100; d ogs; rats Not mutagenic, genotoxic, or carcinogenic US National Library of Medicine, 2002 Acute Raphidocelis subcapitata (microalgae) Growth inhibition (cell density) 72 h EC50 0.5 mg/L Minguez et al., 2014 Growth inhibition (cell density) 96 h 0.019 mg/L Harada et al., 2008 Acute Fish (Rainbow trout) ( Oncorhynchus mykiss ) Growth, survival, and functioning EC50 84 mg/L Vestel et al., 2016 Chronic Fish (Rainbow trout) ( Oncorhynchus mykiss ) Growth, survival, functioning, and reproduction NOEC 4.6 mg/L Vestel et al., 2016 Chronic Soil microbes Overall respiration and gene function, 90 d NOEC 3.2 mg/kg Sidhu, Chapter 4 Chronic Worms ( Eisenia foetida ) Growth, survival, functioning, 28 d NOEC 0.16 mg/kg Sidhu, Chapter 5 Chronic Microcystis aeruginosa (cyanobacteria) Growth and survival (cell density) NOEC 0.00019 mg/kg Vestel et al., 2016 Chronic Green algae (Chlorophyta) Growth and survival (cell density) NOEC 0.0018 mg/kg Vestel et al., 2016 Acute Xenopus laevis (South African clawed frog) Mortality and embryo malformation, 96 h NOEC 10 mg/L Harada et al., 2008

PAGE 181

181 Table 6 3. Ciprofloxacin toxicity data Study Species Description Endpoint Results Reference Chronic studies Rats/mice Join bearing, histopathological effects, cardiotoxicity, reproduction, development NOEC 20 mg/kg d US National Library of Medicine, 2002 Acute Buteo jamaicensis (Red tailed hawks) Blood disposition and pharmacokinetics, > 12 h (observation end time not provided) NOEC 50 mg/kg Isaza et al., 1993 Acute Daphnia magna Immobilization, 48 h NOEC 60 mg/L Halling Srensen et al., 2000 EC50 10 mg/L Robinson et al., 2005 Chronic Daphnia magna Chronic effects on reproduction, growth, and survival NOEC 12.8 mg/L Martins et al., 2012 Acute Danio rerio (zebrafish) Reproduction, immobilization, growth, and survival, 72 h NOEC 100 mg/L Halling Srensen et al., 2000 Genotoxicity/ Mutagenicity/ Carcingogenicity / Teratogenicity Ames Salmonella typhimurium test strains TA102; d ogs; rats Probably mutagenic to bacteria Mamber et al., 1993 Negative for rats, dogs, and humans US National Library of Medicine, 2002; Simon et al., 2013 Acute Raphidocelis subcapitata Growth inhibition (cell density) up to 96 h EC50 2.97 mg/L Halling Srensen et al., 2000 18.7 mg/L Robinson et al., 2005 4.83 Martins et al., 2012

PAGE 182

182 Table 6 3. Continued Study Species Description Endpoint Results Reference Acute Lemna Minor (Duckweed) Growth inhibtion (no. of fronds) 7 EC50 0.2 mg/L Robinson et al., 2005 Chronic Lemna Minor (Duckweed) Reproduction (no. of fronds) 7 d NOEC 0.05 mg/L Martins et al., 2012 Chronic Fish (Fathead minnow) ( Pimephales promelas ) Growth and larvae survival, 7 d NOEC 10 mg/L Robinson et al., 2005 Chronic Vibrio fischeri Luminescence and growth NOEC 0.3 mg/L Vasconcelos et al., 2009 Chronic Soil biota Overall respiration and gene function, 90 d NOEC 36.1 mg/kg Sidhu, Chapter 4 Chronic Worms ( Eisenia foetida ) Growth, survival, functioning, 28 d NOEC 1.8 mg/kg Sidhu, Chapter 5 Acute Microcystis aeruginosa Growth inhibition (cell density), up to 120 h EC50 0.005 mg/L Halling Srensen et al., 2000 0.017 mg/L Robinson et al., 2005 Acute Potamopyrgus Antipodarum (Snail) Body size, shape, development and growth rates, and antioxidant enzymes, 96 h NOEC 0.01 mg/kg Peltzer et al., 2017

PAGE 183

183 Table 6 4. Selected C IP and AZ references dose (RfD) (mg/kg d ) for various HEI groups. HEI RfD CIP AZ Human (mg/kg d ) 0.004 0.0 67 Birds (mg/kg d ) 1.25 0. 5 Mammals (mg/kg d ) 2.5 2.5 Soil organisms (mg/kg soil)* 1.8 0.16 Soil microbes (mg/kg soil)** 36.1 3.2 Aquatic organisms ( Microcystis aeruginosa ) (mg/L) 0.0005*** 0.00019 *Soil organism RfD i s based on negligible response of Eisenia fetida to greatest chemical concentrations tested in biosolids amended soils (Sidhu, Chapter 5) *Soil microbial RfD i s based on minimal response of overall microbial health to greatest concentrations tested in biosolids (Sidhu, Chapter 4). *** NOEC for Microcyst is aeruginosa exposure to CIP i s estimated using an uncertainty factor of 10 to convert acute toxicity (Table 6 3) to chronic toxicity Surface Water Chemical Concentration Calculations Ciprofloxacin and AZ are frequently detected TOrCs in surface waters, particularly in waters close to WWTPs and/or wastewater discharge area (Hughes et al., 2013; Petrie et al., 2015). Chemical concentrations up to 0.002 mg AZ/L and 6.5 mg CIP/L have been detected in rivers and ponds, but the median values are ~0.0002 mg AZ/L and 0.16 mg CIP/L (Hughes et al., 2013). Even the median CIP and AZ values represent potential for risks to aquatic organisms (e.g., 0.0002 mg AZ/L water results in a HQ greater than 1 for cyanobacteria). However, the majority of CIP and AZ in surface water bodies is contributed by release of contaminated wastewater effluents (Hughes et al., 2013; Petrie et al., 2015) and are not applicable to biosolids borne chemicals.

PAGE 184

184 Extensive and st rong retention and limited desorption (Sidhu, Chapter 2; Nowara et al., 1997; Ericson, 2007) of biosolids borne CIP and AZ limit chemical release into the surface water bodies from biosolids runoff/erosion events. Similarly, the limited mobility restricts vertical movement of the chemicals into drainage and groundwater (US National Library of Medicine, 2002; Gottschall et al., 2012; von Hellens, 2015). Measured data pertaining to biosolids borne CIP and AZ contamination of surfa ce waters are lacking but im portant for sound risk assessment purposes. Therefore, concentrations of biosolids borne CIP and AZ in surface waters under a worst case scenario a re estimated herein using a series of equations from the Part 503 Biosolids Rule making risk assessment. The equations account for sedimentation from biosolids amended fields and CIP/AZ (reversible) partitioning into the aqueous phase from the solid phase. The concentrations of CIP and AZ in soil eroding from amended land (C sma ) are calculated from the equation: C sma = (P a f ero ) / ME sms CF where: P a = annual application rate of CIP or AZ (mg/ha y), f ero = fraction of total CIP or AZ loss by erosion, ME sma = estimated rate of soil loss for the SMA, SMA = biosolids management area, CF = unit conversion factor. Assuming 100% of the eroded material to be biosolids that contains 36.1 mg CIP/kg or 3.2 mg AZ/kg, the C sma values a re 36.1 mg CIP/kg or 3.2 mg/kg.

PAGE 185

185 A dilution factor (DF) accounts for the dilution of the eroded biosolids with non contaminated soil prior to entering a surface water body. The DF is calculated from the following equation: DF = (A sma S sma ) / [(A sma S sma ) + ((A ws A sma ) S ws )] where: A sma = area affected by land applicat ion of biosolids (SMA), S sma = sediment delivery ratio for the SMA, A ws = area of the watershed (ha), S ws = sediment delivery ratio for the watershed. The USEPA Part 503 risk assessment (1995) assumed the following conservative values: 1074 ha (A sma ), 0.4 6 (S sma ), 440,300 ha (A ws ), and 0.17 (S ws ), resulting in a DF value of 0.0066. The amount of CIP and AZ in sediment entering surface water (C sed ) i s calculated by multiplying C sma by the DF; the resulting C sed values a re 0.24 mg CIP/kg and 0.02 mg AZ/kg. T he surface water concentration (C sw ) resulting from the calculated C sed values i s estimated using the following equation: C sw = C sed / [KD sw where: KD sw = partition coefficient between solids and liquids within the s tream (assuming reversible partitioning), Pl = percent liquid in the water column, Ps = percent solids in the water column,

PAGE 186

186 The USEPA Part 503 risk assessment used a value of 62,500 as a ratio o f Pl to Ps, and assumed the surface water density to be 1 kg/L. Using the KD sw values of 3 60 L/kg (CIP) and 430 L/kg (AZ) ( Sidhu, Chapter 2; because the compounds are coming from biosolids sw i s estimated to be 3.8 x 10 6 mg CIP per L and 3.2 x 10 7 mg AZ per L. Data U sed in R isk A ssessment Description of data, along with calculations and assumptions involved, employed in the IRA are listed in Table 6 5. Results First Tier (Screening Level) Risk Assessment Tables 6 6 and 6 7 list the HQ values obtained for HEIs in each of the 16 pathways. Based on the screening assessment, the following pathways and receptors of concern were chosen for subsequent tier assessment: Pathways of concern for AZ 9. Biosolids soil soil organism ; 10. Biosolids soil soil organism predator Pathways of concern for CIP 8. Biosolids soil plant 9. Biosolids soil soil organism 10. Biosolids soil soil organism predator

PAGE 187

187 Table 6 5. Parameters and assumptions for calculating screening level hazard quotient (HQ) values. Abbreviation Parameter definition Value Assumptions/ Explanation Pathway the parameter was used in Reference (AZ) (CIP) AC Lifetime inhalation of chemical in the air in vicinity of biosolids amended soil. (mg) BAR 1 = 1.9 x 10 19 BAR 2 = 4 .0 x 10 19 BAR 3 = 1.5 x 10 1 8 BAR 1 = 2.6 x 10 11 BAR 2 = 6.2 x 10 11 BAR 3 = 2.2 x 10 10 AC = CAR 70 y IR 365 d 13 CAR calculated from vapor pressure and constant estimates BAF Bioaccumulation factor (BAF) is the ratio of chemical concentration in an organism divided by the chemical concentration in the surrounding (growth) media, and encompasses chemical uptake by all possible routes (Suter, 2007) BAFp = 0. 02 (plant, w .w.) BAF ww = 1.4 (worm, w.w.) BAFww = 4 (un depurated worms) BAFdw = 7 (worm, d.w.) BAF m = 7 (mammals and poultry) BAFp = 0.0 02 (plant, w .w.) BAFww = 0.8 (worm, w.w.) BAF w w = 4 (un depurated worms) BAFdw = 4 (worm, d.w.) BAF m = 4 (mammals and poultry) Greatest BAF p (wet weight, plants) and BAF dw (dry weight, worms) obtained used. Conservatively, BAF m assumed to be same as that for worms (d.w ). 1, 2, 4, 6 5, 10 5 Sidhu Chapters 3 and 5. BAR 1 Biosolids application rate One time land application of biosolids containing 95 th percentile chemical concentration at 10 0 Mg/ha (d.w.) Worst case scenario; one time, high dose application of severely contaminated biosolids. 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 14 USEPA, 2003 BAR 2 40 y of lan d application of biosolids containing average chemical concentration at 2 0 Mg/ha y (d.w.) Typical biosolids; applied annually for a 40 y high end application. BAR 3 40 y of land application of biosolids containing 95 th percentile chemical concentration at 2 0 Mg/ha y (d.w.) Severely contaminated biosolids; applied annually for a 40 y high end application.

PAGE 188

188 Table 6 5 continued Abbreviation Parameter definition Value Assumptions/ Explanation Pathway the parameter was used in Reference (AZ) (CIP) BCF Bioconcentration factor (BCF) is the ratio of chemical concentration in an organism divided by the chemical concentration in water, and encompasses only direct uptake from solution (Suter, 2007) BCF fs = 1.2 (whole fish, w.w.) BCF fs = 1.0 (whole fish, w.w.) Whole fish BCF values used for animal and human consumption calculations (entire organism consumed). 12, 15 12, 15 USEPA EPI Suite, 2011 BW Body weight (live weight) (kg) Adult: 70 Child: 15 American woodcock: 0.19 Coopers Hawk: 0.405 Cow: 590 Deer Mouse: 0.0196 Osprey: 1.6 Red Fox: 4.53 River otter: 8.1 Red tailed Hawk: 1.13 Short tailed Shrew: 0.015 Mean Mean Mean Mean Mean Mean Mean Mean Mean Mean Mean 1, 2, 3, 4, 5, 6, 7, 10, 11, 12, 13, 14, 15 Sample et al., 1997; USEPA, 1993; USEPA, 2003; USEPA, 2008; USEPA 2011 CA Chemical concentration in consumed animal (mg/kg) CA ani.pl (from plant consumption) CA ani.so (from soil consumption) CA fish (whole fish; w.w.) BAF m PC BAF m SC FS BCF fs SWC 4 5 12, 15 USEPA, 1995

PAGE 189

189 Table 6 5. Continued. Abbreviation Parameter definition Value Assumptions/ Explanation Pathway the parameter was used in Reference (AZ) (CIP) CAR Chemical concentration in air (mg/L) BAR 1 = 3.4 x 10 28 BAR 2 = 7.4 x 10 28 BAR 3 = 2.8 x 10 27 BAR 1 = 5.0 x 10 20 BAR 2 = 1.2 x 10 19 BAR 3 = 4 .0 x 10 19 Calculated from vapor constant using following equations: pchem = H[A]; where pChem = partial pressure of the chemical, H = chemical concentration in soil. Ideal gas equation was used to calculate air concentration in moles/L: 11, 13 11, 13 Vapor pressure and H estimated using USEPA EPI Suite, 2011 CB Chemical concentration in biosolids (mg/kg) (d.w.) BAR 1 = 3.2 BAR 2 = 0.83 BAR 3 = 3.2 BAR 1 = 36.1 BAR 2 = 10.5 BAR 3 = 36.1 95 th percentile (BAR 1 and BAR 3 ) or average (BAR 2 ) concentration in 2009 Targeted National Sewage Sludge Survey 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 14 USEPA, 2009 CF Correction factor for unit conversions DC Dust concentration in air (mg/L) 0.010 Concentration used in Part 503 biosolids risk assessment 11 USEPA, 1995 DF w Worms as diet fraction Deer Mouse: 0.375 Red fox: 0.7 Coopers Hawk: 1 Red tailed Hawk: 1 Short tailed Shrew: 0.81 American Woodcock: 0.94 High end values used by USEPA for risk assessment 10 USEPA, 2003

PAGE 190

190 Table 6 5. Continued Abbreviation Parameter definition Value Assumptions/ Explanation Pathway the parameter was used in Reference (AZ) (CIP) EC d Chemical concentration in earthworm (mg/kg) BAR 1 = 1.12 BAR 2 = 2.38 BAR 3 = 8.96 BAR 1 = 7.2 BAR 2 = 16.8 BAR 3 = 57.6 BAFdw SC 9, 10 9, 10 Sidhu, Chapter 5 FC Fish consumption (kg WW/ d ) Adult: 0.026 (all ages) High end values used by USEPA for risk assessment 12 USEPA, 2011 FI Food ingestion rate (kg WW /d) American woodcock: 0. 12 Coopers Hawk: 0.213 Cow: 2 5 Deer Mouse: 0.0177 Osprey: 0.31 Red Fox: 1.55 River otter: 1.5 Red tailed Hawk: 0.415 Short tailed Shrew: 0.0142 High end values used by USEPA for risk assessment 6, 7, 10, 15 USEPA, 1993; USEPA, 2003 FS Soil fraction of diet American woodcock: 0.104 Coopers Hawk: 0.01 Cow: 0.025 Deer Mouse: 0.02 Red Fox: 0.028 Red tailed Hawk: 0.01 Short tailed Shrew: 0.01 High end values used by USEPA for risk assessment 7 7 7 USEPA, 1993; USEPA, 1995; USEPA, 2003 FVC Combined fruit and vegetable consumption (kg WW/kg BW/ d ) Adult: 0.023 Child: 0.025 High end values used by USEPA for risk assessment 1, 2 USEPA, 2011 HFS Hectare furrow slice mass 2.2 x 10 6 kg Soil bulk density = 1.3 g/cm 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 14 Weil and Brady, 2016

