Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effects of pulp mill wastewater treatment on phytosterol biotransformation and genomic response in rainbow… Miskelly, Andrea 2009

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2009_spring_miskelly_andrea.pdf [ 3.14MB ]
Metadata
JSON: 24-1.0063112.json
JSON-LD: 24-1.0063112-ld.json
RDF/XML (Pretty): 24-1.0063112-rdf.xml
RDF/JSON: 24-1.0063112-rdf.json
Turtle: 24-1.0063112-turtle.txt
N-Triples: 24-1.0063112-rdf-ntriples.txt
Original Record: 24-1.0063112-source.json
Full Text
24-1.0063112-fulltext.txt
Citation
24-1.0063112.ris

Full Text

 EFFECTS OF PULP MILL WASTEWATER TREATMENT ON PHYTOSTEROL BIOTRANSFORMATION AND GENOMIC RESPONSE IN RAINBOW TROUT    by  ANDREA MISKELLY  B.Sc., The University of Victoria, 2003      A THESIS SUBMITTED IN PARTIAL FULFUILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF APPLIED SCIENCE   in   THE FACULTY OF GRADUATE STUDIES   (Civil Engineering)             THE UNIVERSITY OF BRITISH COLUMBIA  April 2009  © Andrea Miskelly, 2009  ii ABSTRACT  Phytosterols are ubiquitous, naturally occurring plant steroids that are released into solution during pulping processes.  These chemicals have formerly been identified as endocrine disrupting compounds.  Phytosterols are known to be degraded by microorganisms during biological wastewater treatment, but the metabolites produced during this process and their subsequent effects on fish remain unclear.   The objectives of the present study were to determine whether phytosterol biotransformation products can be identified in pulp mill effluents and to establish a relationship between wastewater treatment conditions, concentrations of phytosterols and their metabolites, and genomic response in fish.  To accomplish this, a benchscale biological wastewater treatment system was operated using aerobic and microaerophilic treatment trains with a range of solids retention times (SRTs).  By varying these operational parameters, four independently treated effluents were produced.  Treated effluents were characterized for concentrations of phytosterols and their potential biotransformation products, including thirteen estrogenic and three androgenic compounds.  Results showed microaerophilic effluents to contain sterols concentrations at levels two to three orders of magnitude higher than aerobic effluents.  Sterols biotransformation products were identified in each effluent and included cholesterol, desmosterol, dihydrocholesterol and estrone.  Estrone was the most ecologically relevant metabolite identified and was detected in final treated aerobic effluent at a maximum concentration of 623 ng/L.  Following chemical characterization, underyearling rainbow trout were exposed to the effluents at 1% and 10% g/g for 96 hours.  Liver tissues were dissected, persevered and analyzed by quantitative polymerase chain reaction.  Genes selected for examination of molecular level responses were VTG1, VEPB, VEPG, NME, AR and THRa.  Results demonstrated a strong estrogenic and thyroid response with a clear pattern among wastewater treatment conditions.  Microaerophilic effluent generated with a short SRT induced the strongest response and was able to do so at both 10 % and 1 % concentrations. Aerobic effluents induced a significant response at 10 % effluent concentration but only for NME, which proved the most sensitive of all genes tested.  This study highlights the use of genomic analysis as a predictive indicator for endocrine disrupting effects and provides insight into the operation of wastewater treatment systems for reduction of biological effects.    iii TABLE OF CONTENTS  ABSTRACT ........................................................................................................................................................ ii  TABLE OF CONTENTS .............................................................................................................................. iii  LIST OF TABLES ..........................................................................................................................................vii  LIST OF FIGURES ...................................................................................................................................... viii  LIST OF ACRONYMS AND ABBREVIATIONS................................................................................... ix  ACKNOWLEDGEMENTS ........................................................................................................................... x  1 INTRODUCTION .......................................................................................................................... 1  2 LITERATURE REVIEW ............................................................................................................... 3  2.1 THE ENDOCRINE DISRUPTION PROBLEM ..................................................................... 3 2.1.1 The Endocrine System ........................................................................................................... 3 2.1.2 Endocrine Disrupting Compounds ..................................................................................... 7 2.1.2.1 Composition of pulp mill effluents ................................................................................. 8 2.1.3 Endocrine Disruption in Fish Exposed to Pulp Mill Effluents .................................... 10 2.1.4 Impacts of Biological Wastewater Treatment .................................................................. 12  2.2 PHYTOSTEROLS ......................................................................................................................... 14 2.2.1 Nature and Source of Phytosterols .................................................................................... 14 2.2.2 Phytosterols as EDCs .......................................................................................................... 16 2.2.3 Impact of Biological Wastewater Treatment on Phytosterols ....................................... 17 2.2.4 Phytosterol Biotransformations ......................................................................................... 21  2.3 LITERATURE REVIEW SUMMARY ...................................................................................... 25  3 OBJECTIVES ................................................................................................................................ 26  4 MATERIALS AND METHODS............................................................................................... 28  4.1 STUDY SITE .................................................................................................................................. 28 4.1.1 Mill Characteristics ............................................................................................................... 28 4.1.2 Wastewater Treatment ......................................................................................................... 28 4.1.3 Mill Sampling ......................................................................................................................... 29  4.2 BENCHSCALE BIOLOGICAL WASTEWATER TREATMENT SYSTEM .................. 30 4.2.1 System Characteristics .......................................................................................................... 30 4.2.1.1 Oxygen, nutrients and pH ............................................................................................. 32 4.2.2 Operating Conditions .......................................................................................................... 33 4.2.2.1 DO .................................................................................................................................... 33  iv 4.2.2.2 SRT.................................................................................................................................... 34 4.2.3 Sampling ................................................................................................................................. 34 4.2.3.1 Sample group 1 ................................................................................................................ 35 4.2.3.2 Sample groups 2 and 3 ................................................................................................... 36 4.2.3.3 Sample group 4 ................................................................................................................ 38 4.2.4 BBWTS Operational Parameters ....................................................................................... 38 4.2.4.1 DO, temperature and pH .............................................................................................. 39 4.2.4.2 HRT .................................................................................................................................. 39 4.2.4.3 TSS .................................................................................................................................... 39 4.2.4.4 SRT and wasting .............................................................................................................. 40 4.2.4.5 COD ................................................................................................................................. 41 4.2.4.6 Nutrients........................................................................................................................... 42 4.2.4.7 QA/QC ............................................................................................................................ 43  4.3 QUANTIFICATION OF PLANT STEROLS AND POTENTIAL BIOTRANSFORMATION PRODUCTS ................................................................................. 43 4.3.1 Analytical Schedule and Methods ...................................................................................... 43 4.3.2 QA/QC .................................................................................................................................. 46 4.3.2.1 Surrogates ......................................................................................................................... 46 4.3.2.2 Replicates .......................................................................................................................... 46 4.3.2.3 Blanks ................................................................................................................................ 47 4.3.2.4 Direct spike ...................................................................................................................... 47 4.3.3 Organic Chemistry Data Analysis ...................................................................................... 47  4.4 GENOMIC RESPONSE OF RAINBOW TROUT TO EXPOSURE TO TREATED PULP MILL EFFLUENTS .......................................................................................................... 48 4.4.1 Fish Exposure Bioassays ..................................................................................................... 48 4.4.1.1 Bioassay test conditions ................................................................................................. 50 4.4.1.2 Dissection and preservation .......................................................................................... 50 4.4.1.3 QA/QC ............................................................................................................................ 51 4.4.2 Genomic Analytical Methods ............................................................................................. 51 4.4.2.1 Extraction of total RNA ................................................................................................ 52 4.4.2.1.1 QA/QC ...................................................................................................................... 53 4.4.2.2 cDNA synthesis .............................................................................................................. 54 4.4.2.2.1 QA/QC ...................................................................................................................... 55 4.4.2.3 QPCR ................................................................................................................................ 55 4.4.2.3.1 QA/QC ...................................................................................................................... 57 4.4.2.3.1.1 NTC .................................................................................................................... 58 4.4.2.3.1.2 Dissociation curves........................................................................................... 58 4.4.2.3.1.3 Reference sample .............................................................................................. 59 4.4.2.4 Gene selection ................................................................................................................. 59 4.4.2.5 Method development ..................................................................................................... 62 4.4.2.6 Final sample preparation ................................................................................................ 62 4.4.2.7 Bioinformatics ................................................................................................................. 63  v 5 RESULTS AND DISCUSSION ................................................................................................. 65  5.1 BENCHSCALE BIOLOGICAL WASTEWATER TREATMENT SYSTEM .................. 65 5.1.1 DO, Temperature and pH .................................................................................................. 65 5.1.2 HRT ........................................................................................................................................ 69 5.1.3 SRT ......................................................................................................................................... 69 5.1.4 MLSS and TSS ...................................................................................................................... 71 5.1.5 COD ....................................................................................................................................... 74 5.1.6 Nutrients ................................................................................................................................ 76 5.1.7 BBWTS Summary ................................................................................................................ 79  5.2 QUANTIFICATION OF PLANT STEROLS AND POTENTIAL BIOTRANSFORMATION PRODUCTS ................................................................................. 80 5.2.1 Compatibility of AD, ADD, T and Pr with the Environment Canada Sterols Method ................................................................................................................................... 80 5.2.2 Influent Characteristics ........................................................................................................ 81 5.2.3 Effluent Characteristics ....................................................................................................... 85 5.2.4 Sterols Removal Efficiencies............................................................................................... 87 5.2.5 Sterols Biotransformation ................................................................................................... 89 5.2.5.1 Cholesterol (chole) .......................................................................................................... 91 5.2.5.2 Dihyrocholesterol (DHC) .............................................................................................. 92 5.2.5.3 Desmosterol (desmo) ..................................................................................................... 93 5.2.5.4 Estrone (E1) .................................................................................................................... 94 5.2.6 Statistical Comparison of Treatment Systems .................................................................. 96 5.2.7 Phase-Separation Analysis and Sterols Mass Balance ..................................................... 99 5.2.7.1 Phase-Separation Analysis ............................................................................................. 99 5.2.7.2 Mass Balance .................................................................................................................. 102 5.2.8 Summary of Sterols Removal and Biotransformation .................................................. 105  5.3 GENOMIC RESPONSE OF RAINBOW TROUT TO EXPOSURE TO TREATED PULP MILL EFFLUENTS ........................................................................................................ 107 5.3.1 Fish Exposure Bioassays ................................................................................................... 107 5.3.2 Genomic Analysis ............................................................................................................... 109 5.3.2.1 Gene selection ............................................................................................................... 109 5.3.2.2 Sample preparation ....................................................................................................... 112 5.3.2.3 Bioinformatics ............................................................................................................... 112 5.3.2.3.1 Method development ............................................................................................. 113 5.3.2.3.2 Final data analysis ................................................................................................... 113 5.3.2.3.2.1 Estrogenic Response ...................................................................................... 115 5.3.2.3.2.2 Androgenic Response .................................................................................... 117 5.3.2.3.2.3 Thyroid Hormone Receptor Response ....................................................... 117 5.3.2.4 Summary of Genomic Analysis .................................................................................. 118  6 CONCLUSIONS AND RECOMMENDATIONS.............................................................. 120  REFERENCES ............................................................................................................................................ 127  vi APPENDIX A:  Distribution Plots for Sterols Chemistry .................................................................... 138 A1: β-Sitosterol Q-Q Plot ........................................................................................................................ 138 A2: Campesterol Q-Q Plot ...................................................................................................................... 139 A3: Stigmasterol Q-Q Plot ....................................................................................................................... 140 A4: Cholesterol Q-Q Plot ........................................................................................................................ 141 A5: Dihydrocholesterol Q-Q Plot .......................................................................................................... 142 A6: Estrone Q-Q Plot .............................................................................................................................. 143 A7: Coprostanol Q-Q Plot ...................................................................................................................... 144 A8: Desmosterol Q-Q Plot ...................................................................................................................... 145  APPENDIX B:  Animal Care Approval ................................................................................................... 146  APPENDIX C:  Data Tables for BBWTS Operation and Sterols Chemistry .................................... 147 C1: Temperature and Dissolved Oxygen ............................................................................................... 147 C2: HRT and SRT ..................................................................................................................................... 151 C3: Suspended Solids ................................................................................................................................ 154 C4: Chemical Oxygen Demand ............................................................................................................... 156 C5: Nutrients .............................................................................................................................................. 158 C6: Sterols and Potential Biotransformation Products ........................................................................ 160  APPENDIX D:  QA/QC ........................................................................................................................... 162 D1: Benchscale Biological Wastewater Treatment System ................................................................. 162 D2: Quantification of Plant Sterols and Potential Biotransformation Products ............................. 165 D2.1: Surrogates .................................................................................................................................... 165 D2.2: Replicates .................................................................................................................................... 165 D2.3: Blanks .......................................................................................................................................... 170 D2.4: Direct spike ................................................................................................................................. 171 D3: Genomic Response of Rainbow Trout to Exposure to Treated Pulp Mill Effluents ............. 172 D3.1: Bioassays ..................................................................................................................................... 172 D3.2: Genomic Analysis ...................................................................................................................... 172  APPENDIX E:  Standard Curves for Genomic Analysis ...................................................................... 173 E1: L8 Standard Curve ............................................................................................................................. 173 E2: AR Standard Curve ............................................................................................................................ 174 E3: VTG1 Standard Curve ...................................................................................................................... 175 E4: VEPb Standard Curve ....................................................................................................................... 176 E5: VEPg Standard Curve ....................................................................................................................... 177 E6: NME Standard Curve ........................................................................................................................ 178 E7: THRa Standard Curve ....................................................................................................................... 179         vii LIST OF TABLES  Table 2.1: Overview of wood extractives ....................................................................................................... 9 Table 2.2:  Steroid uses in medicine .............................................................................................................. 21 Table 4.1:  Sample group details. ................................................................................................................... 35 Table 4.2:  Filtration volumes for calculation of TSS. ................................................................................ 40 Table 4.3:  Limits of quantitation for parameters in Environment Canada’s sterols method (LOQ). .......................................................................................................................................... 44 Table 4.4:  Analytical schedule for organic parameters, grouped by steroid classification. .................. 45 Table 4.5:  Bioassay treatment and control groups. .................................................................................... 49 Table 4.6:  Select test conditions for the 96 hour rainbow trout bioassays. ............................................ 50 Table 4.7:  QPCR thermocycle profile.  This profile was used for all QPCR plates analyzed. ............ 56 Table 4.8:  Genes screened in select test samples ....................................................................................... 61 Table 5.1: Summary statistics for BBWTS operational parameters .......................................................... 66 Table 5.2: Spike analysis for AD, ADD, T and Pr ...................................................................................... 81 Table 5.3: Summary statistics for influent sterols ....................................................................................... 82 Table 5.4: Statistical comparison of sterols inputs between experimental runs ..................................... 84 Table 5.5: Summary statistics for effluent sterols and biotransformation products .............................. 86 Table 5.6: Average removal efficiencies for organic parameters across treatments............................... 87 Table 5.7:  Net metabolite production across treatments .......................................................................... 91 Table 5.8:  Summary of treatment system performance p values ............................................................. 97 Table 5.9:  Summary of treatment system performance p values during the 5 d SRT .......................... 98 Table 5.10: Summary statistics for sterols and metabolite concentrations in soluble and particulate phases .......................................................................................................................................... 100 Table 5.11: Sterols mass flows ..................................................................................................................... 103 Table 5.12: Biotransformation products mass flows ................................................................................ 104 Table 5.13: Bioassay chemistry summary ................................................................................................... 108 Table 5.14: Fish physical measurements..................................................................................................... 109 Table 5.15: Gene screening results .............................................................................................................. 110 Table 5.16: Standard curve summary data.................................................................................................. 113 Table 5.17: Student’s t-test p values for genomic data analysis (control vs. treatment) ...................... 115  viii LIST OF FIGURES  Figure 2.1:  Simplification of the endocrine system ...................................................................................... 4 Figure 2.2:  Steroid nucleus (sterane) .............................................................................................................. 5 Figure 2.3:  Steroid biosynthesis in teleost fish ............................................................................................. 6 Figure 2.4:  Conceptualization of trade offs between levels of measurement ........................................ 11 Figure 2.5:  Common plant sterols ................................................................................................................ 15 Figure 2.6:  Conceptualization of biotransformation from β-sitosterol to androstenedione ............... 22 Figure 2.7:  Phytosterol biotransformation reactions used commercially to produce steroids...…….......23 Figure 4.1:  Schematic of the wastewater treatment system at the study mill. ........................................ 29 Figure 4.2:  Schematic of benchscale biological wastewater treatment system. ...................................... 32 Figure 4.3:  Conceptualization of BBWTS treatment conditions. ............................................................ 33 Figure 4.4:  Example electrophoresis results for RNA quality check ...................................................... 53 Figure 4.5:  Conceptualization of cDNA synthesis. ................................................................................... 54 Figure 4.6:  Example amplification plot. ...................................................................................................... 57 Figure 4.7:  Example dissociation curves. .................................................................................................... 59 Figure 5.1:  Reactor dissolved oxygen .......................................................................................................... 65 Figure 5.2:  Reactor temperature ................................................................................................................... 68 Figure 5.3:  Hydraulic retention time ............................................................................................................ 69 Figure 5.4:  Solids retention time ................................................................................................................... 70 Figure 5.5:  Mixed liquor suspended solids .................................................................................................. 71 Figure 5.6:  Influent and effluent suspended solids .................................................................................... 72 Figure 5.7:  Total system solids ...................................................................................................................... 74 Figure 5.8:  Dissolved chemical oxygen demand ........................................................................................ 75 Figure 5.9:  Effluent phosphorus levels ........................................................................................................ 77 Figure 5.10:  Aerobic effluent nitrogen levels .............................................................................................. 78 Figure 5.11:  Microaerophilic effluent nitrogen levels ................................................................................ 78 Figure 5.12:  Average influent sterols concentrations ................................................................................ 83 Figure 5.13:  Average influent sterols concentrations during the 5 d SRT ............................................. 85 Figure 5.14:  Average removal efficiencies for β-sito, campe, and stigma. ............................................. 89 Figure 5.15:  Average biotransformation product generation across treatments ................................... 90 Figure 5.16:  Simplification of the biotransformation of β-sito to chole................................................. 92 Figure 5.17:  Simplification of the biotransformation of chole to DHC ................................................. 93 Figure 5.18:  Simplification of the biotransformation of β-sito to chole with desmo as an intermediary .............................................................................................................................. 94 Figure 5.19:  Simplification of the biotransformation of β-sito to estrone with AD as an intermediary .............................................................................................................................. 96 Figure 5.20:  Fold change diagrams ............................................................................................................. 114    ix LIST OF ACRONYMS AND ABBREVIATIONS AD – androstenedione ADD – androstadienedione AR – androgen receptor gene AS – activated sludge ASB – aerated stabilization basin APHA – American Public Health Association BBWTS – benchscale biological wastewater                  treatment system BC – British Columbia BP – base pairs cDNA – complementary DNA Campe – campesterol Chole – cholesterol COD – chemical oxygen demand Copro – coprostanol Ct – cycle at threshold CV – coefficient of variation dCOD – dissolved COD Desmo – desmosterol DHC – dihydrocholesterol / cholestanol DNA – deoxyribonucleic acid dNTPs – deoxynucleotides DO – dissolved oxygen E1 – estrone E2 – estradiol E3 – estriol EDC – endocrine disrupting compound EEM – environmental effects monitoring ERa – estrogen receptor alpha ERb – estrogen receptor beta GC-MS – gas chromatography-mass                 spectrometry HRT – hydraulic retention time FIA – flow injection analysis 11-KT – 11-ketotestosterone LC50 – lethal concentration resulting in death of 50% of test species LOQ – limit of quantitation MLSS – mixed liquor suspended solids mRNA – messenger RNA N – nitrogen NME – nucleoside diphosphate kinase gene NTC – non-threshold control P – phosphorus PCR – polymerase chain reaction PESC – Pacific Environmental Science Centre PME – pulp mill effluent PPLER – pulp and paper mill liquid effluent regulations Pr – progesterone QA/QC – quality assurance / quality control QPCR – qualitative polymerase chain reaction RAS – return activated sludge RNA – ribonucleic acid RPM – revolutions per minute SIM – selective ion monitoring β-Sito – beta-sitosterol SRT – solids retention time Stigma – stigmasterol T – testosterone tCOD – total COD TIE – toxicity identification evaluation THRa – thyroid releasing hormone alpha gene TMP – thermomechanical pulping TSS – total suspended solids UBC – University of British Columbia UNOX – pure oxygen activated sludge process VEP – vitelline envelope protein gene VTG – vitellogenin gene  x ACKNOWLEDGEMENTS  This project was dependant on the efforts of many people including Dr. Eric Hall at the University of British Columbia (UBC) and the UBC Environmental Engineering Lab staff.  Funding was provided through the Natural Sciences and Engineering Research Council of Canada, the UBC Bridge Program and the Province of British Columbia’s Pacific Leaders Program.  Environment Canada provided significant in-kind contributions to this project through the Pacific Environmental Science Center (PESC) in North Vancouver, BC, largely through the efforts of Graham van Aggelen, Head of Environmental Toxicology.  Contributions were provided from both the environmental toxicology and organic chemistry divisions of PESC, and included expertise and guidance, lab staff support, and the use of facilities, equipment and analytical instruments.  Cooperation of the local pulp mill, which provided inputs to this study, was also a factor in the project’s success.  1 1 INTRODUCTION For at least the last two decades, endocrine disruption effects have been observed in fish living in aquatic environments that receive pulp mill effluents (PMEs) (Andersson et al. 1988; Munkittrick et al. 1997; Rickwood et al. 2006b).  This occurrence is now widely recognized by scientists, government and industry.  The extent of the situation is such that Environment Canada has mandated long term biological monitoring of all pulp mill effluent receiving environments in Canada through the environmental effects monitoring (EEM) program to assess the adequacy of their pulp and paper mill liquid effluent regulations (PPLER).  Results of the EEM program, from 1992 to present, have indicated a national pattern of endocrine disruption in fish exposed to PMEs (Lowell et al. 2005).  However, it has proved challenging to link these effects with a specific cause such as tree species used for pulping, pulping process, wastewater treatment process, or receiving water characteristics.  Now, even with decades of research and improvements in quality of PMEs, the chemicals and conditions causing endocrine disruption in fish remain unclear.  Throughout the 1990s, upgrades were made to wastewater treatment systems in Canadian pulp mills in order to comply with new federal PPLER; these were meant to decrease effluent acute toxicity and discharges of biochemical oxygen demand and dioxins and furans into the aquatic receiving environment (Munkittrick et al. 1997).  As a result of the new regulations, all mills installed biological wastewater treatment and moved toward elemental chlorine free bleaching practices.  These changes resulted in dramatic improvements in final effluent quality from reduced concentrations of compounds producing acute toxicity (Hewitt et al. 2006).  A decrease in sublethal toxicity was also demonstrated through the EEM program (Lowell et al. 2005).  Yet, even though acute and sublethal toxicity have improved, evidence of endocrine disruption still exists in fish exposed to biologically treated PMEs.  Today, experts suggest that the effects of PME biological treatment on endocrine disruption in fish remains unclear and needs to be resolved (Hewitt et al. 2008).  One of the problems in identifying the causes of endocrine disruption is that PMEs are both complex and dynamic.  The chemical make up is variable in time (at any one mill) and in space (between mills).  Interactions between compounds abound, making it virtually impossibly to relate the effect of a single compound determined under controlled laboratory conditions to the effects of a PME in nature.  Complexities of fish reproduction exacerbate the difficulty of identifying specific  2 endocrine disrupting compounds (EDCs) in PMEs, as does the fact that different fish species exhibit a range of sensitivities to EDCs (MacLatchy et al. 2006).  One source of EDCs in PMEs is the wood itself.  Phytosterols, or plant sterols, make up a group of compounds in wood that have been proven to induce endocrine disruption in fish through controlled laboratory experiments (MacLatchy & Van der Kraak 1995; Mellanen, et al. 1996; Tremblay & Van der Kraak 1998; Tremblay & Van der Kraak 1999; Nakari & Erkomaa 2003; Christianson-Heiska et al. 2007).  Phytosterols are especially interesting contaminants because they are a type of steroid and, therefore, are capable of mimicking the natural steroids present in fish.  There have been a number of studies evaluating the fate of phytosterols during biological wastewater treatment of PMEs in recent years, as evidence has accumulated regarding the ability of phytosterols to act as EDCs.  This research has shown that phytosterols are largely removed through biological wastewater treatment, due, at least in part, to biodegradation (Mahmood-Khan & Hall 2003;  Kostamo et al. 2004; Xavier et al. 2009).  This finding is interesting because as phytosterols act as EDCs, so can their degradation and oxidation products (Denton et al. 1985; Howell & Denton 1989; Christianson-Heiska et al. 2007).  Despite findings such as these, very little is known about metabolites of phytosterols produced during biological wastewater treatment.  It is plausible that microbial, or bio-, transformations may be producing other EDCs from phytosterols during the treatment process.  Indeed, phytosterol biotransformations may generate other steroidal compounds that have equal or more pronounced effects on exposed fish.  This hypothesis merits investigation, especially in light of the challenges in pinpointing compounds of concern in PMEs and the uncertainty of the effects of biological wastewater treatment on endocrine disruption.   3 2 LITERATURE REVIEW 2.1 THE ENDOCRINE DISRUPTION PROBLEM In order to investigate potential causes of endocrine disruption, it is necessary to have a basic understanding of the problem.  This section provides a brief introduction to the components and functions of the endocrine system, characteristics of endocrine disrupting compounds and PMEs, and a summary of endocrine disruption effects observed in fish exposed to PMEs.  2.1.1 The Endocrine System The endocrine system regulates metabolism, growth, development and reproduction and is crucial to maintain homeostasis within an organism (Watson & Miller 2004).  It functions by sending chemical messengers, called hormones, through the body to carry out cellular functions.  Hormones are produced and secreted by endocrine glands including the pituitary gland, thyroid, gonads (ovaries and testes) and adrenal glands in mammals, or interrenal glands in teleost (boney) fish (Thain & Hickam 1996; Kime 1998).  Control of the endocrine system occurs in the brain, where the hypothalamus sends signals to the pituitary gland to release or to stop releasing hormones.  Pituitary hormones travel through the bloodstream to other endocrine glands, stimulating them, in turn, to produce their own hormones (Watson & Miller 2004).  When levels of these hormones reach sufficient levels in the bloodstream, a negative feedback loop is activated and the hypothalamus signals the pituitary to stop sending out releasing hormones.  Communication also occurs between the thyroid, adrenal or interrenal glands and gonads, so that endocrine functions are coordinated throughout the body.  For example, interaction between the interrenal gland and gonads maintains a balance between reproduction and stress response in fish (Kime 1998).  A schematic of the endocrine system is provided in Figure 2.1, showing the glands and the pathways through which hormones travel.   4 Hypothalamus Pituitary Thyroid Adrenal / Interrenal Gonads   Figure 2.1:  Simplification of the endocrine system.  Solid black lines show pathways from the brain, dotted lines show negative feedback loops, grey lines show crosstalk between glands.  Each endocrine gland has separate functions and produces a unique set of hormones.  In fish, the thyroid gland primarily regulates metabolic rate, affecting activity and growth.  The thyroid is also involved in other functions and interacts with gonadal hormones and growth hormone, which is secreted by the pituitary gland (Kime 1998).  The interrenal gland of fish is homologous to the mammalian adrenal gland, with a major difference being that the interrenal gland is dispersed within the kidney, whereas the adrenal gland comprises a distinct structure.  The interrenal gland produces corticosteroids, which function in carbohydrate metabolism and stress response, much like in mammalian systems (Kime 1998).  Fish gonadal hormones control secondary sex characteristics, behaviour and courtship, gamete development and maturation, and spawning (Kime 1998).  Steroids are one group of hormones and are of particular interest to endocrine disruption in fish. Steroids that are released by the gonads and interrenal glands include sex steroids, which regulate sex differentiation and reproduction, and corticosteroids, which regulate metabolism, electrolyte balance and stress response (Watson & Miller 2004).  Many of these functions have been disrupted in fish exposed to pulp mill effluents, as described in the following sections.  As such, it is useful to take a closer look at steroid structure, synthesis and function.  5  All steroids share the same base structure consisting of three six-carbon rings and one five-carbon ring, which on their own make up the compound sterane.  This base structure, also called the steroid nucleus, is illustrated in Figure 2.2.   6   Figure 2.3:  Steroid biosynthesis in teleost fish.  Reprinted from Figure 4.3 (pg 86) in Kime (1998), with kind permission from Springer Science and Business Media (formerly Kluwer Academic Publishers).  Enzymes responsible for the transformations are provided on the arrows.  7 of potential viable offspring that a female can produce (Kime 1999).  E2 also plays a role in oogonial proliferation in fish, although this process is also directed by androgens (Young et al., 2005).  E2 is produced during steroidogenesis from the androgens testosterone and androstenedione.  The livers of male fish also contain receptors for E2, even though E2 is not normally present in males, making them vulnerable to environmental estrogens (Kime 1998).  11-KT is produced from testosterone in male fish.  It is primarily responsible for gonadal development, sperm production, secondary sexual characteristics and sexual behaviour (Kime 1998, Young et al. 2005, Thomas 2008).  Testosterone (T), is produced from androstenedione and is present in fish of both sexes.  The functions of T in fish have not yet been fully defined (Kime 1998), but it is at minimum a precursor to other sex steroids.  For comparison, in mammals T is responsible for sexual differentiation, physiological changes during puberty, sperm production and sex drive (Watson & Miller 2004).  Sex steroids are synthesized at specific times during a fish’s lifecycle.  Several months are required for gonadal development to progress to the production of viable gametes (Kime 1999).  After this time, E2 and 11-KT production ceases and steroidogenesis switches to progestogen production for the final maturation of oocytes and sperm (Kime 1999).   Since the levels and actions of circulating sex steroids in fish vary with sex and stage of sexual maturity, measurements of sex steroid levels are commonly used in tests for endocrine disruption.  Early development stages appear to be especially sensitive to endocrine disruption, such as the period of phenotypic sex determination, gonadal sex differentiation and the onset of puberty (Thomas 2008).  For this reason, juvenile fish are commonly used in exposure bioassays.  2.1.2 Endocrine Disrupting Compounds As can be seen from the previous section, the endocrine system is a complex configuration of tissues and hormones, all of which interact to maintain life processes and homeostasis.  Because of the intricacies of this system, there are many opportunities for inference and disruption from environmental toxins.  Chemicals in the environment that hamper endocrine processes are termed endocrine disrupting compounds (EDCs).  EDCs may interfere with the synthesis, secretion, transport, binding, action and/or elimination of hormones (US Environmental Protection Agency  8 2006), which may ultimately lead to changes in the endocrine functions of metabolism, growth, development and reproduction.  2.1.2.1 Composition of pulp mill effluents Pulp mill effluents are complex solutions containing natural chemicals from wood as well as those added or formed during pulping and paper making procedures.  It is useful to review the constituents of PMEs, in order to obtain an appreciation for their complexities and to evaluate the sources of chemical inputs and potential EDCs.  Wood is predominantly made up of the fibrous material cellulose and hemicellulose, and the amorphous inter-fibre bonding agent lignin (Fengel & Wegener 1983).  In addition to these compounds, a large variety of chemicals exists in wood in small quantities.  Extractives generally make up 3 - 8% of the wood mass and are defined as the material that can be removed from wood by polar and non-polar solvents (Hocking 1998).  Examples of wood extractives are provided in Table 2.1.  Although they account for a small fraction of wood, extractives can greatly influence effluent qualities (Hocking 1998).  The amount and type of extractives present in wood varies widely depending on the species, and even among species, extractives composition is variable by geographic location and season (Fengel & Wegener 1983).  The chemical additives in PMEs vary with pulping process.  The goal of pulping is to release wood fibres into water, removing as much lignin as possible.  As a result of this process, extractives are released into the pulping solution and make their way into the waste stream.  Pulping can be accomplished by either physical or chemical means, or a combination thereof.  Thermomechanical pulping has low chemical inputs as it utilizes high temperatures and grinding to produce high yield, low strength paper, such as newsprint (Hocking 1998).  Chemical pulping processes generally produce lower pulp yields than mechanical processes but can produce pulp with little degradation of cellulose fibres, which is suitable for high strength paper.  The kraft process is the most common method of chemical pulping both in Canada and worldwide and utilizes a solution of sodium hydroxide and sodium sulphide for lignin removal (Hocking 1998).  Acid sulfite pulping is another chemical pulping process used in Canada, though it is an older technology and is less common today than the kraft process.   9 Table 2.1: Overview of wood extractives.  Summarized from (Fengel & Wegener 1983). Extraction Medium Chemical Classes Example Chemicals Non-polar solvent terpenes diterpenes (including resin acids), triterpenes (including steroids), mono-, sesqui-, di-, tetra-, and polyterpenes  phenolic compounds tannins, phlobaphenes, stilbenes, lignans, flavinoids  aliphatic acids fatty acids  alcohols aliphatic alcohols, sterols Polar solvent Minerals potassium, calcium, magnesium  Macromolecules carbohydrates, proteins  Another important chemical addition to PMEs occurs during bleaching.  Bleaching is conducted following pulping in order to create maximum brightness while maintaining minimum yield loss. Mills use beaching sequences that involve a range of stages, for example, chlorine, hypochlorite, chlorine dioxide, peroxide, oxygen and caustic extraction (Hocking 1998).  Due to the range of chemicals that may be present in PMEs, the specific chemicals responsible for endocrine disruption effects are difficult to identify.  This is exacerbated by interactions between compounds, formation of degradation products, and the temporally dynamic nature of PMEs (Lister & Van der Kraak, 2001; MacLatchy et al. 2006).  As a result, some researchers have suggested that identification of chemical classes, rather than specific compounds, is sufficient for effects-directed chemical analysis (Hewitt & Marvin 2005).  Toxicity identification evaluation (TIE) has been employed to systematically identify compounds of interest, in order to simplify EDC identification. TIE involves fractionation of effluent samples into less complex mixtures in order to measure the toxicity of the fractionated groups of compounds.  A recent review by Hewitt et al. (2008) summarizes evidence for specific EDCs in PMEs.  This assessment showed that juvabione, dehydrojuvabione and manool have been identified as compounds of concern from TIE studies, and that β-sitosterol, abietic-based resin acids, pinosylvin, betulin, retene and genistein have been identified as compounds of concern using investigations with individual compounds (Hewitt et al. 2008  and references there within).  However, the authors also suggest that there are bioactive substances in PME that remain unidentified, and imply that the problem of elucidating compounds responsible for endocrine disruption in PMEs is far from over.   10 2.1.3 Endocrine Disruption in Fish Exposed to Pulp Mill Effluents Effects of EDCs on fish living in pulp mill effluent receiving environments are well documented and responses have been extensively summarized in literature reviews such as those by McLeay (1987), Munkittrick et al. (1998), McMaster et al. (2006), Kovacs et al. (2006), and Hewitt et al. (2008). Many different types of endocrine disruption effects have been reported in fish living in PME receiving environments.  Work was initially focused on measuring the activity of ethoxyresorufin-O- deethylase (EROD), a catalytic measurement of cytochrome P4501A induction, as this became a well documented response to dioxin exposure (Hewitt et al. 2006).  Many studies have shown pulp mill effluents to induce EROD activity, such as Munkittrick et al. 1992, Soimasuo et al. 1998, Lappivaara 2001, and Orrego et al. 2006.  However, since the mid 1990s, effects-based research has strived to elucidate more complex reproductive effects (Hewitt et al. 2006).  Some of the most common reproductive effects reported are depressions in circulating sex steroid levels, predominantly testosterone and estradiol (Munkittrick et al. 1992; McMaster et al. 1996; Karels et al. 2001; Sepulveda et al 2001; Fentress et al. 2006), induced vitellogenin (egg yolk protein precursor) production in male fish (Lappivaara 2001; Oakes et al. 2005; Orrego et al. 2006), reduced gonad size (Shaughnessy et al. 2007), increased gonad somatic index (Orrego et al. 2006), development of secondary sexual characteristics (Howell et al. 1980; Cody & Bortone 1997; Parks et al 2001; Ellis et al. 2003; Parrott et al. 2004; Orlando et al 2007), decreased egg production (Karels et al. 2001; Parrott et al. 2004; Kovacs et al. 2005; Rickwood, et al. 2006a), and decreased spawning frequency (Rickwood et al. 2006a).  In addition, pulp mill effluents are often described as being estrogenic or androgenic in nature, referring to their ability to induce the characteristics of one or the other sexes, contributing to an overall androgyny among exposed fish populations.  Estrogenic effluents are largely labelled as such due to their ability to induce vitellogenin production in male fish.  Comparatively, anti-estrogenic responses have been identified from exposure to components of PMEs through inhibition of vitellogenin production (Kukkonen et al. 1999; Christianson-Heiska & Isomaa 2008).  Effluents are termed androgenic where, in female fish, they lead to the development of a gonopodium, an elongated anal fin used to inject sperm into females during mating.  In addition, some effluents are capable of producing a variety of these types of effects (Parrott et al. 2004; Oern et al. 2006).  These ways of classifying PMEs demonstrate the complexity of both their chemical make up and of the responses they elicit.  11 Due to the range of effects that may be present, debate exists around which are the most useful biological endpoints for monitoring.  Ideally, the focus should be on effects that are ecology relevant and that allow useful comparison between studies.  However, it is also important that measurement allows some degree of risk assessment, so that the occurrence of severe effects can be prevented. Clearly, there are trade-offs between the level of measurement and the type of information obtained. A schematic of this relationship is provided in Figure 2.4.  Parrott et al. (2006) suggested the most sensitive and biologically relevant endpoint is reproductive success, including age at first spawn, total number of eggs and eggs per female per time (Parrott 2006).  Lifecycle analysis is required to measure reproductive success and this can be relatively time consuming and costly.  As a result, short term laboratory bioassays measuring steroid and vitellogenin levels are often used as indicators of greater reproductive effects.  However, in most cases, the linkage between such bioassays and reproductive effects in wild fish are undefined (Parrott 2006).  On the other hand, observations of significant reproductive effects in the wild may not be the most useful endpoints from a regulatory or risk assessment perspective, because these are the exact responses that regulators are trying to minimize.  Measurement on the molecular level may act as a predictive tool so that individual and population level effects can be avoided (Miracle & Ankley 2005).  Clearly, this is most feasible where molecular changes are linked to greater reproductive effects.  Molecular e.g. – alterations in gene expression Biochemical e.g. – changes in receptor binding Cellular e.g. – changes in cell proliferation Organ e.g. – functional changes Individual e.g. – altered development Population e.g. – decreased hatching Increasing ecological relevance Increasing predictive utility  Figure 2.4:  Conceptualization of trade offs between levels of measurement.  Adapted from Miracle & Ankley 2005.  Ecotoxicogenomics is a relatively new field of environmental toxicologically that uses molecular science to measure the changes in gene expression patterns in response to exposure to toxins in the environment.  Gene sequencing in an organism of interest can target specific genes known to  12 respond to specific chemicals, or can be conducted using the genome, measuring thousand of genes at once with cDNA microarray technology.  Measuring for endocrine disruption at this level is extremely powerful in its ability to screen for a vast number of effects, rather than a few singular endpoints that are usually measured at higher biological levels.  Such a tool is useful for analysis of effects of PMEs and other complex effluents capable of inducing a wide variety of effects (Francois et al. 2003).  Currently, response patterns are being mapped out using known EDCs to identify groups of endocrine responsive genes.  For example, an estrogenic response pattern has been defined by exposing fish to natural and synthetic estrogens.  Such bioassays have led to induction of genes associated with vitellogenin, vitelline envelope proteins and estrogen receptors (Flouriot et al. 1995; Denslow et al. 2001; Larkin et al. 2002; Filby et al. 2007).  The use of genomic analysis in PME exposure bioassays is relatively new and only a few studies exist in the literature.  One such investigation found PME exposure led to upregulation of the vitellogenin gene (VTG) in juvenile whitefish (Mellanen et al. 1999), while in another led to downregulation of VTG in female perch (Karels et al. 1998).  This shows that PME may lead to both estrogenic and anti-estrogenic effects.  A broad scale toxicogenomic study was conducted by Environment Canada whereby underyearling rainbow trout were exposed to effluent from nine BC pulp mills and analysis was conducted using DNA microarray technology (Environment Canada 2005a).  Here it was found that effluent from six of the mills induced upregulation of estrogenic genes, including VTG and three vitelline envelope protein genes.  Chemical and metal detoxification and stress response gene classes were also included on the microarray and effluents were chemically characterized for phytosterols, resin and fatty acids, absorbable organic halides, dioxin toxicity equivalency and metals.  Although a range of genomic responses were detected, one conclusion of the study was that chemical composition of the effluents was complex and difficult to correlate to biological effect.  Using this conclusion, it is suggested in the report that microarray data provide a more accurate indication of biological effects than chemistry data.  This, as well as the ability of genomic analysis to monitor wide ranges of effects, demonstrates the utility of genomic analysis.  2.1.4 Impacts of Biological Wastewater Treatment From the above section, it can be concluded that endocrine disruption occurs in fish exposed to PMEs on a variety of biological levels.  Although effects are still present today, there can be no doubt that the use of biological wastewater treatment has greatly benefited the Canadian aquatic  13 receiving environment.  The most common biological wastewater treatment technologies used in pulp mills are activated sludge (AS) systems and aerated stabilization basins (ASB), also known as aerated lagoons (Cook et al. 1997).  Approximately one third of Canadian pulp mills utilize ASBs (Mahmood & Paice 2006), presumably meaning that the other two thirds use AS.  The key difference between these types of systems is that in AS, sludge is recycled through the system and routinely wasted to maintain a specific sludge age, whereas sludge age is unknown or infinite under ideal conditions in ASBs.  Also, AS is generally an aerobic process, whereas ASBs utilize both aerobic and anaerobic processes.  The implementation of biological wastewater treatment of PMEs in Canada was associated with reductions in acute toxicity and some forms of sublethal toxicity in fish living in PME receiving waters (Munkittrick et al. 1997; Lowell et al. 2005). However, studies specific to endocrine disruption effects from exposure to untreated and biologically treated PMEs have indicated less promising results.  Martel et al. (2008) showed that ASB treatment of a bleached kraft mill effluent removed reproductive effects in fathead minnows that were present from exposure to untreated effluent, including decreases in egg production and hatching, but had no effect on vitellogenin induction in male fish.  The same study also evaluated the effects of AS treatment on thermomechanical PME. Here, effects associated with egg production, spawning events and steroid depressions were present from exposure to untreated effluent and absent from exposure to biotreated effluent.  However, effects on gonad size and vitellogenin production persisted with treatment.  Other studies have shown biotreatment to result in little to no abatement of reproductive effects.  Research conducted on several species of wild fish in Jackfish Bay, Lake Superior, where steroid depressions were observed for several years prior to installation of biological treatment, showed this upgrade had no detectable effect on steroid depressions in wild fish (McMaster et al. 1996).  Effects on secondary sexual characteristics have also been shown to persist with biotreatment.  A study of New Zealand PME tested the occurrence of androgenic responses, following exposures to treated and untreated effluents.  It was found that the frequency of gonopodial development (masculinization) was reduced after biological wastewater treatment, but that the occurrence was significant from exposure to both treated and untreated effluents (Ellis et al. 2003).  Additional evidence of endocrine disruption following the implementation of biological wastewater treatment can be gleaned from Environment Canada’s EEM program.  Here, the persistence of effects through the 1990s and the current decade has been demonstrated at the national level.  The results of the EEM program were  14 most recently published in 2005 and indicated that there is a national pattern of response for fish exposed to PMEs whereby they are fatter and faster growing, with larger livers and smaller gonads, compared to reference fish (Lowell et al. 2005).  The EEM publication also indicated there was no improvement in fish gonad size through the late 1990s to 2004 and suggested that the persistence of this effect may be linked to changes in sex steroid production.  Due to the persistence of reproductive effects, and inconsistent findings of endocrine disruption pre- and post-wastewater treatment, experts suggest that the effects of biological treatment of PMEs on endocrine disruption requires further study in order to be fully elucidated (Hewitt et al. 2008). One way to help solve this problem is to focus on specific compounds suspected as EDCs in PMEs and investigate their fate during treatment.  2.2 PHYTOSTEROLS Phytosterols, or plant sterols, are known EDCs in PMEs and are interesting because they have been proven to be estrogenic when tested in isolation, and also because there is potential for these compounds to transform into other EDCs during biological wastewater treatment.  This concept has been little explored, but could prove to be a significant source of endocrine disruption in fish exposed to biologically treated PMEs.  2.2.1 Nature and Source of Phytosterols Phytosterols are a natural component of wood extractives (Fengel & Wegener 1983).  They are non- polar organic lipids consisting of a steroid nucleus (Figure 2.2) with a hydroxyl group at carbon atom C3 and an aliphatic side chain of eight or more carbon atoms at C17 (Thain & Hickam 1996).  Many sterols also have a double bond between carbon atoms C5 and C6, but saturated sterols also exist, and although they are included in the overarching sterols category, they are termed stanols.  Sterols have structural functions in plants as components of cell membranes, regulating diffusion and transport across the membrane.  They also have metabolic functions and are precursors to plant growth hormones, brassinosteroids, and other plant steroids (Wojciechowski 1991).  Sterols composition in plants varies by species, plant part and developmental stage, and over 200 different sterols have been identified (Akihisa & Tamura 1991).   The principal sterol in most plants is β- sitosterol (β-sito), followed by stigmasterol (stigma) and campesterol (campe) (Akihisa & Tamura  15 1991).  Cholesterol (chole) is also produced by plants, and although it is largely thought of as an animal sterol, it is in fact ubiquitous in nature and is found in all eukaryotes, including animals, plants and fungi, and even in some prokaryotes (Moncecchi et al. 1991).  An illustration of β-sito, stigma, campe and chole is provided in Figure 2.5.  β-Sitosterol Campesterol Stigmasterol Cholesterol  Figure 2.5:  Common plant sterols.  Sterols are ubiquitous in PMEs and are predominantly released during the pulping and debarking processes (Fardim & Duran 2003; Kostamo et al. 2004).  A study in the late 1990’s of 22 US pulp mills using biological wastewater treatment found mass discharge rates of total phytosterols ranged from 0.2 to 25.2 g per air dried ton of pulp (Cook et al. 1997).  In that study, β-sito was identified as the major component of phytosterols in PMEs, and it was present in all samples collected from each of the 22 mills over the 13 month study period.  