PAGE 191

191 Table 6 5. Continued. Abbreviation Parameter definition Value Assumptions/ Explanation Pathway the parameter was used in Reference (AZ) (CIP) IR Inhalation rate (m 3 / d ) Adult: 21 Child: 25 High end values used by USEPA for risk assessment 11 11 USEPA, 2011 MC Meat consumption (kg WW /kg BW d) Adult: 0.005 Child: 0.01 High end values used by USEPA for risk assessment 4, 5 USEPA, 2008 USEPA, 2011 PC Chemical concentration in plant tissue (mg/kg) ( w .w.) BAR 1 = 0.0 032 BAR 2 = 0.0 068 BAR 3 = 0. 026 BAR 1 = 0.0 036 BAR 2 = 0.0 084 BAR 3 = 0. 029 PC = BAF p SC 1, 2, 4, 6 Sidhu, Chapter 4 RfD Reference dose: lifetime daily intake dose of a chemical that is without a considerable risk over the entire lifespan (mg/kg d ), (mg/kg soil), or (mg/L water) American Woodcock: 0.5 Coopers Hawk: 0. 5 Red Tailed Hawk: 0. 5 Short Tailed Shrew: 2.5 Terrestrial Plants: 3.2 Deer Mouse: 2.5 Red Fox: 2.5 Soil Biota: 3.2 Aquatic Community: 0.00019 Fish: 4.6 American Woodcock: 1.25 Coopers Hawk: 1.25 Red Tailed Hawk: 1.25 Short Tailed Shrew: 2.5 Terrestrial Plants: 1.1 Deer Mouse: 2.5 Red Fox: 2.5 Soil Biota: 36.1 Aquatic Community: 0.0005 Fish: 10 1, 2, 3, 4, 5, 7, 11, 14 6, 10 See RfD calculations SC Chemical concentration in soil (mg/kg) (d.w.) BAR 1 = 0. 16 BAR 2 = 0. 34 BAR 3 = 1.28 BAR 1 = 1.8 BAR 2 = 4.2 BAR 3 = 14.4 Assuming one time or cumulative (over 40 y) biosolids application and mixing with upper 0.15 m soil weighing 2000 Mg/ha. USEPA, 2009

PAGE 192

192 Ta ble 6 5. Continued Abbreviation Parameter definition Value Assumptions/Explanation Pathway the parameter was used in Reference (AZ) (CIP) SI Soil ingestion (kg/d) Adult: 0.00005 Child: 0.0002 Pica child: 0.001 High end values used by USEPA for risk assessment 3 USEPA, 2011 SWC Surface water concentration (mg/L) 3.2 x 10 7 3.8 x 10 6 Based on equations and assumptions used by USEPA 12, 15, 16 See Surface water concentration calculations. WI Water intake (consumption) (L/d) Adult: 3.1 Child: 1.1 American woodcock: 0.019 Coopers Hawk: 0.0322 Cow: 5 Deer Mouse: 0.00288 Osprey: 0.07 Red Fox: 0.386 River otter: 0.79 Red tailed Hawk: 0.0641 Short tailed Shrew: 0.00354 High end values used by USEPA for risk assessment 12, 14 15 USEPA, 1993; USEPA, 2003; USEPA, 2011 WI sw Water intake while swimming (L/d) Adult: 0.071 Child: 0.12 Assuming an hour of swimming each d ay 12 USEPA, 2011

PAGE 193

193 Table 6 6. Azithromycin hazard quotient (HQ) values, and equations and assumptions to calculate HQ values. Pathway Hazard Quotient Equation BAR 1 BAR 2 BAR 3 Comments/ assumptions 1. and 2. Biosolids soil plant human Adult: 0. 00 1 Child: 0 00 1 Adult: 0. 0 02 Child: 0. 0 02 Adult: 0. 0 08 Child: 0.0 1 All diet involves 100% of produce grown in biosolids amended soils. 3. Biosolids soil human Adult: 1. 7 x 10 6 Child: 7.2 x 10 5 Pica child: 0.0 002 Adult: 3.6 x 10 6 Child: 6.8 x 10 5 Pica child: 0.0 003 Adult: 1.3 x 10 5 Child: 0.0 003 Pica Child: 0. 001 4. Biosolids soil plant animal human Adult: 3.4 x 10 6 Child: 6.8 x 10 6 Adult: 3.6 x 10 6 Child: 1.5 x 10 5 Adult: 2.6 x 10 5 Child: 0.00 06 100% of animal diet consists of plants grown in biosolids amended soil and the animal makes up 100% of human diet. 5. Biosolids soil animal human Adult: 3.0 x 10 5 Child: 5.9 x 10 5 Adult: 6.3 x 10 5 Child: 0.0001 Adult: 0.0 002 Child: 0.0 005 100% of human diet is the animal. 6. Biosolids soil plant animal / bird Cow: 3.6 x 10 5 Deer Mouse: 0.00 1 Red fox: 0.00 06 Coopers Hawk: 0.0 02 Red tailed Hawk: 0.0 02 Short tailed Shrew: 0.0 06 American Woodcock: 0.0 02 Cow: 0.000 08 Deer Mouse: 0.00 1 Red fox: 0.00 08 Coopers Hawk: 0.0 06 Red tailed Hawk: 0.0 06 Short tailed Shrew: 0.0 1 American Woodcock: 0.0 08 Cow: 0.00 02 Deer Mouse: 0.0 08 Red fox: 0.0 02 Coopers Hawk: 0. 02 Red tailed Hawk: 0.0 1 Short tailed Shrew: 0. 06 American Woodcock: 0. 02 100% of diet consists of plants grow n in biosolids amended soil

PAGE 194

194 T able 6 6. Continued Pathway Hazard Quotient Equation BAR 1 BAR 2 BAR 3 Comments/ assumptions 7. Biosolid s soil animal / bird Cow: 4.4 x 10 4 Deer Mouse: 0.00 1 Red fox: 0.000 4 Coopers Hawk: 0.00 1 Red tailed Hawk: 0.00 1 Short tailed Shrew: 0.000 5 American Woodcock: 0.0 1 Cow: 9.6 x 10 4 Deer Mouse: 0.00 3 Red fox: 0.00 1 Coopers Hawk: 0.00 3 Red tailed Hawk: 0.00 3 Short tailed Shrew: 0.00 1 American Woodcock: 0.0 3 Cow: 0.000 1 Deer Mouse: 0.0 0 7 Red fox: 0.00 4 Coopers Hawk: 0.0 1 Red tailed Hawk: 0.0 0 7 Short tailed Shrew: 0.00 4 American Woodcock: 0. 1 5 8. Biosolids soil plant 0.0 6 0. 1 0. 4 0 9. Biosolids soil soil organism Eisenia fetida : 0 .99 Soil microbes: 0.0 6 Eisenia fetida : 2.1 Soil microbes: 0. 10 Eisenia fetida : 8 Soil microbes: 0. 4 0 RfD based on the greatest dose tested (Sidhu, Chapters 4 and 5) 10. Biosolids soil soil organism predator Deer Mouse: 0. 43 Red fox: 0. 1 7 Coopers Hawk: 1.1 Red tailed Hawk: 0. 7 8 Short tailed Shrew: 0. 41 American Woodcock: 1. 4 Deer Mouse: 0. 84 Red fox: 0. 37 Coopers Hawk: 2. 4 Red tailed Hawk: 1. 7 Short tailed Shrew: 0. 8 7 American Woodcock: 2. 9 Deer Mouse: 3.2 Red fox: 1. 2 Coopers Hawk: 9.4 Red tailed Hawk: 6.6 Short tailed Shrew: 3. 4 American Woodcock: 1 1 100% of diet consists of earthworms growing in biosolids amended soil 11. Biosolids soil airborne dust human Adult: ~ 10 24 Child: ~ 10 24 Adult: ~ 10 24 Child: ~ 10 24 Adult: ~ 10 23 Child: ~ 10 23 Exposed to maximum concentration of biosolids dusts for 24 h/d

PAGE 195

195 Table 6 6. Continued Pathway Hazard Quotient Equation BAR 1 BAR 2 BAR 3 Comments/ assumptions 12. Biosolids soil surfac e water human Adult: 1.3 x 10 5 Child: 0.0002 100% of water intake consists of contaminated surface water. 100% of fish consumption consists of fish caught in contaminated surface water. 1 h/d of swimming activity. 13. Biosolid s soil air human ~ 10 15 ~ 10 15 ~ 10 15 Life of 100% air inhalation is from vicinity of biosolids amended soils. 14. Biosolids soil groundwater human Not important pathway. Demonstration via rudimentary RA = Adult: 0.00 6 Child: 0.00 8 Adult: 0.0 1 Child: 0.0 2 Adult: 0.0 4 Child: 0.0 6 Kd of biosolids = 430 L/kg Kd of sand = 10 L/kg (Sidhu, Chapter 2) Concentration reaching groundwater, GC = SC/ (431*11) All chemical not sorbed, assumed to reach groundwater and no dilution of the chemical. 15. Biosolids soil surfac e water animal Osprey: 2.0 x 10 7 River otter: 1.7 x 10 6 100% of diet consists of fish in contaminated surface water. 100% water intake is from contaminated surface water. 16. Biosolids soil surfac e water aquatic organism Microcystis aeruginosa : 0.002 Microcystis aeruginosa used as the most sensitive species, NOEC = 0.0002 mg/L. Lifetime exposure to contaminated surface water assumed.

PAGE 196

196 Table 6 7. Ciprofloxacin hazard quotient (HQ) values, and equations and assumptions to calculate HQ values. Pathway Hazard Quotient Equation BAR 1 BAR 2 BAR 3 Comments/ assumptions 1. and 2. Biosolids soil plant human Adult: 0. 02 Child: 0 02 Adult: 0. 05 Child: 0. 05 Adult: 0. 17 Child: 0. 18 All diet involves 100% of produce grown in biosolids amended soils. 3. Biosolids soil human Adult: 0.000 4 Child: 0.00 6 Pica child: 0.0 4 Adult: 0.000 8 Child: 0.0 1 Pica child: 0.0 8 Adult: 0.00 2 Child: 0.0 4 Pica Child: 0. 24 4. Biosolids soil plant animal human Adult: 0.0 2 Child: 0. 04 Adult: 0. 04 Child: 0. 08 Adult: 0. 15 Child: 0.28 100% of animal diet consists of plants grown in biosolids amended soil and the animal makes up 100% of human diet. 5. Biosolids soil animal human Adult: 0. 005 Child: 0. 01 Adult: 0. 01 Child: 0. 03 Adult: 0. 05 Child: 0.09 100% of human diet is the animal. 6. Biosolids soil plant animal / bird Cow: 0.000 06 Deer Mouse: 0.0 0 1 Red fox: 0.00 06 Coopers Hawk: 0.0 02 Red tailed Hawk: 0.0 0 1 Short tailed Shrew: 0.0 02 American Woodcock: 0.0 02 Cow: 0.000 2 Deer Mouse: 0.0 04 Red fox: 0.0 0 1 Coopers Hawk: 0.0 04 Red tailed Hawk: 0.0 02 Short tailed Shrew: 0.0 06 American Woodcock: 0.0 04 Cow: 0.00 06 Deer Mouse: 0. 0 1 Red fox: 0.0 04 Coopers Hawk: 0. 0 1 Red tailed Hawk: 0.0 08 Short tailed Shrew: 0. 02 American Woodcock: 0. 0 2 100% of diet consists of plants growing in biosolids amended soil

PAGE 197

197 Table 6 7. Continued. Pathway Hazard Quotient Equation BAR 1 BAR 2 BAR 3 Comments/ assumptions 7. Biosolid s soil animal / bird Cow: 0.000 8 Deer Mouse: 0.0 1 Red fox: 0.0 05 Coopers Hawk: 0.0 1 Red tailed Hawk: 0.0 05 Short tailed Shrew: 0.0 05 American Woodcock: 0. 1 0 Cow: 0.00 1 Deer Mouse: 0.0 3 Red fox: 0.0 2 Coopers Hawk: 0.0 2 Red tailed Hawk: 0.0 1 Short tailed Shrew: 0.0 2 American Woodcock: 0. 2 Cow: 0.00 6 Deer Mouse: 0. 1 0 Red fox: 0. 06 Coopers Hawk: 0. 06 Red tailed Hawk: 0.0 4 Short tailed Shrew: 0. 06 American Woodcock: 0.80 8. Biosolids soil plant 1.6 2. 3 13 9. Biosolids soil soil organism Eisenia fetida : 0. 99 Soil microbes: 0.03 Eisenia fetida : 2.1 Soil microbes: 0. 12 Eisenia fetida : 8 Soil microbes: 0. 4 0 RfD based on the greatest dose tested (Sidhu, Chapters 4 and 5) 10. Biosolids soil soil organism predator Deer Mouse: 2.6 Red fox: 0.99 Coopers Hawk: 3 .0 Red tailed Hawk: 2 .1 Short tailed Shrew: 2.7 American Woodcock: 3.6 Deer Mouse: 6.0 Red fox: 2.3 Coopers Hawk: 7.1 Red tailed Hawk: 4.9 Short tailed Shrew: 6.4 American Woodcock: 8.5 Deer Mouse: 21 Red fox: 7.9 Coopers H awk: 24 Red tailed Hawk: 17 Short tailed Shrew: 22 American Woodcock: 29 100% of diet consists of earthworms growing in biosolids amended soil 11. Biosolids soil airborne dust human Adult: 2.6 x 10 15 Child: 2.0 x 10 16 Adult: 6 .0 x 10 15 Child: 4.8 x 10 16 Adult: 2 .0 x 10 14 Child: 1.6 x 10 1 5 Exposed to maximum concentration of biosolids dusts for 24 h/d

PAGE 198

198 T able 6 7. Continued. Pathway Hazard Quotient Equation BAR 1 BAR 2 BAR 3 Comments/ assumptions 12. Biosolids soil surface water human Adult: 5.6 x 10 5 Child: 0.0001 100% of water intake consists of contaminated surface water. 100% of fish consumption consists of fish caught i n contaminated surface water. 1 h/d of swimming activity. 13. Biosolid s soil air human 6.6 x 10 9 1.5 x 10 8 5.4 x 10 8 Lifetime of 100% air inhalation is from vicinity of biosolids amended soils. 14. Biosolids soil groundwater human Not important pathway. Demonstration via rudimentary RA = Adult: 0.00 1 Child: 0.00 2 Adult: 0.00 4 Child: 0.00 6 Adult: 0.0 1 Child: 0.0 2 Kd of biosolids = 360 L/kg Kd of sand = 37 L/kg (Sidhu, Chapter 2) Concentration reaching groundwater, GC = SC/ (361*38) All chemical not sorbed, assumed to reach groundwater and no dilution of the chemical. 15. Biosolids soil surface water animal Osprey: 1.4 x 10 6 River otter: 1. 7 x 10 6 100% of diet consists of fish in contaminated surface water. 100% water intake is from contaminated surface water. 16. Biosolids soil Surface water aquatic organism Microcystis aeruginosa : 0.008 Microcystis aeruginosa used as the most sensitive species, NOEC = 0.0005 mg/L. Lifetime exposure to contaminated surface water assumed.