Another study confirmed these results to be relevant to British Columbia (BC) PMEs.  Here, the effluents of two BC pulp mills were investigated, showing total sterols concentrations ranging from 100 to 700 µg/L in biologically treated effluents (Mahmood-Khan & Hall 2003).  In this study, β-sito, campe and stigmastanol accounted for 70% of the total sterols released.  A study of three Finnish pulp mills found biologically treated PMEs to contain sterols concentrations of 21.4 to 60.5 µg/L, consisting of 70%  16 β-sito and 25% campe, and also detected a spike of sterols concentrations of 1305 µg/L during an operational change (Kostamo et al. 2004).  A study of a New Zealand PME reported an average sterols concentration of 255 µg/L following biological wastewater treatment, consisting of β-sito, stigma, campe, chole and β-sitostanol (van den Heuvel et al. 2002).  Additionally, another study estimated that pulp and paper wastewater in Portugal alone could generate over 90 tons of sterol concentrate per year (Dias et al. 2002).  This evidence shows that phytosterols are present in PMEs worldwide at a relatively high range of concentrations and that β-sito makes up a large portion of most PME sterol mixtures.  2.2.2 Phytosterols as EDCs Research over the past three decades has proven that phytosterols cause a variety of endocrine disruption effects including reduced steroidogenesis, embryological and developmental consequences and feminization.  Feminization is perhaps the most popularized effect of phytosterols, likely due to the surge of environmental estrogens detected in aquatic environments in recent years.  Laboratory studies have shown β-sito to produce estrogenic effects in exposed fish via increases in vitellogenin levels in rainbow trout (Tremblay & Van der Kraak 1998; Tremblay & Van der Kraak 1999)  and zebrafish (Nakari & Erkomaa 2003; Christianson-Heiska et al. 2007).  This effect has also been proven on the molecular level, as Mellanen et al. (1996)  found β-sito exposure led to induction of the vitellogenin gene in rainbow trout.  Studies have also shown phytosterol exposure to cause reductions in circulating sex steroids and reduced gonadal steroid production.  Reduced levels of plasma testosterone and estradiol have been demonstrated in goldfish (Maclatchy & Van der Kraak 1995)  and blenny (Mattsson et al. 2001), and reduced levels of testosterone, pregnenolone and cholesterol have been demonstrated in rainbow trout (Tremblay & Van der Kraak 1998; Tremblay & Van der Kraak 1999).  The conversion of cholesterol to pregnenolone is the first step in steroidogenesis, and as such, the effect of sterol exposure on this pathway has been investigated in an attempt to elucidate the mode of action of sterols-induced steroid depressions.  One such study showed that, although chole levels were depressed in goldfish and brook trout from exposure to β-sito, the activity of the enzyme that converts chole to pregnenolone (P450) was not affected (Gilman et al. 2003).  Another study found that β-sito exposure did not impact P450 activity in goldfish but did decrease the size of the gonadal mitochondrial pool of reactive chole, suggesting β-sito was impeding chole transfer across the  17 mitochondrial membrane (Leusch & MacLatchy 2003).  This finding is similar to that of a study of PME by Oakes et al. (2005), who found that exposure to PME caused impairments in pregnenolone production relative to chole availability.  These studies together show that exposure to phytosterols alone, as well as to PMEs (which have been consistently proven to contain phytosterols), can impair steroidogenesis in fish populations.  The significance of phytosterol exposure is further exemplified by the occurrence of multigenerational effects, which occur as phytosterols are transferred from parents to offspring (Lehtinen et al. 1999; Mattsson et al. 2001).  Lehtinen et al. (1999)  showed that pre-exposure of male and female brown trout to β-sito induced effects in the subsequent generation from egg through larval stages, including increased egg mortality, smaller egg size, lower mean weight of egg sac stage larvae and increased prevalence of deformed or diseased larvae.  Similar findings were evidenced in blenny in a study conducted by Mattsson et al. (2001), while another study showed phytosterol exposure significantly shortened the hatching time of grayling eggs (Honkanen et al. 2005). Additionally, exposure of multiple generations of zebrafish to phytosterols produced a shift toward predominance of females (Nakari & Erkomaa 2003), reaffirming the feminization pattern evidenced through increases in vitellogenin production.  In addition to sterols themselves, phytosterols degradation products have also been linked to endocrine disruption effects in fish.   Research conducted in the 1980s demonstrated that fish exposed to a mixture of phytosterols incubated with the bacteria Mycobacterium smegmatis, led to masculinization of the anal fin in 100% of the female specimens (Denton et al. 1985; Howell & Denton 1989), whereas fish exposed to phytosterols or Mycobacterium alone exhibited no masculinization (Denton et al. 1985).  Also, more recently, oxidized phytosterols were shown to interfere with spermatogenesis, oogenesis and gonadal steroidogenesis in zebrafish (Christianson- Heiska et al. 2007).  These studies have been useful in showing that derivatives of phytosterols can be problematic to fish; however, the compounds generated from the phytosterols were not identified.  2.2.3 Impact of Biological Wastewater Treatment on Phytosterols Removal of sterols from PMEs is desirable primarily due to their inherent endocrine disruption potential, and as implied above, their ability to transform into other EDCs.  Biological wastewater  18 treatment can provide a high degree of sterols removal, and removal efficiencies and mechanisms have been studied over recent years.  Both activated sludge (AS) systems and aerated stabilization basins (ASBs) have been proven capable of a high degree of phytosterols removal.  In a survey of 22 US pulp mills (Cook et al. 1997) AS systems were found to decrease the concentrations of all phytosterols analyzed, with removal efficiencies of 70 to 95%.  Slightly lower values were reported in studies of BC and Finnish PMEs treated with AS, with average sterol removal efficiencies of 66 to 72% and 61 to 79%, respectively (Mahmood-Khan & Hall 2003; Kostamo et al. 2004).  The highest sterols removal efficiencies reported in the literature are from a study conducted on an AS high-efficiency compact reactor in Norway, at 96 % removal (Magnus et al. 2000), and a study on AS treated PME in Chile, at 100% removal (Xavier et al. 2009).  Other evidence for sterols removal through AS systems can be gleaned from sediment core analysis of a Finnish lake that has been receiving pulp mill effluent for over 100 years.  Here researchers found that peak sterol concentrations corresponded to the period prior to initiation of wastewater treatment, and that sterols concentrations decreased dramatically after the establishment of AS treatment, despite coinciding increases in pulp production (Lahdelma & Oikari 2006).  Results of sterols treatment in ASBs are more variable than those from AS systems.  In the survey of 22 US pulp mills, Cook et al. (1997) found ASBs capable of achieving total sterol removal efficiencies of 56 to 78 %.  This result is lower than that of the AS systems in the same study (70 to 95 %).  Also, the authors found major discrepancies between removal efficiencies of different sterols in ASBs.  Most importantly, the concentration of stigmasterol increased across all ASBs investigated. This suggests that conditions in ASBs promote the transformation of other sterols to stigmasterol. In addition, the concentrations of β-sitosterol and stigmastanol increased following ASB treatment at one mill investigated.  Comparatively, the study of Chilean PME demonstrated near 100% removal of sterols in both AS systems and ASBs (Xavier et al. 2009).  Hence it appears that sterols removal in ASBs is more variable than in AS systems, and generally, sterols removal efficiencies are higher in AS systems.  Although it is clear that biological wastewater treatment is capable of high degrees of sterols removal, many of the studies mentioned above also found that removal efficiencies were highly  19 variable over time.  For example, Mahmood-Khan and Hall (2003) found the range of sterols removal in pure oxygen AS systems to be from 20 to 90 % over the study period.  No direct causes were linked to the variability, showing that the conditions acting on sterols removal have not yet been fully identified.  Theoretically, changes in sterols removal may be due to changes in wood chips, treatment conditions or interactions between compounds.  Clarification, or settleability, may also play an important role in sterols removal, as one study found that 79 % of the sterols released in biologically treated effluent were associated with the particulate phase (Kostamo et al. 2004).  Hence improved clarification would likely reduce the amount of sterols in the effluent, increasing sterols removal.  The source of wood fibre also plays a role in the amount of sterols in effluents, as Cook et al. (1997) found that mills using recycled fibres had lower estimated mass discharge rates for phytosterols than those using raw wood.  This is logical, as the majority of sterols in PMEs are released into solution during debarking and pulping of wood chips, which is not required when processing recycled fibres.  Another factor influencing sterols removal during treatment is the composition of the influent sterols, as it has been found that AS treatment leads to different degrees of removal for different sterols.  For example, Mahmood-Khan and Hall (2003) found more β-sito was removed than campe, as β-sito made up the largest fraction of sterols in primary influent, whereas campe was the dominant sterol in the final effluent.  In order to further understand sterols removal during biological wastewater treatment, an investigation of the underlying mechanisms is required.  Biodegradation and bioadsorption are the two main of mechanisms for sterols removal during biological wastewater treatment described in the literature.  It appears that these are competing processes, and the conditions that make one mechanism favoured over the other are complex and not yet fully elucidated.  Three studies were found in which sterols biodegradation and bioadsorption were investigated in full-scale pulp mill wastewater treatment systems.  In the study involving two BC pulp mills using pure oxygen, AS treatment, mass balance analysis indicated little to no evidence of sterol degradation, but showed extensive sterols accumulation in the mixed liquor suspended solids (MLSS).  Here, the secondary sludges contained 31 to 72 % of the influent sterols (Mahmood-Khan & Hall 2003).  In contrast, the study of three Finnish pulp mills, using AS treatment, showed as little as 7.3 to 37 % of influent sterols absorbed to the secondary sludge, leading the authors to conclude that AS treatment results in a high degree of biodegradation of sterols (Kostamo et al. 2004).  Finally, the Chilean study involving an AS system and an ASB showed sterols removal was mainly by degradation in AS, and  20 mainly by adsorption in the ASB (Xavier et al. 2009).  Together these studies demonstrate conflicting results, illustrating a need for a clearer understanding of the conditions affecting sterols biodegradation and bioadsorption processes.  In attempt to elucidate the conditions affecting the competing sterols removal processes, Mahmood- Khan (2005) conducted tests using lab-scale AS bioreactors to treat PME spiked with various levels of sterols.  This work indicated that biodegradation of sterols is sensitive to normal operating ranges of pH, hydraulic retention time (HRT) and solids retention time (SRT).  Biodegradation was found to be optimized at a pH of 6.7, an HRT of 11 to 12 hr and an SRT of 11 to 12 d (Mahmood-Khan 2005).  Under these conditions biodegradation contributed 80% of total sterols removal, accommodating up to 4500 μg/L of influent sterols.  This study also found that sterols biodegradation was affected by the sterols concentration in the system solids, as increased sterols concentrations in the MLSS and secondary sludge were associated with decreased biodegradation of influent sterols, and decreased sterols removal efficiency in general.  In the same study, the author indicated that when conditions became unfavourable to biodegradation, bioadsorption prevailed as the principal mechanism of sterols removal (Mahmood-Khan 2005).  It was further found that when bioadsorption was the principal sterols removal mechanism, the amount of sterols in the sludge increased by two to four orders of magnitude.  It is clear from the studies described above that phytosterols are removed during biological wastewater treatment, at least in part by biodegradation.  However, no studies could be found in the literature that attempt to identify the metabolites of this process.  This is a potentially important point because sterols contain the steroid nucleus (sterane) in their chemical structure.  Theoretically, any degradation of a phytosterol, whereby the sterane structure is maintained intact, essentially forms a new steroid molecule.  If steroid production occurs in this manner during treatment of PMEs, the observation of this phenomenon may help shed some light on occurrence of endocrine disruption in fish following exposure to biologically treated PMEs.  The need to examine phytosterol biotransformations during biological treatment of PMEs is further exemplified by the fact that fish exposed to decomposed or oxidized phytosterols also exhibit endocrine disruption.     21 2.2.4 Phytosterol Biotransformations Although it is currently unclear if, and how, phytosterols are transformed into other steroids during biological treatment of pulp mill effluents, much information on phytosterol biotransformations can be gleaned from the steroid biosynthesis literature.  This field of study is largely focused on the production of commercially viable steroids via microbial reactions, and includes an extensive variety of bacteria and metabolic pathways. Bacteria used in these reactions are capable of transformations similar to those of the enzymes involved in vertebrate steroidogenesis (Figure 2.3).  However, where cholesterol is the starting product in vertebrate steroidogenesis, microbial biotransformations utilize a variety of phytosterols for steroid production.  β-Sito, stigma and campe are the most commonly used sterols for this purpose, due in part to their ease of transformation into steroid intermediates (Fernandes et al. 2003).  PMEs are one source of phytosterol starting products in the commercial steroid industry, as sterols can be isolated from pulp mill tall oil with a yield of 70 % and a purity of 96 to 98 % (Dias et al. 2002).  The range of compounds that are produced through microbial biotransformation reactions is very broad and over 5000 commercially useful steroid compounds have been identified (Fernandes et al. 2003).  The most common commercially synthesized steroids are progestogens and corticosteroids and others include estrogens, androgens, anabolic agents and diuretic agents (Fernandes et al. 2003). A summary of the uses of commercially produced steroids is provided in Table 2.2.  Table 2.2:  Steroid uses in medicine. Compiled from Fernandes et al. 2003. pharmaceutical uses disease treatment disease prevention neurological agents anti-inflammatory breast cancer coronary heat disease memory enhancement immunosuppressive prostate cancer HIV infections neuroprotectors progestational osteoporosis fungal infections alteration of stress responses diuretic adrenal insufficiencies anti-obesity agents anabolic HIV / AIDS  anxiolytic agents contraceptive herpes  anticonvulsants  spinal cord injury  antidepressants  myelin repair   Although the types of steroids produced are vast, the majority of them are generated from the intermediaries androstenedione (AD), androstadienedione (ADD) and progesterone (Pr) (Fernandes  22 et al. 2003).  Phytosterols are used to produce the intermediaries, via selective microbial degradation of their side chains.  In these reactions, carbon-carbon bonds in the saturated side chain are oxidized by microorganisms, causing the side chain to break (Mahato & Garai 1997).  A number of other microbial conversions are required to transform phytosterols into intermediaries, depending on the specific starting and end products. For reference, conceptualization of the transformation from β- sito to AD is provided in Figure 2.6.  β-Sitosterol Androstenedione Dehydrogenation of the C3 hydroxyl group to produce a ketone Oxidative side chain cleavage at C20 to produce a ketone Isomerization of the C5-6 double bond  Figure 2.6:  Conceptualization of biotransformation from β-sitosterol to androstenedione.  Pure phytosterols mixtures have been found to undergo 90% conversion to AD and ADD (Stefanov et al. 2006), illustrating the efficiency of this metabolic pathway.  Comparatively, AD and ADD can be produced directly from pulp mill tall oil, with yields between 16 and 55 % (Conner et al. 1976; Dias et al. 2002).  Additionally, Szykula et al. (1991)  showed that tall oil isolates containing 84% total sterols can achieve conversion to steroid precursors at a yield similar to that of pure β- sitosterol.  Conversion of phytosterols to intermediaries and steroid end products has been extensively summarized in the literature, for example by Ahmad et al. (1992), Mahato & Garai (1997), Fernandes et al. (2003), and Malaviya & Gomes (2008).  An illustration of steroid production from phytosterols to intermediaries to end products is provided in Figure 2.7.  It is important to note that this figure is an overview of reactions, and does not include all steps in the conversion pathways.  The figure also  23                       Figure 2.7: Phytosterol biotransformation reactions used commercially to produce steroids.  Reprinted with minimal alterations from  Figure 2 in Fernandes et al. (2003), with kind permission from Elsevier.   23  24 shows a number of transformations that are conducted using chemical rather than microbial means, mostly in the conversion of intermediaries to end products.  It is conceivable, however, that bacteria do exist that are capable of the same transformations, that either have not yet been identified or are not efficient enough for commercial use.  An interesting biotransformation reaction not detailed in Figure 2.7 is the conversion of phytosterols to testosterone (T).  A mutant bacteria of Mycobacterium sp. has been identified which can transform phytosterols into T via oxidation to AD followed by reduction to T, with a maximum conversion of 31% (Lo et al. 2002).  There is reason to suspect that sterol biotransformations may occur during biological wastewater treatment of PMEs.  Mahato and Garai (1997) list over 40 sterol biotransformation reactions, mostly involving bacteria of the genera Mycobacterium, Rhodococcus, and Arthrobacter.  Activated sludge systems have been known to contain bacteria of these genera (Schuppler et al. 1995; Kampfer et al. 1996; Stainsby et al. 2002) and quite possibly other unidentified bacteria capable of steroidal transformations.  However, the details of phytosterols biotransformations during treatment of PMEs have not yet been elucidated.  This is partially because the range of compounds that may potentially be formed through this process is wide, and effluent chemistry analysis is confined to a limited number of parameters.  However, the steroidal intermediaries AD, ADD and progesterone (Pr) have been detected in the Fenholloway River in Florida, downstream of a PME discharge (Jenkins et al. 2001; Durhan et al. 2002; Jenkins et al. 2003), suggesting the occurrence of phytosterol biotransformations in PME.   The same studies have shown that masculinization of female mosquitofish is prominent in these waters, and is attributed to the presence of the androgens (AD and ADD).  It has been suggested that biotransformations that form the androgens are taking place in the sediment of the receiving water (Jenkins et al. 2003), and that the precursors of the androgens are not phytosterols, but instead Pr which is naturally present in the wood (Carson et al. 2008). Regardless, phytosterol biotransformation during effluent biotreatment as the source of these compounds cannot be ruled out, as no publications investigating this exists in the literature. Findings from another Florida waterway show that an overall reduction in the degree of mosquitofish masculinization was associated with the initiation of tall oil recovery (Cody & Bortone 1997), which suggests that the androgens were formed in the effluent rather than the receiving water.  Hypothetically, the reduction of androgenic effects may be due to a reduction in androgenic phytosterol biotransformation products in the discharge.   25 The presence of the steroidal intermediaries AD, ADD and Pr in PME receiving waters leads to the question of whether these compounds promote other transformations to produce other chemically active steroid compounds.  However, very few studies exist in the literature that evaluate the presence of active animal steroids in PME.  One study of this nature was found, which attempted to measure 30 estrogenic organic compounds in biologically treated PME and municipal wastewaters in BC.  Parameters analyzed included sterols, natural and synthetic estrogens, and plasticizers, among others.  Notably, the natural steroidal estrogens estrone (E1) and estradiol (E2) were detected in the PME analyzed in this study, at maximum concentrations of 49 ng/L and 17 ng/L respectively (Fernandez et al. 2007).  The source of these compounds is uncertain, and no studies are known to date that have attempted to measure these compounds in raw PME, thus it cannot be ascertained whether these compounds were generated during treatment.  In municipal effluent however, where steroids of animal origin are ubiquitous, E1 and E2 levels have been shown to increase in concentration through biological treatment (Servos et al 2005; Fernandez et al. 2007).  This information shows it is possible for animal steroids to be generated during biological wastewater treatment, but does not speak to the mechanism.  In the case of PME, where there is no clear source of animal steroids, it can be speculated that phytosterol biotransformation plays a role in their generation.  2.3 LITERATURE REVIEW SUMMARY Endocrine disruption in fish exposed to biologically treated PMEs continues to be documented in the literature.  This occurrence leads one to question the fate of EDCs during treatment. Phytosterols are interesting in this regard as it is clear they are capable of transforming into other types of steroids, via reactions with microorganisms.  This may pose a problem to the pulp and paper industry, because bacteria present during biological wastewater treatment may be changing phytosterols into other EDCs.  This hypothesis, though plausible, has not yet been evaluated in the literature.  If the hypothesis is correct, it may be that pulp mills that utilize biological wastewater treatment do indeed reduce sterols concentrations in their final effluent, but with limited impact on the harmful EDC effects in their receiving environments.  In this way, research on phytosterols biotransformations in PMEs may help elucidate the effects of biological treatment on endocrine disruption in fish populations exposed to those effluents.    26 3 OBJECTIVES The literature described in the previous chapter has led to the hypothesis that phytosterols can be transformed into active animal steroids during biological wastewater treatment of pulp mill effluents, contributing to endocrine disruption in exposed organisms.   The objectives developed to explore this hypothesis were:  1)  To assess whether phytosterol biotransformation products can be identified in biologically treated pulp mill effluents, and 2)  To determine if wastewater treatment conditions affect the levels of phytosterols and their biotransformation products in treated effluents.  Objective (2) was included because by evaluating a range of conditions, there exists a greater opportunity to identify phytosterol biotransformation products.  In addition, objective (2) is needed to help to gauge the types of biotransformation products that may reasonably be expected to be present at mills using different treatment conditions.  It was also determined that, in order for this research to be ecologically significant, it was necessary to couple the phytosterols biotransformation research to a study of biological effects.  Without this inclusion, the study would only be an evaluation of chemistry, which would have merit in and of itself, but would lack an important connection to biological effects, which truly define the endocrine disruption problem.  To this end, it was determined that novel toxicogenomic methods would be employed, in order to address endocrine disruption from a predictive perspective.  The following objectives were developed in this regard:  3)  To assess the genomic response in fish exposed to biologically treated pulp mill effluents using a range of wastewater treatment conditions, and 4)  To determine whether a relationship exists between the genomic response in exposed fish and (a) the concentrations of phytosterols and their biotransformation products in test effluents, or (b) the wastewater treatment conditions used to produce the test effluents.   27 Objective (3) was meant to determine whether genomic effects could, in fact, be detected, whereas objective (4) was an attempt to link genomic response to a cause.  In setting the objectives, it was hoped that genomic response could be linked directly to levels of phytosterols biotransformation products, but it was also recognized that this would be unlikely given the history of challenges in correlating endocrine disruption effects with specific compounds.  In this regard it was important to also evaluate whether effects could be linked to specific wastewater treatment conditions.  Given the uncertainty of the effects of biological wastewater treatment on endocrine disruption, tying a genomic response to wastewater treatment conditions could prove extremely useful.  28 4 MATERIALS AND METHODS A study was designed encompassing three phases in order to address the objectives listed in the previous chapter.  In the first phase, a benchscale biological wastewater treatment system (BBWTS) was constructed and operated using effluent and biomass obtained from a local pulp mill.  The benchscale system was used to produce final treated effluent under four sets of operating conditions, producing four independently treated effluents.  In the second phase, treated effluents were characterized for plant sterols and their potential metabolites.  In the third phase, treated effluents were used as feed into bioassays to test the genomic response of exposed fish.  4.1 STUDY SITE This study was conducted in cooperation with a local pulp and paper mill (the mill), which provided the primary treated pulp mill effluent and sludge samples required to run the BBWTS.  4.1.1 Mill Characteristics The mill produces kraft pulp and newsprint utilizing softwood fibres from wood chips primarily of Western Red Cedar, Western Hemlock, Douglas Fir, White Spruce and Lodgepole Pine.  Newsprint is produced using a thermomechanical pulping process (TMP).  Kraft pulp is generated by the kraft process, utilizing a solution of sodium hydroxide and sodium sulfide (white liquor) for wood chip digestion.  The wood fibres remain intact during digestion and all other portions of the wood are dissolved into solution (black liquor).  Pulp is bleached using oxygen, chlorine dioxide and caustic extraction stages and the bleaching sequence is elemental chlorine free (ODEopDnD).  4.1.2 Wastewater Treatment Wastewater is generated in both the TMP and Kraft mills, and is combined with storm water runoff generated on-site.  Before treatment, wastewater is high in chemical oxygen demand (COD) and total suspended solids (TSS), and contains a complex mix of organic compounds.  The wastewater is treated to conform with Pulp and Paper Mill Effluent Regulations enacted through the federal Fisheries Act.  In accordance with these regulations, the final effluent discharged from the mill is not acutely lethal to juvenile rainbow trout after 96 hours of exposure.   29 A schematic of the wastewater treatment process utilized at the mill is provided in Figure 4.1. Wastewater enters the treatment system into a primary clarifier where the majority of TSS is settled out.  Next it is sent to a cooling tower prior to entering the biological treatment system.  This system utilizes a pure oxygen activated sludge process (UNOX).  Finally the effluent is sent to one of two secondary clarifiers where biosolids are settled.  Biomass is then returned to the UNOX system through a return activated sludge (RAS) line, or is dewatered and burned in the power boiler.  Primary Clarifier UNOX Bio-Reactor Secondary Clarifier Cooling Tower Influent Final Effluent Return Activated Sludge Waste Activated Sludge Primary Sludge Sample Point  Figure 4.1:  Schematic of the wastewater treatment system at the study mill.  4.1.3 Mill Sampling Primary treated (primary) effluent and RAS were collected from the mill covering the period of July 2007 to March 2008 for use in the BBWTS.  Primary treated effluent was collected every four to six weeks over the sample period.  Approximately three months into the study, it was determined that five weeks was the maximum holding time, since after this duration, black precipitate began forming in the lines and storage containers.  RAS was collected at the beginning of the sample period, and periodically throughout.  Sample locations at the mill are provided in Figure 4.1.  Primary effluent was collected at the base of the wastewater cooling tower and was poured into carboys from a hose.  The hose was allowed to run for approximately two minutes prior to sample collection.  RAS was obtained from a sample  30 point in the UNOX system.  The RAS line that was sampled had a very high flow, therefore the line was allowed to run only briefly before collecting a sample.  RAS was poured into carboys until they were approximately two thirds full, allowing head space for an oxygen reservoir for bacterial respiration.  All samples were immediately transported to the UBC Environmental Engineering Laboratory and stored at approximately 4o C.   4.2 BENCHSCALE BIOLOGICAL WASTEWATER TREATMENT SYSTEM A benchscale biological wastewater treatment system (BBWTS) was set up and operated at the UBC Environmental Engineering Laboratory, using samples collected from the mill.  The system was based on that used by Mahmood-Khan (2005), using optimal conditions for sterols removal determined during that study.  To be consistent with the previous work, activated sludge (AS) was selected as the system process.  AS and aerated stabilization basins (ASBs) are common at pulp mills across North America, and AS has been shown to achieve better sterols removal efficiencies than ASBs (Cook et al. 1997), as described in Chapter 2.  It was suspected that there is more opportunity for biotransformation where removal is higher, therefore AS was the preferred treatment process.  4.2.1 System Characteristics The BBWTS contained two treatment trains running in parallel.  A schematic of the system is provided in Figure 4.2.  Each train contained a double walled 3.25 L plexiglass activated sludge reactor with a diameter of 155 mm and an overflow height of 172 mm, and a 2.2 L plexiglass clarifier with a diameter of 100 mm and an overflow height of 323 mm.  The system was inoculated by filling the reactors with RAS collected from the mill and initiating treatment as per the characteristics described.  The system was heated to approximately 35oC by passing water from a water bath through the reactors’ double walls.  The clarifiers were not heated but were wrapped in Reflectix® foil insulation.  Reactors were continuously stirred using three pronged paddles attached to variable speed Arrow 1750 motors, with maximum RPM of 2000 (Arrow Engineering, Hillside, NJ).  Motor speeds were set to low, roughly near 100 RPM in the microaerophilic system and near 300 RPM in the aerobic system. Speeds were set and adjusted according to the desired dissolved oxygen levels in the reactors.   