PAGE 199

199 Second Tier Assessment P athways 8 (CIP), 9 (for both CIP and AZ), and 10 (for both CIP and AZ) remained of concern even after bioavailability (50%) and half life (10 y) adjustments and adjusting predator diet fractions (Tables 6 8 and 6 9). After second tier adjustments to the I RA the HQ values for AZ in earthworms and predators were < 1 in the BAR 1 (one time heavy application of severely contaminated biosolids) and the BAR 2 (40 y land application of typical biosolids annually at agronomic rate ) scenarios T he HQ values for AZ for predators were greater than 1 only in the BAR 3 (40 y land application of severely contaminated biosolids annually at agronomic rate) scenario and only for Coppers Hawk (1. 7 ) and American Woodcock ( 2 .1) (Tables 6 8 and 6 9). T he HQ values for CIP fo r terrestrial plants (2.5) and earthworms (1.3) were greater than 1 only in BAR 3 scenario (Table 6 8) The HQ values for predator s ranged from less than 1 to less than 2 in the BAR 1 and BAR 2 scenarios (Table 6 9). Only i n the unrealistic BAR 3 scenario, wer e the HQ values for some predators greater than 2; i.e., for Coopers Hawk ( 3.8 ), Red tailed Hawk ( 2.1 ), Short tailed Shrew ( 4.2 ), and American Woodcock ( 3.5 ) (Table 6 9). Third Tier Assessment Assuming a 30% chemical bioavailability, reduced all the HQ values for all the pathways to less than 1 in the BAR 1 and BAR 2 scenarios, indicating negligible risks under real world biosolids management practices. The HQ values for AZ were only marginally greater than 1 and only for American Woodcock (1.2) in the BAR 3 scenario (Table 6 10) Similarly, the HQ values for CIP remained greater than 1 only for plants

PAGE 200

200 (1.5), Coopers Hawk (2.3), Red tailed Hawk (1.3), Short tailed Shrew (2.1), and American Woodcock (2.1) in the BAR 3 scenario (Table 6 10). Table 6 8. Second tier risk assessment hazard quotient (HQ) values for plants (pathway 8) and earthworms (pathway 9). Organism HQ value AZ HQ value CIP Terrestrial plants BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 BAR 1 = <1 BAR 2 = <1 BAR 3 = 2.5 Earthworms BAR 1 = <1 BAR 2 = <1 BAR 3 = 1.3 BAR 1 = <1 BAR 2 = <1 BAR 3 = 1.3 Table 6 9 Second tier risk assessment hazard quotient (HQ) values for predators (pathway 10) Predator HQ value AZ HQ value CIP Deer Mouse BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 Red fox BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 Coopers Hawk BAR 1 = <1 BAR 2 = <1 BAR 3 = 1.7 BAR 1 = 1.5 BAR 2 = 1.3 BAR 3 = 3.8 Red tailed Hawk BAR 1 = <1 BAR 2 = <1 BAR 3 = 1.0 BAR 1 = 1.1 BAR 2 = <1 BAR 3 = 2.1 Short tailed Shrew BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 BAR 1 = 1.1 BAR 2 = <1 BAR 3 = 3.6 American Woodcock BAR 1 = <1 BAR 2 = <1 BAR 3 = 2.1 BAR 1 = 1.7 BAR 2 = 1.1 BAR 3 = 3.5

PAGE 201

201 Table 6 10. Third tier risk assessment hazard quotient (HQ) values Predator HQ value AZ HQ value CIP Terrestrial plants BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 BAR 1 = <1 BAR 2 = <1 BAR 3 = 1.5 Deer Mouse BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 Red fox BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 Coopers Hawk BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 BAR 1 = <1 BAR 2 = <1 BAR 3 = 2.3 Red tailed Hawk BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 BAR 1 = <1 BAR 2 = <1 BAR 3 = 1.3 Short tailed Shrew BAR 1 = <1 BAR 2 = <1 BAR 3 = <1 BAR 1 = <1 BAR 2 = <1 BAR 3 = 2.1 American Woodcock BAR 1 = <1 BAR 2 = <1 BAR 3 = 1.3 BAR 1 = <1 BAR 2 = <1 BAR 3 = 2.1 Putting t he HQ Values i nto P ersp ective The IRA suggests that biosolids borne AZ and CIP pose negligible human and ecological risk under upper end (i.e., cumulative application at 20 Mg/ha y for 40 y) of realistic (BAR 2 ) and a worst case ( BAR 1 ) scenarios. The HQ values from even the unrealistic BAR 3 scenario are only ~2, suggesting minimal risks. The HQ values for CIP for plants are based on a study where sand was directly spiked with CIP (no biosolids; Sidhu, Chapter 3) Phytotoxicity was obse rved at concentration of ~1 mg CIP per kg soil. However, in the same study, no phytotoxicity was observed from silt loam spiked directly with CIP (no biosolids), even at an

PAGE 202

202 extraordinarily high chemical concentration of 36.1 mg CIP per kg soil (100 fold gr eater than environmentally relevant chemical concentration). Thus, biosolids borne CIP is not likely to be phytotoxic even in BAR 3 scenario. Ciprofloxacin and AZ concentrations in modern USA biosolids appear to mimic median concentrations reported in the USEPA (2009) targeted national sewage sludge survey (Table 1 2; Youngquist et al., 2014; Kennedy/Jenks Consultants, 2015 ; and Starnes, 2016 ). Annual l and application of biosolids containing 95 th percentile concentrations (USEPA, 2009) over long te rm is, therefore, highly unlikely Further, b iosolids are typically land applied every 2 3 years at agronomic rates of 20 Mg/ha or less ( USEPA, 2000), reducing the associated human and ecological health risks even more D epurated earthworm BAF values obtai ned in Chapter 5 a re used in the IRA based on previous risk assessments (USEPA Part 503 Biosolids Rule making risk assessment, 1995; Synder et al., 2013) The BAF values likely over estimate risk s because of several factors First, dry weight BAF values a r e used, which are about 5 times the wet weight values (predators consume worms on wet weight basis). Use of dry weight BAF values and wet weight predator food ingestion rates overestimate risks by 3 5 times and partially compensate s for using depurated (in stead of un depurated) BAF values. Second, t he BAF values were generated under laboratory conditions with freshly spiked biosolids under maximum bioaccessibility conditions whereas a ctual field BAF values are usually less (Sidhu, Chapter 5; Higgins et al., 2010; Pannu et al., 2012; Hoke et al., 2016; van den Brink et al., 2016).

PAGE 203

203 The IRA assume s the same chemical BAF values for higher organisms (mammals and birds) as the earthworms, conservatively. Also, upper percent ile exposure rates (e.g., diet fractions) for predators (USEPA, 1993; USEPA, 2003) a re used in the IRA. Adjusting the exposure rates to mean values further decreases the already small risk potential. A conservative half life value of 10 y i s used for both compounds reducing the CIP half life to ~ 8 y minimizes risks in the BAR 3 scenario. Also, c onsidering that CIP and AZ forms non extractable residues of negligible bioaccessibility over time (Sidhu, Chapter 4; Ericson et al., 2007; Walters et al., 2010; Gir ardi et al., 2011 ; Sabourin et al., 2012; Aristilde and Sposito, 2013; Cui et al., 2014 ) persistence does not equate to bioavailability. Thus, over time (certainly years), the majority of CIP (greater than the 70% used herein) is expected to be non bioava ilable reducing risks even more. The IRA assesse s risk s to HEIs rather than to the general population that have less exposure to biosolids or foods grown on biosolid s amended soils. Further, the risk assessment considers an unlikely scenario of daily lif etime (chronic) exposure to the biosolids borne chemicals ; considering biosolids are applied to ~ 0.1% of total agricultural land in the USA on annual basis (NRC, 2002). Also, t he RfD values for acute exposures are at least 10 fold greater than chronic exp osure s and, thus, the IRA is protective of acute exposures. C ollectively, the IRA indicates negligible human and ecological health risks from biosolids borne CIP and AZ under real world based biosolids use and management practices.

PAGE 204

204 Calculation of B iosolids borne CIP and AZ P ollutant L imits The four types of pollutant limits established by USEPA Part 503 Biosolids Rule (1995) for land applied biosolids are: c umulative pollutant loading rate (CPLR), a nnual pollutant loading rate (APLR), c eiling concen tration limit, and p ollutant concentration limit. Cumulative P ollutant L oading R ate (CPLR) Limit The CPLR represents the maximum cumulative concentration of a chemical in amended soils that remains protective of human and ecological health. The IRA identi fied Coopers Hawk and American Woodcock as the most sensitive species to biosolids borne CIP and AZ respectively Therefore, the CPLR values a re calculated for Coopers Hawk (CIP) and American Woodcock (AZ) because rates protective of the most sensitive sp ecies are also protective of humans and other exposed organisms. The CPLR limits a re calculated by setting the Pathway 10 HQ equation to 1, adding a multiplication factor of 0. 5 ( to account for the assum ed 50 % bioavailability ; USEPA, 1995 ) and solving for chemical concentration in the soil: 1 = ((SC* BAF dw FI/BW) 1 )/(RfD*CF) where: SC = CIP or AZ concentration in soil (mg CIP or AZ per kg amended soil) BAF d w = bioaccumulation factor in worm (d.w.) FI = food ingestion rate of preda tor (kg per d ay d.w. ; ~0.0 25 (American Woodcock) and 0.043 (Coopers Hawk), assuming 100% diet is earthworm that contains 80 % moisture by weight (Sidhu, Chapter 5 ; Suter, 2007 ) ).

PAGE 205

205 The calculated CPLR values are 5.9 mg CIP per kg amended soil (or 12 kg CIP per ha) and 1.1 mg AZ per kg amended soil (or 2.2 kg AZ per ha). Based on the CPLR values, biosolids containing 95 th percentile CIP or AZ can be land applied at the rate of 320 Mg/ha (CIP) or 70 0 Mg/ha (AZ) with minimal human and ecological health risks More typical biosolids (containing median CIP and AZ concentrations) can be land applied at 2 0 Mg/ha y for over 10 0 y (CIP) and 360 y (AZ) assuming no chemical attenuation. Annual P ollutant L oading R ate (APLR) Limit Annual pollutant loading rate (A PLR) applies to commercially sold biosolids and identifies the maximum mass of pollutant that can be land applied in one y ear such that CPLR limits are not reached before 20 y of annual biosolids application. Thus, APLR values are calculated by dividing th e CPLR by 20 (i.e., assuming no chemical loss). The APLR values protective of all terrestrial and aquatic populations are 0. 6 kg CIP/ha y and 0. 11 kg AZ/ha y. Ceiling C oncentration L imit The ceiling concentration limit is the maximum concentration of a pollutant in biosolids that are land applied, and is intended to prevent low quality highly contaminated biosolids from being land applied. A ceiling limit is the greater of 99 th percentile pollutant concentration in biosolids or the biosolids pollutant limit calculated from the risk assessment. The CPLR values from the IRA are 5.9 mg CIP and 1.1 mg AZ per kg amended soil correspond to half life (assumed 10 y) adjusted concentratio ns of ~ 95 mg CIP/kg and ~ 26 mg AZ/kg in the BAR 3 scenario The 99 th percentile chemical concentrations in the targeted national sewage sludge survey (USEPA, 2009) were 79.6

PAGE 206

206 mg CIP and 8.7 mg AZ per kg biosolids. The ceiling concentration limits for biosoli ds borne chemicals are, thus, 95 mg CIP and 26 mg AZ per kg biosolids. Pollutant C oncentration L imit The pollutant concentration limit is the most stringent guideline for biosolids borne pollutants. Biosolids meeting the pollutant concentration limit can b e land applied freely, provided they meet Part 503 Biosolids Rule pathogen and vector control requirements. The pollutant concentration limit considers land application of a cumulative total of 1000 Mg of biosolids over a 100 y period. Thus, the pollutant concentration limit i s calculated by dividing the CPLR (in kg/ha) value by 100 y (assuming an application rate of 10 Mg/ha y). The resultant pollutant concentration limits are 12 mg CIP and 2.2 mg AZ per kg biosolids. Thus, biosolids containing < 12 mg CIP and < 2.2 mg AZ per kg can be land applied freely without having to track annual additions The pollutant concentration limits are two (CIP ) to eight (AZ) times greater than the median chemical concentrations (5.4 mg CIP/kg and 0.26 mg AZ/kg) detected in USA biosolids (USEPA, 2009). Based on the more recent chemical concentration in biosolids data (Table 1 2 ; Youngquist et al., 2014; Kennedy/Jenks Santiago et al., 2016 ), pollutant load tracking is not r equired for the majority of USA biosolids. R efining /validating the IRA with additional, especially field and predator toxicity, data is advised before recommending modifications to current land application regulations Sources of U ncertainty and IRA L imitations Unavailability of data on CIP and AZ toxicity to predators: U ncertainty factors applied to minimize the uncertainty due to limited toxicity data appear appropriate A study on raptors (Red tailed Hawk) (Isaza et al., 1993) showed no toxicity eve n at an

PAGE 207

207 oral dose of 50 mg CIP /kg BW Also, minimal CIP and AZ accumulation is expected in birds due to fast elimination rates on order of few hours (Frazier et al., 1995). Thus, RfD calculations are likely conservative despite the general lack of data in the predator pathway. BAF values : BAF values for mammals and birds are largely absent from the literature but are likely no greater than the earthworm values generated here. The values generated herein are f or freshly spiked maximum exposure scenario s (Si dhu, Chapter 5). Mammals and birds have less contact with the contaminated soil than worms, thus, accumulate less CIP and AZ than suggested by the earthworm BAF values. Dry weight BAF values, which are around 5 times greater than wet weight values, were us ed, even though animals and birds consume organisms on wet weight basis. Also, CIP and AZ have low Log ; pH up to 8 for AZ) (Ross et al., 1992; McFarland et al., 1997). Such c hemicals likely have BAF values ~1 because the aqueous phase of organism body becomes the dominant phase of chemical partitioning (Jager, 1998) and chemical elimination is fast ( Abada et al., 1994 ; Flores Miranda et al., 2012). Based on human data, AZ and CIP are not likely to biomagnify to a considerable extent (Food and Drug Administration, 2004 a,b). Therefore, the BAF values used herein likely resulted in overestimation of risks. BCF values : BCF values for fish are missing but even if the fish BCF is 10 00 (unlikely, based on quantitative structure activity relatio nships; USEPA EPI Suite, 2011) risk from biosolids borne CIP and AZ would be negligible. High concentrations of the chemicals (from wastewater effluents; Hughes et al., 2013; Petrie et al., 2015) could

PAGE 208

208 pose potential risk to aquatic organisms and pathways 12 and 15, but are not applicable to biosolids borne chemicals. S urface water chemical concentrations : S urface water chemical concentrations resulting from biosolids borne CIP and AZ were estimated using a set of conservative equation s utilized in the Par t 503 Biosolids Rule risk assessment. The estimate assume s th at the entire runoff/eroded material is biosolids contain ing 95 th percentile CIP or AZ concentrations. The estimate ignores c hemical photodegradation despite the fact that CIP and AZ are highly p hotodegradable ; half lives rang e from a few hours to less than a d ay (Burhenne et al., 1997 a,b; Ericson, 2007; Tong et al., 2011; Maier and Tjeerdema, 2018). The most sensitive aquatic species (cyanobacteria and algae) live on or near water surface, so ig noring photolysis and its effect on organisms exaggerates exposure. Actual risks are likely significantly less than estimated in the IRA. Metabolites and transformation products: Risk assessments that focus exclusively on parent compounds can underestimate risks associated with metabolites and transformation products. Ciprofloxacin and AZ can degrade under stress (e.g., acid or base stress) and specific conditions (e.g., white rot fungi degrades CIP) (Wetzstein et al., 1999; Parshikov et al., 2000; Sunderland et al., 2001; US National Library of Medicine, 2002; Bel et al., 2009). Similarly, photodegradation (CIP and AZ) and hydrolysis (AZ) can yield chemical metabolites (US National Library of Medicine, 2002). More than 20 CIP and AZ metabolites have been identified and characterized under specific conditions. Activities of the metabolites are reportedly less than or equal to the parent compounds (Sunderland et al., 2001; US National Library of Medicine, 2002; Bel

PAGE 209

209 et al., 2009). Considering chemical persistence in the biosolids systems, only minute concentrations of the metabolites are expected in the environment. Small concentrations and expected limited activity, suggest that the metabolites of biosolids borne CIP and AZ are of negligible concern. Plants and (likely) animals can transform accumulated CIP and AZ and store the transformation products (Coleman et al., 1997; Huber et al.; 2009; Bartha et al., 2010). The transformation products can revert to the parent compoun ds in human and animal guts (Coleman, 1997), increasing potential chemical exposure (Wu et al., 2015). The BAF values used in the IRA for worm and predator pathways (pathways 5, 7, 9, and 10) were determined using 3 H labeled compounds that incidentally inc lude transformation and degradation products. Also, calculations of HQ values for plant pathways (1, 2, 4, and 6) assumed greatest ( wet weight basis) BAF values of 0.0 02 (CIP) and 0. 02 (AZ) Based on the IRA, risks would be negligible, e ven if concentratio ns of the transformation products in plants a re 20 fold the parent compound concentrations M etabolites and transformation products of biosolids borne CIP and AZ are likely of minimal concern in the IRA of biosolids borne CIP and AZ Antibiotic resistance: B iosolids can be a source of antibiotic resistance determinants (i.e., resistant bacteria, resistance genes, and mobile genetic elements) in the environment (Clarke and Smith, 2011 ; Larrson, 2014; Mao et al., 2015 ) Literature, however, is ambivalent on the extent of trans fer of antibiotic resistance determinants present in biosolids to biosolids amended soils ( Halling Srensen et al., 2002; Sengelv et al., 2003 ; Brooks et al., 2007; Eriksen et al., 2009; Munir and Xagoraraki, 2011 ; Marti

PAGE 210

210 et al., 2013; L arsson, 2014 ; Chen et al., 2016 ; Rahube et al., 2016 ; Singer et al., 2016 ). Further risks associated with biosolids borne antibiotic resistance are unknown. Antibiotic resistance gene expression and possible enrichment of resistant bacteria observed in t his study were minor (~10 4 copies/g or less). Thus, our data are insufficient to support, but qualitatively back literature suggested biosolids borne antibiotic resistance development and spread concern s Several data gaps need to be filled to adequately assess the consequences of antibiotic resistance determinants in biosolids. To that end, long term field investigations are necessary, as is a quantitative framework to address risks from biosolids borne antibiotic resistance. Conclusions A tiered integra ted risk assessment for biosolids borne CIP and AZ was conducted using the framework developed by WHO (2001). Where applicable, data measured in environmentally relevant scenarios were employed. Limitations of missing data on target organism toxicity and u ptake were minimized by using conservative reference doses derived by applying appropriate uncertainty factors. A highly conservative screening level risk assessment involved three biosolids application scenarios: a) BAR 1 : a single heavy (at the rate of 10 0 Mg/ha) application of biosolids containing 95 th percentile concentrations of CIP or AZ, b) BAR 2 : 40 y of annual land application of biosolids containing average CIP or AZ concentrations, and c) a highly unlikely scenario: BAR 3 : 40 y of annual land applic ation of biosolids containing 95 th percentile concentrations found in USA biosolids (USEPA, 2009) The initial screening identified three pathways of concern: 8) Biosolids soil plant; 9) Biosolids soil soil organism; and 10) Biosolids soil soil organism predator. A second tier refinement of the risk assessment parameters conservatively assumed 50% bioavailabilit ies and 10 y

PAGE 211

211 half lives for both compounds and reduced HQ values in the three pathways of concern A t hird tier assessment assumed 30% chemical bi oavailability based on evidence presented in Chapter 4 and resulted in negligible estimates of risks in all pathways to human and ecological health under real world biosolids application scenarios. Even the BAR 3 scenario pose d minimal risk s. The IRA resul ts are consistent with our hypothesis of limited bioavailability of biosolids borne CIP and AZ E xtensive sorption onto, and minimal desorption from, biosolids minimizes human and ecological health risks associated with the target TOrCs Preliminary pollut ant limits were calculated based on the most sensitive organism s in the most limiting ( predator ) pathway following USEPA protocol The calculations suggest that application scenarios BAR 1 and BAR 2 are without appreciable human and ecological risks. Preliminary p ollutant concentration limits are 12 mg CIP/kg and 2.2 mg AZ/kg and suggest that majority of modern USA biosolids can be freely land applied without CIP and AZ load tracking requirements. Field validations, measured data on mammalian an d avian toxicities and data on runoff and erosion potential of biosolids amended soils c ould reduce uncertainty in the risk assessment. However, mo st important are needed improvements to assessment of antibiotic resistance risk s Assessment of biosolids b orne antibiotic resistance determinants requires long term field investigations and developing frameworks to quantitatively assess associated risks.

PAGE 212

212 CHAPTER 7 SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS Introduction Uncertainty regarding risks from biosolids borne ciprofloxacin (CIP) and azithromycin (AZ) and the dearth of data on certain aspects of their environmental fate served as the foundation of this dissertation project. The ultimate objective of the project was to conduct an integrated human and envir onmental health risk assessment of biosolids borne CIP and AZ to guide regulatory policies on the safe and sustainable use of biosolids A literature review identified gaps in r etention/release, plant, earthworm, and microbial response data on the biosolid s borne TOrCs, all of which were addressed in various intermediate objectives The central hypothesis was that limited bioaccessibility of the sorbed CIP and AZ minimizes risks to human and environmental health. Retention/release studies were conducted to assess bioaccessibility of the target TOrCs. Assessing bioaccessibility is valuable for predict ing potential chemical bioavailability, but bioavailability (and associated risks) are ultimately determined by the exposed organism. Thus, subsequent studies assessed bioavailability (in terms of toxicity, bioaccumulation, etc.) of the biosolids borne target TOrCs to plants, microbes, and earthworms. The data generated from the intermediate objectives supplemented data obtained from the pertinent literature and that generated by Co PIs on a Water Environment & Reuse Foundation (WE & RF) funded project The risk assessment utilized an integrated risk assessment (IRA) framework devised by World Health Organization (WHO, 2001).

PAGE 213

213 This chapter summarizes the findings of intermediate objectives, ultimate objective, and provides overall conclusions, limitations, and fut ure work recommendations. Intermediate Objectives Intermediate objective 1. To assess retention release behavior of CIP and AZ Both CIP and AZ extensively and strongly sorb to biosolids and soils at environmentally relevant as well as unrealistically high chemical concentrations. Partitioning coefficient s (Kd values ) of ~360 L/kg (CIP) and ~430 L/kg (AZ) were obtained for the biosolids. Des orption was highly hysteretic ( h ysteresis coefficients less than 0.003 ). Desorption of biosolids borne chemicals was to various soils Strong linearity of sorption and desorption isotherms over the tested concentration range suggests minimal bioaccessibility o f target TOrCs even when severely contaminated biosolids are land applied over several years. The study formed the basis for our central hypothesis of limited bioaccessibility of biosolids borne CIP and AZ minimiz ing human and environmental health risks I ntermediate o bjective 2. To assess plant responses to biosolids borne CIP and AZ Environmentally relevant concentrations of biosolids borne CIP and AZ are neither phytotoxic nor significantly phyto accumulate d P lant uptake of, and toxicity from, biosolids borne CIP and AZ were minimal even in a worst case scenario (i.e., sand as a growth medium) and in the presence of the uppermost end of environmentally relevant chemical concentrations. Phytotoxicity experiments (without biosolids) yielded no observed adverse effect concentrations (NOAEC) of 3.2 mg/kg for AZ, and 0.36

PAGE 214

214 mg/kg (lettuce) and 1.1 mg/kg (radish and fescue) for CIP. The NOAEC value for AZ represent s accumulation of AZ from land application of a severely contaminated biosolids over 100 years. The NOAEC values for CIP are also greater than environmentally relevant concentrations of CIP in biosolids amended soils. S everal years of land application of biosolids containing typical (~median) concentrations of the target TOrCs, at the typical 1% agronomic or greater application rates, would pos e trivial risks to plants; even assuming no chemical attenuation. Even severely AZ or CIP contaminated biosolids pose insignificant risks to plants. Point estimates of bi oaccumulation factors were 0.1 (AZ) and 0.01 (CIP), and suggest minimal introduction of the two compounds into human and ecological food chains. Intermediate objective 3. To assess microbial responses to biosolids borne CIP and AZ Environmentally relevant concentrations of biosolids borne CIP and AZ are bioavailable to at least some microbes but likely pose minimal risks to overall microbial activity The target TOrCs a dverse ly affected AOB (only initially) in biosolids and (to much less extent) in biosol ids receiving soils but impacts on microbial respiration and over all N and P cycling were negligible Long term l and application of typical biosolids appears to pose negligible risks to overall microbial activity from an agronomic viewpoint (i.e., N and P cycling). S everely contaminated biosolids pose minimal risks, even when applied for several years. Expected chemical attenuation between biosolids applications w ould further abate risks to overall microbial health from agronomic viewpoint The two biosoli ds borne antibiotics induced adverse impacts on some microbes but only initially M inor increases in expression of some resistance genes possible

PAGE 215

215 maintenance of antibiotic resistance and possible enrichment of some r esistant bacteria also occurred The data are insufficient to fully document but qualitatively support biosolids facilitated antibiotic resistance concerns suggested in the literature Long t erm field investigations using various biosolids (including Class A and Class B) are necessary to va lidate the laboratory based results and to better assess environmental fate of biosolids borne antibiotic resistance. Intermediate objective 4. To assess earthworm responses to biosolids borne CIP and AZ Laboratory based earthworm bioaccumulation factors were ~4 (CIP) and ~ 7 (AZ) in depurated worms and ~20 ( both CIP and AZ) in un depurated worms. Thus, contaminated earthworms are a potential pathway of biosolids borne TOrC entry into ecological food web. Field sample data w ere too limited for definitive i nterpretation, so we conservativel y utilize d laboratory based data to estimate risks Ultimate Objective : To C onduct H uman and E cological H ealth R isk A ssessment of B iosolids borne CIP and AZ Biosolids borne CIP and AZ pose negligible human and environmental risks from the 16 exposure pathways (Table 6 1) under common ly utilized land application scenarios Even severely contaminated biosolids regularly applied, long term, pose minimal risks Based on CPLR values of 12 kg CIP and 2.2 kg AZ per ha biosolids containing 95 th percentile CIP or AZ concentrations can be land applied at the rate of 320 Mg/ha (CIP) or 700 Mg/ha (AZ) with minimal human and ecological risks. More typical biosolids (containing median CIP and AZ concentrations) can be land a pplied at 20 Mg/ha y for more than 100 y (CIP) and 360 y (AZ), even with out chemical attenuation.

PAGE 216

216 From a (USEPA) regulatory viewpoint, t he most stringent criteria for biosolids borne pollutants are the pollutant concentration limits Biosolids meeting the pollutant concentration limit can be land applied without having to t rack pollutant load s, provided pathogen, vector control and other Part 503 Biosolids Rule requirements are met. Estimated pollutant concentration limits a re 12 mg /kg for CIP and 2.2 mg /kg for AZ and suggest that the majority of modern USA biosolids do not require load tracking. Field validations and measured predator toxicity data could reduce uncertainty in the risk assessment. T he IRA, due to a lack of data, is insufficient to assess a n additional (17 th ) pathway biosolids borne antibiotic resistance of potential human and ecological risks. L ong term field investigations are required to assess the environmental fate of biosolids borne antibiotic resistance, as is developing framework s to quantitatively assess associated risks. Future R esearch P riorit ies The IRA identified some of the same knowledge /data gaps th at hampered the original USEPA Part 503 Biosolids Rule making risk assessment (USEPA, 1995) and more recent biosolids borne T OrC risk assessments (e.g., Snyder et al., 2013). V aluable data are contributed herein to close some of the knowledge gaps, but a more refined (and realistic) risk assessment requires addressing the following issues : # 1 Antibiotic resistance: We utilized Class A biosolids and observed only marginal increases in biosolids borne antibiotic resistance determinants. Class B biosolids have greater potential for antibiotic resistance development and spread than Class A materials and warrant similar i nvestigations. Literature suggests that biosolids (especially Class B) can be a source of antibiotic resistance development and potential spread in the environment ( Halling Srensen et al., 2002; Sengelv et al., 2003; Munir

PAGE 217

217 and Xagoraraki, 2011; Marti et al., 2013; Larsson, 2014; Mao et al., 2015; Chen et al., 2016; Rahube et al., 2016; Singer et al., 2016 ) However, s everal data gaps /issues require addressing before risks from biosolids borne antibiotic resistance can be assessed. In lieu of increasing ev idence on biosolids borne antibiotic resistance, human and environmental risks from expression and transfer of biosolids borne resistance genes need to be a focus of future investigations. L ong term f ield studies should include quantification of antibi otic resistance determinants in different biosolids (including Class A and B), biosolids amended soils, and the spread of the determinants to other ecological compartments (e.g., plants, animals). Also, frameworks to quantitatively associate exposure of biosol ids borne antibiotic resistance to human and ecological risks need develop ing # 2 All inclusive field studies : In typical field settings, plants, microbes, worms, etc., all interact with each other and with the biosolids borne TOrCs. L aboratory studies o n individual organisms are valuable because they represent risks under controlled scenarios However, responses of organisms as a consortium under field conditions may be different from those revealed by individual organism studies and/or lab oratory studie s. For instance, l aboratory studies usually over estimate risks to earthworms and plants (Higgins et al., 2010; Pannu et al., 2012; Hoke et al., 2016 ; van den Brink et al., 201 6 ) and represent conservative assessment s Data from long term f ield studies (simulating real world scenarios) w ould better represent the risks associated with biosolids borne TOrCs Field studies should include assessment of: 1. different biosolids (including Class A and Class B), 2. different biosolids amended soils, 3. run off, erosion, and leaching potential of amended soils,

PAGE 218

218 4. plants grown in amended soils, 5. earthworms and other invertebrates living in amended soils, 6. animals grazing amended soils 7. captive birds (where possible) frequenting the amended soils, and 8. antibiotic resist ance determinants in various compartments (i.e., biosolids, soils, leachates, invertebrates, plants, animals, and birds). S uch studies often require multiple collaborations cost millions of dollars and are rarely undertaken ; but are necessary to validat e risk assessment models The wide range of chemical concentrations (from environmentally relevant to unrealistically high) used in this work suggest that responses of many organism to CIP and AZ in different biosolids should be similar to found herein. Ho wever, s tudies using a variety of biosolids of varying physico chemical and biological properties are needed to address biosolids borne antibiotic resistance concerns # 3 Mammalian and avian response data: The IRA applied a ppropriate uncertainty factors to mammalian and avian reference doses and utilized conservative estimates of bioaccumulation factors. However, a more refined risk assessment should include the field measurements of mammalian and avian bioaccumulation and toxicity data.

PAGE 219

219 APPENDIX A AXYS METHOD MLA 075: ANALYSIS OF AZITHROMYCIN AND CIPROFLOXACIN IN SOLID, AQUEOUS, AND TISSUE BY LC MS/MS Extraction Surrogate standards (AZ 13 C 3 Trimethoprim; CIP 13 C 3 15 N Ciprofloxacin) a re added to all samples prior to extraction at a pH of 2. Solid and tissue samples a re sonicat ed with aqueous buffered acetonitrile and with pure acetonitrile. Extracts a re concentrated by rotary evaporation and diluted with double deionized water to 200 mL. The acidic extract i s treated with EDTA, filtered, cleaned u p by solid phase extraction (SPE), and analyzed by LC MS/MS. Analysis Analysis of the extracts i s performed on a high performance liquid chromatograph coupled to a triple quadrupole mass spectrometer. The LC MS/MS i s run in MRM (Multiple Reaction Monitorin g) mode and quantification i s performed by recording the peak areas of the applicable parent ion/daughter ion transitions. Analytes a re analyzed in the ESI positive mode. Calibration Appropriate calibration standards of at least 5 consecutive calibration levels (including spiked surrogates) a re prepared in 75:25 methanol: 0.1% formic acid buffer to construct calibration curves and determine range, efficiency, and accuracy of analysis. A known standard i s analyzed to confirm accuracy during runs every 12 h or every 20 samples, whichever occur s first. Analyte I dentification Positive identification of target trace organics, surrogate standards, and recovery standards requires :

PAGE 220

220 transition, on condition that the result is above the lowest calibration standard level. For surrogate and recovery standards: > 10:1 for parent ion to daughter ion transition. Quantitation Concentrations of the target compounds a re calculated either by isotope dilution quantification against the surrogate standard or by internal standard quantification against the recovery standard with linear regression calibration. Reporting limits Sample specific detection limits (SDLs) a re calculated by QuanLynx software usi ng 3 times the signal to noise ratio in the target channel converted to an equivalent sample concentration. Concentrations and detection limits for the target analytes a re reported. The lower reporting limit for each target compound is defined as the conce ntration equivalent to the lowest calibration standard analyzed or the SDL, whichever is greater. Quality A ssurance/ Q uality C ontrol All samples a re analyzed in batches with the following composition: Batch Size: Each batch consist s of up to twenty samples and additional QC samples. Blanks: One procedural blank i s analyzed for each batch. The procedural blank i s prepared by spiking an aliquot of the surrogate standard solution into a clean matrix. The procedural blank i s extracted and analyzed using the same procedures as the test samples in the analysis batch. On going Precision and Recovery (OPR) Samples : On going Precision and Recovery (OPR) i s demonstrated by the analysis of a spiked reference matrix (SPM)

PAGE 221

221 analyzed with each batch. The OPR sample i s p repared by spiking an aliquot of the spiking solution into an accurately weighed in house reference matrix (known to contain low background levels of target analytes). The OPR sample i s extracted and analyzed using the same procedures as the test samples i n the analysis batch. Duplicates: 5% of the test samples within a batch (containing 7 or more test samples) a re analyzed in duplicate. Limitations to P erformance S oil S amples The 13 C 3 15 N Ciprofloxacin surrogate can yield recoveries from soil samples tha t do not meet method criteria. The exact reason is not known, as recoveries are in the normal range for other matrices including biosolids samples that undergo identical processing, and for aqueous samples as well. The interaction of dissolved inorganic co mponents of the matrix with the analytes and the material in the Oasis HLB cartridge is the most likely cause for low recovery.