31 Primary treated effluent obtained from the mill was pumped into each bioreactor from a storage bin at a rate of 4.5 to 4.9 mL/min to maintain a hydraulic retention time (HRT) of 11 to 12 d; this is the optimum HRT for sterols removal as per Mahmood-Khan (2005).  The feed bin was topped up daily to minimize the amount of time that the primary effluent was kept out of the refrigerator. Mixed liquor overflowed from the reactors to secondary clarifiers.  Clarifier walls were scraped slowly to enhance settling, at a rate of 1 - 4 RPM using 1 mm stainless steal wire attached to single speed, gear operated motors. Sludge was concentrated on the bottom of the clarifiers and RAS was pumped from their base back to the reactors at roughly double the flow rate of the feed.  Waste activated sludge was removed from the reactors daily, via siphon or syringe.  Clarified secondary treated (final) effluent overflowed from the clarifiers to effluent collection basins.  Final effluent volumes were measured daily and effluent was sampled and/or discarded.  When mixed liquor settleability was noted to be poor, it was necessary to further settle the final effluent after it was collected, in order to retain biomass in the system for maintenance of SRT.  During daily maintenance at these times, effluent was siphoned off the collection bins, leaving roughly a litre of effluent and mixed liquor suspended solids (MLSS); this was then transferred to a graduated cylinder to settle.  Supernatant was siphoned off, solids volumes were recorded and solids were transferred back to the reactors.   32 Microaerophilic Reactor Aerobic Reactor Final Treated Aerobic Effluent Final Treated Microaerophilic Effluent P Clarifier Clarifier Acid Acid D D Air Air N & P Water Bath Feed Basin P P P D Effluent line RAS line Recirculating hot water line Continuous pump Dosing pump Air control valve A A A P  Figure 4.2:  Schematic of benchscale biological wastewater treatment system.  4.2.1.1 Oxygen, nutrients and pH Primary treated effluent collected from the mill was slightly basic and required acid addition in order to maintain a pH of 6.7 ± 0.1, as per the conditions for optimal sterols degradation determined by Mahmood-Khan (2005).  HANNA DP 7916 pH controllers (HANNA Instruments, Laval, QC) were used to achieve the desired pH, by dosing a 0.2 N sulfuric acid solution.  Nutrients were added directly to the primary effluent in the feed basin in a ratio of approximately 100:3.5:1 as dissolved COD:N:P, as per Mahmood-Khan (2005).  Nitrogen was provided as ammonium chloride (NH4Cl) and phosphorus was provided as potassium phosphate monobasic (KH2PO4).  Dissolved oxygen (DO) was maintained in the reactors via continuous air flow through porous rock diffusers.  DO levels were controlled using Cole-Parmer flow meters with valves (Cole-Parmer Canada Inc., Montreal, QC).  Differences in DO concentrations were further and more finely controlled by adjusting the stirring intensity of the reactor mixers.   33 4.2.2 Operating Conditions The BBWTS was operated using four different combinations of treatment conditions.  This was done to cover a spectrum of conditions used in pulp mills and to test the influence of operational parameters on sterols transformations and biological effects.  Two different DO concentrations and solids retention times (SRTs) were selected in this regard.  The operating conditions used for each of the four treatments are summarized in Figure 4.3 and described in more detail below.  Solids Retention Time (SRT), days D is so lv ed  O xy ge n (D O ), m g/ L 5 20 0.2 4.0 Aerobic, low SRT: DO = 4 mg/L SRT = 5 days Microaerophilic, low SRT: DO = 0.2 mg/L SRT = 5 days Aerobic, high SRT: DO = 4 mg/L SRT = 20 days Microaerophilic, high SRT: DO = 0.2 mg/L SRT = 20 daysD is so lv ed  O xy ge n (D O ), m g/ L  Figure 4.3:  Conceptualization of BBWTS treatment conditions.  4.2.2.1 DO DO level is one of the primary influences on microbial metabolism and is of great importance in determining the products of biodegradation (Tchobanoglous et al. 2003).  DO concentration varies in wastewater treatment systems in Canadian pulp mills, generally ranging from anaerobic levels in ASBs to supersaturated in pure oxygen AS systems.  To cover a portion of this range, the treatment trains were operated at two different DO concentrations.  One train was operated as an aerobic system, with DO concentrations maintained around 4 mg/L.  The second train was operated as a microaerophilic system, with DO concentrations maintained around 0.2 mg/L.    34 4.2.2.2 SRT SRT was the second parameter varied during operation of the BBWTS.  SRT was selected for this purpose because it is the basic design and operating parameter for AS and it directly influences the bacterial populations present in the system (Tchobanoglous et al. 2003).  For this reason, SRT was deemed likely to influence sterols biotransformation.  The average SRT in pure oxygen AS systems in Canadian pulp mills is 13 d and varies from 4 to 27 d (Paice et al. 2003).  Through discussions with Paprican staff, it was determined that air-based AS systems in Canadian pulp mills have a similar range of SRTs, with an average of 10 d (Paice, M., Environment Program, Paprican Division, FPInnovations, Pointe-Claire, QB, personal communication, August 2007).  Comparatively, ASBs have a theoretically infinite SRT under ideal conditions.  Hence a large range in SRTs is used in pulp mill wastewater treatment systems across Canada.  To capture this variation, the treatment trains were operated at a 20 d and a 5 d SRT over two distinct periods of time.  The 20 d SRT run was maintained for 75 d from August to December 2007 and the 5 d SRT run for 58 d from January to March 2008.  In both cases, a minimum of 3 SRTs were allowed to pass prior to organic chemistry and biological effects analysis to maximize the acclimation and turnover of biomass.  4.2.3 Sampling Samples were collected from the BBWTS for four different purposes.  These were to: 1) monitor the operational parameters that define the integrity of the BBWTS; 2) compare sterols removal efficiencies and biotransformations between treatments; 3) collect water for conducting biological effects analysis; and 4) conduct a sterols mass balance on the soluble and particulate phases in the system.  The phase-separation mass balance was added towards the end of the operation of the BBWTS, in order to help explain the removal processes acting on sterols in the system.  Samples collected for this purpose were analyzed for the same suite of organic parameters as for sample group 2 but were collected in a different fashion.  The sample locations, frequencies and types for all sample groups are summarized in Table 4.1, and sampling techniques are described below.   35 Table 4.1:  Sample group details.  Notes:  “-” indicates not sampled 1)Operational    Parameters 2)Sterols and    Biotrans-    formation    products 3)Biological    Effects 4)Phase-    Separation    Mass    Balance Twice weekly for entire operation of BBWTS One batch per treatment One batch per treatment One batch Feed basin 1 d composite 1 d composite and 4 to 5 d composite - 1 d composite Aerobic final effluent collection bin 1 d composite 1 d composite and 4 to 5 d composite 4 to 5 d composite 1 d composite Microaerophilic final effluent collection bin 1 d composite 1 d composite and 4 to 5 d composite 4 to 5 d composite 1 d composite Aerobic reactor Grab - - Grab Microaerophilic reactor Grab - - Grab Aerobic clarifier Grab - - - Microaerophilic clarifier Grab - - - Aerobic effluent settled solids 1 d composite - - - Microaerophilic effluent settled solids 1 d composite - - - Sample Group Sample Frequency Sa m pl e Lo ca tio n an d T yp e  4.2.3.1 Sample group 1 Samples collected for analysis of BBWTS operational parameters were collected twice weekly from five points in the system: the feed basin, the aerobic and microaerophilic reactors, and the aerobic and microaerophilic final effluent collection bins for calculation of total system solids during periods of poor settleability.  Additional samples were sometimes required from the clarifier and the settled  36 solids from the final effluent collection bins.  All samples for operational parameters were collected in plastic bottles that had been washed with detergent and air dried.  Samples from the feed basin were collected by first vigorously mixing the basin, then submerging the sample bottles under the surface of the water.  The feed basin was filled once per day, therefore these can be considered one day composite samples, capturing the feed entering the benchscale system over a full day.  Samples from the reactor were collected as grabs.  A syringe was used to draw mixed liquor directly from the reactors while they were being continuously mixed.  Final effluent samples were collected as composites, accumulated over the duration of a day.  The final effluent was continuously discharged to collection bins which were sampled and/or emptied daily. The duration of collection for one day composites varied from approximately 17 to 30 hr, depending on the time of day the effluent collection bins were emptied.   To sample the clarifier, a piece of tubing was attached to a syringe and was lowered approximately two thirds down the clarifier and a grab sample was extracted.  To verify that this method captured the average suspended solids concentrations in the clarifiers, a short test was run in which the clarifiers were plugged, stirred and sampled from the middle.  Since it was found that these two methods were in approximately 90% agreement of each other, it was decided that stirring the clarifiers was unnecessary.  To sample the final effluent settled solids the slurry was mixed and poured into a sample bottle.  4.2.3.2 Sample groups 2 and 3 Final effluent compositing events were required over a duration of 4 to 5 d in order to collect sufficient water (25 to 30 L) for the biological effects analysis.  This was conducted once for each of the four BBWTS treatment scenarios, following a minimum BBWTS operating period of three SRTs.  Samples were also collected during this time for analysis of organic parameters to compare sterols and biotransformation products between effluents.  In this way, the effluent used for the biological effects analysis was chemically characterized.  Effluent accumulated during the 4 to 5 d compositing events was held in 10 and 20 L glass jugs for collection and storage purposes, respectively.  Prior to use, this glassware was cleaned using the following sequence: wash with detergent, rinse with distilled water, rinse with 10 % hypochlorous acid, rinse with distilled water, rinse with acetone.  The jugs were then inverted and allowed to air-  37 dry overnight.  Final effluent was collected continuously from the BBWTS into the 10 L jugs, which were stored on-ice, in portable coolers at the base of the system.  Each day of the compositing events, effluent from the collection jugs was transferred via siphon into the storage jugs, which were kept in a walk-in cooler at 4oC.  On the last day of each compositing event, all jugs would be near full and were transported to PESC where effluents from the same treatment were combined, well mixed and portioned for biological effects analysis.  Effluent collected in this manner was sampled for organic parameters using two different methods, as indicated in Table 4.1.  First, 1 d composite samples were taken from the collection jugs twice per compositing event, prior to transferring effluent to the storage jugs.  This was conducted by mixing the effluent, then siphoning off a sample from the middle of the collection jug.  Second, after the effluents were combined at PESC, 4 to 5 day composite samples were obtained in triplicate while continuously mixing the effluent.  Feed water was also composited during the final effluent compositing events, to obtain organics analysis from the feed that could be compared to that of the final effluent.  On each day of final effluent composition, the feed basin was stirred vigorously and 400 mL of influent was portioned into a 4 L glass jug that was stored in the walk-in cooler at 4oC.  Prior to this, the 4 L jug was washed using the same procedure as for the final effluent collection and storage jugs.  Both 1 d composite and 4 to 5 d composite samples were obtained from the feed.  The 400 mL portion was sampled twice per compositing event to make up 1 d composite samples, and the 4 L jug was sampled in triplicate at the end of final effluent composition to make up the 4 to 5 d composite samples.  All samples collected for analysis of sterols and biotransformation products were kept in 1 L glass amber bottles that were purchased pre-cleaned.  Sample bottles were later reused after washing with detergent, rinsing with acetone followed by deionized water, and baking at 400oC for 4 hours. Teflon lids were rinsed with acetone followed by deionized water.   For collection, 100 mL of sample water was measured and poured into sample bottles.  The samples were then topped up to 400 mL with distilled water, acidified with sulphuric acid to pH ≤ 2.0, and the extraction process was initiated immediately.    38  4.2.3.3 Sample group 4 The feed, mixed liquor and final effluent were sampled for phase-separation analysis, as summarized in Table 4.1.  Data obtained for mixed liquor was used to represent waste activated sludge (WAS), as WAS was removed directly from the reactors.  Sample collection techniques were identical to those used for the operational parameters, with feed and final effluent collected as one day composite samples and mixed liquor collected as grab samples.  All samples for the phase-separation mass balance were collected in glass bottles prepared as per those used for Sample Group 2.  Each sample collected for the phase-separation mass balance was split between a whole water sample and a sample for centrifugation.  Centrifugation was necessary to split the sample between soluble (supernatant) and particulate (settled) fractions.  Centrifugation was conducted for 30 min at approximately 2800 RMP (1750 x g relative centrifugal force) in 200 mL glass tubes.  This glassware was washed using the same method as for the sample bottles for organic parameters.  Centrifuged supernatant was gently poured off into graduated cylinders, volume was measured, and a maximum of 100 mL of supernatant was transferred to sample bottles.  Best efforts were made to prevent particulates from being entrained into the supernatant.  For all whole water feed and effluent phase-separation mass balance samples, 100 mL of sample was measured and poured into sample bottles.  For whole water mixed liquor, 10 mL was measured and poured into samples bottles.  All samples were topped up to 400 mL with distilled water.  Samples were acidified with sulphuric acid to pH ≤ 2.0 and the extraction process was initiated immediately.  4.2.4 BBWTS Operational Parameters A number of parameters were monitored to ensure the integrity of the biological processes responsible for wastewater treatment.  These parameters included DO, temperature, pH, HRT, TSS, SRT, COD, ammonia, nitrates, nitrites and orthophosphate.  Where possible, these parameters were maintained at the levels determined to be optimal for sterol degradation according to Mahmood- Khan (2005).    39  4.2.4.1 DO, temperature and pH DO, temperature and pH were measured with probes placed directly into the reactors.  DO and temperature were measured daily in both reactors using a single YSI 5740 probe attached to a YSI 54A oxygen meter (YSI Inc., Yellow Springs, OH).  This instrument and probe were calibrated before each use.  The YSI probe was on a monthly maintenance schedule which included washing out and replacing the filling solution and replacing the membrane.  The pH of the two reactors was maintained with HANNA DP 7916 pH controllers (HANNA Instruments, Laval, QC) that each had an accupHast refillable pH probe (Fisher Scientific, Pittsburgh, PA).  Readings for pH were obtained continuously by the controllers. The probes were on a weekly maintenance schedule which included calibration and washing and replacement of the filling solution when the probes were less than half full.  4.2.4.2 HRT Actual HRT was measured on most days along with final effluent collection.  Collection start and stop time was noted and the duration of collection as well as the total collection volume were used to calculate flow through the system using Equations 1 and 2.  Pump speeds from the feed basin to the reactors were adjusted when necessary in order to maintain a theoretical value of 11 to 12 hours. This is the HRT considered ideal for plant sterol degradation as per Mahmood-Khan (2005).  HRT = V  (1)  Q  HRT =  volume reactor  * length of time final effluent collected   (2)   volume final effluent collected  4.2.4.3 TSS Total suspended solids were measured twice weekly using standard method 2540 D (American Public Health Association (APHA) et al. 1998).  Samples were filtered through 1.6 μm fibreglass filters and dried at 103 to 105oC for a minimum of 2 hours.  Filtration volume and residual weights  40 were used to calculate TSS.  Filtration volumes varied between sample types and are provided in Table 4.2. Table 4.2:  Filtration volumes for calculation of TSS.  These are maximum values as lower volumes were sometimes used depending on the filtration rate. Sample Type Volume (mL) Feed 100 Aerobic System:      Treated Effluent 100      Reactor 15      Clarifier 20      Final Effluent Settled Solids  5 Microaerophilic System:      Treated Effluent 100      Reactor 10      Clarifier 15      Final Effluent Settled Solids 5  4.2.4.4 SRT and wasting Actual solids retention time was calculated twice weekly using the results of the TSS analysis.  The equation used for this calculation is provided in Equation 3.  Under ideal conditions, the system would demonstrate good settleability and reactor solids would suffice to represent total system solids.  This was practiced for the first 79 d of BBWTS operation (approximately half way through the 20 d SRT experimental run), until it was noticed that settleability was variable, and at times, quite poor.  Following, total system solids were calculated as a sum of solids concentrations in the reactor, clarifier, and when necessary, final effluent settled solids.  41  SRT =  total system solids  (3)  losses in effluent + losses in wasting Where: Total system solids  = reactor TSS * reactor volume, (mg) ; or = (reactor TSS * reactor volume)+(clarifier TSS * clarifier volume), (mg); or = (reactor TSS * reactor volume)+(clarifier TSS * clarifier volume)+(final effluent settled solids TSS * final effluent settled solids volume), (mg)  Losses in effluent = final effluent TSS / effluent collected per day, (mg/d)  Losses in wasting = reactor TSS / volume wasted per day, (mg/d)  SRT was controlled to maintain the theoretical SRTs of 5 d or 20 d, depending on the experimental run, by varying wasting volumes.  Wasting was conducted daily from the reactors, based on the semi-weekly TSS calculations.  After measuring TSS, Equation 3 was rearranged to solve for the required losses in wasting.  This is shown in Equation 4.  Wrequired =  total system solids  -  losses in effluent 1 (4)  SRT   reactor TSS  Where: Wrequired = wasting required (mL/d)  4.2.4.5 COD Total and soluble COD were measured twice weekly using the Closed Reflux, Colorimetric Method, as per Standard Method 5220D (APHA et al. 1998).  Here, dichromate is oxidized to produce the chromic ion (Cr3+), which strongly absorbs light at 600 nm.  Absorption was measured using a Hach DR 2800 spectrophotometer (Hach Company, Loveland, CO).  At the very beginning of the study, effluent samples were measured for chloride concentration, as chloride is the most common interferent in the COD method (APHA et al. 1998).  A Horiba D-22 pH/mV meter (Anachemia  42 Science, Richmond, BC),  was used to measure electrical signal (mV), and three standards with known chloride concentrations were tested to produce a calibration curve.  Following, samples from the feed, aerobic effluent and microaerophilic effluent were tested.  Chloride concentrations ranged from approximately 3500 to 4500 mg/L in the test samples, therefore mercuric sulfate (HgSO4) was added to the COD digestion reagent.  Samples were digested using a COD Reactor (Hach Company, Loveland, CO).  4.2.4.6 Nutrients Nitrogen (N) was measured as ammonia-N and nitrite plus nitrate-N and phosphorus (P) was measured as orthophosphate-P.  Each of these parameters was measured by flow injection analysis (FIA) using a Lachat Quikchem 8000 (Lachat Instruments, Milwaukee, WI).  Nutrient samples were filtered prior to analysis using 0.45 μm membrane filters.  Ammonia-N Ammonia-N was measured using Standard Method 4500-NH3- H (APHA et al. 1998).  Here, ammonia and ammonium cations react with reagents in the FIA carrier stream to produce a blue indophenol dye.  This colour is intensified and absorbance is measured at 630 nm.  Nitrite plus nitrate-N Nitrite plus nitrate-N was measured using the Cadmium Reduction Flow Injection Method, as per Standard Method 4500-NO3- I (APHA et al. 1998).  Initially, the sample is passed through a copperized cadmium column to reduce nitrate to nitrite.  Following, a sulfanilamide colour reagent diazotizes the nitrite to a magenta dye and absorbance is measured at 540 nm.  Orthophosphate-P Orthophosphate-P was measured using Standard Method 4500-P G (APHA et al. 1998).  Here, a complex is formed through reaction with ammonium molybdate and antimony potassium tartrate under acidic conditions.  Ascorbic acid is then used to produce a blue complex and absorbance is measured at 880 nm.     43 4.2.4.7 QA/QC Approximately one in ten COD and nutrient samples were collected in duplicate to assess the error in the measurement of operational parameters.  Coefficient of variation (CV) was calculated for each duplicate pair, and duplicate measurements were averaged to provide experimental values.  CV values were used to account for human error incurred during sampling and analysis, as well as random instrumentation error.   4.3 QUANTIFICATION OF PLANT STEROLS AND POTENTIAL BIOTRANSFORMATION PRODUCTS Concentrations of plant sterols and their potential biotransformation products were measured at Environment Canada’s Pacific Environmental Science Centre (PESC) in North Vancouver, BC.  All samples analyzed for organic parameters were analyzed in accordance with the methods described in this section, including samples collected for comparisons of sterols and biotransformation products, and for samples collected for the phase-separation mass balance.  4.3.1 Analytical Schedule and Methods The target organic parameters were quantified according to Environment Canada’s Sterols Method V 1.1 (Environment Canada 2005b) using the procedure for unfiltered effluent samples.  There are a total of 18 parameters included in the method; these are listed along with their limits of quantitation (LOQ) in Table 4.3.   This method was selected because it included the most abundant phytosterols as well as a number of natural and synthetic estrogenic and progestational animal steroids which may potentially be generated by phytosterol biotransformation.  Four of the estrogenic hormones included in this list (17α-estradiol, 17β-estradiol, estrone and mestranol) have been previously identified in biologically treated pulp mill effluent (Fernandez et al. 2007).  The method includes a liquid-liquid extraction using dichloromethane, derivitization by acetylation, and measurement by gas chromatography-mass spectrometry (GC-MS) in selective ion monitoring (SIM) mode.  Samples were evaporated and brought up to 1 mL with internal standard (0.5 µg/mL p-terphenyl-d14) prior to running on the instrument.  Tasks conducted by PESC staff included  44 preparation of reagents and standards, operation of the gas chromatograph and mass spectrometer, instrumental QA/QC, and cleaning of glassware.  In addition to the parameters listed in Table 4.3, PESC staff added four target parameters to the method for the purposes of this study.  These were: progesterone (Pr), androstenedione (AD), androstadienedione (ADD) and testosterone (T). The three androgens (AD, ADD and T) are controlled substances and were imported into Canada from Sigma-Aldrich (Saint Louis, MO) for the purposes of this project by Health Canada.  Table 4.3:  Limits of quantitation for parameters in Environment Canada’s sterols method (LOQ). From: (Environment Canada 2005b). Target Compound LOQ (µg/L) Campesterol  0.005 Cholesterol 0.005 Coprostanol 0.005 Desmosterol 0.005 Dihydrocholesterol 0.005 Epicoprostanol 0.005 Equilin 0.005 Equol 0.2 17α-Estradiol 0.005 17β-Estradiol 0.005 Estriol 0.02 Estrone 0.005 17α-Ethynylestradiol 0.02 Mestranol 0.005 Norgestrel 0.2 Norethindrone 0.2 β-Sitosterol 0.005 Stigmasterol 0.005  The four additional compounds were included in the study primarily because there is evidence that each of these compounds can be generated by biotransformation of phytosterols.  AD, ADD and Pr are known intermediaries between phytosterols and other compounds (Marsheck et al. 1972; Mahato  45 & Garai 1997; Fernandes et al. 2003; Perez et al. 2006; Stefanov et al. 2006)  and have all been previously identified in pulp mill effluent receiving waters (Jenkins et al. 2001; Jenkins et al. 2003) as described in Chapter 2.  Testosterone, though not documented in pulp mill effluent receiving waters, can be formed under controlled conditions through biotransformation of β-sitosterol via oxidation to AD, followed by reduction to T (Lo et al. 2002).  It was decided that reducing conditions in the microaerophilic system deemed it useful to examine the possibility of T production.  It was also advantageous to add AD, ADD and T to the study because their inclusion broadened the analytical package to include androgenic compounds.  This is significant given the androgenic nature of some pulp mill effluents, as described in Chapter 2.  With the inclusion of the four additional compounds, a total of 22 organic parameters were analyzed.  These can be grouped into four categories: plant steroids, estrogenic animal steroids, progestational animal steroids and androgenic animal steroids.  A list of the organic parameters grouped in this manner is provided in Table 4.4.  Background information about these compounds is provided in Chapter 2.  Table 4.4:  Analytical schedule for organic parameters, grouped by steroid classification. Plant Steroids Estrogenic Animal Steroids Progestational Animal Steroids Androgenic Animal Steroids Campesterol  Equilin Norgestrel Androstendione Cholesterol Equol Norethindrone Androstadienedione Coprostanol 17α-Estradiol Progesterone Testosterone Desmosterol 17β-Estradiol Dihydrocholesterol Estriol Epicoprostanol Estrone β-Sitosterol 17α-Ethynylestradiol Stigmasterol Mestranol  The compatibilities of the four additional parameters with the Environment Canada Sterols Method were determined by PESC staff, by first analyzing high concentration (4 µg/mL) standards by GC- MS in full scan mode.  Resultant total ion chromatograms were verified using the reference library in the National Institute of Standards and Technology database.  The ion with the highest mass to charge ratio (m/z) intensity was chosen as the target ion to correlate with concentration for each  46 parameter.  The two ions with the next highest m/z intensities were chosen as qualifying ions.  The retention times of the target ions were checked against those of the other parameters in the method to confirm there was no overlap.  Following this, a range of standards were generated and four point calibration curves were produced in SIM mode.  The method was validated for each of the four additional parameters by analyzing high and low spikes in distilled water, using concentrations of 2 µg/mL and 0.05 µg/mL respectively in final 1 mL samples prepared for GC-MS analysis.  These corresponded to concentrations of 5 µg/L and 0.125 µg/L respectively in 400 mL samples (as collected during BBWTS sampling) prior to preparation for GC-MS.  Three sets of spikes were analyzed to test the effects of extraction and derivitization on compound recoveries.  The first set was treated identically to experimental samples, the second set was derivitized but not extracted and the third set was extracted but not derivitized. LOQs were not determined for these parameters.  4.3.2 QA/QC Numerous QA/QC procedures were carried out for samples analyzed for organic parameters, including analysis of surrogates, replicates, blanks and one direct spike.  4.3.2.1 Surrogates As per Environment Canada’s Sterols Method (2005) 17β-estradiol-d3 was used as a surrogate analyte at a concentration of 2 µg/mL to monitor extraction efficiencies.  Each sample was injected with 250 µL of surrogate prior to extraction to produce a theoretical surrogate concentration of 0.5 µg/mL in the final 1 mL sample prepared for GC-MS.  Theoretical and actual surrogate concentrations were compared to calculate percent surrogate recovery for each sample analyzed.  4.3.2.2 Replicates Replicate samples were collected in order to capture variation in the experimental design and the analytical methods.  For each 4 to 5 d effluent compositing event, the two 1 d composite samples represented experimental replicates and were collected to show variation due to changes in the wastewater treatment system.   The triplicate 4 to 5 d composite samples represented method  47 replicates and were collected to show variation due to changes in the analytical procedures or instruments.  Phase separation samples were collected in triplicate for the feed and final effluents and in duplicate for the reactor samples in order to capture experimental variation.  No method replicates were collected during this portion of the study because of constraints on the number of samples that could be processed, and because many method replicates had been previously collected.  It was determined to be more beneficial to collect experimental replicates for the phase-separation samples, in order to evaluate effect of variation in the wastewater treatment system on analyte phase.  4.3.2.3 Blanks Each batch of samples analyzed using the Environment Canada Sterols Method included one blank, which was prepared at the time of sample collection and treated identically to the samples.  Blanks were made up of distilled water and were used to test for contamination of samples from time of collection through sample preparation and GC-MS analysis.  4.3.2.4 Direct spike One direct spike was analyzed during the study to test the effect of derivitization on compound recoveries.  Here, 25 μL of sterols solution was analyzed, which contained all compounds listed in Table 4.3 in known concentrations from 20 to 100 μg/mL,  The direct spike was aliquoted at the time of sample preparation and was treated identically to the samples, excepting extraction steps.  4.3.3 Organic Chemistry Data Analysis Removal efficiencies were calculated for each set of effluent and influent concentrations using Equation 5, and were then averaged by parameter, for each treatment.  For removal efficiency calculations, values that were less than the detection limits were converted to zero.  Using this approach, concentrations were biased low and removal efficiency may have been biased high. Removal efficiency was not calculated for cases in which effluent concentration was higher than influent concentration.  In these instances, metabolite production was assumed to have occurred. Metabolite (byproduct) generation values were calculated as effluent concentration less influent concentration and were averaged by parameter across treatments.  48  Removal Efficency = 1 -  Effluent Concentration X 100% (5)   Influent Concentration  Statistical analyses were performed using R software (R Foundation for Statistical Computing, 2008). The non-parametric Wilcoxon Rank Sum Test was used to compare the significance of differences in sterols removal and biotransformation between treatments.  Data were assumed to be non- normally distributed following review of Q-Q plots, in which data deviated from linearity.  Q-Q plots are provided in Appendix A.  The Wilcoxon Rank Sum Test requires data to be independent; however in this case, concentrations in final effluents were dependant on the concentrations in the influents.  The effect of the influent was accounted for by converting effluent concentrations to the net change in concentration (i.e. effluent concentration less influent concentration) prior to calculation of p values.   