PAGE 222

222 APPENDIX B MICROBIAL DNA QPCR DATA Figure B 1. AOB DNA quantification from controls and 95 th percentile TOrC treatments on d ay 0 and d ay 90. Letters (a,b) represent significant differences (according to student t d ay 0 and d ay 90 for a particular chemical, media, and treatment. The AOB DNA data suggest no difference betwee n populations of ammonia oxidizing bacteria during incubation, and confirm a bacteriostatic effect of biosolids borne CIP and AZ. Figure B 2. qnrS DNA quantification from controls and 95 th percentile TOrC treatments on d ay 0 and d ay 90. Letters (a,b) represent significant differences (according to student t 5 ) between d ay 0 and d ay 90 for a particular media and treatment. The qnrS DNA data suggest significant increase s in qnrS copies and possible (but marginal ) enrichment of CIP resistant bacteria

PAGE 223

223 Figure B 3. ermB DNA quantification from controls and 95 th percentile TOrC treatments on d ay 0 and d ay 90. Letters (a,b) represent no differences (according to student t 5 ) between d ay 0 and d ay 90 for a particular media and treat ment. The ermB DNA data suggest no change in e rm B copies and (likely) AZ resistant bacteria. Figure B 4. mefE DNA quantification from controls and 95 th percentile TOrC treatments on d ay 0 and d ay 90. Letters (a) represent significant differences (accor ding to student t d ay 0 and d ay 90 for a particular media and treatment. The mefE data suggest that AZ did not affect mefE copy numbers Possibly, the mefE expression in the two incubation media was largely due to some unknown factor and was not treatment dependent.

PAGE 224

224 Figure B 5. 16S rRNA quantification from controls and 95 th percentile TOrC treatments on d ay 0 and d ay 90. Letters (a) represent significant differences (according to student t day 0 and day 90 for a particular media and treatment The negligible impacts of biosolids borne CIP and AZ on microbial populations (as evident from 16S r RNA data) confirm bacteriostatic nature o f the target TOrCs. Figure B 6 Representative figure showing KCl extractable NH 4 N and NOx, and water extractable P values over time for various CIP and AZ treatments in the biosolids

PAGE 225

225 R esults for CIP and AZ treatments are combined in Figures B 6 to B 8 because there w ere no significant difference s in N and P data for a particular CIP and AZ treatment at a particular time The i ncrease in NOx is attributed to nitrification of NH 4 N. The apparent decrease in water extractable P in al l solid matrices (Figures B 6 to B 8) over time could be due to P fixation by microbes and/or changes in P extractability over time. Confirmation, however, requires additional studies, which a re beyond the scope of this work. Figure B 7 Representative f igure showing KCl extractable NH 4 N and NOx, and water extractable P values over time for various CIP and AZ treatments in the manured sand Figure B 8 Representative figure showing KCl extractable NH 4 N and NOx, and water extractable P values over tim e for various CIP and AZ treatments in the biosolids amended manured sand.

PAGE 226

226 LIST OF REFERENCES Abada, A.R., Aramayona, J.J., Muoz, M.J., Delfina, J.M., Saez, M.P., Bregante. M.A., 1994. Disposition of ciprofloxacin following intravenous administration in dogs. J. Vet. Pharmacol. Ther. 17, 384 388. Agyin and TCS in soils and sediments. Environ. Toxicol. Chem. 29, 1925 1933. Al Ahmad, A., Daschner, F.D., Kummerer, K., 1999. Biodegradability of cefotiam, ciprofloxacin, meropenem, penicillin G., and sulfamethoxazole and inhibition of wastewater bacteria. Arch. Environ. Contam. Toxicol. 37, 158 163. Alexander, D.E., 1999. Bioaccumulation, bioconcentration, biomagnification. Envi ronmental Geology. Encyclopedia of Earth Science. Springer, Dordrecht, pp. 43 44. Alexander, M., 2000. Aging, bioavailability, and overestimation of risk from environmental pollutants. Environ. Sci. Technol. 34, 4259 4265. Altshuler, B., Pasternack, B., 19 63. Statistical measures of the lower limit of detection of a radioactivity counter. Health Physics. 9, 293 298. Ambrose, K.D., Nisbet, R., Stephens, D.S., 2005. Macrolide efflux in Streptococcus pneumoniae is mediated by a dual efflux pump ( mel and mef ) a nd is erythromycin inducible. Antimicrob. Agents Chemother. 49, 4203 4209. 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. American Societ y of Agronomy, Madison, WI, pp. 842 842. Andrade, R.J., Tulkens, P.M., 2011. Hepatic safety of antibiotics used in primary care. J. Antimicrob. Chemother. doi:10.1093/jac/dkr159. Appel, C., Ma, L.Q., Rhue, R.D., Reve, W., 2008. Sequential sorption of lead and cadmium in three tropical soils. Environ. Pollut. 155, 132 40. Aristilde, L., Sposito, G., 2008. Molecular modeling of metal complexation by a fluoroquinolone antibiotic. Environ. Toxicol. Chem. 27, 2304 2310. Aristilde, L., Melis, A., Sposito, G., 2 010. Inhibition of photosynthesis by a fluoroquinolone antibiotic. Environ. Sci. Technol. 44, 1444 1450. Aristilde, L., Sposito, G., 2013. Complexes of the antimicrobial ciprofloxacin with soil, peat, and aquatic humic substances. Environ. Toxicol. Chem. 3 2, 1467 1478. Atli, O., Ilgin, S., Altuntas, H., Burukoglu, D., 2015. Evaluation of azithromycin induced cardiotoxicity in rats. Int. J. Clin. Exp. Med. 8, 3681 3690.

PAGE 227

227 Banning, N.C., Maccarone, L.D., Fisk, L.M., Murphy, D.V., 2015. Ammonia oxidizing bacteri a not archaea dominate nitrification activity in semi arid agricultural soil. Scientific Reports 5, Article number: 11146 Barnard, F.M., Maxwell, A., 2001. Interaction between DNA gyrase and quinolones: effects of alanine mutations at GyrA subunit residues Ser(83) and Asp(87). Antimicrob. Agents Chemother. 45, 1994 2000. Bartha, B., Huber, C., Harpaintner, R., Schroeder, P., 2010. Effects of acetaminophen in Brassica juncea L. Czern .: investigation of uptake, translocation, detoxification, and the induced d efense pathways. Environ. Sci. Pollut. Res. 17, 1553 1562. Becic, E., Imamovic, B., Dedic, M., Sober, M., 2014. SPE extraction and TLC identification of tetracycline and fluoroquinolone in surface water. Bulletin of the Chemists and Technologists of Bosnia and Herzegovina. 43, 35 40. Bel, E., Dewulf, J., Witte, B.D., Van Langenhove, H., Janssen, C., 2009. Influence of pH on the sonolysis of ciprofloxacin: biodegradability, ecotoxicity and antibiotic activity of its degradation products. Chemosphere.77, 291 295. Bengtsson Palme, J., Larsson, D.G., 2016. Concentrations of antibiotics predicted to select for resistant bacteria: proposed limits for environmental regulation. Environ. Int. 86, 140 149. Bergkemper, F., Kublik, S., Lang, F., Krger, J., Vestergaard G., Schloter, M., Schulz, S., 2016. Novel oligonucleotide primers reveal a high diversity of microbes which drive phosphorous turnover in soil. J. Microbiol. Methods. 125, 91 97. Berhane, T.M., Levy, J., Krekeler, M.P.S., Danielson, N.D., 2016. Adsorptio n of bisphenol A and ciprofloxacin by palygorskite montmorillonite: effect of granule size, solution chemistry and temperature. Appl. Clay Sci. 132 133, 518 527. Braker, G., Fesefeldt. A., Witzel, K.P., 1998. Development of PCR primer systems for amplifica tion of nitrite reductase genes ( nirK and nirS ) to detect denitrifying bacteria in environmental samples. Appl. Environ. Microbiol. 64, 3769 3775. Brami, C., Glover, A.R., Butt, K.R., Lowe, C.N., 2017. Avoidance, biomass and survival response of soil dwell ing (endogeic) earthworms to OECD artificial soil: potential implications for earthworm ecotoxicology. Ecotoxicology. 26, 576 579. Brooks, J.P., Maxwell, S.L., Rensing, C., Gerba, C.P., Pepper, I.L., 2007. Occurrence of antibiotic resistant bacteria and e ndotoxin associated with the land application of biosolids. Can. J. Microbiol. 53, 616 622. Brown, S.P., Treadwell, D., Stephens, J.M., Webb, S., 2015. Florida vegetable gardening guide. Extension Data Information Source, University of Florida Extension, E DIS, 1998. Available at http://edis.ifas.ufl.edu/vh021 accessed on 1/15/2018.

PAGE 228

228 Burhenne, J., Ludwig, M., Nikoloudis, P., Spiteller, M., 1997a. Photolytic degradation of fluoroquinolone carboxylic acids in aqueo us solution, Part I: primary photoproducts and half lives. Environ. Sci. Pollut. Res. 4, 10 15 Burhenne, J., Ludwig, M., Spiteller, M., 1997b. Photolytic degradation of fluoroquinolone carboxylic acids in aqueous solution, Part II: isolation and structural elucidation of polar photometabolites. Environ. Sci. Pollut. Res. 4, 61 67. Calabrese, E.J., Baldwin, L., 1997. The dose determines the stimulation (and poison): development of a chemical hormesis database. Int. J. Toxicol. 16, 545 559. Calabrese, E.J., B lain, R.B., 2009. Hormesis and plant biology. Environ. Pollut. 157, 42 48. Cardoza, L.A., Knapp, C.W., Larive, C.K., Belden, J.B., Lydy, M.J., Graham, D.W., 2005. Ciprofloxacin attenuation rates and mechanisms in aquatic field systems. Water Air Soil Poll. 161, 383 398. Carmosini, N., Lee, L., 2009. Ciprofloxacin sorption by dissolved organic fraction from reference and bio waste materials. Chemosphere. 77, 813 820. Carrasquillo, A.J., Bruland, G.L., MacKay, A.A., Vasudevan, D., 2008. Sorption of ciproflox acin and oxytetracycline zwitterions to soils and soil minerals: influence of compound structure. Environ. Sci. Technol. 42, 7634 7642. Carrillo, M., Braun, G.C., Siebe, C., Amelung, W., Siemens, J., 2016. Desorption of sulfamethoxazole and ciprofloxacin f rom long term wastewater irrigated soils of the Mezquital Valley as affected by water quality. J. Soils Sediments. 16, 966 975. Castela Papin, N., Cai, S., Vatier, J., Keller, F., Souleau, C.H., Farinotti, R., 1999. Drug interactions with diosmectite: a st udy using the artificial stomach duodenum model. Int. J. Pharm. 182, 111 119. Cattoir, V., Poirel, L., Rotimi, V., Soussy, C.J., Nordmann, P., 2007. Multiplex PCR for detection of plasmid mediated quinolone resistance qnr genes in ESBL producing enterobact erial isolates. J. Antimicrob. Chemother. 60, 394 397 Chander, Y., Kumar, K., Goyal, S.M., Gupta, S.C., 2005. Antibacterial activity of soil bound antibiotics. J. Environ. Qual. 34, 1952 1957. Chen, H., Ma, L.Q., Gao, B., Gu, C., 2013. Effects of Cu and Ca cations and Fe/Al coating on ciprofloxacin sorption onto sand media. J. Hazard. Mater. 252 253, 375 381. Chen, H., Gao, B., Yang, L Y, Ma, L., 2015. Montmorillonite enhanced ciprofloxacin transport in saturated porous media with sorbed ciprofloxacin show ing antibiotic activity. J. Cont. Hydrol. 173, 1 7.

PAGE 229

229 Chen, M., Wang, W X., 2001. Bioavailability of natural colloid bound iron to marine plankton: influences of colloidal size and aging. Limnol. Oceanogr. 46, 1956 1967. Chen, Q., An, X., Li, H., Su, J., Ma, Y., Zhu, Y.G., 2016. Long term field application of sewage sludge increases the abundance of antibiotic resistance genes in soil. Environ. Int. 92 93, 1 10. Chenier, M.R., Beaumier, D., Roy, R., Driscoll, B.T., Lawrence, J.R., Greer, C.W., 2003. Impact o f seasonal variations and nutrient inputs on nitrogen cycling and degradation of hexadecane by replicated river biofilms. Appl. Environ. Microbio. 69, 5170 5177. Chenxi, W., Spongberg, A.L., Witter, J.D., 2008. Determination of the persistence of pharmaceu ticals in biosolids using liquid chromatography tandem mass spectrometry. Chemosphere. 73, 511 518. Chhalotiya, U.K., Patel, N.M., Shah, D.A., Mehta, F.A., Bhatt, K.K., 2017. Thin layer chromatography method for the simultaneous quantification and stabilit y testing of alprazolam and mebeverine in their combined pharmaceutical dosage form. J. Taibah University Sci. 11, 66 75. Clarke, B.O., Smith, S.R., 2011. Review of 'emerging' organic contaminants in biosolids and assessment of international research prior ities for the agricultural use of biosolids. Environ. Int. 37, 226 247. Clarke, R., Healy, M.G., Fenton, O., Cummins, E., 2017. Quantitative risk assessment of antimicrobials in biosolids applied on agricultural land and potential translocation into food. Food Res. Int. doi.org/10.1016/j.foodres.2017.12.072. Coleman, J., Blake Kalff, M., Davies, E., 1997. Detoxification of xenobiotics by plants: chemical modification and vacuolar compartmentation. Trends Plant Sci. 2, 144 151. Collins, C.D., Martin, I., Dou cette, W., 2011. Plant uptake of xenobiotics. In: Schrder, P., Collins, C. (eds.) Organic xenobiotics and plants. Plant Ecophysiology, vol 8. Springer, Dordrecht, pp. 3 16. Colucci, M.S., Topp, E., 2002. Dissipation of part per trillion concentrations of estrogenic hormones from agricultural soils. Canadian J. Soil Sci. 82, 335 340. Conkle, J.L., Lattao, C., White, J.R., Cook, R.L., 2010. Competitive sorption and desorption behavior for three fluoroquinolone antibiotics in a wastewater treatment wetland so il. Chemosphere. 80, 1353 1359. Cui, H., Wang, S P., Fu, J., Zhou, Z Q., Zhang, N., Guo, L., 2014. Influence of ciprofloxacin on microbial community structure and function in soils. Biol. Fertil. Soils. 50, 939 947.