4.4 GENOMIC RESPONSE OF RAINBOW TROUT TO EXPOSURE TO TREATED PULP MILL EFFLUENTS Composited effluent from each of the four wastewater treatments was used for biological effects analysis consisting of fish exposure assays and toxicogenomic analysis.  PESC staff assisted with all aspects of this phase of the project.  Approval for the fish exposure work was obtained from Environment Canada and UBC Animal Care, and documentation is provided in Appendix B.  4.4.1 Fish Exposure Bioassays Underyearling rainbow trout (Oncorhynchus mykiss) were exposed to the effluents at 1% and 10% g/g for 96 hr under controlled conditions in the PESC laboratory.  The bioassays were carried out using Environment Canada’s Biological Test Method: RM 9 Acute Lethality Test Using Rainbow Trout (Environment Canada 2007).  Following the exposure period, the fish were euthanized and liver and muscle tissues were dissected, persevered in RNA Later® and retained for genomic analysis.  Bioassays were carried out using aerobic and microaerophilic effluents generated with a 20 d SRT from December 7th to 11th 2007, and with a 5 d SRT from February 8th to 12th 2008.  Composited effluent was first brought to PESC where it was well mixed and portioned into glass fish tanks  49 within a climate controlled chamber (15 ºC ± 1).  Tanks were filled to 65 L using a combination of effluent and groundwater from the PESC freshwater well.  Nine tanks were used per treatment, including three containing 1 % effluent, three containing 10 % effluent and three containing 100 % well water (controls).  Effluent was measured by weight so that the 1 % and 10 % tanks contained 0.65 kg and 6.5 kg of effluent, respectively.  Given that two concentrations were used per effluent, in total there were 8 treatment groups and 2 control groups of test fish.  Exposure groups are summarized in Table 4.5.  Table 4.5:  Bioassay treatment and control groups. Treatment Code BBWTS Treatment Conditions Assay Batch 1: December 7-11th, 2007 Control 1 n.a. - 100% PESC well water AB1 1% aerobic effluent generated using a 20 d SRT AB1 10% aerobic effluent generated using a 20 d SRT MP1 1% microaerophilic effluent generated using a 20 d SRT MP1 10% microaerophilic effluent generated using a 20 d SRT Assay Batch 2: February 8-12th, 2008 Control 2 n.a. - 100% PESC well water AB2 1% aerobic effluent generated using a 5 d SRT AB2 10% aerobic effluent generated using a 5 d SRT MP2 1% microaerophilic effluent generated using a 5 d SRT MP2 10% microaerophilic effluent generated using a 5 d SRT  Ten fish were added to each bioassay tank, and best attempts were made to include those of similar size with no obvious deformations or other abnormalities.  The fish used in the bioassays were obtained by PESC staff from Sun Valley Trout Farm in Mission, BC, on September 24th, 2007 and were reared at PESC until the time of exposure.  The trout farm holds a Fish Health Certificate issued by the Department of Fisheries and Oceans, indicating the fish are free of disease and disease agents.  In addition, prior to their use in the bioassays, the cohort of fish was tested to ensure their quality, as described in Section 4.4.1.3, below.     50 4.4.1.1 Bioassay test conditions The test conditions used during the 96 hr exposure period followed Environment Canada’s Biological Test Method: Acute Lethality Test Using Rainbow Trout (Environment Canada 2007).  A summary of these conditions is provided in Table 4.6.  Table 4.6:  Select test conditions for the 96 hour rainbow trout bioassays.  From Environment Canada (2007). Test Parameter Target Value Temperature 15 ± 1 oC Photo period 16 ± 1 hr light:8 ± 1 hr dark DO ≥ 70% saturation pH 5.5 - 8.5   Tanks were filled and aeration was initiated the day before the fish were added, allowing for a pre- aeration period.  Just prior to adding the fish, tank water was measured for conductivity, pH, DO and temperature.  These parameters, excepting conductivity, were measured again at the end of the test period, prior to removing the fish.  Also, throughout the test period, fish were monitored for mortalities and other abnormalities, relative to controls.  4.4.1.2 Dissection and preservation Following the 96 hr exposure period, fish were removed from the tanks and dissected.  Each fish was individually euthanized in accordance with the Canadian Council of Animal Care protocols. Ethical approval is documented in Appendix B.  After being removed from the tank, fish were anesthetised in tricaine methanesulfonate.  Following, they were measured for fork length and weight and were then decapitated.  Observations were recorded to note any disfigurations or discolourations.  All liver tissues and a portion of muscle tissue were removed from each fish and preserved separately in RNA Later™ in 1.5 mL certified nuclease-free plastic tubes.  Tubes were labelled with fish and treatment ID and date of dissection and were stored at -80oC.     51 4.4.1.3 QA/QC Samples of the test effluents and fish were analyzed prior to the bioassays to ensure method requirements were met.  Aerobic and microaerophilic effluent were tested by PESC staff for acute toxicity to provide evidence that the effluent would not be acutely toxic to the fish. This was conducted because genomic analysis requires the fish to survive the exposure period.  Effluent acute toxicity was tested on Microtox® using Environment Canada’s Biological Test Method: Toxicity Test Using Luminescent Bacteria (Environment Canada 1992).   The results of this test are known to correlate well with the acute toxicity to rainbow trout (Bulich & Isenberg 1981).  A sample of the fish cohort was tested using a standard reference toxicant (phenol) to ensure the sensitivity of the test population was within the charted value for the species. This test was carried out by PESC staff as per Environment Canada’s Biological Test Method: Acute Lethality Test Using Rainbow Trout (Environment Canada 2007).  Using this method, the LC50 was measured after 96 hr of phenol exposure and this value was compared to the historic geometric mean and its 2- standard deviation warning limits.  4.4.2 Genomic Analytical Methods RNA was extracted from liver tissues and was analyzed to compare the expression of select genes between fish exposed to effluents and control fish.  To summarize, RNA was extracted from the livers, transcribed into complementary DNA (cDNA) and measured by quantitative polymerase chain reaction (QPCR).  Six genes were selected for this analysis following an initial gene screening process.  In addition, a reference gene was analyzed, which does not change expression with exposure to toxicants, to provide a baseline of comparison.  Thus, in total, liver samples were used to measure the expression of seven genes.  Nine fish (n of 9) were used for this analysis from each of the 8 treatment and 2 control groups, totalling 90 samples analyzed per gene.  This value exceeds an n of 6, which is the minimum acceptable value in PESC for statistical analysis of gene expression data using rainbow trout (Osachoff, H., Toxicogenomic Technologist, Environmental Toxicology Section, PESC, personal communication, May 2008).  An n of 9 was chosen to maximize the number of samples analyzed within the time constraints of the method.  Each QPCR run required approximately 2 hr, therefore  52 four plates could be run over the duration of a work day.  Also, it was essential to run all four replicates of a single gene on the same day, using a consistent QPCR reagent mixture made fresh on the day of analysis.  Therefore, one replicate for a single gene had to fit on a plate.  Each plate contains 96 wells, therefore an n of 9 was the maximum possible.  4.4.2.1 Extraction of total RNA Genes are expressed by the action of a series of proteins for which synthesis is initiated by messenger RNA (mRNA).  In this way, the presence of a specific mRNA sequence represents the expression of an associated gene on the molecular level.  Total RNA was extracted from the fish livers using Environment Canada’s Standard Operating Procedure, Extraction of Total RNA from Tissues Stored in RNA Later™ (Environment Canada 2006a).  To be conservative, RNA extractions were conducted using an n of 12 (totalling 120 samples), in hopes that nine from each group would be suitable for further analysis.  The 120 samples were selected from all the fish used in the bioassays using observations taken during dissections.  The heaviest fish from each treatment and control batch were chosen, excepting those with notes of discolouration, disfiguration or other such abnormalities.  The heaviest fish were assumed to have the largest livers and hence the most RNA available for analysis.  The RNA extraction procedure consisted of first weighing out 30 to 40 mg of tissue per liver sample.  Weights were recorded and samples were homogenized to lyse the liver cells, freeing RNA from cell structures.  Homogenization was conducted by inserting a 3 mm stainless steal bead in each sample tube and shaking the tubes for a minimum of 6 min in a Mixer Mill 300 (Qiagen Inc., Mississauga, ON) at 25 Hz.  Tubes were inspected for uniformity and put back on the Mixer Mill where solids were still observed.  When homogenization appeared complete, the lysate was centrifuged and the supernatant transferred to RNeasy® MINI spin columns (Qiagen Inc., Mississauga, ON).  The samples were further centrifuged and washed several times, the column membrane was dried, and total RNA was eluted off the column using RNase-free, sterile water. Following extraction, the concentration of total RNA was measured spectrophotometrically using a NanoDrop ND-1000 with software ND-1000 (Fisher Scientific Limited, Nepean, ON).     53 4.4.2.1.1 QA/QC The overall quality of total RNA samples was tested by electrophoresis following the Agilent RNA 6000 Nano Kit Guide (Agilent Technologies 2006), along with an RNA 6000 Nano Kit, 2100 Bioanalyzer and 2100 Expert Software (Aglient Technologies, Mississauga, ON).  The bioanalyzer is located at the British Columbia Institute for Technology (BCIT, Burnaby, BC).  The kit includes chips with a series of interconnected channels that are used to separate nucleic acid fragments by size.  Chips are placed in the Bioanalyzer and the samples are electrophoretically driven through the channels.  The software produces electropherograms which show peaks for the two most abundant fragments of ribosomal RNA.  Ribosomal RNA is used to indicate the quality of the total RNA because ribosomal is the most abundant form of RNA and, hence, the easiest to measure and plot. Results showing examples of good quality and poor quality RNA samples are provided in Figure 4.4.   Figure 4.4:  Example electrophoresis results for RNA quality check.  a) Electropherograms showing a good quality sample (top) and degraded sample (bottom).  b) Electrophoresis summary showing good quality samples and a degraded sample (4).  Electropherograms were examined and nine samples from each treatment and control group were selected for cDNA synthesis.  The selection criteria were that the peaks for the two most abundant fragments of ribosomal RNA (18s and 20s) had to be clearly distinct and had to occur in a position consistent with the other samples.  Where more than nine samples in a treatment or control group a) b) 18s 20s 18s 20s  54 had good quality results, samples with the highest RNA concentrations were selected for further processing.  4.4.2.2 cDNA synthesis Synthesis of cDNA converts single stranded mRNA sequences into their double stranded DNA counterparts, forming gene transcripts.  The transcription process is conceptualized in Figure 4.5. This conversion was necessary because QPCR only measures double stranded products.  mRNA AAA RTGGG Reverse Transcriptase AAA RT GGG CCC TTT AAAGGG CCC TTT cDNA 1) 2)  Figure 4.5:  Conceptualization of cDNA synthesis. 1) Reverse transcriptase produces a strand of nucleic bases complementary to the mRNA sequence.  2) Base pairs connect forming a strand of cDNA (the gene transcript).  Synthesis of cDNA was conducted using the Invitrogen kit SuperScriptTM III First-Strand Synthesis SuperMix (Invitrogen Canada Inc., Burlington, ON) and associated protocol (Invitrogen Life Technologies 2004).  The kit includes various reagents including reverse transcriptase and polymerase enzymes. Reagents were mixed with RNA samples and reacted by incubation using Environment Canada’s polymerase chain reaction (PCR) instrument, MJ Research DNA Engine PTC-200 (MJ Research Inc., Waltham, MA).  The volume of RNA used in the cDNA synthesis reactions depended on the RNA concentration of each sample, so that resultant cDNA samples had identical concentrations.   55 4.4.2.2.1 QA/QC The products of the cDNA synthesis reactions were tested for the presence of the L8 reference gene to determine whether the reactions were successful.  L8 is normally present in fish liver samples and expression of L8 does not change due to toxicity (Osachoff, H., Toxicogenomic Technologist, Environmental Toxicology Section, PESC, personal communication, May 2008). Reactions were said to be successful where the L8 gene transcript was present.  This test was conducted using QPCR methods as described below.  4.4.2.3 QPCR Complimentary DNA samples were analyzed by QPCR as per Osachoff (2008) using Environment Canada’s Stratagene Mx3000P with software MxPro v3.0 (Stratagene, Cedar Creek, TX).  QPCR was first used first to analyze a selection of the test samples to evaluate the activity of 28 endocrine- related genes.  Following this, six of the most active genes and one reference gene were selected for analysis and the QPCR method was adapted to each.  After the method development was complete, all 90 samples were analyzed by QPCR for each of the seven genes in quadruplicate.  The main theorem of QPCR is that a reaction takes place to exponentially amplify a targeted section of DNA (termed amplicon), which is quantified by fluorescence on each amplification cycle. Preparation of cDNA samples for QPCR involved generating a reaction mixture, aliquoting it into QPCR tubes, then adding cDNA samples to each of the tubes.  The reaction mixture included primers (short pieces of DNA made to target a specific gene), reagents to amplify the target gene sequence (polymerase enzyme, deoxynucleotides (dNTPs) and buffer with magnesium), and a fluorescent dye (SYBR Green 1) and sterile, RNase-free water.  The fluorescent dye binds to the target gene sequence, thus fluorescence is directly proportional to the amount of PCR product produced.  For this project, the buffer reagent was modified from that used in the reference method by Osachoff (2008).  The buffer was made up of 10 mM Tris-HCl (pH 7.5 at 20oC), 50 mM potassium chloride, 3 mM magnesium chloride, 0.01% Tween 20, 0.8% glycerol, 40,000-fold dilution of SYBR Green I, 200 μM dNTPs, and 83.3 nM ROX reference dye.  For each QPCR reaction, the total volume of reagents and cDNA was made to equal 15 μL.  The ratio of reagents : cDNA was varied by adjusting the amount of water in the reagent mixture.  The specific volumes of  56 cDNA used for gene selection, method development and final analysis are described the associated sections, below.  The QPCR instrument operates by cycling between temperatures to exponentially amplify the target gene transcript.  The thermocycle profile used for all QPCR runs for this project is provided in Table 4.7.  Table 4.7:  QPCR thermocycle profile.  This profile was used for all QPCR plates analyzed.  Cycle Number of Cycles per QPCR Run Time  Temperature  (oC) Pre- Amplification 1  2 min 50    9 min  95 Amplification 40  20 sec 95    30 sec 58   40 sec 72 Dissociation 1 1 min 95   30 sec 55   30 sec 95  Each amplification cycle theoretically doubles the amount of the target gene (the amplicon), so the amount produced is 2n, where n is the number of cycles (Schefe et al. 2006).  The QPCR instrument outputs amplification plots which graph cycle number against fluorescence.  An example amplification plot is provided in Figure 4.6.  These graphs show cycle at threshold (Ct) values, which represent the value at which a target gene begins amplification.  Threshold is a relative value calculated by the software and is equivalent to 10 times the signal to background noise ratio. Threshold was calculated for each target gene individually during method development (see below). The threshold was applied to experimental samples and Ct values were assigned where the amplification curves cross the threshold, as illustrated in Figure 4.6   57 Cycle Fl uo re sc en ce  (d R n) 10 40 0.1 0.2 0.3 0.4 0.0 Threshold NTC Experimental Samples Example Ct values (cycle at threshold) 20 30 Fl uo re sc en ce  (d R n)  Figure 4.6:  Example amplification plot.  This plot was generated from one QPCR plate and includes 90 experimental samples and one non-threshold control (NTC).  Copy numbers are the final data used for QPCR analysis and are calculated from Ct values using calibration curves.  In theory, the copy number is the number of copies of a gene transcript in the original tissue.  In this way, copy number represents gene expression.  Ct values are inversely proportional to copy numbers, as tissues with higher copy numbers will amplify more quickly and therefore, have lower Ct values.  4.4.2.3.1 QA/QC A number of QA/QC measures were carried out during QPCR analysis to confirm that the instrument was working well and that the correct product was being measured.  These measures consisted of analysis of non-threshold controls, dissociation curves and reference samples.      58 4.4.2.3.1.1 NTC A non-threshold control (NTC), containing all QPCR reaction reagents except for cDNA, was included on each QPCR plate analyzed.  The NTC acts as a method control to ensure the fluorescent signal in experimental samples is due to the target gene only.  This is necessary because SYBR dye binds to all double-stranded DNA.  If the QPCR reaction reagents are contaminated with non-specific DNA fragments, the NTC will show lower than expected Ct values on the amplification plots.  It is generally accepted that the NTC must appear at minimum 2 Ct values higher than the samples for it to be acceptable (Osachoff, H., Toxicogenomic Technologist, Environmental Toxicology Section, PESC, personal communication, May 2008).  An acceptable NTC is illustrated in Figure 4.6 with a Ct value 9 cycles higher than the highest test sample.  4.4.2.3.1.2 Dissociation curves Dissociation curves were generated for all samples analyzed by QPCR in order to verify that the reaction was specific to a single gene sequence.  The thermocycle profile used to generate dissociation curves is provided in Table 4.7. The QPCR instrument was set to run through a dissociation cycle following generation of amplification plots.  During the dissociation cycle, the signal is measured while temperature is gradually increased.  At a specific temperature the double stranded amplification products dissociate, producing a spike in fluorescence (the melting point). Each unique gene sequence has a unique melting point, therefore, a homogeneous mixture of amplification products will result in a single spike on the plot.  Example dissociation curves, showing amplification of a single product, are provided in Figure 4.7.           59 Temperature (oC) Fl uo re sc en ce  (- R ’(T )) 90 1200 NTC 0 400 800 8070 Experimental Samples Fl uo re sc en ce  (- R ’(T ))  Figure 4.7:  Example dissociation curves.  This plot includes curves for 90 experimental samples and one non-threshold control (NTC), representing one QPCR plate.  4.4.2.3.1.3 Reference sample A reference sample of mouse cDNA was analyzed by QPCR at a minimum of once daily while the instrument was in use.  These samples were used as instrument controls to ensure that the QPCR machine was working correctly.  The mouse cDNA is a standardized reference sample used at PESC and is known to have a well defined dissociation curve, with a melting point near 90oC.  If the instrument was not working correctly, it is assumed that the dissociation curve would appear irregular.  4.4.2.4 Gene selection During cDNA synthesis, all mRNA sequences were converted to their DNA counterparts, therefore, all the genes expressed in the liver samples were transcribed.  In order to choose 6 genes for analysis, a subset of 12 of the 90 test samples was evaluated for the activity of 28 endocrine- related and reference genes.  The 28 screening genes, listed in Table 4.8, were chosen with PESC staff based on the following criteria:  1) the gene sequence must have been previously determined by  60 PESC staff, or be publicly available in  the GenBank® database (National Center for Biotechnology Information 2008); 2) the gene must be related to the endocrine system either directly or through cross-talk between pathways; and 3) the gene may be suspected to illustrate a response from exposure to pulp mill effluents.  The 12 samples chosen for analysis for gene selection purposes included two samples from each of the following bioassay groups: Control 1, AB1 10%, MP1 10%, Control 2, AB2 10%, MP2 10%.  It was assumed that by analyzing samples from the control and 10% effluent groups the extreme changes in gene expression would be captured, thus showing which of the 28 screening genes were the most active.  QPCR tubes were prepared by aliquoting 14.5 μL of QPCR reagent mixture, then adding 0.5 μL of cDNA.  The primers used for each gene are listed in Table 4.8.  Following QPCR analysis for gene selection purposes, amplification products were separated by agarose gel electrophoresis.  This was conducted by first mixing 9 μL of sample with 1 μL of 10X loading buffer (30% sterile glycerol, 0.25 % bromophenol blue, 60 mM EDTA, 40 mM tris-acetate). The mixtures were run through a 2 % agarose gel at 50 V for 50 min in Sub-cell GT electrophoresis boxes (Bio-Rad Laboratories, Hercules, CA).  The first well of each column on the gel was filled with DNA ladder as a reference to estimate the size of the amplification products, in base pairs (bp). Following electrophesis, gels were photographed using a Gene Genius Bio-Imaging System with software GeneSnap v6.05 (Syngene, Frederick, MD).  Photographs were examined and used to confirm the validity of the QPCR products.  Results were considered indicative of the expression of the target gene if both of the following criteria were met: 1) one single clear and bright band was present and 2) the size of the products in the band coincided with the size of the target gene sequence.  The results of QPCR and electrophoresis were compiled and examined in detail with PESC staff in order to select seven genes for analysis, including six experimental and one reference gene.  The following criteria were required for a gene to be selected: 1) the NTC had to be a minimum of 2 Ct values greater than the samples; 2) dissociation curves had to display one melting point, and all curves had to fit the same line tightly; and 3) gels had to display one clean, bright band with products of the same size as the target gene sequence.   61 Table 4.8:  Genes screened in select test samples.  The activity of these genes was evaluated to select 6 target genes and 1 reference gene for analysis of all 90 test samples. Gene Symbol Gene Name PESC Primer ID Size (bp) Primer 1 Primer 2 Endocrine Gene Class ERA short estrogen receptor α – short 228 229 502 ERA long estrogen receptor α – long 591 592 237 VEPA vitelline envelope protein α  252 253 490 VEPB vitelline envelope protein β 326 327 537 VEPG vitelline envelope protein γ  254 255 465 VTG 1 vitellogenin 1 219 220 488 VTG 2 vitellogenin 2 593 594 445 AR androgen receptor α + β (Arboth) 21 22 471 GNRH1 gonadotropin releasing hormone 1 264 265 419 GNRH2 gonadotropin releasing hormone 2 266 267 463 GNRHR2 gonadotropin releasing hormone receptor 2 268 269 535 StAR steroidogenic acute regulatory protein 373 374 534 FSHR follicle stimulating hormone receptor 538 539 502 GH1 growth hormone 1  85 86 515 INSRB insulin receptor β 107 108 546 THRA thyroid hormone receptor α 246 247 543 Metabolism Gene Class CYP11A1 cyt p450 cholesterol sidechain cleavage (cytp450scc) 312 313 480 CYP11B1 11-β-hydroxylase (11-β-hyd) 278 279 506 CYP17A1 17-α-monooygenase (17-α-mo) 280 281 501 CYP19B cyt p450 aromB-I (AROMB) 454 455 527 HSD3B1 3-β-hydroxysteroid dehydrogenase (3β-HDE) 5 6 513 PTGS2 cyclooxygenase2 (COX-2) 288 289 502 NME nucleoside diphosphate kinase (NucDiKin) 512 513 402 Signal Transduction Gene Class AHR aryl hydrocarbon receptor α + β (arylHCRα+β) 23 24 522 ARNT arylHCR translocator (aryltransα+β) 284 285 514 HSPCA heat shock protein 90 (hsp90) 388 389 496 Reference Gene Class RPL23A ribosomal protein 60S  428 429 278 L8 ribosomal protein L8 (UL8A) 599 602 270   62 4.4.2.5 Method development Once the six analytical genes and one reference gene were selected, standard curves were produced for each in order to correlate Ct values with copy numbers.  Standards with known copy numbers were made from plasmid vectors, circular molecules of specific DNA sequences isolated from bacteria.  Plasmids were generated by PESC staff using the pCR®2.1 TOPO® vector from the Invitrogen TOPO TA Cloning® Kit (Invitrogen Canada Inc, Burlington, ON), in conjunction with the PESC operating procedure, Cloning from Excised Bands (Environment Canada 2006b). Plasmids generated using this method were serially diluted to produce eight standards with copy numbers from 1,000 to 1,000,000.  QPCR tubes were prepared by aliquoting 13 μL of QPCR reagent mixture, then adding 2 μL of standard, and samples were analyzed using the QPCR methods described above.  Following, instrument output was analyzed to determine whether the curves met the following criteria:  1) The correlation coefficient (R2) of the line had to be greater than or equal to 0.95; 2) the doubling efficiency of the amplicon, calculated from the slope of the line, had to be between 60 and 120%; and 3) the threshold value had to be less than 0.020.  The threshold was unique to each target gene and was calculated by the software based on the fluorescent signal between cycles 4 to 12, which is considered ideal by PESC staff for evaluating background noise. The threshold calculation included a safety factor of 10, ensuring Ct values were above the background noise levels.  4.4.2.6 Final sample preparation Following method development, all 90 experimental samples were analyzed by QPCR in quadruplicate for each of the seven genes. The methods used are described in the QPCR section above, apart from the reaction volumes and concentrations.  For final analysis, all QPCR tubes were prepared by aliquoting 13 μL of QPCR reagent mixture, then adding 2 μL of cDNA.  It was considered essential to maintain the ratio of QPCR reagents to cDNA, in order to produce consistent results.  In addition, this same ratio was used during production of standard curves; therefore, the final data was well suited to the method.  Although the same volume was used for each reaction, the concentration of cDNA in the 2 μL aliquot was dependant on the gene being analyzed.  This was necessary as two of the seven genes selected for analysis indicated lower than ideal Ct values during the gene screening process.  63 According to PESC procedures, Ct values were required to appear between 12 to 33 cycles because, in this range, there is confidence that background interference is minimized and reagents are have not yet run out.  For the two selected genes that displayed lower than ideal Ct values during gene screen analysis (the “low-expressers”), high larger quantity of cDNA was required per reaction to bring Ct values into the ideal range (recall that for the gene selection QPCR analysis 0.5 μL of undiluted cDNA was used per reaction).  For final analysis for the the low-expresser genes, 2 μL of undiluted cDNA was added to the QPCR reagent mixture, increasing the quantity of cDNA four- fold from that used during the screening process.  For the five target genes that did exhibit Ct values in the ideal range during the gene screening process, 0.5 uL of cDNA per reaction was maintained. Here, four-fold dilutions were required to bring the reaction volume up to 2 μL for final analysis.  4.4.2.7 Bioinformatics QPCR results were analyzed as per Osachoff (2008), including references to Burns et al. (2005) and Applied Biosystems (2004).  Ct values obtained from amplification plots were first exported from the Mx3000 Pro software into Microsoft Excel, where all analyses were conducted.  Results for quadruplicate technical replicates were compared against each other in order to identify and remove outliers for individual samples.  This was conducted by calculating 25th and 75th percentiles and converting them into upper and lower limits, using an outlier coefficient of 1.5.  All Ct values outside the limits were removed and mean Ct values were obtained for each set of technical replicates.  Copy numbers were generated from mean Ct values using the calibration curve equations produced during method development.  Copy numbers for each sample for each experimental gene were normalized by comparison to the data for the reference gene (L8).  Each sample was designated a factor by dividing the average L8 copy number for all samples by the L8 copy number for each individual sample.  This factor represents the degree of derivation of each sample from the average condition.  Copy numbers for the six experimental genes were multiplied by the factor, producing normalized data.  Following this, factored copy numbers for every fish liver sample were compared for each gene in order to identify and remove outliers among bioassay groups.  This was conducted first using percentiles testing as described above, and second using the Grubbs test.  Data would only be removed if they failed both tests.  Grubbs was considered more robust, therefore, if an outlier was identified through the percentile test, but the Grubbs test showed it to be within the expected range, the datum would be maintained.  After outliers were removed, average factored  64 copy numbers were produced for bioassay groups for each gene.  Using these values, exposure groups were compared to controls to test whether the differences in average factored copy number were significant.  Both Welch’s t-test and Student’s t-test were used to calculate p values for this purpose.  The values obtained using the two tests were virtually identical; therefore, it was determined that the data were normally distributed.  As such, p values generated by the Student’s t- test were used for the final genomic data analysis to compare results between exposure groups and controls.    65 5 RESULTS AND DISCUSSION 5.1 BENCHSCALE BIOLOGICAL WASTEWATER TREATMENT SYSTEM The BBWTS was carefully monitored to ensure the integrity of wastewater treatment.  Summary statistics for BBWTS operational parameters are provided in Table 5.1, and complete data tables are provided in Appendix C.  QA/QC results are summarized and discussed in Appendix D.  5.1.1 DO, Temperature and pH Dissolved oxygen (DO) concentration in the aerobic reactor averaged 4.7 mg/L during the 20 d SRT experimental run and 4.2 mg/L during the 5 d SRT experimental run.  In the microaerophilic reactor, DO averaged 0.2 mg/L during both the 20 and 5 d SRT runs.  Figure 5.1 shows DO levels for the duration of BBWTS operation, with a substantial and relatively consistent difference in DO concentrations between the aerobic and microaerophilic systems.  These conditions met the requirements set out in the experimental design for DO.  The variation in DO concentration was satisfactory, with a maximum CV of 21 % in the aerobic system and 53 % in the microaerophilic system, as shown in Table 5.1.  It should be noted, however, that higher variation in the microaerophilic system is a reflection of the low concentration, as the range in DO concentrations in this system was 0.1 to 0.6 mg/L.  0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Day Co nc en tra tio n (m g O 2/L ) Aerobic Microaerophilic RUN 2: 5 d SRT RUN 1: 20 d SRT SET UP: 11 d SRT  Figure 5.1:  Reactor dissolved oxygen.  66 Table 5.1: Summary statistics for BBWTS operational parameters. Note that influent values are identical for aerobic and microaerophilic systems.  Asterisk (*) identifies influent TSS values, to highlight that these data are obtained from weekly averages (two samples were collected per week).  “Run 2” signifies the full duration of the second experimental run, “Run 2a” signifies the first half of the second experimental run (up to changes in COD and nitrification) and “Run 2b” signifies the second half. n X min max σ CV n X min max σ CV Dissolved oxygen (mg/L) 1 73 4.7 0.9 6.7 1.0 21% 73 0.2 0.1 0.6 0.1 51% 2 58 4.2 2.8 5.7 0.6 14% 58 0.2 0.1 0.6 0.1 53% Temperature (oC) 1 73 34.4 22.0 37.0 2.9 8% 34 33.9 25.0 36.5 2.6 8% 2 53 35.2 32.0 36.0 0.7 2% 60 34.5 31.0 35.0 0.6 2% HRT (hr) 1 38 12.0 10.2 14.5 0.8 7% 39 11.8 10.5 14.4 0.8 7% 2 41 11.6 10.4 13.1 0.5 4% 43 11.6 10.4 13.0 0.4 4% SRT (d) 1 21 20.4 8.9 30.8 3.7 18% 23 15.9 9.7 21.6 3.7 23% 2 18 5.0 5.0 5.3 0.1 2% 18 5.0 5.0 5.1 0.0 1% Solids MLSS (mg/L) 1 22 5774 3600 8720 1229 21% 23 2777 1330 4110 910 33% 2 18 2423 1040 5040 1094 45% 18 1389 880 2280 398 29% Influent TSS (mg/L) * 1 13 510 95 1216 420 82% equal to aerobic values 2 8 463 304 694 151 33% equal to aerobic values Effluent TSS (mg/L) 1 22 62 14 400 87 139% 23 96 46 148 28 29% 2 18 48 18 111 25 53% 18 77 36 128 24 31% System TSS (mg/L) 1 22 19865 11700 28340 4188 21% 23 9900 6498 13358 2203 22% 2 18 9621 5551 17778 3427 36% 18 5890 3940 8574 1276 22% COD Influent tCOD (mg/L) 1 22 1858 1328 3028 368 20% equal to aerobic values 2 18 2397 1775 3255 415 17% equal to aerobic values 2a 9 2220 1775 3173 440 20% equal to aerobic values 2b 9 2575 2207 3255 319 12% equal to aerobic values Influent dCOD (mg/L) 1 22 1382 965 1677 182 13% equal to aerobic values 2 17 1684 1331 2226 242 14% equal to aerobic values 2a 9 1493 1331 1650 105 7% equal to aerobic values 2b 8 1899 1750 2226 145 8% equal to aerobic values Effluent tCOD (mg/L) 1 22 690 507 1158 164 24% 22 1117 695 1394 167 15% 2 18 777 598 1015 139 18% 17 1286 1004 1561 218 17% 2a 9 683 598 1015 132 19% 9 1104 1004 1389 114 10% 2b 9 871 813 1004 58 7% 8 1491 1418 1561 51 3% Operational Parameter Run Aerobic System Microaerophilic System   67  Table 5.1 (continued).  n X min max σ CV n X min max σ CV COD (continued) Effluent dCOD (mg/L) 1 22 597 420 1006 131 22% 22 962 822 1194 104 11% 2 18 698 551 845 98 14% 17 1086 837 1304 177 16% 2a 9 625 551 845 90 14% 9 940 837 1194 104 11% 2b 9 770 732 804 24 3% 8 1251 1213 1304 30 2% tCOD removal effeciency 1 22 62% 22% 81% 0.1 19% 22 38% 18% 66% 0.1 30% 2 17 67% 52% 81% 0.1 11% 17 45% 26% 67% 0.1 22% 2a 9 68% 52% 81% 9% 12% 9 49% 26% 67% 12% 24% 2b 8 65% 57% 75% 5% 8% 8 41% 35% 53% 6% 16% dCOD removal effeciency 1 22 56% 16% 69% 0.1 22% 22 29% 6% 49% 0.1 42% 2 17 58% 37% 65% 0.1 11% 17 36% 28% 44% 0.0 13% 2a 9 58% 37% 65% 9% 15% 9 37% 28% 44% 5% 13% 2b 8 59% 55% 65% 3% 5% 8 34% 29% 43% 4% 13% Nutrients Influent NH3-N (mgN/L) 1 23 42.4 33.7 53.8 5.6 13% equal to aerobic values 2 17 38.8 33.1 44.8 3.5 9% equal to aerobic values Influent NOx-N (mgN/L) 1 23 0.0 0.0 0.0 0.0 0.0 equal to aerobic values 2 16 0.0 0.0 0.0 0.0 0.0 equal to aerobic values Influent PO4-P (mgP/L) 1 23 15.1 11.7 18.0 1.2 8% equal to aerobic values 2 13 13.5 12.1 14.7 0.7 6% equal to aerobic values Effluent NH3-N (mgN/L) 1 22 4.0 0.8 24.5 5.5 138% 23 37.4 27.1 44.2 4.2 11% 2 17 4.3 0.6 13.2 3.5 82% 17 30.1 22.7 33.5 2.7 9% 2a 9 1.6 0.6 3.1 0.7 46% 9 31.3 28.8 33.5 1.4 5% 2b 8 7.4 4.7 13.2 2.8 37% 8 28.9 22.7 32.8 3.3 11% Effluent NOx-N (mgN/L) 1 22 19.6 0.6 28.4 7.1 36% 23 0.0 0.0 0.0 0.0 0.0 2 16 8.9 0.2 19.8 7.6 86% 16 0.0 0.0 0.0 0.0 0.0 2a 9 14.9 10.4 19.8 3.6 24% 9 0.0 0.0 0.0 0.0 0.0 2b 7 1.1 0.2 3.6 1.2 112% 7 0.0 0.0 0.0 0.0 0.0 Effluent PO4-P (mgP/L) 1 22 13.8 9.9 16.5 1.3 10% 23 14.4 10.7 15.7 0.9 6% 2 13 12.3 11.0 13.0 0.6 4% 13 12.5 11.9 13.2 0.4 3% Aerobic System Microaerophilic SystemOperational Parameter Run    68 Temperature was largely consistent near 35 oC during the 5 and 20 d SRT experimental runs for both the aerobic and microaerophilic systems.  There was, however, an upset in temperature in both systems beginning near day 55 of operation and lasting nearly three weeks, as shown in Figure 5.2. The upset was determined to be caused by a block in flow of hot water to the reactor water jackets. Flow proportioning between the two systems was difficult due to pressure differences resulting from an improper seal on the inner walls of the microaerophilic reactor. A ball value was used to control the flow between the two systems to maintain a constant temperature in each.  It was discovered at day 75 that the ball valve was broken and was blocking flow out of the water bath. Temperature stabilized after the valve was removed and was maintained near 35 oC for the remainder of the study, with the microaerophilic reactor consistently 1 oC cooler than the aerobic reactor.  The 1 oC difference in temperature was not intentional, but the impact to the treatment trains was considered negligible.  Even with the upset, variation in temperature was low, with a maximum CV of 8%, as shown in Table 5.1.  0 5 10 15 20 25 30 35 40 45 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Day Te m pe ra tu re  (o C)  Aerobic Microaerophilic RUN 2: 5 d SRT RUN 1: 20 d SRT SET UP: 11 d SRT  Figure 5.2:  Reactor temperature.  Reactor pH was maintained continuously at 6.7 by pH controllers.  Several small pH upsets occurred during system operation due to excessive dosing, when probes were accidentally removed from the reactors without first powering down the controllers.  These brief upsets were quickly  69 neutralized using dilute sodium hydroxide.  There were no extended periods of abnormal pH during the study.  5.1.2 HRT Hydraulic retention time remained relatively constant during each experimental run with average values within the target range of 11 to 12 hr for microaerophilic and aerobic systems during both SRT periods, as shown in Table 5.1 and Figure 5.3.  Variation in HRT values was low, with a range in CV of 4 to 7 % throughout the study.  0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Day HR T (h r) Aerobic Microaerophilic RUN 2: 5 d SRT RUN 1: 20 d SRT  Figure 5.3:  Hydraulic retention time.  5.1.3 SRT System SRT values were maintained near the targets of 20 and 5 d for both treatment trains during the respective experimental runs.  It was, however, difficult to reach the target SRT under microaerophilic conditions during the 20 d SRT period, as shown in Figure 5.4 and Table 5.1. Under this set of operating conditions, SRT averaged 15.9 d despite best attempts to reach 20 d.  All consideration of the 20 d SRT experimental run from here on should consider this as the target value, rather than the actual value obtained.  It was difficult to reach a 20 d SRT in the  70 microaerophilic system because the biomass exhibited poor settleability in the microaerophilic system; this resulted in excessive solids losses in final effluent, evidenced by a constant build up of solids in the final effluent collection basin.  This became an operational issue as biomass losses threatened washout of the system.   To solve the problem, a daily solids recovery program was implemented consisting of settling and decanting final effluents and returning settled solids to the reactors.  Solids recovery efforts are described in more detail in Section 4.2.3.  Despite the challenges in maintaining a 20 d SRT in the microaerophilic system, there was a consistent and well defined difference in SRT between the first and second experimental runs, in both the aerobic and microaerophilic systems, as shown in Figure 5.4.  The SRT requirements of the experimental design were, therefore, met.  In addition, variation in SRT was low among systems and experimental runs, even with the swings observed in the microaerophilic system. Data listed in Table 5.1 show that SRT exhibited a maximum CV of 23 % in the microaerophilic system during the 20 d SRT experimental run, and CVs as low as 1 and 2 % during the 5 d SRT experimental run, for the microaerophilic and aerobic systems, respectively.  0 5 10 15 20 25 30 35 40 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Day SR T (d ) Aerobic Microaerophilic RUN 2: 5 d SRT RUN 1: 20 d SRT SET UP: 11 d SRT  Figure 5.4:  Solids retention time.  Biweekly measurements are plotted using a single exponential smoothing model (National Institute of Standards and Technology & SEMATECH 2006) to represent the slowly-changing nature of SRT.  71 5.1.4 MLSS and TSS MLSS was consistently lower in the microaerophilic reactor than in the aerobic reactor, as shown in Table 5.1 and Figure 5.5, with average concentrations near 2800 and 5800 mg/L, respectively, during the 20 d SRT run and 1400 and 2400 mg/L, respectively, during the 5 d SRT run.  This observation is indicative of reduced substrate removal and perhaps a lower microbial growth yield under the oxygen-limited conditions of the microaerophilic system. Figure 5.5 shows that MLSS levels decreased relatively continuously throughout operation of the microaerophilic system, whereas in the aerobic system, MLSS remained high during the 20 d SRT  run and then decreased markedly after the implementation of the 5 d SRT.  In all cases, MLSS levels were moderately variable, with CVs ranging from 21 to 45 % among treatments, as shown in Table 5.1.   0 2000 4000 6000 8000 10000 12000 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Day M LS S (m g/ L) Aerobic Microaerophilic RUN 2: 5 d SRT RUN 1: 20 d SRT SET UP: 11 d SRT  Figure 5.5:  Mixed liquor suspended solids.  Influent TSS values displayed in Table 5.1 and Figure 5.6 are averaged over each week of BBWTS operation, to better represent the quantity of suspended solids entering the system.  This was required as suspended solids levels in the feed basin were variable depending on the time of the day and the day of the week that the sample was collected.  The feed basin was filled daily and suspended solids that settled to the bottom of the basin were collected at the end of each week and  72 added to the reactors.  Therefore, if a grab sample was collected late in the week and/or late in the day, samples could contain disproportionate amounts of solids compared to what was entering the BBWTS over the whole week.  Therefore, weekly averaged concentrations were deemed more representative of influent TSS values than the discreet influent grab sample concentrations.  Even with averaging, influent TSS data displayed high variability during the 20 d SRT experiment run, with a CV of 82 %, as shown in Table 5.1.  Influent TSS variation was lower during the 5 d SRT run, with a CV of 33 %.  The changes in influent TSS were directly related to the feed stock collected from the mill.  Feed was collected in batches lasting for approximately four to six weeks each, as described in Chapter 4.  Changes in feed batch are evident in Figure 5.6, where drastic changes in influent TSS are demonstrated near days 30, 79, and 110 of BBWTS operation.  These days coincide with a change in primary effluent collected from the mill.  Additionally, there was a final change in feed batch around day 135 of BBWTS operation which did not result in a noticeable change in influent TSS.  Overall, the variations in influent TSS do not seem to have affected suspended solids levels in the effluents or mixed liquors, as seen in Figures 5.5 and 5.6, or effluent COD concentrations, as discussed below.  0 200 400 600 800 1000 1200 1400 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Day TS S (m g/ L) Aerobic Effluent Microaerophilic Effluent Influent RUN 2: 5 d SRT RUN 1: 20 d SRT SET UP: 11 d SRT  Figure 5.6:  Influent and effluent suspended solids.  Influent concentrations are averaged by week and effluent concentrations are from discreet samples obtained twice per week.  Effluent TSS measurements were taken following settling and transfer of heavier solids back to the reactors.  73 Microaerophilic effluent displayed higher TSS values than aerobic effluent, as shown Figure 5.6.  It is believed this was due to poor settleability of the microaerophilic biomass.  This observation was further evidenced by low filtration rates for TSS measurements of microaerophilic process streams. It is not surprising that effluent TSS was high in the microaerophilic system, as high TSS discharges are a common problem in ASBs (Mahmood & Paice 2006), suggesting low DO concentrations are conducive to poorly settling sludge.  When the SRT was decreased from 20 to 5 days, the average effluent TSS decreased from 95 to 75 mg/L in the microaerophilic system and from 60 to 45 mg/L in the aerobic system, as shown in Table 5.1.  Qualitative observations also noted increases in filtration rates and decreases of effluent solids losses during the 5 d SRT experimental run.  These differences indicate that biomass settleability was higher in the aerobic system than the microaerophilic system, and that the 5 d SRT was more conducive to biomass settleability compared to the 20 d SRT, over the range of oxygen conditions.  As shown in Figure 5.6, effluent TSS values were relatively constant after day 17 of BBWTS operation in the microaerophilic system, and beyond day 50 in the aerobic system.  The greater consistency in measured effluent TSS values coincided with the time at which suspended solids recovery from final effluent was initiated.  This change was implemented in the set up phase for the microaerophilic system, but not until the beginning of the first experimental run for the aerobic effluent.  The timing of effluent TSS stabilization is reflected in CV values shown in Table 5.1. Here, CV for effluent TSS was 139 % in the aerobic system during the 20 d SRT, and ranged from 29 to 53 % for all other treatments.  Another TSS issue was identified in the maintenance of SRT.  It was decided that total system suspended solids inventories were more appropriate than MLSS for use in calculation of wasting volumes, and hence maintenance of SRT.  This is because the clarifiers were not operating ideally, as they were not heated with hot water jackets like the bioreactors.  Rather, the clarifiers were only insulated.  As such, warm MLSS entering the clarifiers would tend to rise, limiting the ability of the clarifiers to settle the biomass.  As clarification was not ideal, MLSS concentration in the reactors was not representative of the system; therefore, total system suspended solids concentrations were used for maintenance of SRT.  Total system suspended solids values were calculated by utilizing TSS measurements from the mixed liquors, the clarifiers and from the final effluents after settling and return of the settled solids to the bioreactors.  Total system suspended solids values are plotted in  74 Figure 5.7 and listed in Table 5.1.  These data show that biomass inventories for the microaerophilic system were about 50% of those for the aerobic system for the duration of the study, and also that the inventories in both systems were also reduced by about 50 % after the SRT was changed from 20 to 5 days.  0 5000 10000 15000 20000 25000 30000 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Day M as s (m g)  Aerobic Microaerophilic RUN 2: 5 d SRT RUN 1: 20 d SRT SET UP: 11 d SRT  Figure 5.7:  Total system solids.  5.1.5 COD COD values were consistently lower in the effluents than in the influent, proving that the BBWTS was successful in wastewater treatment.  COD removal patterns were generally similar for both total and dissolved COD (tCOD and dCOD), although influent and effluent tCOD levels were more erratic than the coinciding dCOD levels, likely due to changes in influent TSS discussed above. Results for both tCOD and dCOD are summarized in Table 5.1 and dCOD levels are plotted in Figure 5.8.  Dissolved COD was chosen for discussion purposes due to the more consistent nature of this parameter compared to tCOD.  As shown in Figure 5.8, effluent COD was also consistently lower in the aerobic system than in the microaerophilic system, indicating greater COD removal under fully aerobic conditions.  Average influent and effluent COD values were also higher for both systems during the 5 d SRT run than the 20 d SRT run, which is assumed to be due to a change in feed characteristics.  This is shown in Figure 5.8 as a sudden and sustained increase in influent and  75 effluent COD around day 135 of system operation.  This change coincided with a change in feed batch collected from the mill.  As a result, influent dCOD was approximately 300 mg/L higher over the 5 d SRT period on average, compared to the 20 d SRT period, and tCOD was approximately 600 mg/L higher, on average.  Despite these changes, variation was satisfactory for influent and effluent COD concentrations, with CV values ranging from 7 to 24 % over the entire experimental program, as shown in Table 5.1.  Removal efficiencies for total and dissolved COD were consistently higher in the aerobic system and during the 5 d SRT.  As shown in Table 5.1, average tCOD removal efficiencies were near 65 % for the aerobic system and near 40 % for the microaerophilic system during the 5 d SRT experimental run.  Removal efficiencies were lower for dCOD over the same period, with values of 58 and 36 % for the aerobic and microaerophilic systems, respectively.  COD removal generally followed a similar trend between the two treatment trains, with about 20 % more total and dissolved COD removal in the aerobic system than in the microaerophilic system, throughout the study.  COD removal efficiencies do not appear to have been affected by the increase in the influent and effluent COD concentration during the 5 d SRT period..   0 500 1000 1500 2000 2500 3000 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Day CO D (m g/ L) Aerobic Effluent Microaerophilic Effluent Influent RUN 2: 5 d SRT RUN 1: 20 d SRT SET UP: 11 d SRT  Figure 5.8:  Dissolved chemical oxygen demand.  76 5.1.6 Nutrients Nitrogen and phosphorus were both abundantly present in final effluents, as shown in Figures 5.9 to 5.11, indicating that nutrient concentrations should not have been limiting biological activity in the BBWTS.  An evaluation of NH3-N and NOx-N concentrations shows that nitrification occurred in the aerobic system but not in the microaerophilic system, which was expected given that nitrification is an aerobic process.  As shown in Figure 5.10, NOx was present in aerobic effluent, with an average concentration of 20 mg N / L during the 20 d SRT run.  There was one period of reduced nitrification activity during this run, which can be seen in Figure 5.10 near day 70.  This incident coincided with a temperature upset (discussed in Section 5.1.1 and illustrated in Figure 5.2) and nitrification recovered rapidly upon temperature stabilization.  This nitrification upset can therefore be considered to be a discreet event that did not significantly impact the integrity of treatment during the aerobic 20 d SRT run.  During the 5 d SRT run, nitrification in the aerobic system decreased around day 135, coinciding with a change in feed stock and the associated increase in COD levels discussed above and illustrated in Figure 5.8.  Table 5.1 shows that for the first half of the 5 d SRT run, aerobic effluent NOx levels averaged 15 mg N / L, whereas after the feedstock was changed near day 135, these values averaged only 1 mg N / L, and effluent NH3-N levels rose accordingly.  The loss of nitrification coinciding with the change in feed batch suggests something may have been introduced to the system that was toxic to the nitrifying bacteria.  This is supported by the abruptness in the change in N levels at this time.  It is also possible that the 5 d SRT may not have been sufficiently long to sustain nitrifying bacteria; however, research has shown that at the mesophilic temperatures used in this study, nitrifiers can be maintained at SRTs as low as 4 d (US Environmental Protection Agency 1975).  Additionally, if the nitrifying bacteria were being washed out of the system, it is more likely that the observed changes in effluent N levels would be gradual rather than abrupt.  Thus, it appears most likely that the loss of nitrification during the 5 d SRT was due to changes in the characteristics of the influent associated with the change in feed batch around day 135.  Comparatively, Figure 5.11 shows NOx was never present in the microaerophilic effluent and that NH3 levels remained high throughout the experimental program, indicating an absence of nitrification under microaerophilic conditions.  Variation in effluent NH3-N levels was low in this system, with a maximum CV of 11 %, as shown in Table 5.1.  In comparison, the variation  77 associated with NH3-N and NOx-N concentrations in the aerobic system was high, reflecting the changes in nitrification, with maximum CV values of 138 % and 112 % for NH3-N and NOx-N, respectively.  0 2 4 6 8 10 12 14 16 18 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Day C on ce nt ra tio n (m g P / L ) PO4 residual - aerobic effluent PO4 residual - microaerophilic effluent SET UP: 11 day SRT RUN 1: 20 day SRT target RUN 2: 5 day SRT target  Figure 5.9:  Effluent phosphorus levels.  78  0 5 10 15 20 25 30 35 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Day Co nc en tra tio n (m g N / L ) NOx-N  - Aerobic Effluent NH3-N - Aerobic Effluent RUN 2: 5 d SRT RUN 1: 20 d SRT SET UP: 11 d SRT NOx-N NH3-N  Figure 5.10:  Aerobic effluent nitrogen levels.  0 5 10 15 20 25 30 35 40 45 50 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Day Co nc en tra tio n (m g N / L ) NOx-N - Microaerophilic Effluent NH3-N - Microaerophilic Effluent NOx-N NH3-N RUN 2: 5 d SRT RUN 1: 20 d SRT SET UP: 11 d SRT  Figure 5.11:  Microaerophilic effluent nitrogen levels.  79 5.1.7 BBWTS Summary Operation of the BBWTS was successful, as biological activity was consistently evidenced through COD removal in both treatment trains throughout the duration of the experimental program. Nitrification in the aerobic system provided further evidence of biological activity.  Some level of biological disruption was apparent in the aerobic system during the 5 d SRT experimental run, as indicated by variations in nitrogen levels and associated changes in nitrification.  The loss of nitrification in this system, during the 5 d SRT run, was significant as it was sustained throughout the remainder of the experimental run.  However, there were no associated changes in COD removal during this period.  The requirements of the experimental design were met as there were consistent and substantial differences among the two varied operational parameters, DO and SRT, between treatments. However, the target SRT of 20 d could not be met in the microaerophilic system, as a result of poor settleability of the microaerophilic sludge, coupled with low MLSS concentrations in the microaerophilic reactor.  The settleability problems appeared to be a result of highly dispersed microbial growth under oxygen-limited conditions, and as such, an average SRT of 16 d was all that was obtainable in this system, even with the implementation of an intensive solids management program.  Settleability improved markedly in both the microaerophilic and aerobic systems with the implementation of a 5 d SRT, as evidenced by decreases in effluent TSS levels.  Apart from the crash in nitrification and problems with settleability in the microaerophilic system, the BBWTS generally operated well, with minor variation in operational parameters between and within the treatment trains and experimental runs, except where planned based on experimental design.  A few exceptions were noted for which variation was high, most notably in the aerobic effluent nitrogen levels and effluent TSS, as would be expected based on the above discussion.  In addition to the poor settleability of the microaerophilic sludge, variation in effluent TSS was likely a result of poorly operating clarifiers for both the aerobic and microaerophilic systems.  It is anticipated that a higher degree of settling would have occurred in the clarifiers if they were heated to the same temperatures as the reactors; identical temperatures in the reactors and clarifiers would surely reduce rising sludge and therefore enhace settling.  Variation was also high for influent TSS levels compared to other operational parameters.  This was a result of changes in feed batch collected from the mill, and these changes did not detract from the overall stability of the system.  80 5.2 QUANTIFICATION OF PLANT STEROLS AND POTENTIAL BIOTRANSFORMATION PRODUCTS This section describes the levels of phytosterols measured in the various process streams of the BBWTS, in terms of their concentrations, removal efficiencies and likely biotransformation products.  Summary statistics are provided throughout this section, and data for individual samples are listed in Appendix C.  QA/QC results are summarized and discussed in Appendix D.  5.2.1 Compatibility of AD, ADD, T and Pr with the Environment Canada Sterols Method Spike analysis was conducted using androstenedione (AD), androstadienedione (ADD), testosterone (T) and progesterone (Pr) to test the compatibility of these compounds with the Environment Canada Sterols Method (Environment Canada 2005b).  Spikes were prepared using different combinations of sample preparation techniques to evaluate the impact of the method components on analyte recoveries.  Results of these experiments are listed in Table 5.2.  These data show that all four of the additional compounds were detectable using the Environment Canada Sterols Method, with compound recoveries in the range of 43 to 201 % when spikes were prepared following all method steps, including extraction and derivitization.  A minimum recovery of 23 % was associated with the AD spike at 2 μg/mL, and a maximum recovery of 201 % was associated with the T spike at 0.05 μg/mL.  Percent recoveries were comparable for spikes that were derivitized, whether or not they were extracted.  Recoveries were also comparable for spikes that were not extracted nor derivitized, except in the case of T, which was not detected when not derivitized.  This is logical, given that it is the only compound of the four that contains hydroxyl groups, which are acetylated during derivitization.   81 Table 5.2: Spike analysis for AD, ADD, T and Pr.  Concentrations are in μg/mL.  “n.d.” indicates not detected.  Compound Analytical Method Extracted, Derivitized Not Extracted, Derivitized Not Extracted, Not Derivitized Conc. Recovery Conc. Recovery Conc. Recovery Spike 1 (0.05 μg/mL) AD 0.084 167% 0.090 180% 0.084 168% ADD 0.090 179% 0.095 190% 0.081 163% T 0.100 201% 0.104 208% n.d. 0 Pr 0.090 180% 0.093 185% 0.072 144% Spike 2 (2 μg/mL) AD 0.865 43% 1.35 68% 1.72 86% ADD 1.29 65% 1.60 80% 1.72 86% T 1.48 74% 1.71 86% n.d. 0 Pr 1.50 75% 1.69 84% 1.56 78%  These results show that the additional compounds can be detected and tentatively quantified using the Environment Canada Sterols Method at levels that served the purposes of the study.  However, future method development requires determination of LOQs and analysis of matrix effects. Without determination of LOQs, low concentrations are associated with high uncertainty.  For reference, other parameters tested in this method have LOQs from 0.2 to 0.005 μg/L (Table 4.3).  5.2.2 Influent Characteristics Influent to the BBWTS was sampled from the feed basin as described in Chapter 4. Five influent samples were analyzed for sterols and potential biotransformation products during the 20 d SRT experimental run, and 8 were analyzed during the 5 d SRT experimental run.  These samples were analyzed for a total of 19 sterols and potential biotransformation products, including all 18 parameters in the Environment Canada Sterols Method, as well as progesterone.  Out of these 19 analytes, six were detected in influent samples, all of which were phytosterols.  No animal hormones were detected in these samples, although it should be noted that they were not tested for AD, ADD or T.   The parameters identified, in order of abundance, were β-sitosterol (β-sito), campesterol (campe), stigmasterol (stigma), cholesterol (chole), coprostanol (copro) and dihydrocholesterol (DHC), also known as cholestanol.  Influent chemistry results are summarized in Table 5.3 and listed in Appendix C.   It should be noted that since the same influent was pumped to both treatment trains, influent characteristics were identical for the aerobic and microaerophilic systems.  82 Table 5.3: Summary statistics for influent sterols.  All concentrations are in μg/L; “n.d.” indicates not detected; (1) - MDLs varied from 0.005 ug/L to the value shown, based on final sample dilution for GC-MS analysis; (2) Sum of averages calculated as the total of average concentrations for all sterols detected during the experimental run.  n x min max σ CV n x min max σ CV Campesterol 0.1 (1) 5 156 107 249 58.0 37% 8 200 104 320 70.6 35% Cholesterol 0.05 (1) 5 14.2 0.881 67.3 29.7 209% 8 2.23 0.69 3.66 0.991 44% Coprostanol 0.05 (1) 5 0.138 n.d. 0.69 0.31 224% 8 0.944 n.d. 1.61 0.634 67% DHC 0.05 (1) 5 1.09 n.d. 5.44 2.43 224% 8 0.357 n.d. 1.93 0.714 200% β-Sitosterol 0.2 (1) 5 659 454 1010 228 35% 8 1081 508 1557 361.0 33% Stigmasterol 0.005 5 5.11 3.13 10.1 2.86 56% 8 36.9 17.9 67.8 18.8 51% 836 1322Sum of averages (2) 20 day SRT (run 1) 5 day SRT (run 2) Parameter MDL   According to the data in Table 5.3, measured influent concentrations of individual analytes ranged from 1010 μg/L of β-sito to non-detectable for copro and DHC during the 20 d SRT run.  During the 5 d SRT run, measured influent concentrations ranged from 1557 μg/L of β-sito to non- detectable for copro and DHC.  Calculated total average sterol concentrations were lower during the 20 d SRT experimental run, at 836 μg/L, compared to 1322 μg/L during the 5 d SRT experimental run.  Variation in concentrations of individual analytes was comparable in both runs, with a range in CV of 37 to 224 % during the 20 d SRT and 33 to 200 % during the 5 d SRT.  In both cases, high CVs were associated with analytes of very low concentrations; for example, a CV of 224 % was associated with a range in copro from non-detectable to 0.690 μg/L.  It should be noted than non- detectable values were replaced with zeros for the purpose of CV calculations.  Using this method, variation is assumed to be biased high.  Mean concentrations of the detected influent analytes are illustrated in Figure 5.12.  The figure shows that the measured influent concentrations for individual analytes were of similar orders of magnitude in both experimental runs.  The only exceptions in this regard were chole and stigma, as average measured chole concentrations were 14.2 and 2.23 μg/L during the 20 and 5 d SRT experimental runs, respectively; average measured stigma concentrations were 5.11 and 36.9 μg/L during the 20 and 5 d SRT experimental runs, respectively.  As demonstrated in Figure 5/12, β-sito was the most abundant sterol present in influent, with average concentrations near 1000 μg/L in both runs.  Campe was the next most abundant, followed by stigma.  Concentrations of chole,  83 copro and DHC were notably lower than those of the three most abundant sterols over both experimental runs, with values near or below 1 μg/L.  0.1 1 10 100 1000 10000 1 2 C on ce nt ra tio n (μ g/ L) β-Sito Stigma DHC Copro Chole Campe 20 d SRT (run 1) 5 d SRT (run 2)  Figure 5.12:  Average influent sterols concentrations.  For run 1, n = 5; for run 2, n=8.  Differences in measured influent sterols concentrations were further analyzed to test for statistical significance.  The Wilcoxon Rank Sum test was used to calculate p values, which are summarized in Table 5.4.  These data show that there were significant differences in measured influent levels of copro and stigma in the two experimental runs, and that measured differences between the other influent sterols were not significant.  It is important to note that copro and stigma made up a small fraction of the total measured influent sterols concentrations, with a combined contribution of 0.62 % of the total influent sterols load during the 20 d SRT and 2.9 % during the 5 d SRT, on average. There was no statistical difference between measured concentrations of influent β-sito, which made up the majority of sterols in all influent samples.   84 Table 5.4: Statistical comparison of sterols inputs between experimental runs.  P values determined by the Wilcoxon Rank Sum Test.  Bold indicates statistical significance (p ≤ 0.05).  Parameter p value Run 1 (n = 5) vs Run 2 (n = 8) Run 2a (n = 5) vs Run 2b (n = 3) Campesterol 0.354 1.000 Cholesterol 0.418 0.036 Coprostanol 0.038 0.283 Dihydrocholesterol 0.546 0.329 β-Sitosterol 0.065 1.000 Stigmasterol 0.002 0.571  The measured influent sterols concentrations were also compared over the first and second half of the 5 d SRT experimental run, in an attempt to identify changes associated with the increase in COD and the loss of nitrification that occurred as a new batch of influent wastewater was fed to the system near day 135 of BBWTS operation (as described throughout Section 5.1).  