PAGE 230

230 D'Angelo, E., Starnes, D., 2016. Desorpt ion kinetics of ciprofloxacin in municipal biosolids determined by diffusion gradient in thin films. Chemosphere. 164, 215 224. Del Grosso, M., Northwood, J.G., Farrell, D.J., Pantosti, A., 2007. The macrolide resistance genes erm(B) and mef(E) are carried by Tn2010 in dual gene Streptococcus pneumoniae isolates belonging to clonal complex CC271. Antimicrob. Agents Chemother. 51, 4184 4186. Demoling, L.A., Baath, E., Greve, G., Wouterse, M., Schmitt, H., 2009. Effects of sulfamethoxazole on soil microbial communities after adding substrate. Soil Biol. Biochem. 41, 840 848. Ding, Y., Zhang, W., Gu, C., Xagoraraki, I., Li, H., 2011. Determination of pharmaceuticals in biosolids using accelerated solvent extraction and liquid chromatography/tandem mass spectr ometry. J. Chromatogr. A. 1218, 10 16. Ding, G.C., Radl, V., Schloter Hai, B., Jechalke, S., Heuer, H., Smalla, K., Schloter, M., 2014. Dynamics of soil bacterial communities in response to repeated application of manure containing sulfadiazine. PLoS One. 9, e92958. Dodgen, L.K., Li, J., Parker, D., Gan, J., 2013. Uptake and accumulation of four PPCP/EDCs in two leafy vegetables. Environ. Pollut. 182, 150 156. Dodgen, L.K., Ueda, A., Wu, X., Parker, D.R., Gan, J., 2015. Effect of transpiration on plant accumulation and translocation of PPCP/EDCs. Environ. Pollut. 198, 144 153. Dorfman, M.S., Wagner, R.S., Jamison, T., Bell, B., Stroman, D.W., 2008. The pharmacodynamic properties of azithromycin in a kinetics of kill model and implications for bacterial c onjunctivitis treatment. Adv. Ther. 25, 208 217. Droge, S.T., Goss, K.U., 2013. Development and evaluation of a new sorption model for organic cations in soil: contributions from organic matter and clay minerals. Environ. Sci. Technol. 47, 14233 14241. Du arte Davidson, R., Jones, K.C., 1996. Screening of the environmental fate of organic contaminants in sewage sludge applied to agricultural soils: II. the potential for transfers to plants and grazing animals. Sci. Total Environ. 185, 59 70. Ebert, I., Bach mann, J., Khnen, U., Kster, A., Kussatz, C., Maletzki, D., Schlter, C., 2011. Toxicity of the flouoroquinolone antibiotics enrofloxacin and ciprofloxacin to photoautotrophic aquatic organisms. Environ. Toxicol. Chem. 30, 2786 2792.

PAGE 231

231 ECETOC, 2013. Under standing the relationship between extraction technique and bioavailability. European Centre for Ecotoxicology and Toxicology of Chemicals, Technical Report No. 117, Brussels, Belgium. Available at http://www.ecetoc.org/wp content/uploads/2014/08/ECETOC TR 117 Understanding the relationship between extraction technique and bioavailability.pdf acces sed on 1/10/2018. Eggen, T., Asp, T.N., Grave, K., Hormazabal, V., 2011. Uptake and translocation of metformin, ciprofloxacin and narasin in forage and crop plants. Chemosphere. 85, 26 33. Eltahawy, A.T., 1993. In vitro activity of ciprofloxacin and sixte en other antimicrobial agents against blood culture isolates. J. Chemother. 5, 94 102. Epstein, E., 2003. Land application of sewage sludge and biosolids. CRC Press, Boca Raton, Florida. Ericson, J.F., 2007. An evaluation of the OECD 308 water/sediment sys tems for investigating the biodegradation of pharmaceuticals. Environ. Sci. Technol. 41, 5803 5811 Eriksen, G. S., Amundsen, C.E., Bernhoft, A., Eggen, T., Grave, K., Halling Srensen, B., Kllqvist, T., Sogn, T., Sverdrup, L., 2009. Risk assessment of con taminants in sewage sludge applied on Norwegian soils. Opinion of the panel on contaminants in the Norwegian Scientific Committee for Food Safety. Norwegian Scientific Committee for Food Safety (VKM). 05/511 22 final. Fbrega, J.R., Jafvert, C.T., Li, H., Lee, L.S., 2001. Modeling competitive cation exchange of aromatic amines in water saturated soils. Environ. Sci. Technol.35, 2727 2733. Fang, H., Han, Y., Yin, Y., Pan, X., Yu, Y., 2014. Variations in dissipation rate, microbial function and antibiotic res istance due to repeated introductions of manure containing sulfadiazine and chlortetracycline to soil. Chemosphere. 96, 51 56. Flores Miranda, B.M., Espinosa Plascencia, A., Gmez Jimnez, S., Lpez Zavala, A.A., Gonzlez Carrillo, H.H., Bermdez Almada, M .C., 2012. Accumulation and elimination of enrofloxacin and ciprofloxacin in tissues of shrimp Litopenaeus vannamei under laboratory and farm conditions. ISRN Pharm. 374212. doi: 10.5402/2012/374212. Food and Drug Administration, 2004a. 08918468, R. O. rep ort on Ciprofloxacin. Available at www.accessdata.fda.gov/drugsatfda_docs/label/2005/019537s057,020780s019lb l.pdf accessed on 2/1/18

PAGE 232

232 Food and Drug Adminis tration, 2004b. Zithromax. Available at www.accessdata.fda.gov/drugsatfda_docs/label/2013/050710s039,050711s036, 050784s023lbl.pdf accessed on 2 /1/18 Franklin, A.M., Williams, C.F., Andrews, D.M., Woodward, E.E., Watson, J.E., 2016. Uptake of three antibiotics and an antiepileptic drug by wheat crops spray irrigated with wastewater treatment plant effluent. J. Environ. Qual. 45, 546 554. Frazier, D.L., Jones, M.P., Orosz, S.E., 1995. Pharmacokinetic considerations of the renal system in birds: part II review of drugs excreted by renal pathways J. Avian Medicine Surgery 9, 104 121 Garcia Santiago, X., Franco Uria, A., Omil, F., Lema, J.M., 2016. Risk assessment of persistent pharmaceuticals in biosolids: dealing with uncertainty. J. Hazard. Materials. 302, 72 81. Garza Ramos, G., Xiong, L., Zhong, P., Mankin, A., 2001. Binding site of macrolide antibiotics on the ribosome: new resistance mut ation identifies a specific interaction of ketolides with rRNA. J. Bacteriol. 183, 6898 6907. Gillman, G.P., Sumpter, E.A., 1986. Modification to the compulsive exchange method for measuring exchange characteristics of soils. Aust. J. Soil Res. 24, 61 66. Girardi, C., Greve, J., Lamshft, M., Fetzer, I., Miltner, A., Schffer, A., Kstner, M., 2011. Biodegradation of ciprofloxacin in water and soil and its effects on the microbial communities. J. Hazard. Materials. 198, 22 30. Gbel, A., Thomsen, A., McArde ll, C.S., Joss, A., Giger, W., 2005. Occurrence and sorption behavior of sulfonamides, macrolides, and trimethoprim in activated sludge treatment. Environ. Sci. Technol. 39, 3981 3989. Goldstein, M., Shenker, M., Chefetz, B., 2014. Insights into the uptake processes of wastewater borne pharmaceuticals by vegetables. Environ. Sci. Technol. 48, 5593 5600. Golet, E.M., Alder, A.C., Giger, W., 2002a. Environmental exposure and risk assessment of fluoroquinolone antibacterial agents in wastewater and river water of the Glatt Valley watershed, Switzerland. Environ. Sci. Technol. 36, 3645 3651. Golet, E.M., Strehler, A., Alder, A.C., Giger, W., 2002b. Determination of fluoroquinolone antibacterial agents in sewage sludge and sludge treated soil using accelerated so lvent extraction followed by solid phase extraction. Anal. Chem. 74, 5455 5462. Golet, E.M., Xifra, I., Siegrift, H., Alder, A.C., Giger, W., 2003. Environmental exposure assessment of fluoroquinolone antibacterial agents from sewage to soil. Environ. Sci. Technol. 37, 3243 3249.

PAGE 233

233 Gong, W., Liu, X., He, H., Wang, L., Dai, G., 2012. Quantitatively modeling soil water distribution coefficients of three antibiotics using soil physicochemical properties. Chemosphere. 89, 825 831. Gordillo, M.E., Singh, K.V., Mur ray, B.E., 1993. In vitro activity of azithromycin against bacterial enteric pathogens. Antimicrob. Agents Chemother. 37, 1203 1205. Gottschall, N., Topp, E., Metcalfe, M., Edwards, M., Payne, M., Kleywegt, S., Russell, P., Lapen, D.R., 2012. Pharmaceutica l and personal care products in groundwater, subsurface drainage, soil, and wheat grain, following a high single application of municipal biosolids to a field. Chemosphere. 87, 194 203. Goulas, A., Haudin, C S., Bergheaud, V., Dumeny, V., Ferhi, S., Nelieu S., Bourdat Deschamps, M., Benoit, P., 2016. A new extraction method to assess the environmental availability of ciprofloxacin in agricultural soils amended with exogenous organic matter. Chemosphere. 165, 460 469. Grote, M., Schwake Anduschus, C., Miche l, R., Stevens, H., Heyser, W., Langenkmper, G., Betsche, T., Freitag, M., 2007. Incorporation of veterinary antibiotics into crops from manured soil. Landbauforschung Vlkenrode. 57, 25 32. Halling Sorensen, B., Lutzhoft, H C., Andersen, H.R., Ingerslev, F., 2000. Environmental risk assessment of antibiotics: comparison of mecillinam, trimethoprim and ciprofloxacin. J. Antimicrob. Chemo. 46, 53 58. Halling Srensen, B., Sengelvm, G., Tjrnelund, J., 2002. Toxicity of tetracyclines and tetracycline degrad ation products to environmentally relevant bacteria, including selected tetracycline resistant bacteria. Arch. Environ. Contam. Toxicol. 42, 263 271. Harada, A., Komori, K., Nakada, N., Kitamura, K., Suzuki, Y., 2008. Biological effects of PPCPs on aquatic lives and evaluation of river waters affected by different wastewater treatment levels. Water Sci. Technol. 58, 1541 1546. Hari, A.C., Paruchuri, R.A., Sabatini, D.A., Kibbey, T.C.G., 2005. Effects of pH and cationic and nonionic surfactants on the adsor ption of pharmaceuticals to a natural aquifer material. Environ. Sci. Technol. 39, 2592 2598. Herklotz, P.A., Gurung, P., Vanden Heuvel, B., Kinney, C.A., 2010. Uptake of human pharmaceuticals by plants grown under hydroponic conditions. Chemosphere. 78, 1 416 1421. Hernando, M.D., De Vettori, S., Martnez Bueno, M.J., Fernndez Alba, A., 2007. Toxicity evaluation with Vibrio fischeri test of organic chemicals used in aquaculture. Chemosphere 68, 724 730.

PAGE 234

234 Hendricks, S.B., 1941. Base exchange of the clay mine ral montmorillonite for organic cations and its dependence upon adsorption due to van der Waals forces. J. Phys. Chem. 45, 65 81 McAvoy, D., 2010. Trace organic chemical s in biosolids amended soils. State of the science review. Water Environment Research Foundation (WERF), IWA Publishing, London. Hilpert, R., Winter, J., Hammes, W., Kandler, O., 1981. The sensitivity of archaebacteria to antibiotics. Zentralblatt fr Bakt eriologie Mikrobiologie und Hygiene: I. Abt. Originale C: Allgemeine, angewandte und kologische Mikrobiologie. 2, 11 20. Hochmuth, G., Hanlon, E., Snyder, G., Nagata, R., Schueneman, T., 1996. Fertilization of sweet corn, celery, romaine, escarole, endive and radish on organic soils in Florida. Extension Data Information Source, University of Florida Extension, EDIS, 1998. Available at https://edis.ifas.ufl.edu/cv008 accessed on 1/15/2018. Hoke, R., Huggett, D., Brasfield, S., Brown, B., Embry, M., Fairbrother, A., Kivi, M., Paumen, M.L., Prosser, R., Salvito, D., Scroggins, R., 2016. Review of laboratory based terrestrial bioaccumulation assessment approaches for organic chemicals: current status and future possibilities. Integr. Environ. Assess. Manag. 12, 109 22. Hooper, D.C., 1999. Mode of action of fluoroquinolones. Drugs. 58, 6 10. Horvath, C., Melander, R.W., Molnar, W., 1976. Solvophobic interactions in liquid chromatography with nonpolar stationary ph ases. J. Chromatograph. 125, 129 156. Horvath, C., Melander, W., Molnar, I., 1977. Liquid chromatography of ionogenic substances with nonpolar stationary phases. Anal. Chem. 49, 142 154. Huang, R., Wen, B., Pei, Z., Shan, X.Q., Zhang, S., Williams, P.N., 2 009. Accumulation, subcellular distribution and toxicity of copper in earthworm ( Eisenia fetida ) in the presence of ciprofloxacin. Environ. Sci. Technol. 43, 3688 3693. Huber, C., Bartha, B., Harpaintner, R., Schroeder, P., 2009. Metabolism of acetaminophe n (paracetamol) in plants two independent pathways result in the formation of a glutathione and a glucose conjugate. Environ. Sci. Pollut. Res. 16, 206 213. Huber, C., Bartha, B., Schroeder, P., 2012. Metabolism of diclofenac in plants hydroxylation is followed by glucose conjugation. J. Hazard. Mater. 243, 250 256. Hughes, S.R., Kay, P., Brown, L.E., 2013. Global synthesis and critical evaluation of pharmaceutical data sets collected from river systems. Environ. Sci. Technol. 47, 661 677.

PAGE 235

235 Hyland, K.C., Blaine, A.C., Dickenson, E.R., Higgins, C.P., 2015. Accumulation of contaminants of emerging concern in food crops part 1: edible strawberries and lettuce grown in reclaimed water. Environ. Toxicol. Chem. 34, 2213 2221. Iatrou, E.I., Stasinakis, A.S., Tho maidis, N.S., 2014. Consumption based approach for predicting environmental risk in Greece due to the presence of antimicrobials in domestic wastewater. Environ. Sci. Pollut. Res. Int. 21, 12941 12950. Isaza, R., Budsberg, S.C., Sundlof, S.F., Baker, B., 1 993. Disposition of ciprofloxacin in Red Tailed Hawks ( Buteo jamaicensis ) following a single oral dose. J. Zoo Wildlife Medicine. 24, 498 502. Jacoby, G.A., 2005. Mechanisms of resistance to quinolones. Clin. Infect. Dis. 41, S120 126. Jager, T., 1998. Mechanistic approach for estimating bioconcentration of organic chemicals in earthworms ( oligochaeta ). Environ. Toxicol. Chem. 17, 2080 2090. Jager, T., van der Wal, L., Fleuren, R.H., Barendregt, A., Hermens, J.L., 2005. Bioaccumulation of organic chemica ls in contaminated soils: evaluation of bioassays with earthworms. Environ. Sci. Technol. 39, 293 298. ribosome binding antimicrobials. Antibiotics. 5, 29. Jensen, J., Ingv ertsen, S.T., Magid, J., 2012. Risk evaluation of five groups of persistent organic contaminants in sewage sludge. Danish Ministry of the Environment. Environmental Protection Agency. Environmental Project No. 1406 2012. Jiang, W.T., Chang, P.H., Wang, Y.S ., Tsai, Y., Jean, J.S., Li, Z., Krukowski, K., 2013. Removal of ciprofloxacin from water by birnessite. J. Hazard. Materials. 250 251, 362 369. Jones Lepp, T.L., Sanchez, C.A., Moy, T., Kazemi, R., 2010. Method development and application to determine po tential plant uptake of antibiotics and other drugs in irrigated crop production systems. J. Agri. Food Chem. 58, 11568 11573. Kennedy/Jenks Consultants. 2015. Biosolids risk analysis (K/J Project No. 1476009.00). Seattle, WA: NW Biosolids. Khelaifia, S., Drancourt, M., 2012. Susceptibility of archaea to antimicrobial agents: applications to clinical microbiology. Clinical Microbio. Infect. 18, 841 848. Kinney, C.A., Furlong, E.T., Kolpin, D.W., Burkhardt, M.R., Zaugg, S.D., Werner, S.L., Bossio, J.P., Beno tti, M.J., 2008. Bioaccumulation of pharmaceuticals and other anthropogenic waste indicators in earthworms from agricultural soil amended with biosolids or swine manure. Environ. Sci. Technol. 42, 1863 1870.