The influent sterols concentrations measured before and after the new batch of primary effluent was received are plotted in Figure 5.13 and a statistical comparison is provided in Table 5.4.  As can be seen in the figure, sterols levels were virtually identical in both halves of the second experimental run.  Copro was present in the first half of the run but was not detected in the second half, but this difference was not statistically significant, given that measured concentrations of this analyte above detection limits were so low.  The only significant difference in measured influent sterols concentrations between the first and second half of the 5 d SRT run was in influent chole concentrations.  Overall, there were enough similarities to conclude that sterols inputs were not significantly affected by the change in wastewater feed applied during the 5 d SRT experimental period.   85 0 1 10 100 1000 10000 1 2 C on ce nt ra tio n (μ g/ L) β-Sito Stigma DHC Copro Chole Campe 5 d SRT (run 2a) 5 d SRT (run 2b)  Figure 5.13:  Average influent sterols concentrations during the 5 d SRT.  For run 2a, n = 5; for run 2b, n=3.   5.2.3 Effluent Characteristics Final treated effluent from the aerobic and microaerophilic systems was sampled as described in Section 4.3.3. From each treatment train, five effluent samples were analyzed for sterols and potential biotransformation products during the 20 d SRT experimental run, and eight were analyzed during the 5 d SRT experimental run.  These samples were collected to correspond with influent samples, to allow for calculation of sterols removal efficiencies and net biotransformation product generation.  All effluent samples were analyzed for the 18 parameters in the Environment Canada Sterols Method and progesterone (Pr), and three samples from each treated effluent were also analyzed for androstenedione (AD), androstadienedione (ADD) and testosterone (T).  Out of the 23 compounds for which effluent samples were analyzed, eight were detected.  Results of this analysis are summarized in Table 5.5 and listed in Appendix C.  Seven of the compounds  86 detected in aerobic and microaerophilic effluents were phytosterols, and the eighth, estrone, is an animal steroid.  AD, ADD, T and Pr were not detected in any of the effluent samples.  Table 5.5: Summary statistics for effluent sterols and biotransformation products.  All concentrations are in μg/L; “n.d.” indicates not detected; (1) - MDLs varied from 0.005 μg/L to the value shown, based on final sample dilution for GC-MS analysis; (2) sum of averages was calculated as the total of average concentrations for all sterols and metabolites detected in each effluent. Parameter MDL n x min max σ CV n x min max σ CV Campesterol 0.005 5 1.71 0.833 4.46 1.54 90% 8 6.75 1.88 15.1 5.62 83% Cholesterol 0.005 5 0.585 0.350 1.26 0.386 66% 8 0.727 0.422 1.38 0.319 44% DHC 0.05 (1) 5 0.375 0.286 0.431 0.056 15% 8 0.060 n.d. 0.241 0.111 185% Estrone 0.05 (1) 5 0.067 n.d. 0.334 0.149 224% 8 0.133 n.d. 0.623 0.250 189% β-Sitosterol 0.005 5 9.00 4.19 23.9 8.35 93% 8 41.6 12.2 85.3 30.3 73% Stigmasterol 0.005 5 0.301 0.194 0.656 0.199 66% 8 1.85 0.454 3.89 1.62 88% 12.0 51.1 Parameter MDL n x min max σ CV n x min max σ CV Campesterol 0.1 (1) 5 93.0 65.1 114 21.2 23% 8 74.8 52.0 102 18.2 24% Cholesterol 0.005 5 2.09 0.919 5.88 2.12 102% 8 0.989 0.540 1.53 0.321 32% Coprostanol 0.05 (1) 5 n.d. n.d. n.d. n.d. n.d. 8 0.037 n.d. 0.298 0.105 283% Desmosterol 0.05 (1) 5 n.d. n.d. n.d. n.d. n.d. 8 0.075 n.d. 0.602 0.213 283% DHC 0.05 (1) 5 0.526 0.336 1.08 0.314 60% 8 0.163 n.d. 0.496 0.227 139% β-Sitosterol 0.1 (1) 5 430 305 549 99.0 23% 8 385 300 489 70.7 18% Stigmasterol 0.005 5 3.01 2.12 3.72 0.658 22% 8 13.2 8.57 22.3 5.05 38% 528 474 20 day SRT (run 1) 5 day SRT (run 2) Aerobic Effluent Microaerophilic Effluent Sum of averages (2) Sum of averages (2)   As indicated in Table 5.5, the phytosterols and their biotransformation products identified in final effluent samples were β-sito, campe, stigma, chole, copro, DHC, desmosterol (desmo), and estrone (E1).  Copro and desmo were only detected in microaerophilic effluent during the 5 d SRT period, and E1 was only detected in aerobic effluent, but during both SRT runs.  Calculated total sterol concentrations were lowest in effluent samples collected from the aerobic system, with values of 12.0 μg/L and 51.1 μg/L over the 20 d SRT and 5 d SRT runs, respectively.  Comparatively, total sterol concentrations calculated from samples collected from the microaerophilic effluent were 528 μg/L and 474 μg/L over the 20 d SRT and 5 d SRT runs, respectively.  87 Measured concentrations of individual analytes in aerobic effluent ranged from 85.3 μg/L for β-sito (during the 5 d SRT run) to non-detectable for a variety of compounds during both runs. Comparatively, measured concentrations of individual analytes in microaerophilic effluent ranged from 549 μg/L for β-sito (during the 5 d SRT) to non-detectable for a variety of compounds during both runs.  Variation in concentrations of individual analytes was comparable in all four effluents, as shown in Table 5.5. These data indicate a range in CV of 15 to 224 % in the aerobic system over both experimental runs, and a range in CV of 18 to 283 % in the microaerophilic system, over the same period.  As with the influent, the highest CVs were associated with analytes of the lowest concentrations.  Effluent data are compared to influent data in the sections below.  5.2.4 Sterols Removal Efficiencies Removal efficiencies were calculated for all six phytosterols detected in the influent.  A summary of these values is provided in Table 5.6.  Table 5.6: Average removal efficiencies for organic parameters across treatments.  Asterisk (*) indicates parameters not included in average and CV calculations, as these compounds were inconsistently removed and generated (i.e. at times their concentration was higher in the effluent than the influent).  Parameter Treatment Conditions Aerobic 20 d SRT (n= 5) Aerobic 5 d SRT (n= 5) Microaerophilic 20 d SRT (n= 8) Microaerophilic 5 d SRT (n= 8) Campesterol 99% 96% 37% 60% Cholesterol * 64% * 51% * 2% * 50% * Coprostanol 100% 100% 100% 95% Dihydrocholesterol * 93% * 100% * 80% * 77% * β-Sitosterol 99% 95% 31% 61% Stigmasterol 94% 94% 33% 60% Average 98% 96% 50% 69% CV 3% 3% 67% 25%  These data show that a high degree of sterols removal occurred during treatment, with close to 100 % removal observed for the three major sterols and coprostanol in the aerobic system, at both SRTs.  Removal efficiencies were much lower for the microaerophilic system, with an average removal efficiency of β-sito, campe, stigma and coprostanol combined of 50 % during the 20 d SRT  88 run and 69 % during the 5 d SRT run.  Average removal efficiencies for chole appeared low because this compound was occasionally detected at higher concentrations in the final treated effluent than in the influent.  Hence, the change in chole concentrations through treatment indicated net removal at some times, and at other times indicated net generation. DHC exhibited a similar tendency; however, removal efficiencies of this compound appeared substantially higher than those of chole. This can be explained by the very low concentrations of DHC that were observed in influent samples, as shown in Table 5.3.  Hence, 100 % removal of DHC was obtainable by removal of very low concentrations.  Removal efficiencies for the three most abundant sterols, β-sito, campe and stigma, are plotted in Figure 5.14.  This figure shows a clear pattern of sterols removal throughout the experimental program, with the largest removal efficiencies associated with the aerobic system over both SRTs, followed by the microaerophilic system at a 5 day SRT, and finally, the microaerophilic system at a 20 d SRT.  It is logical that the aerobic system exhibited the highest removal efficiencies for phytosterols, in parallel with the relatively high level of biological activity observed in this system through COD and nitrification measurements.  However, in the case of the microaerophilic system, it is somewhat counterintuitive that the long SRT resulted in lower sterol removal efficiencies than did the short SRT.  This may be due to the poor settleability and higher TSS of the microaerophilic effluent at the 20 d SRT, compared to that at the 5 d SRT (as described in Section 5.1.4).  It has previously been shown that the majority of sterols discharged in full scale treatment systems are attached to the solid phase (Kostamo et al. 2004).  Following this logic, higher effluent TSS concentrations in the microaerophilic system during the 20 d SRT run could have contributed to the decreased sterol removal efficiencies also observed.  It is interesting to note that in the aerobic system, sterols removal efficiencies were virtually identical at the two experimental SRTs, with only slightly higher removal during the 20 d SRT than at the 5 d SRT.  Also, as shown in Table 5.6, variation in removal efficiency was low for the aerobic system. Together, these observations demonstrate the robustness of the aerobic treatment system. Comparatively, the microaerophilic system operating under a 20 d SRT displayed the highest variation in removal efficiencies, with a CV of 67 % for all analytes, further reflecting the lower biological activity of this treatment system.   89  0% 20% 40% 60% 80% 100% β-Sitosterol Campesterol Stigmasterol aerobic system microaerophilic system 5 d SRT n = 8 20 d SRT n = 5 5 d SRT n = 8 20 d SRT n = 5   Figure 5.14:  Average removal efficiencies for β-sito, campe, and stigma.  The analysis of phytosterols removal efficiencies generally indicates that through the BBWTS, a large proportion of the influent phytosterols was removed, equivalent to about 1 mg/L. Theoretically, at least a portion of this material was microbially degraded and, therefore, potentially available for generation of sterols biotransformation products.  5.2.5 Sterols Biotransformation Evidence of sterols biotransformation can be gleaned from the influent and effluent data, as effluent concentrations were higher for some analytes compared to corresponding influent concentrations; further, some of the compounds detected in effluent samples were never detected in influent samples at all.  Net concentration changes indicated that four organic compounds were produced during wastewater treatment, comprising cholesterol (chole), dihyrocholesterol (DHC), desmosterol (desmo), and estrone (E1).  The net quantities of compounds generated in each treatment system are compared in Figure 5.15.   This figure shows that E1 was generated at both SRTs but only under aerobic conditions.  DHC, on the other hand, was generated under all DO and SRT conditions tested.  Chole and desmo production displayed no obvious trends across treatments, with chole  90 generated under opposing experimental conditions, and desmo generated under microaerophilic conditions at a 5 d SRT only.  Aerobic system Microaerophilic system 5 d SRT n = 8 20 d SRT n = 5 5 d SRT n = 8 20 d SRT n = 5 0 .0 0 .1 0 .2 0 .3 0 .4 0 .5 0 .6 0 .7 0 .8 Cholesterol Dihydro- cholesterol Desmosterol Estrone N et  c on ce nt ra tio n ch an ge  (μ g/ L) 14.2 2.23 0.357 0.357 1.09 1.09 < 0.05 < 0.05 < 0.05 N et  c on ce nt ra tio n ch an ge  (μ g/ L)  Figure 5.15:  Average biotransformation product generation across treatments.  Net concentration change was calculated as effluent concentration less influent concentration.  Values above bars indicate the average concentration in the influent for the parameter and experimental run indicated (in μg/L).  As illustrated in Figure 5.15, phytosterol biotransformation products were identified in effluents produced using each set of wastewater treatment conditions applied in this study.  The sums of net metabolite concentrations for each treatment are provided in Table 5.7.  These data indicate that the largest net concentration of sterols biotransformation products was generated in the aerobic system during the 5 d SRT experimental period, with a total of 1.47 μg/L generated on average. Comparatively, the smallest net concentration of metabolites was generated in the microaerophilic system during the 20 d SRT experimental period, with a total of 0.580 μg/L generated on average. In addition, there was approximately twice as much biotransformation product formation during the 5 d SRT than the 20 d SRT, for both DO conditions.  These findings suggest sterols  91 transformations occur most readily under aerobic conditions and at shorter SRTs.  Each of the four sterols biotransformation products measured in this study are discussed in detail below.   Table 5.7:  Net metabolite production across treatments.  All concentrations in μg/L; Concentrations for individual parameters calculated as effluent less influent concentrations, the averaged by treatment conditions; n.d. indicates not detected. Parameter Treatment Conditions Aerobic 20 d SRT (n=5) Aerobic 5 d SRT (n=8) Microaerophilic 20 d SRT (n=5) Microaerophilic 5 d SRT (n=8) Cholesterol n.d. 0.697 0.193 n.d. Desmosterol n.d. n.d.  n.d. 0.602 Dihydrocholesterol 0.369 0.241 0.387 0.444 Estrone 0.334 0.530 n.d. n.d. TOTAL 0.703 1.47 0.580 1.05   5.2.5.1 Cholesterol (chole) There was a net generation of cholesterol (chole) under aerobic conditions during the 5 d SRT run and under microaerophilic conditions during the 20 d SRT run.  These two treatments combined opposite DO concentrations and SRTs, and resulted in the highest and lowest phytosterols removal efficiencies, respectively.  This suggests that chole formation occurred independently of treatment conditions.  Although this phenomenon was not explained in the present study, it is interesting that a wide range of conditions can produce this compound, given its ubiquitousness in nature, as discussed in Section 2.2.1.  Chole is very close in structure to β-sito, as shown in Figure 5.16.  The only structural difference between the two compounds is the presence of two additional carbon atoms (and associated hydrogen atoms) in the aliphatic side chain of β-sito.  Both compounds are common starting products used in commercial steroid production (Fernandes et al. 2003), which is likely attributable to their structural similarities.  As shown in Figure 5.16, to convert β-sito to chole, side chain cleavage at carbon atom 24 is the only change required.   92 Cholesterolβ-Sitosterol Side chain cleavage at C24    Figure 5.16:  Simplification of the biotransformation of β-sito to chole.  As discussed in Chapter 2, chole is a component of all eukaryotic cells. Therefore, it is naturally present in both fish and wood.  In vertebrate steroidogenesis, chole is the starting substance for the generation of sex steroids (Moncecchi et al. 1991).  Also as described in Chapter 2, β-sito exposure can lead to decreased chole levels in fish ( Tremblay & Van der Kraak 1998; Tremblay & Van der Kraak 1999), reduced steroidogenesis relative to chole availability (Oakes et al. 2005), and reduced transfer of chole into the mitochondria, where steroidogenesis is initiated (Leusch & MacLatchy 2003).  Thus it can be seen that β-sito exposure impedes the function of chole in fish.  However, no studies were found in the literature that evaluate the effect of chole exposure on fish.  Despite this knowledge gap, it can be seen from the work of Monechecchi et al. (1991), Bird et al.  (2002), Leusch & MacLatchy (2003), and Oakes et al. (2005), that increased chole availability does not imply an increase is steroid production, even though chole is the natural starting product for this process. This is an important note when considering the steroid depressions caused by β-sito exposure, because it suggests that chole exposure will not counteract steroidogenesis impediments.  5.2.5.2 Dihyrocholesterol (DHC) DHC, also known as cholestanol, was generated under all wastewater treatment conditions applied in this study, with the highest concentrations produced under microaerophilic conditions and a 5 d SRT.  As can been seen in Figure 5.17, chole and DHC have identical aliphatic side chains, but DHC has the characteristic stanol structure, with fully saturated carbon rings.  In plants, DHC can be produced as a metabolite of cholesterol (Nakajima et al. 2002).  In order for this transformation to occur, the double bond between carbon atoms 5 and 6 requires hydrogenation, as illustrated in  93 Figure 5.17.  No information could be found on DHC function in fish, or on effects of DHC exposure to fish.  Cholesterol Dihydrocholesterol / Cholestanol Hydrogenation of the C5-6 double bond   Figure 5.17:  Simplification of the biotransformation of chole to DHC.  5.2.5.3 Desmosterol (desmo) Desmo was produced under microaerophilic conditions with a 5 d SRT only.  Unlike chole and DHC, desmo was never detected in the influent, suggesting that its presence in effluent was fully due to transformations during treatment.  Desmo is a common intermediate in the conversion of β-sito, campe, stigma and other 24-alkyl sterols to cholesterol (Svoboda 1997).  The transformation to desmo occurs through dealkylation in the aliphatic side chain, forming a double bond at carbon 24 and hydrogenation of this bond then produces cholesterol (Svoboda 1997).  There are multiple intermediaries along this pathway, but desmo is the common terminal intermediary (Svoboda 1997). A simplified version of this biotransformation pathway is illustrated in Figure 5.18.     94 β-Sitosterol Desmosterol Cholesterol Side chain cleavage at C24 Formation of C24-25 double bond Hydrogenation of C24-25 double bond   Figure 5.18:  Simplification of the biotransformation of β-sito to chole with desmo as an intermediary.  It is possible that sterols biotransformation was limited under the 5 d SRT in the microaerophilic system, such that conversion of β-sito to chole was not complete, but instead ended at the generation of desmo.  As desmo was not produced during any other treatment conditions, this specific limitation in biotransformation may have only existed in the microaerophilic system at the 5 d SRT.  This hypothesis is supported by the fact that chole was not detected in effluent treated under these conditions (as shown in Figure 5.15).  However, this is just one possible explanation of how desmo could have been generated, and it is recognized that there are likely other possibilities, given the range and complexities of sterols biotransformation pathways.  No information could be found on desmo exposure to fish.  However, in humans, a rare genetic disease exists called desmosterolosis, where desmo makes up an abnormally high percentage of total sterols in the brain, kidneys and liver (Rozman & Waterman 2002).  This disease is associated with miscarriage, stillborns, and abnormal development of brains and genitals (Rozman & Waterman 2002).  Thus, although no information could be found on desmo exposure, the ill effects of desmo in humans suggests a precautionary approach should be used for effluents containing desmo, at least until such a time as effects levels are established in fish.  5.2.5.4 Estrone (E1) Estrone was produced in the aerobic system at mean concentrations of 67 ng/L and 133 ng/L during the 20 d and 5 d SRT experimental periods respectively, as shown in Table 5.5.  The maximum E1 concentration detected throughout the study was 623 ng/L during the 5 d SRT.  E1  95 was present in approximately 10 % of the samples analyzed from the aerobic effluent, but was not detected in any influent samples, suggesting it was entirely generated through the treatment process. E1 was not detected in any microaerophilic effluent samples analyzed.  As described in Chapter 2, E1 is one of the three primary estrogens in the animal kingdom.  It has been found to affect fish reproduction at concentrations as low as 484 ng/L (Imai et al. 2007). These authors conducted full lifecycle analysis of E1 exposure on the marine fish Java medaka. Significant differences were demonstrated between control fish and fish exposed to E1 for a variety of effects.  Specifically, significantly lengthened hatching time and increased levels of hepatic vitellogenin were observed in fish exposed to E1 at a concentration of 484 ng/L.  Significant decreases in number of eggs spawned and fertilized were also demonstrated in this study for fish exposed to E1 at concentrations of 1188 ng/L.  The authors suggest these responses are indicative of effects on oogenesis and spermatogenesis.  The lowest no observable effects level determined by Imai et al. (2007) was 198 ng E1 / L.  In all likelihood, the concentrations of E1 detected in the present study would be diluted in the receiving environment below the proven effects levels discussed above.  However, the fact that this natural estrogen was found to be generated during wastewater treatment is important in and of itself. This finding could help to explain the estrogenic nature of some treated pulp mill effluents (described in Chapter 2).  At a minimum, the findings of E1 production in this study warrants investigations of similar biotransformations in full scale pulp mill wastewater treatment systems.  It is not surprising that estrone can be formed during biological wastewater treatment, given the breadth of knowledge of microbial conversion pathways used for steroid biosynthesis (described in Section 2.2.4).  E1 was previously detected in pulp mill effluent at a concentration of 49 ng/L (Fernandez et al. 2007).  However in that study, E1 was not measured in influent, making it difficult to make assumptions regarding the source of E1 in the effluent.  In the present study, however, E1 levels were evaluated in all influent samples and the compound was never detected, confirming that E1 was generated periodically during the wastewater treatment process.  A common pathway for estrone production is from β-sitosterol after an initial conversion to AD (Kieslich 1980; Malaviya & Gomes 2008), as shown in Figure 5.19.  In the present study, no AD was  96 detected in effluent, which suggests that either that the reaction shown in Figure 5.19 went to completion, or that a different pathway was used.  For example, estrone may have been formed from cholesterol.  Estrone was shown to be produced from cholesterol and its derivates as early as the 1960s (Sih et al. 1965).  More recently, estrone has been found to be the major metabolite formed from degradation of the cholesterol derivative acetoxyhydroxycholestene (19-HCA) (Madyastha & Shankar 1994).  This conversion pathway has been studied in detail and Madyastha and Shankar (1994) were able to obtain approximately 12% conversion of 19-HCA to estrone in the presence of Moraxella sp.  Hence there are a number of possible ways that estrone may have been generated during this study, but it appears likely that it was generated through biotransformation of sterols.  Androstenedioneβ-Sitosterol Estrone Aromatization of the A ring Hydrogenation of the ketone  Figure 5.19:  Simplification of the biotransformation of β-sito to estrone with AD as an intermediary.  The transformation of β-sito to AD is illustrated in Figure 2.6.  It is logical that estrone production occurred in the aerobic system but not in the microaerophilic system, because an oxygen molecule is added during the transformation of sterols to E1, as shown in Figure 5.19.  In the microaerophilic system there was likely to be more competition for the limited available oxygen, which would theoretically limit oxidation of AD to E1, if indeed this was the conversion pathway used.  5.2.6 Statistical Comparison of Treatment Systems Data from the four sets of experimental treatment conditions were compared to determine whether changes in DO and / or SRT led to significant changes in sterols removal or biotransformation products generation.  The Wilcoxon Rank Sum Test was used to generate p values comparing the  97 net change in compounds (effluent less influent concentrations) between the treatment systems and experimental runs.  Results of the statistical analysis are summarized in Table 5.8.  Table 5.8:  Summary of treatment system performance p values.  P values determined by the Wilcoxon Rank Sum Test using effluent concentration less influent concentration.  Bold indicates statistical significance (p ≤ 0.05).  Parameter DO Comparison - Aerobic vs Microaerophilic Systems SRT Comparison - 20 d vs 5 d Experimental Run 20 d SRT (run 1) 5 d SRT (run 2) Aerobic System Microaerophilic System Campesterol 0.056 0.105 0.354 0.065 Cholesterol 0.095 0.382 0.284 0.093 Coprostanol 1.000 1.000 0.038 0.038 Desmosterol n.a. 0.382 n.a. 0.527 Dihydrocholesterol 0.841 0.822 0.088 0.414 Estrone 0.424 0.170 0.767 n.a. β-Sitosterol 0.016 0.195 0.127 0.011 Stigmasterol 0.095 0.279 0.002 0.003  The p values in Table 5.8 show that there was a significant difference in removal of β-sito between the aerobic and microaerophilic systems, during the 20 d SRT experimental run.  The p values also show there was a significant difference in removal of β-sito in the microaerophilic system during the 20 day SRT, compared to that during the 5 d SRT.  There were no significant differences in influent β-sito over this time (as shown in Table 5.4). Therefore, it can be concluded that the differences in β-sito removal were due to the variations in treatment conditions.  Differences in β-sito removal are especially important due to the known endocrine disruption effects of this specific sterol (MacLatchy & Van der Kraak 1995; Mellanen, et al. 1996; Tremblay & Van der Kraak 1998; Tremblay & Van der Kraak 1999; Nakari & Erkomaa 2003; Christianson-Heiska et al. 2007), as discussed in Chapter 2.  Also, β-sito was the dominant sterol in all process streams, and therefore β-sito removal likely influenced the formation of sterols biotransformation products. Interestingly, there were no significant differences in β-sito removal in the aerobic system between SRTs, nor were there significant differences in β-sito removal during the 5 d SRT between the microaerophilic and aerobic systems.  These observations demonstrate the robustness of fully aerobic and low SRT treatment conditions.  98 As shown in Table 5.8, there were also significant differences in the removal of copro and stigma between the 5 and 20 d SRT experimental runs, in both the aerobic and microaerophilic systems. However, this is likely accounted for by changes in influent characteristics between the two experimental runs, as there were also significant differences in influent concentrations of copro and stigma over the same period (as shown in Table 5.4).  No other significant differences were observed in the comparison of the net change in compounds among the treatment trains and experimental runs.  It is interesting to note that there were no significant differences in generation of biotransformation products between treatment systems or experimental runs.  This is probably due to the very small concentrations of these compounds that were detected throughout the study.  It was also important to compare treatment performance between the first and second half of the 5 d SRT run because of the changes in the influent wastewater during this time.  Table 5.11 summarizes p values for treatment performance over the 5 d SRT, showing a statistically significant difference in concentrations of chole in the aerobic effluent between the first and second half of the 5 d SRT run. Given that this is the only statistically significant difference between the two halves of this experimental run, and that chole concentrations were very low in every case where it was detected, it can be concluded that the changes during the 5 d SRT did not significantly affect the metabolism of sterols and their derivatives.  Table 5.9:  Summary of treatment system performance p values during the 5 d SRT.  P values determined by the Wilcoxon Rank Sum Test using effluent concentration less influent concentration.  Bold indicates statistical significance (p ≤ 0.05).  Parameter DO Comparison - Aerobic vs Microaerophilic Systems SRT Comparison - 5 d vs 5 d (Run 2a vs Run 2b) 5 d SRT (Run 2a) 5 d SRT (Run 2b) Aerobic System Microaerophilic System Campesterol 0.222 0.700 1.000 1.000 Cholesterol 0.222 1.000 0.036 0.250 Coprostanol 1.000 1.000 0.177 0.177 Desmosterol n.a. 0.505 n.a. 0.302 Dihydrocholesterol 0.737 1.000 0.081 0.525 Estrone n.a. 0.197 0.079 n.a. β-Sitosterol 0.222 0.700 0.786 0.786 Stigmasterol 0.151 0.400 0.571 0.571   99 5.2.7 Phase-Separation Analysis and Sterols Mass Balance A final set of samples was collected during the 5 d SRT experimental run in order to conduct a sterols mass balance and phase-separation analysis.  The samples were collected during the second half of the experimental run, from day 161 of BBWTS operation to day 165, and the resulting data refer to “Run 2b”.  Three samples were collected from each of the influent and final effluents during this time and two samples were collected from each of the mixed liquors.  Information regarding sampling methodology is provided in Section 4.2.3.3 and the results are summarized below.  5.2.7.1 Phase-Separation Analysis Phase-separation analysis was used to examine the degree of sterols partitioning throughout the influent, mixed liquors and final effluents.  Particulate and soluble phases were separated via centrifugation and decantation.  Whole water samples (prior to centrifugation) were collected and separated during the 5 d SRT period.  Chemical analysis was conducted on the decanted water to obtain data for the soluble phase, and data were obtained for the particulate phase through analysis of whole water (prior to centrifugation) and subtraction of coinciding soluble values.  Results of the phase-separation analysis are summarized in Table 5.10, and listed in Appendix C.  100 Table 5.10: Summary statistics for sterols and metabolite concentrations in soluble and particulate phases.  All concentrations in μg/L; %Sol / Par - indicates the fraction of the analyte associated with the soluble (sol) or particulate phase (par), calculated as the concentration in that phase divided by the concentration in whole water, multiplied by 100 %, using the average concentrations of individual analytes within the process stream and phase indicated; n.d. - indicates not detected; (1) - MDLs varied from 0.005 μg/L to the value shown; (2) sum of averages calculated as the total of average concentrations for all sterols and metabolites detected in the process stream and phase indicated; (3) average fraction of whole water calculated as the average % par or % sol of all analytes in the process stream and phase indicated. Parameter MDL n x min max σ CV %Sol n x min max σ CV %Par Campesterol 0.1 (1) 3 73.7 41.8 92.7 27.8 38% 35% 3 137 62.5 227 83.6 61% 65% Cholesterol 0.05 (1) 3 0.55 0.469 0.65 0.091 17% 42% 3 0.77 0.216 1.24 0.515 67% 58% Coprostanol 0.05 (1) 3 0.114 n.d. 0.342 0.197 173% 18% 3 0.503 n.d. 0.774 0.436 87% 82% β-Sitosterol 0.2 (1) 3 347 207 451 126 36% 34% 3 681 301 1106 404 59% 66% Stigmasterol 0.005 3 13.7 7.60 17.9 5.41 39% 28% 3 35.2 10.3 49.9 21.7 62% 72% Sum of averages (2) 435 854 Average Fraction of Whole Water (3) 31% 69% Parameter MDL n x min max σ CV %Sol n x min max σ CV %Par Campesterol 0.005 3 3.44 1.82 5.56 1.92 56% 26% 3 9.88 9.58 10.3 0.39 4% 74% Cholesterol 0.005 3 0.551 0.337 0.903 0.307 56% 60% 3 0.369 0.204 0.479 0.146 39% 40% Dihydrocholesterol 0.05 (1) 3 n.d. n.d. n.d. n.d. n.d. 0 3 0.160 n.d. 0.241 0.139 87% 100% Estrone 0.05 (1) 3 n.d. n.d. n.d. n.d. n.d. 0 3 0.353 n.d. 0.623 0.320 91% 100% β-Sitosterol 0.005 3 17.5 8.29 29.8 11.1 63% 23% 3 57.7 55.5 60.8 2.74 5% 77% Stigmasterol 0.005 3 0.972 0.698 1.44 0.404 42% 26% 3 2.82 2.46 3.10 0.331 12% 74% Sum of averages (2) 19.0 61.4 Average Fraction of Whole Water (3) 22% 78% Parameter MDL n x min max σ CV %Sol n x min max σ CV %Par Campesterol 0.1 (1) 3 45.7 35.1 55 10.1 22% 52% 3 42.6 19.9 67 23.5 55% 48% Cholesterol 0.005 3 0.522 0.414 0.595 0.095 18% 73% 3 0.193 n.d. 0.442 0.231 120% 27% Coprostanol 0.05 (1) 3 n.d. n.d. n.d. n.d. n.d. 0 3 0.099 n.d. 0.298 0.172 173% 100% Desmosterol 0.05 (1) 3 n.d. n.d. n.d. n.d. n.d. 0 3 0.201 n.d. 0.602 0.348 173% 100% Dihydrocholesterol 0.05 (1) 3 n.d. n.d. n.d. n.d. n.d. 0 3 0.131 n.d. 0.392 0.226 173% 100% β-Sitosterol 0.1 (1) 3 227 176 279 51.6 23% 55% 3 183 112 311 111 61% 45% Stigmasterol 0.005 3 10.9 8.91 13.3 2.20 20% 61% 3 7.08 2.12 13.4 5.77 82% 39% Sum of averages (2) 238 191 Average Fraction of Whole Water (3) 34% 66% Microaerophilic Effluent Particulate PhaseSoluble Phase Influent Aerobic Effluent    101 Table 5.10, continued. Parameter MDL n x min max σ CV %Sol n x min max σ CV %Par Campesterol 0.005 2 4.46 3.34 5.58 1.59 36% 1% 2 487 395 579 130 27% 99% Cholesterol 0.005 2 0.466 0.373 0.559 0.13 28% 3% 2 16.0 13.8 18.1 3.06 19% 97% Coprostanol 0.05 (1) 2 n.d. n.d. n.d. n.d. n.d. 0 2 1.02 n.d. 2.04 1.44 141% 100% Dihydrocholesterol 0.05 (1) 2 n.d. n.d. n.d. n.d. n.d. 0 2 4.25 4.19 4.30 0.08 2% 100% β-Sitosterol 0.05 (1) 2 21.2 15.5 27.0 8.10 38% 1% 2 2838 2317 3360 737 26% 99% Stigmasterol 0.005 2 1.43 1.17 1.68 0.36 25% 1% 2 127 121 133 8.42 7% 99% Sum of averages (2) 22.7 2969 Average Fraction of Whole Water (3) 1% 99% Parameter MDL n x min max σ CV %Sol n x min max σ CV %Par Campesterol 0.005 2 127 82.9 172 62.9 49% 19% 2 552 480 624 102 18% 81% Cholesterol 0.005 2 0.706 0.603 0.810 0.147 21% 11% 2 5.53 5.31 5.76 0.316 6% 89% Coprostanol 0.05 (1) 2 0.338 0.282 0.394 0.079 23% 13% 2 2.31 2.12 2.50 0.267 12% 87% Desmosterol 0.05 (1) 2 n.d. n.d. n.d. n.d. n.d. 0 2 3.25 n.d. 6.49 4.59 141% 100% Dihydrocholesterol 0.05 (1) 2 0.23 n.d. 0.45 0.32 141% 100% 2 n.d. n.d. n.d. n.d. n.d. 0 β-Sitosterol 0.05 (1) 2 596 423 769 245 41% 17% 2 2856 2677 3036 254 9% 83% Stigmasterol 0.005 2 26.8 17.2 36.4 13.6 51% 17% 2 132 124 139 10.4 8% 83% Sum of averages (2) 623 2991 Average Fraction of Whole Water (3) 25% 75% Microaerophilic Mixed Liqour Aerobic Mixed Liquor Soluble Phase Particulate Phase   The results tabulated above indicate that averages of 66 to 99 % of analytes in all process streams were associated with the particulate phase. It is logical that the particulate phase accounted for the majority of analytes, given the low solubilities and lipophilic nature of sterols.  