PAGE 236

236 Kipper, K., Herodes, K., Lillenberg, M., Nei, L. Haiba, E., Litvin, S.V., 2010. Plant uptake of some pharmaceuticals commonly present in sewage sludge compost. 2nd International Conference on Chemical, Biological and Environmental Engineering (ICBEE 2010). 5. High performance thin layer chromatography with densitometry for the determination of ciprofloxacin and impurities in drugs. J. AOAC Int. 88,1530 1536. densitometric determination of azithromycin in pharmac eutical preparations. J. Planar Chromatogr. Mod. TLC. 21, 177 181. Kumar, K., Gupta, S.C., 2016. A framework to predict uptake of trace organic compounds by plants. J. Environ. Qual. 45, 555 564. Lam, M.W., Tantuca, K., Mabury, S.A., 2003. Photofate: a new approach in accounting for the contribution of indirect photolysis of pesticides and pharmaceuticals in surface waters. Environ. Sci. Technol. 37, 899 907. Larsson, D.G.J., 2014. Antibiotics in the environment. Ups J. Med. Sci. 119, 108 112. Lawrence, sampling earthworms. Soil Bio. Biochem.34, 549 552. LeBel, M., 1993. Pharmacokinetic properties of clarithromycin: a comparison with erythromycin and azithromycin. Can. J. Infect. D is. 4, 148 152. Lee, B T., Shin, K H., Kim, J Y., Kim, K W., 2008. Progress in earthworm ecotoxicology. In: Kim, Y.J., Platt, U., (eds.). Advanced environmental monitoring. Springer, Berlin, Germany, pp. 248 258. Lillenberg, M., Litvin, S.V., Nei, L., Roas to, M., Sepp, K., 2010. Enrofloxacin and ciprofloxacin uptake by plants from soil. Agronomy Res. 8, 807 814. Logan, T.J., Henry, C.L., Schnoor, J.L., Overcas h, M., McAvoy D.C., 1999. An a ssessment of h ealth and e nvironmental r isks of t race e lements and t o xic o rganics in l and a pplied m unicipal s olid w aste c ompost Compost Sci. Util. 7, 38 53. Lyman, W.J., Reehl, W.F., Rosenblatt, D.H., 1990. Handbook of chemical property estimation methods. Washington, DC: Amer. Chem. Soc., pp. 74 75. Ma, L., Zhong, H., Wu, Y.G., 2015. Effects of metal soil contact time on the extraction of mercury from soils. Bull. Environ. Contam. Toxicol. 94, 399 406. Maier, M.L.V., Tjeerdema, R.S., 2018. Azithromycin sorption and biodegradation in a simulated California river system. Che mosphere. 190, 471 480.

PAGE 237

237 Maggioli, C., Santi, L., Zaccherini, G., Bevilacqua, V., Giunchi, F., Caraceni, P., 2011. A case of prolonged cholestatic hepatitis induced by azithromycin in a young woman. Case Reports in Hepatology. Article ID 314231. Mamber, S.W ., Kolek, B., Brookshire, K.W., Bonner, D.P., Fung Tomc, J., 1993. Activity of quinolones in the Ames Salmonella TA102 mutagenicity test and other bacterial genotoxicity assays. Antimicrob. Agents Chemother. 37, 213 217. Mao, D., Yu, S., Rysz, M., Luo, Y., Yang, F., Li, F., Hou, J., Mu, Q., Alvarez, P.J., 2015. Prevalence and proliferation of antibiotic resistance genes in two municipal wastewater treatment plants. Water Res. 85, 458 466. Marchandin, H., Jean Pierre, H., Jumas Bilak, E., Isson, L., Drouill ard, B., Darbas, H., Carrire, C., 2001. Distribution of macrolide resistance genes erm(B) and mef(A) among 160 penicillin intermediate clinical isolates of Streptococcus pneumoniae isolated in southern France. Pathol. Biol. 49, 522 527. Margulies, L., Rozen, H., Nir, S., 1988. Model for competitive adsorption of organic cations on clays. Clays and Clay Minerals. 36, 270 276. Marti, R., Scott, A., Tien, Y.C., Murray, R., Sabourin, L., Zhang, Y., Topp, E., 2013. Impact of manure fertilization on the abundance of antibiotic resistant bacteria and frequency of detection of antibiotic resistance genes in soil and on vegetables at h arvest. Appl. Environ. Microbiol. 79, 5701 5709. Martins, N., Pereira, R., Abrantes, N., Pereira, J., Gonalves, F., Marques, C.R., 2012. Ecotoxicological effects of ciprofloxacin on freshwater species: data integration and derivation of toxicity threshol ds for risk assessment. Ecotoxicology. 21, 1167 1176. Matamoros, V., Caldern Preciado, D., Dominguez, C., Bayona, J.M., 2012. Analytical procedures for the determination of emerging organic contaminants in plant material: A review. Anal. Chim. Acta. 722, 8 20. Mayne, J., Walsh, A., Johnson, N., Roesler, A., Shepard, R., Tachibana, M., 1996. Preclinical toxicology studies with azithromycin: acute and subchronic studies in rodents. Oyo Yakuri. 51, 53 63 McFarland, J.W., Berger, C.M., Froshauer,S.A., Hayashi ,S.F., Hecker, S.J., Jaynes, B.H., Jefson, M.R., Kamicker, B.J., Lipinski, C.A., Lundy, K.M., Reese, C.P., Vu, C.B., 1997. Quantitative structure activity relationships among macrolide antibacterial agents: in vitro and in vivo potency against Pasteurella multocida J. Med. Chem. 40, 1340 1346. McLain, J.E., Rock, C.M., Gerba, C.P., 2017. Environmental antibiotic resistance associated with land application of biosolids. In: Keen P.L., Fugre, R. (eds). Antimicrobial resistance in wastewater treatment proces ses. John Wiley and Sons, Inc. Hoboken, NJ, pp. 241 252.

PAGE 238

238 Migliore, L., Cozzolino, S., Fiori, M., 2003. Phytotoxicity to and uptake of enrofloxacin in crop plants. Chemosphere. 52, 1233 1244. Miller, E.L., Nason, S.L., Karthikeyan, K.G., Pedersen, J.A., 201 6. Root uptake of pharmaceuticals and personal care product ingredients. Environ. Sci. Technol. 50, 525 541. Minguez, L., Pedelucq, J., Farcy, E., Ballandonne, C., Budzinski, H., Halm Lemeille, M.P., 2016. Toxicities of 48 pharmaceuticals and their freshwa ter and marine environmental assessment in northwestern France. Environ. Sci. Pollut. Res. Int. 23, 4992 5001. Mohapatra, D.P., Cledn, M., Brar, S.K., Surampalli, R.Y., 2016. Application of wastewater and biosolids in soil: occurrence and fate of emergin g contaminants. Water Air Soil Pollut. 227, 77. Munir, M., Xagoraraki, I., 2011. Levels of antibiotic resistance genes in manure, biosolids, and fertilized soil. J. Environ. Qual. 40, 248 255. Mustaev, A., Malik, M., Zhao, X., Kurepina, N., Luan, G., Oppeg ard, L.M., Hiasa, H., Marks, K.R., Kerns, R.J., Berger, J.M., Drlica, K., 2014. Fluoroquinolone gyrase DNA complexes: two modes of drug binding. J. Bio. Chem. 289, 12300 12312. Nannipieri, P., Giagnoni, L., Landi, L., Renella, G., 2011. Role of phosphatase enzymes in soil. In: Phosphorus in action. Soil Biology 26. Bunemann et al. (eds.). Springer, Berlin, Germany, pp. 215 243. National Research Council (NRC), 2002. Biosolids applied to land: advancing standards and practices. United States Environmental P rotection Agency, Committee on Toxicants and Pathogens in Biosolids Applied to Land. National Academies Press, Washington, D.C. Nguyen, M.C.P., Woerther, P.L., Bouvet, M., Andremont, A., Leclercq, R., Canu, A., 2009. Escherichia coli as reservoir for macro lide resistance genes. Emerg. Infect. Dis. 15, 1648 1650. Noble, C.G., Barnard, F.M., Maxwell, A., 2003. Quinolone DNA interaction: sequence dependent binding to single stranded DNA reflects the interaction within the gyrase DNA complex. Antimicrob. Agents Chemother. 47, 854 862. Nowara, A., Burhenne, J., Spiteller, M., 1997. Binding of fluoroquinolone carboxylic acid derivatives to clay minerals. J. Agric. Food Chem. 45, 1459 1463 amended soils and their po tential for uptake by plants. Sci. Total Environ. 185, 71 81.

PAGE 239

239 O'Reilly, A., Smith, P., 1990. Development of methods for predicting the minimum concentrations of oxytetracycline capable of exerting a selection for resistance to this agent. Aquaculture. 180, 1 11. OECD, 2010. OECD test no. 317: bioaccumulation in terrestrial oligochaetes Available at OECD 317: http://www.oecd ilib rary.org/environment/test no 317 bioaccumulation in terrestrial oligochaetes_9789264090934 en accessed on 2/12/18. Ortiz de Garca, S.A., Pinto, G., Garca Encina, P.A., Irusta Mata, R., 2014. Ecotoxicity and environmental risk assessment of pharmaceutical s and personal care products in aquatic environments and wastewater treatment plants. Ecotoxicol. 23, 1517 1533. biosolids borne triclosan in terrestrial organisms. Environ.Tox icol. Chem. 31, 646 653. bioaccumulation of biosolids borne triclosan in food crops. Environ. Toxicol. Chem. 31, 2130 2137. Papageorgiou, M., Kosma, C., Lambropoulou, D., 2016. Seaso nal occurrence, removal, mass loading and environmental risk assessment of 55 pharmaceuticals and personal care products in a municipal wastewater treatment plant in central Greece. Sci. Total Environ. 543, 547 569. Park, I., Zhang, N., Ogunyoku, T., Youn g, T., Scow, K., 2013. Effects of triclosan and biosolids on microbial community composition in an agricultural soil. Water Environ. Res. 85, 2237 2242. Parnham, M.J., Haber, V., Giamarellos Bourboulis, E.J., Perletti, G., Verleden, G.M., Vos. R., 2014. Az ithromycin: mechanisms of action and their relevance for clinical applications. Pharmacol. Ther. 143, 225 245. Parshikov, I.A., Freeman, J.P., Lay, Jr., J.O., Beger, R.D., Williams, A.J., Sutherland, J.B., 2000. Microbiological transformation of enrofloxa cin by the fungus Mucor ramannianus Appl. Environ. Microbiol. 66, 2664 2667. Peltzer, P.M., Lajmanovich, R.C., Attademo, A.M., Junges, C.M., Teglia, C.M., Martinuzzi, C., Curi, L., Culzoni, M.J., Goicoechea, H.C., 2017. Ecotoxicity of veterinary enrofloxa cin and ciprofloxacin antibiotics on anuran amphibian larvae. Environ. Toxicol. Pharmacol. 51, 114 123. Peterson, J.W., O'Meara, T.A., Seymour, M.D., Wang, W., Gu, B., 2009. Sorption mechanisms of cephapirin, a veterinary antibiotic, onto quartz and feldspar minerals as detected by Raman spectroscopy. Environ. Pollut. 157, 1849 56.

PAGE 240

240 Petrie, B., Barden, B., Kasprzyk Horderna, B., 2015. A review on emerging contaminants in wastewaters and the environment: current knowledge, understudied areas and recomme ndations for future monitoring. Water Res. 72, 3 27. Petropoulos, A.D., Kouvela, E.C., Starosta, A.L., Wilson, D.N., Dinos, G.P., Kalpaxis, D.L., 2009. Time resolved binding of azithromycin to Escherichia coli ribosomes. J. Mol. Biol. 385, 1179 1192. Posc henrieder, C., Cabot, C., Martos, S., Gallego, B., Barcel, J., 2013. Do toxic ions induce hormesis in plants? Plant Sci. 212, 15 25. Prosser, R.S., Sibley, P.K., 2015. Human health risk assessment of pharmaceuticals and personal care products in plant tis sue due to biosolids and manure amendments, and wastewater irrigation. Environ. Int. 75, 223 233. Radjenovic, J., Petrovic, M., Barcelo, D., 2009. Fate and distribution of pharmaceuticals in wastewater and sewage sludge of the conventional activated sludge (CAS) and advanced membrane bioreactor (MBR) treatment. Water Res. 43, 831 841. Rahube, T.O., Marti, R., Scott, A., Tien, Y.C., Murray, R., Sabourin, L., Duenk, P., Lapen, D.R., Topp, E., 2016. Persistence of antibiotic resistance and plasmid associated g enes in soil following application of sewage sludge and abundance on vegetables at harvest. Can. J. Microbiol. 62, 600 607. Redshaw, C.H., Wootton, V.G., Rowland, S.J., 2008. Uptake of the pharmaceutical fluoxetine hydrochloride from growth medium by Brass icaceae Phytochem. 69, 2510 2516. Reid, B.J., Jones, K.C., Semple, K.T., 2000. Bioavailability of persistent organic pollutants in soils and sediments -a perspective on mechanisms, consequences and assessment. Environ. Pollut. 108, 103 12. Robinson, A.A., Belden, J.B., Lydy, M.J., 2005. Toxicity of fluoroquinolone antibiotics to aquatic organisms. Environ. Toxicol. Chem. 24, 423 30. Rosendahl, I., Siemens, J., Kindler, R., Groeneweg, J., Zimmermann, J., Czerwinski, S., Lamshft, M., Laabs, V., Wilke, B.M., Vereecken, H., Amelung, W., 2012. Persistence of the fluoroquinolone antibiotic difloxacin in soil and lacking effects on nitrogen turnover. J. Environ. Qual. 41, 1275 1283. Ross, D. L., Elkinton, S. K., Riley, C. M., 1992. Physicochemical properties of t he fluoroquinolone antimicrobials. IV. 1 octanol/water partition coefficients and their Rotthauwe, J H., Witzel, K P., Liesack, W., 1997. The ammonia monooxygenase structural gene amoA as a functional marker: molecular fine scale analysis of natural ammonia oxidizing populations. Appl. Environ. Microbiol. 63, 4704 4712.

PAGE 241

241 Sabourin, L., Duenk, P., Bonte Gelok, S., Payne, M., Lapen, D.R., Topp, E., 2012. Uptake of pharmaceuticals, hormones and parabens int o vegetables grown in soil fertilized with municipal biosolids. Sci. Total Environ. 431, 233 236. Sakurai, M., Wasaki, J., Tomizawa, Y., Shinano, T., Osaki, M., 2008. Analysis of bacterial communities on alkaline phosphatase genes in soil supplied with org anic matter. Soil Sci. Plant Nut. 54, 62 71. Samanta I., Bandyopadhyay S. 2017. Infectious diseases. In: Samanta, I., Bandyopadhyay, S. (eds.). Pet b ird d iseases and c are Springer, Berlin, Germany, pp. 13 166. Sample, B.E., Aplin, M.S., Efroymson, R.A., Suter, G.W., Welsh, C.J.E., 1997. Methods and tools for estimation of the exposure of terrestrial wildlife to contaminants. ORNL/TM 13391. Prepared for U.S. Department of Energy, Office of Environmental Policy and Assistance. Prepared by Oak Ridge National Laboratory, Oak Ridge, TN. Sassman, S.A., Lee, L.S., 2005. Sorption of three tetracyclines by several soils: assessing the role of pH and cation exchange. Environ. Sci. Technol. 39, 7452 7459. Schroll, R., Scheunert, I., 1992. A laboratory system to dete rmine separately the uptake of organic chemicals from soil by plants and by leaves after vaporization. Chemosphere. 24, 97 108 Sengelv, G., Agers, Y., Halling Srensen, B., Baloda, S.B., Andersen, J.S., Jensen, L.B., 2003. Bacterial antibiotic resistance levels in Danish farmland as a result of treatment with pig manure slurry. Environ. Int. 28, 587 595. Senta, I., Terzic, S., Ahel, M., 2013. Occurrence and fate of dissolved and particulate antimicrobials in municipal wastewater treatment. Water Res. 47, 705 714. Shen, L.L., Pernet, A.G., 1985. Mechanism of inhibition of DNA gyrase by analogues of nalidixic acid: the target of the drugs is DNA. Biochemistry. Proc. Nati. Acad. Sci. USA. 82, pp. 307 311. Shenker, M., Harush, D., Ben Ari, J., Chefetz, B., 20 11. Uptake of carbamazepine by cucumber plants a case study related to irrigation with reclaimed wastewater. Chemosphere. 82, 905 910. Silva, F., Lourenco, O., Queiroz, J.A., Domingues, F.C., 2011. Bacteriostatic versus bactericidal activity of ciproflox acin in Escherichia coli assessed by flow cytometry using a novel far red dye. J. Antibiot. 64, 321 325. Simon, N., Bochman, M.L., Seguin, S., Brodsky, J.L., Seibel, W.L., Schwacha, A., 2013. Ciprofloxacin is an inhibitor of the Mcm2 7 replicative helicase Biosci. Rep. 33, e00072.

PAGE 242

242 Singer, A.C., Shaw, H., Rhodes, V., Hart, A., 2016. Review of antimicrobial resistance in the environment and its relevance to environmental regulators. Front. Microbiol. 7, 1728. Snyder, E.H., O'Connor, G.A., McAvoy, D.C., 2010 Fate of 14 C triclocarban in biosolids amended soils. Sci. Total Environ. 408, 2726 2732. Snyder, E.H., O'Connor, G.A., McAvoy, D.C., 2011. Toxicity and bioaccumulation of biosolids borne triclocarban (TCC) in terrestrial organisms. Chemosphere. 82, 460 4 67. Snyder, E.H., O'Connor, G.A., 2013. Risk assessment of land applied biosolids borne triclocarban (TCC). Sci. Total Environ. 442, 437 444. Sone, M., Kishigami, S., Yoshihisa, T., Ito, K., 1997. Roles of disulfide bonds in bacterial alkaline phosphatase. J. Biol. Chem. 272, 6174 6178. Sposito, G., 1984. The surface chemistry of soils. Oxford Univ. Press, Oxford, England. Sterngren, A. E., Hallin, S., Bengtson, P., 2015. Archaeal ammonia oxidizers dominate in numbers, but bacteria drive gross n itrification in N amended grassland soil. Front. Microbiol., 6, 1350. lactams and florfenicol antibiotics remain bioactive in soils while ciprofloxacin, neomycin, and tetracycline are neutrali zed. Appl. Environ. Microbiol. 77, 7255 7260. Sunderland, J., Tobin, C.M., Hedges, A.J., MacGowan, A.P., White, L.O., 2001. Antimicrobial activity of fluoroquinolone photodegradation products determined by parallel line bioassay and high performance liquid chromatography. J. Antimicrob. Chemother. 47, 271 275. Sutcliffe, J., Grebe, T., Tait Kamradt, A., Wondrack, L., 1996. Detection of erythromycin resistant determinants by PCR. Antimicrob. Agents Chemother. 40, 2562 2566. Suter, G.W., Efroymson, R.A., Samp le, B.E., Jones, D.S., 2000. Ecological risk assessment for contaminated sites. CRC Press, Boca Raton, Florida. Suter, G.W., 2007. Ecological risk assessment. CRC Press, Boca Raton, Florida. Takacs Novak, K., Jozan, M., Hermeczi, I., Szasz, G., 1992. Lipop hilicity of antibacterial fluoroquinolones. Int. J. Pharm. 79, 89 96. Taylor, A.E., Zeglin, L.H., Wanzek, T.A., Myrold, D.D., Bottomley, P.J., 2012. Dynamics of ammonia oxidizing archaea and bacteria populations and contributions to soil nitrification pote ntials. ISME J. 6, 2024 2032.

PAGE 243

243 Theerachat, M., Virunanon, C., Chulalaksananukul, S., Sinbuathong, N., Chulalaksananukul, W., 2011. nirK and nirS nitrite reductase genes from non agricultural forest soil bacteria in Thailand. World J. Microbiol. Biotechnol. 27, 999 1003. Thiele Bruhn, S., 2003. Pharmaceutical antibiotic compounds in soils a review. J. Plant Nutr. Soil Sci. 166, 145 167. Thiele Bruhn, S., Beck, I C., 2005. Effects of sulfonamide and tetracycline antibiotics on soil microbial activity and mi crobial biomass. Chemosphere. 59, 457 465. Tong, L., Eichhorn, P., Prez, S., Wang, Y., Barcel, D., 2011. Photodegradation of azithromycin in various aqueous systems under simulated and natural solar radiation: kinetics and identification of photoproducts Chemosphere. 83, 340 348. Topp, E., Renaud, J., Sumarah, M., Sabourin, L., 2016. Reduced persistence of the macrolide antibiotics erythromycin, clarithromycin and azithromycin in agricultural soil following several years of exposure in the field. Sci. To tal Environ. 562, 136 144. Tourna, M., Freitag, T.E., Nicol, G.W., Prosser, J.I., 2008. Growth, activity and temperature responses of ammonia oxidizing archaea and bacteria in soil microcosms. Environ. Microbiol. 10, 1357 1364. Trapp, S., 2000. Modelling u ptake into roots and subsequent translocation of neutral and ionizable organic compounds. Pest Manag. Sci. 56, 767 778. Trapp, S., 2009. Bioaccumulation of polar and ionizable compounds in plants. In: Devillers, J. (eds). Ecotoxicology modeling. Springer, Berlin, Germany, pp. 299 353. Troxler, R.F., Brown, A.S., Kost, H P., 1978. Quantitative degradation of radiolabeled phycobiliproteins. Chromic acid degradation of C Phycocyanin. Eur. J. Biochem. 87, 181 189. Uivarosi, V., 2013. Metal complexes of quinolon e antibiotics and their applications: an update. Molecules. 18, 11153 11197. US National Library of Medicine, 2002. PubChem Open chemistry database. CID 2764. Available at https://PubChem.ncbi.nlm.nih.gov/compound/ciprofloxacin#section=Top accessed on 1/10/2018. US National Library of Medicine, 2002. PubChem Open chemistry database. CID 2764. Available at https://PubChem.ncbi.nlm.nih.gov/compound/447043 accessed on 1/10/2018.

PAGE 244

244 USEPA, 1983. Method 150.1. In: Methods for chemical analysis of water and wastes. EPA 600/4 79/020, pp. 150.1 3. USEPA, 1 991. Regional guidance on handling chemical concentration data near the detection limit in risk assessments. United States Environmental Protection Agency, Region 3. Hazardous Waste Management Division Office of Superfund Programs. Available at https://www.epa.gov/risk/regional guidance handling chemical concentration data near detection limit risk assessments assessed on 1/1 1/2018. USEPA, 1993. Wildlife exposure factors handbook. Volumes I and II. EPA/600/R 93/187. Office of Health and Environmental Assessment and Office of Research and Development, Washington, DC. December. USEPA, 1994. Method 200.7, Revision 4.4. Determinat ion of metals and trace elements in water and wastes by inductively coupled plasma atomic emission spectrometry. EPA 600/R 94 111. USEPA, 1995. A guide to the biosolids risk assessments for the EPA Part 503 Rule. EPA/832 B 93 005. USEPA, Washington, DC. US EPA, 2000. Biosolids technology fact sheet. EPA 832 F 00 064 Office of Water, Washington, DC. Available at https://www3.epa.gov/npdes/pubs/land_application.pdf accessed on 2/13/18. USEPA, 2003. Multimedia, multipathway, and multireceptor risk assessment (3MRA) modeling system: Volume II: site based, regional, and national Data. SAB Review Draft. EPA530 D 03 001b. Office of Research and Development, Athens, GA, and Office of Solid Was te, Washington DC. Available at: http://www.epa.gov/osw/hazard/wastetypes/wasteid/hwirwste/risk03.htm accessed on 2/1/18. USEPA, 2007. USEPA Method 1694: Pharmaceuticals and personal care products in water, soil, sediment, and biosolids by HPLC/MS/MS. EPA 821 R 08 002 Washington DC Office of Water, United States Environmental Protection Agency, Washington, DC. USEPA, 2008. Child specific exposure factors handbook. EPA 600/ R 06 096F. National Center for Environmental Assessment, Cincinnati, OH. Available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=199243 accessed on 2/1/18 USEPA, 2009. Bio solids: targeted national sewage sludge survey report overview. EPA 822 R 08 014. Available at http://water.epa.gov/ scitech/wastetech/biosolids/upload/2009_04_23_biosolids_t nsss overview.pdf accessed on 1/10/18

PAGE 245

245 USEPA, 2011. Exposure factors handbook. EPA/600/R 090/052F. National Center for Environmental Assessment, Office of Research and Development, Washington, DC. A vailable at http://www.epa.gov/ncea/efh/pdfs/efhcomplete.pdf accessed on 2/1/18. USEPA, Estimation Program Interface (EPI) Suite, 2011. Available at http://www.epa.gov/sab/panels/epi_suite_review_panel.htm accessed on 1/4/2018. USEPA, 2012. Ecological effects test guidelines OCSPP 850.3100: earthworm sub chronic toxicity test. EPA 712 C 016. Available at https://nepis.epa.gov/Exe/ZyPDF.cgi/P100IREJ.PDF?Dockey=P100IREJ.PDF assessed on 1/10/2018. van den Brink, N.W., Arblaster, J.A., Bowman, S.R., Conder, J.M., Elliott, J.E., Johnson, M.S. Muir, D.C., Natal da Luz, T., Rattner, B.A., Sample, B.E., Shore, R.F., 2016. Use of terrestrial field studies in the derivation of bioaccumulation potential of chemicals. Integr. Environ. Assess. Manag. 12, 135 45. van Moorleghem, C., Schutter, N.D., Sm olders, E., Merckx, R., 2013. The bioavailability of colloidal and dissolved organic phosphorus to the alga Pseudokirchneriella subcapitata in relation to analytical phosphorus measurements. Hydrobiologia. 709, 41 53. Varanda, F., Pratas de Melo, M.J., Cac o, A.I., Dohrn, R., Makrydaki, F.A., Voutsas, E., Tassios, D., Marrucho, I.M., 2006. Solubility of antibiotics in different solvents. 1. hydrochloride forms of tetracycline, moxifloxacin, and ciprofloxacin. Ind. Eng. Chem. Res. 45, 6368 6374. Vasconcelos, T.G., Henriques, D.M., Knig, A., Martins, A.F., Kummerer, K., 2009. Photo degradation of the antimicrobial ciprofloxacin at high pH: identification and biodegradability assessment of the primary by products. Chemosphere. 76, 487 493. Vasudevan, D., Brulan d, G.L., Torrance, B.S., Upchurch, V.G., MacKay, A.A., 2009. pH dependent ciprofloxacin sorption to soils: interaction mechanisms and soil factors influencing sorption. Geoderma. 151, 68 76. Vikesland, P.J., Pruden, A., Alvarez, P.J.J., Aga, D.S., Buergman n, H., Li, X., Manaia, C.M., Nambi, I.M., Wigginton, K.R., Zhang, T., Zhu, Y G., 2017. Towards a comprehensive strategy to mitigate dissemination of environmental sources of antibiotic resistance. Environ. Sci. Technol. doi: 10.1021/acs.est.7b03623. Verlicchi, P., Zambello, E., 2015. Pharmaceuticals and personal care products in untreated and treated sewage sludge: occurrence and environmental risk in the case of application on soil A critical review. Sci. Total Environ.538, 750 767.

PAGE 246

246 Vestel, J., Ca ldwell, D.J., Constantine, L., D'Aco, V.J., Davidson, T., Dolan, D.G., Millard, S.P., Murray Smith, R., Parke, N.J., Ryan, J.J., Straub, J.O., Wilson, P., 2016. Use of acute and chronic ecotoxicity data in environmental risk assessment of pharmaceuticals. Environ. Toxicol. Chem. 35, 1201 1212. von Hellens, A., 2015. Pharmaceuticals leaching from biosolids amended soils. Faculty of Natural Resources and Agricultural Sciences. Swedish University of Agricultural Sciences, Uppsala, Sweden. Available at https://stud.epsilon.slu.se/8188/11/von_hellens_a_150817.pdf assessed on 1/10/2018. Walters, E., McClellan, K., Halden, R., U., 2010. Occurrence and loss over three years of 72 pharmaceut icals and personal care products from biosolids soil mixtures in outdoor mesocosms. Water Res. 44, 6011 6020. Wang, W X., Guo, L., 2000. Bioavailability of colloid bound Cd, Cr, and Zn to marine plankton. Mar. Ecol. Prog. Ser. 202, 41 49. Weed, S.B., Weber J.B., 1968. The effect of adsorbent charge on the competitive adsorption of divalent organic cations by layer silicate minerals. American Mineralogist. 53, 478 490. Weil, R., Brady, N.C., 2016. The nature and properties of soils. 15th edition. Pearson, N Y. Wen, B., Huang, R., Wang, P., Zhou, Y., Shan, X.Q., Zhang, S., 2011. Effect of complexation on the accumulation and elimination kinetics of cadmium and ciprofloxacin in the earthworm Eisenia fetida Environ. Sci. Technol. 45, 4339 4345. Wetzstein, H G., Stadler, M., Tichy, H V., Dalhoff, A., Karl, W., 1999. Degradation of ciprofloxacin by basidiomycetes and identification of metabolites generated by the brown rot fungus Gloeophyllum striatum Appl. Environ. Microbiol. 65, 1556 1563. Wise, R., Andrews, J. M., Edwards, L.J., 1983. In vitro activity of Bay 09867, a new quinoline derivative, compared with those of other antimicrobial agents. Antimicrob. Agents Chemother. 23, 559 564. World Health Organization (WHO), 2001. Report on integrated risk assessment. WHO/IPCS/IRA/01/12. World Health Organization, Geneva, Switzerland. Wu, C., Spongberg, A.L., Witter, J.D., 2009. Sorption and biodegradation of selected antibiotics in biosolids. J. Environ. Sci. Health. A. Tox. Hazard Subst. Environ. Eng. 44, 454 461

PAGE 247

247 Wu C., Spongberg, A.l., Witter, J.D., Fang, M., Czajkowski, K.P., 2010. Uptake of pharmaceutical and personal care products by soybean plants from soils applied with biosolids and irrigated with contaminated water. Environ. Sci. Technol. 44, 6157 6161. Wu, Q., Li, Z., Hong, H., Li, R., Jiang, W.T., 2013. Desorption of ciprofloxacin from clay mineral surfaces. Water Res. 47, 259 268. Wu, X., Dodgen, L.K., Conkle, J. L., Gan, J., 2015. Plant uptake of pharmaceutical and personal care products from recycled wat er and biosolids: a review. Sci. Total Environ. 536, 655 666. Young, T.M., Ogunyoku, T., Giudice, B., Scow, K., Park, I., Zhang, N., 2011. Antimicrobials and other trace organics in biosolids: effects on soil microbial processes and potential for endocrine disruption. In: Proceedings of the Water Environ. Federation, Residuals and Biosolids, pp. 478 494. Youngquist, C.P., Liu, J., Orfe, L.H., Jones, S.S., Call, D.R., 2014. Ciprofloxacin residues in municipal biosolid s compost do not selectively enrich popul ations of resistant bacteria. Appl. Environ. Microbiol. 80, 7521 7526. Zuckerman, J.M., 2000. The newer macrolides: azithromycin and clarithromycin. Infect. Dis. Clin. North Am. 14, 449 462.

PAGE 248

248 BIOGRAPHICAL SKETCH Harma n Sidhu is a dreamer and sometimes a procrastinator who, from the foothills of Himalayas, travelled to the sunshine state of Florida for his graduate studies. An admirer of mother nature and its wonders, Harman graduated with Ph.D. in environment related s ciences, and along the way grew both personally and academically. Harman is a movie buff, but also likes to learn and see different cultures, traditions, and places. One d ay he plans to ditch responsibilities for a while and travel the entire world. Passio nate and zealo us about life and the wonders of nature, Harman is determined to contribute to the betterment of human civilization and planet E arth