Of all the process streams analyzed, the mixed liquors had highest portions of analytes associated with the particulate phase, with a range of 75 to 95 %.  This compares to range of 66 to 78 % of analytes partitioned onto particulates in the effluents and 69 % in the influent.  High particulate phase values in the mixed liquors were likely related to the comparatively-high suspended solids concentrations in the mixed liqour, as MLSS provides surface area for bioadsorption.  The aerobic system exhibited the highest portion of analytes in the particulate phase, with average portions of 78 and 99 % in the aerobic effluent and mixed liquor, respectively.  Conversely, results for the microaerophilic system demonstrated a lower association with the particulate phase. Average particulate percentage values were 66 and 75 % across all analytes detected in the microaerophilic final effluent and mixed liquor, respectively.  Lower particulate phase percentages in the  102 microaerophilic system could be partially due to sampling error caused by poor settleability of the wastewater in the microaerophilic system, as some solids may have been entrained in the decanted supernatants if centrifugation was not sufficient to separate the phases.  This type of error would cause the values for the soluble phase to appear artificially high.  Poor settleability of microaerophilic process streams was evidenced throughout the study, as described throughout Section 5.1.  Not surprisingly, sterols levels were highest among effluent samples in those collected from the microaerophilic system during the 20 d SRT experimental run, and this set of operating conditions also resulted in the the highest effluent TSS levels measured throughout the study (previously reported in Sections 5.1.4 and 5.2.3 respectively).  The phase-separation analysis was only conducted during the 5 d SRT experimental run, so only indirect comparisons can be made to the 20 d SRT run.   However, it would appear that the high levels of sterols in this effluent were associated with high effluent TSS, due to partitioning of sterols onto the solid phase.  5.2.7.2 Mass Balance A mass balance was calculated using the data obtained for whole water samples (prior to centrifugation) from the phase-separation analysis.  Sterols mass flows were calculated based on the sum of the average concentrations for each sterol measured, by process stream.  The results of mass flow calculations are provided in Table 5.11.  These data indicate average influent sterols loads to the BBWTS were approximately 8.0 mg/d and 8.2 mg/d for the aerobic and microaerophilic systems, respectively.  Influent loads were variable between the two treatment trains as influent was pumped to each reactor with separate pumps.  Hence, loading rates were dependant on pump speeds.  Of the influent loads, approximately 0.5 mg/d (6 %) was discharged in aerobic effluent and approximately 3.0 mg/d (37 %) was discharged in microaerophilic effluent.  The higher value for the microaerophilic system may have been due to the poor biomass settleability observed in this system. An additional 3.3 mg/d (41 %) of the influent sterols load was removed in aerobic WAS and 2.1 mg/d (26 %) was removed in microaerophilic WAS.  For the aerobic system, this indicates that approximately 47 % of the incoming sterols exited the treatment system without undergoing degradation and therefore, approximately 53 % of incoming sterols appeared to be degraded or biotransformed.  These values agree with those obtained from previous studies of full scale PME activated sludge treatment systems (Mahmood-Khan, 2005).  In the microaerophilic system on the other hand, approximately 63 % of incoming sterols exited the treatment works without undergoing  103 degradation, and only 37 % of incoming sterols appeared to be degraded or biotransformed.  This reflects the reduced level of biological activity in the microaerophilic system, and is not surprising given the lower dissolved oxygen concentration maintained in the microaerophilic system.  Table 5.11: Sterols mass flows.  Sterols concentrations are summed for individual samples and resultant values are averaged across process streams.  Analytes that were present in mixed liquor or final effluent that were not present in the influent were not included in mass flow calculations as they were assumed to be fully generated during treatment. Process Stream Sterols Concentration (ug/L) Flow (L/d) Mass Flow (mg/d) % of Influent Load Influent:   Aerobic 1290 6.21 8.01 100%   Microaerophilic 1290 6.35 8.19 100% Mixed Liquor:   Aerobic 3497 n.a. n.a. n.a.   Microaerophilic 4299 n.a. n.a. n.a. Final Effluent:   Aerobic 93 5.28 0.49 6%   Microaerophilic 518 5.86 3.03 37% WAS:   Aerobic 3497 0.932 3.26 41%   Microaerophilic 4299 0.492 2.11 26%   The sterols concentration data provided in Table 5.11 show that mixed liquor samples contained approximately three times higher sterols concentrations than did samples collected from the influent, indicating that a portion of incoming sterols was accumulated in the mixed liquor without undergoing degradation or transformation.  This suggests that bioadsorption was an active process in sterols removal in both the aerobic and microaerophilic systems.  The phase-separation data support this idea, by showing that the majority of analytes in the mixed liquor was associated with the particulate phase.  Using the mass flow data, the degree of biotransformation was estimated.  Table 5.12 shows the mass flows of biotransformation products as percentages of the influent sterols mass flows. Biotransformation product concentrations were calculated as the sum of concentrations of those analytes present in mixed liquor or final effluent that were not detected in the influent.  The analytes  104 that met this criterion during the phase-separation sampling period were desmosterol, dihydrocholesterol and estrone.  Table 5.12: Biotransformation products mass flows.  Analyte concentrations are summed for individual samples and resultant values are averaged across process streams. Process Stream Sterols Concentration (ug/L) Biotransformation Products Concentration Flow (L/d) Mass Flow (mg/d) Biotransformation Products (% of Influent Sterols Load) (ug/L) Influent:   Aerobic 1290 n.a. 6.21 8.01 n.a.   Microaerophilic 1290 n.a. 6.35 8.19 n.a. Mixed Liquor:   Aerobic n.a. 4.25 n.a. n.a. n.a.   Microaerophilic n.a. 6.49 n.a. n.a. n.a. Final Effluent:   Aerobic n.a. 0.771 5.28 0.004 0.05%   Microaerophilic n.a. 0.994 5.86 0.006 0.07% WAS:   Aerobic n.a. 4.25 0.93 0.004 0.05%   Microaerophilic n.a. 6.49 0.49 0.003 0.04%   The biotransformation products leaving the BBWTS in final effluent during the phase-separation sampling period accounted for 0.05 and 0.07 % of the influent sterols load in the aerobic and microaerophilic systems, respectively.  These values represent a very small fraction of the influent sterols.  However, although these values are small, they should not be considered negligible due to two reasons.  First, some of the biotransformation products found in this study are potent EDCs. Specifically, estrone can induce reproductive effects in fish at concentrations as low as 484 ng/L (Imai et al. 2007), as described in Section 5.2.5.  The total concentrations of biotransformation products in the final effluent shown in Table 5.12 exceed this level, but these values are summations of the average metabolites detected in the process streams, and are not representative of the range of any individual parameter.  Estrone was detected in final effluent at a maximum concentration of 623 ng/L as discussed earlier (Table 5.7).  Second, the concentrations and mass loadings of sterols biotransformation products in full scale wastewater treatment systems may be important.  For comparison purposes, a mill with similar effluent concentrations to those listed in Table 5.12, but with a final effluent flow rate of 70,000 m3/d, would discharge sterols biotransformation products at a mass flow rate of approximately 90 g/d.  This value would be even higher using influent sterols  105 concentrations published in earlier studies. For example, Kostamo et al. (2004) showed influent sterols concentrations to be as high as 5614 μg/L during a process change at a Finnish kraft mill. Using 0.1 % conversion to biotransformation products, this could produce an effluent concentration of biotransformation products of 5.6 μg/L, and a mass loading of approximately 400 g/d if using a hypothetical final effluent flow rate of 70,000 m3/d.  Hence, the mass flow of sterols biotransformation products entering the receiving environment from full scale mills may be quite high and are, therefore, worth investigating.  The fact that they may contain animal steroids produced during wastewater treatment is further impetus to full scale study.  The data provided in Table 5.13 may also be interpreted to suggest that some sterols biotransformation products may have been present in the BBWTS that were not measured in the present study.  This is possible as 53 % and 37 % of influent sterols were degraded in the aerobic and microaerophilic systems respectively, but only 0.1 % of influent sterols were measured as biotransformation products in effluent and WAS.  Of course, it is also possible that the remaining degraded sterols were fully mineralized and no other metabolites were generated.  Even if other metabolites were generated, the potential for generation of steroid molecules capable of endocrine disruption is limited by a number of factors.  For example, it has been shown that the presence of certain bacteria limit transformation of sterols into other steroids, such as those that degrade the steroid nucleus or that inhibit side chain degradation (Malaviya & Gomes 2008).  Regardless of these limitations, the discrepancy in the present study between the fraction of influent sterols degraded and the quantities of biotransformation products measured is so large as to suggest that further study may be able to identify other sterols metabolites, apart from those measured in this work.  5.2.8 Summary of Sterols Removal and Biotransformation The aerobic system was able to achieve very high sterols removal efficiencies, with values close to 100 % for the three main plant sterols during both the 5 d and 20 d SRT experimental runs.  Sterols removal efficiencies for the microaerophilic system were substantially lower, with average values of 69 % and 50 % during the 5 d and 20 d SRT experimental periods, respectively.  Removal of β-sito was significantly lower in the microaerophilic system compared to the aerobic system when operating under a 20 d SRT. β-Sito removal was also significantly lower in the microaerophilic system when operating under a 20 d SRT, compared to a 5 d SRT.  Thus, it can be seen that the treatment conditions that led to the lowest extent of sterols removal were low dissolved oxygen  106 concentrations coupled with a high SRT.  Under this combination of treatment conditions, low sterols removal efficiencies were likely associated with poor settleability and increased effluent TSS, which were also observed.  This argument is strengthened by the findings of the phase-separation analysis, which confirmed that the majority of sterols in all process streams were partitioned in the particulate phase.  Biodegradation was the dominant sterols removal mechanism in both the aerobic and microaerophilic systems and accounted for removal of approximately 53 % and 37 % of the incoming sterols in the two systems, respectively.   Comparatively, bioadsorption contributed 41 % of sterols removal in the aerobic system and 26 % of sterols removal in the microaerophilic system. Biotransformation of sterols into other steroid molecules involved only a very small portion of the sterols removed throughout the experimental program, with 0.1 % of the influent sterols measured as biotransformation products in final effluent and WAS in both systems.  The conversion percentages are very low, but nonetheless, small amounts of cholesterol, dihydrocholesterol, desmosterol and estrone were observed to be generated through operation of the BBWTS.  Estrone is the most ecologically significant compound generated, as it is a natural estrogen and can affect fish reproduction at concentrations in the range of ng/L.  Estrone was produced only under aerobic conditions, but was detected in samples collected from both experimental runs at maximum concentrations of 334 and 623 ng/L during the 20 SRT run and 5 d SRT run, respectively.  The formation of 17-keto sex steroids, such as estrone, are most likely limited in the microaerophilic system by the low DO concentrations.  This shows that fully aerobic conditions are more conducive to formation of animal sex steroids, despite the high sterols removal efficiencies observed under the same conditions.   Also, twice as much net generation of biotransformation products was observed during the 5 d SRT run, relative to the 20 d SRT run, in both the aerobic and microaerophilic systems; this appears to suggest that short SRTs are more conducive to the growth of sterols transforming bacteria than are long SRTs.  107 5.3 GENOMIC RESPONSE OF RAINBOW TROUT TO EXPOSURE TO TREATED PULP MILL EFFLUENTS  5.3.1 Fish Exposure Bioassays The bioassays were successful, as no mortalities or abnormalities were observed in the test fish or control fish over the 96 hr exposure period.  Results describing effluent chemistry, test parameters and fish dissections follow.  Effluent used in the bioassays was collected and sampled over 4 to 5 day periods at the end of the 20 d SRT experimental run, from day 103 of BBWTS operation to day 107, and approximately half way through the 5 d SRT experimental run, from day 130 of BBWTS operation to day 134.  These samples make up a portion of all the samples collected and analyzed throughout the entire study (which were reported in Table 5.5).  The samples collected for bioassay purposes comprise an n of 4 for the effluents generated with a 20 d SRT (compared to an n of 5 for those analyzed over the full duration of the run) and an n of 5 for the effluents generated with a 5 d SRT (compared to an n of 8 analyzed over the full duration of the run).  Chemistry results for samples collected specifically for the bioassay effluents are summarized in Table 5.13.  The most notable difference between chemistry results reported for the bioassays (Table 5.13) and those reported for the full duration of the study (Table 5.5) is that no estrone or desmosterol was detected in any samples used for the fish exposures.  Sterols concentrations measured in the bioassay effluents were similar to those reported in Table 5.5, with the highest concentrations found in the microaerophilic effluent generated using a 20 d SRT, with average total sterols of 492 μg/L. The lowest sterols concentrations were found in the aerobic effluent generated using a 20 d SRT, with average total sterols of 7.29 μg/L.  These compare to average sterols concentrations of 528 μg/L and 12.0 μg/L in the same process streams for data reported over the full study duration (Table 5.5).  Variation between samples collected for the bioassays ranged from 6 to 48 % in all cases, except β-sito in aerobic effluent treated with a 5 d SRT (CV = 67 %) and DHC in microaerophilic effluent treated with a 5 d SRT (CV = 138 %).  Variations were lower among bioassay effluent samples than those observed among the samples collected over the full duration of study, which may be expected as the sampling interval was shorter for the bioassay samples.  108 Table 5.13: Bioassay chemistry summary.  All concentrations in μg/L for 100 % effluent; (1) - MDLs varied from 0.005 ug/L to the value shown, based on final sample dilution for GC-MS analysis; (2) - sum of averages calculated as the total of average concentrations for all analytes detected  in the process stream. Parameter MDL n x min max σ CV n x min max σ CV Campesterol 0.005 4 1.02 0.833 1.12 0.13 13% 5 2.80 1.88 5.15 1.33 48% Cholesterol 0.005 4 0.416 0.350 0.56 0.096 23% 5 0.611 0.422 0.960 0.225 37% Dihydrocholesterol 0.05 (1) 4 0.369 0.286 0.431 0.062 17% 5 n.d. n.d. n.d. n.d. n.d. β-Sitosterol 0.005 4 5.28 4.19 5.88 0.75 14% 5 21.5 12.2 46.8 14.3 67% Stigmasterol 0.005 4 0.213 0.194 0.219 0.012 6% 5 0.68 0.45 1.00 0.22 32% 7.29 25.5 Parameter MDL n x min max σ CV n x min max σ CV Campesterol 0.1 (1) 4 87.8 65.1 112 20.4 23% 5 66.8 52.0 87.6 13.6 20% Cholesterol 0.005 4 1.14 0.919 1.36 0.20 18% 5 1.15 0.926 1.53 0.277 24% Dihydrocholesterol 0.05 (1) 4 0.387 0.336 0.447 0.056 14% 5 0.183 n.d. 0.496 0.252 138% β-Sitosterol 0.1 (1) 4 400 305 503 84.6 21% 5 369 300 489 72.2 20% Stigmasterol 0.005 4 2.83 2.12 3.48 0.605 21% 5 10.4 8.57 14.2 2.35 23% 492 448 Aerobic Effluent Microaerophilic Effluent 20 day SRT (run 1) 5 day SRT (run 2) Sum of Averages (2) Sum of Averages   During the bioassays, temperature, pH, DO and conductivity were within the target values for the method (listed in Table 4.6) in each tank used in the experiments.  Length, weight and condition factor data obtained for test organisms during dissections are provided in Table 5.14.  The results are comparable between the two assay batches, although the fish used in Assay Batch 2 were slightly smaller than those used in Assay Batch 1.  Both the test effluents and fish met the QA/QC requirements of the bioassay methods and these results are provided in Appendix D.   109 Table 5.14: Fish physical measurements.   Length  (cm) Weight (g) Condition Factor Loading Density (g/L) Assay Batch 1: December 7-11th, 2007 (20 d SRT) Mean 7.80 5.53 1.16 0.9 SD 0.82 1.95 0.38 Min 6.80 3.21 0.84 Max 9.00 8.82 2.10 Assay Batch 2: February 8-12th, 2008 (5 d SRT) Mean 7.66 4.77 1.06 0.8 SD 0.88 1.54 0.10 Min 6.30 2.58 0.98 Max 8.80 7.20 1.28   5.3.2 Genomic Analysis 5.3.2.1 Gene selection The gene selection process involved screening a subset of samples for a range of genes that may be expected to respond to EDCs in pulp mill effluents.  The suite of 28 screening genes was selected with staff at Environment Canada’s PESC facility, based on previous toxicogenomic experiments involving PMEs and other effluents with known endocrine disrupting properties.  Out of the 28 genes tested during the screening process, L8, VEPb, VEPg, VTG1, NME, AR and THRa displayed the best quality, most consistent results for the test criteria, as indicated in Table 5.15.  L8 was selected as the reference gene to normalize the results of the experimental genes.  All six experimental genes belong to the endocrine gene class and can be classified as estrogenic (VEPb, VEPg, VTG1 and NME), androgenic (AR), and thyroid related (THRa).  The functions of these genes and their uses in toxicogenomics are described in more in the following section.    110 Table 5.15: Gene screening results.  Electrophoresis results were required to display one clean, bright band with products of the same size as the target gene sequence; NTCs were required to be a minimum of 2 Ct values greater than the samples; dissociation curves were required to display one melting point, and all curves had to fit the same line tightly. Gene Symbol Electrophoresis results NTC results Dissociation curve results Notes Endocrine Gene Class ERA short Empty NA NA ERA long Empty NA NA VEPA Good Poor Good VEPB Good Good Good Candidate VEPG Variable intensity Good Good Candidate VTG 1 Faint bands Good Good Candidate VTG 2 Faint bands Poor Good AR Good Good Good Candidate GNRH1 Empty NA NA GNRH2 Empty NA NA GNRHR2 Empty NA NA STAR Multiple bands NA NA FSHR Empty NA NA GH1 Empty NA NA INSRB Faint bands Poor Poor THRA Faint bands Good OK Candidate Metabolism Gene Class CYP11A1 Faint bands Good Poor CYP11B1 Empty NA NA CYP17A1 Wrong size NA NA CYP19B Empty NA NA HSD3B1 Wrong size NA NA PTGS2 Wrong size NA NA NME Good Good Good Candidate Signal Transduction Gene Class AHR Very faint bands Good Good ARNT Wrong size NA NA HSPCA Multiple bands NA NA Reference Gene Class RPL23A Variable intensity Good Good L8 Good Good Good Candidate   VTG1 encodes the egg protein vitellogenin and is involved in sexual development and reproduction. It is a common genomic biomarker used to indicate estrogenic response (Flouriot et al. 1995; Mellanen et al. 1999; Denslow et al. 2001; Larkin et al. 2002; Filby et al. 2007; Benninghoff & Williams  111 2008).  VTG expression has been shown to increase in lacustrine white fish with exposure to pulp mill effluents high in wood extractives, including β-sito (Soimasuo et al. 1998; Mellanen et al. 1999). Although VTG has proven to be a robust indicator of estrogenic effects, recent findings suggest more sensitive biomarkers are required because some reproductive effects can be manifested at concentrations of xenoestrogens too low to induce vitellogenin production (Gunnarsson et al. 2007). It has also been shown that vitellogenin production is mainly mediated through estrogen receptor β (ERb), so those compounds which only exhibit estrogen receptor α (ERa) activity are not detected by VTG analysis (Leanos-Castaneda & Van der Kraak 2007).  VEP genes encode vitelline envelope proteins, which are precursors to vitellogenin and are likewise involved in sexual development and reproduction.  VEP genes respond characteristically to estrogen exposure (Larsson et al. 1994; Westerlund et al. 2001; Larkin et al. 2002).  VEPg specifically has been shown to be significantly up-regulated in rainbow trout by dietary exposure to estradiol (Benninghoff & Williams 2008).  Additionally, VEP genes have been shown to be induced at lower levels of estrogens than VTG (Westerlund et al. 2001).  NME encodes nucleoside dipohosphate kinase (nm23) proteins, which have a diverse variety of functions in growth and development in all levels of organisms (Postel 1998), including modulation of ERa and ERb and their response to estrogen-induced gene transcription in human cells (Lin et al. 2002; Rayner et al. 2008).  In fish, NME expression can be induced using estrogen concentrations too low to induce VTG, yet high enough to have potential to affect gonadal differentiation (Gunnarsson et al. 2007).  Because of its sensitivity, NME is emerging as a novel biomarker for estrogenic effects.  AR encodes androgen receptors which regulate sex steroid action.  Less is known about contaminant modes of action that affect androgen related endocrine genes than their estrogen related counterparts, but a number of fish exposure studies have been conducted to evaluate the effects of contaminants on AR activity.  For example, vinclozolin, a fungicide, and trenbolone, a cattle pharmaceutical, have been found to increase AR mRNA transcripts in fathead minnows (Martinovic et al. 2008).  Also, flutamide, an anti-androgenic cancer drug, has been shown to decrease AR mRNA in fathead minnow females and increase androgen synthesis in males (Filby et al. 2007).  The significance of AR disruption was illustrated by Iguchi et al. (2007), who showed that  112 such changes are associated with altered gonopodium development in mosquitofish fry and anal fin development in adult mosquitofish females.  From these studies, it can be assumed that AR has  the potential to respond to exposure to PME, if that effluent is capable of inducing an androgenic response.  THRa encodes thyroid hormone receptors which regulate thyroid hormone action.  These receptors act on the pituitary gland, causing it to release thyroid stimulating hormones, which in turn stimulates the thyroid gland to release thyroxine, which can be modified into triiodothyronine (De Silva & Anderson 1995).  Therefore, changes in THRa transcription may lead to changes in circulating levels of thyroid hormones.   Thyroid hormones play a variety of functions in fish from growth and development to pigmentation of rod and cone photoreceptors affecting sight (Temple et al. 2008).  Expression of THRa has previously been shown to be down-regulated by exposure to both estrogenic and anti-androgenic compounds (ethinylestradiol and flutamide) (Filby et al. 2007). One study was found in the literature that investigaed the effects of phytosterol exposure on thyroid hormone levels in grayling eggs (Honkanen et al. 2005).  Here, no changes in triiodothyronine or thyroxine levels were found in embryo extracts.  Although it is important to note these findings, they are not necessarily indicative of a response on THRa.  5.3.2.2 Sample preparation Electrophoresis results following extraction of total RNA showed that the extraction procedure was successful, with only three out of the 120 samples extracted displaying degraded or marginal RNA. Nine samples from each exposure and control group were selected for cDNA synthesis predominantly based on RNA concentration.  PCR results evaluating the cDNA synthesis reaction showed that the transcription process was successful for 89 of the 90 samples.  One sample had to be discarded from the bioassay group Control 2 (from the 5 d SRT run), making an n of 8 for this group.  All other groups maintained an n of 9 through to final analysis as planned, thus making a total n of 89 from all bioassay groups.  5.3.2.3 Bioinformatics Final QPCR data was analyzed as per Osachoff (2008) and Burns et al. (2005), with changes to these procedures as described in Sections 4.4.2.5 through 4.4.2.7.  The results of this analysis follow.  113  5.3.2.3.1 Method development Final QPCR data provided Ct values for each test sample (n = 89) for the reference gene and six experimental genes.  These data were converted to mRNA transcript copy numbers by developing standard curves as described in Section 4.4.2.7.  The final standard curves met the criteria defined during method development, as shown in Table 5.16.  The final standard curves are provided in Appendix E.  Table 5.16: Standard curve summary data.  Criteria were: threshold < 0.02; efficiency ≥ 60 % and ≤ 120 %; and R2 ≥ 0.95. Gene Threshold Efficiency R2 L8 0.009 114.9% 0.955 VTG1 0.013 62.2% 0.951 VEPB 0.014 65.6% 0.973 VEPG 0.010 60.8% 0.981 AR 0.010 98.4% 0.977 NME 0.012 82.7% 0.962 THRA 0.013 87.3% 0.967   5.3.2.3.2 Final data analysis Analysis of mRNA copy numbers demonstrated differences in gene expression between test fish and controls.  Both increases and decreases in copy numbers were observed, depending on gene, wastewater treatment, and exposure concentration.  Fold changes in copy numbers between test fish and controls are illustrated in Figure 5.20 and statistical analysis is summarized in Table 5.17. Breifly, Figure 5.20 shows that each of the six genes was up-regulated in fish exposed to effluent MP2 (microaerophilic, 5 d SRT).  This is supported by p values, which show a significant difference in copy numbers between control fish and fish exposed to MP2 10% for each gene, except AR. Additionally, significant differences in NME and THRa copy numbers between test fish and controls were resultant from exposure to effluents other than MP2.  Results for each of the six experimental genes are discussed in detail below.      114 1 2 3 4 5 6 ARFo ld  C ha ng e -6 -5 -4 -3 -2 -1 1 2 3 4 5 6 THRaFo ld Ch an ge -6 -5 -4 -3 -2 -1 1 2 3 4 5 6 VTG1Fo ld Ch an ge -6 -5 -4 -3 -2 -1 1 2 3 4 5 6 VEPgFo ld Ch an ge -6 -5 -4 -3 -2 -1 1 2 3 4 5 6 NMEFo ld Ch an ge -6 -5 -4 -3 -2 -1 AB1 10% AB1 1% AB1 1% AB1 1% AB1 1% AB1 1% AB1 10% AB1 10% AB1 10% AB1 10% MP1 1% MP1 1% MP1 1% MP1 1% MP1 1% MP1 10% MP1 10% MP1 10% MP1 10% MP1 10% AB2 1% AB2 1% AB2 1% AB2 1% AB2 10% AB2 10% AB2 10% AB2 10% AB2 10% MP2 1% MP2 1% MP2 1% MP2 1% MP2 1% MP2 10% MP2 10% MP2 10% MP2 10% MP2 10% 1 2 3 4 5 6 VEPbFo ld Ch an ge -6 -5 -4 -3 -2 -1 AB1 1% MP1 1% MP1 10% AB2 1% AB2 10% MP2 1% MP2 10% a) b) c) d) e) f) ** * * *** * * * ** ** * Fo ld  C ha ng e Fo ld Ch an ge Fo ld Ch an ge Fo ld Ch an ge Fo ld Ch an ge Fo ld Ch an ge  Figure 5.20:  Fold change diagrams.  Error bars indicate standard error of the mean (SEM = standard deviation / square root of n); * indicates p value < 0.05; ** indicates p value < 0.01.  115 Table 5.17: Student’s t-test p values for genomic data analysis (control vs. treatment).  Bold indicates statistical significance (p ≤ 0.05).  Treatment Gene   AR THRa VTG1 VEPb VEPg NME 20 day aerobic 1% 0.564 0.251 0.792 0.732 0.194 0.264 SRT aerobic 10% 0.892 0.802 0.970 0.921 0.508 0.013   microaerophilic 1% 0.763 0.657 0.277 0.307 0.086 0.005   microaerophilic 10% 0.279 0.003 0.251 0.390 0.603 0.041  5 day aerobic 1% 0.547 0.303 0.172 0.380 0.257 0.906 SRT aerobic 10% 0.880 0.119 0.061 0.580 0.168 0.022   microaerophilic 1% 0.905 0.027 0.610 0.727 0.255 0.651   microaerophilic 10% 0.312 0.002 0.016 0.007 0.011 0.032  5.3.2.3.2.1 Estrogenic Response The activity of the estrogen responsive genes VEPb, VEPg and VTG1 was similar. Figure 5.20 shows that exposure to MP2 (microaerophilic effluent generated using a 5 d SRT) at 10 % concentration caused the greatest up-regulation of these three genes.  Fold change between test fish and controls was greatest among all genes and exposure groups for VTG1, with a 4.4 magnitude increase in VTG1 gene transcripts in fish exposed to effluent MP2 at 10 % concentration, compared to control fish.  It is fitting that VTG1 expression displayed the largest increase for any one bioassay, as it is well established biomarker for estrogenic responses.  The three common estrogen related genes were also down-regulated compared to controls in some cases, most consistently with exposure to MP1 (microaerophilic effluent generated using a 20 d SRT) at 1% concentration, with a maximum negative fold change of -1.7 in VEPg transcripts.  However, none of the instances of down-regulation were statistically significant for these three genes. As shown in Table 5.17 and Figure 5.20, differences in copy numbers between treatments and controls were only significant for fish exposed to effluent MP2, at 10% concentration.  The total sterols concentrations in microaerophilic effluents (509 and 448 μg/L for MP1 and MP2, respectively) were much greater than those of the aerobic effluents (7.30 and 25.6 μg/L for AB1 and AB2, respectively), as shown in Table 5.13.  It is logical that exposure to the microaerophilic effluents would more strongly induce estrogen responsive genes, as sterols exposure is linked to estrogenic effects.  However, the highest sterols concentrations were observed in MP1, followed by MP2.  As MP2 was the only effluent to cause significant effects on expression of VEPb, VEPg and VTG1, sterols concentration alone is clearly not enough to predict estrogenicity of the effluents.  116 This observation may be explained in a number of ways.  First, it suggests the presence of an estrogenic compound not measured in this study.  Second, estrogenicity of the effluent may be related to the treatment process SRT.  The results of VEPb, VEPg and VTG1 analysis suggest that effluent generated using a 5 d SRT produced a stronger estrogenic response in test fish, compared to controls, than did effluent generated using a 20 d SRT.  The response in NME was markedly different from that of the other three estrogen responsive genes.  Fold change values comparing test fish to controls were relatively low for NME relative to the other estrogenic genes with a range of fold change limited to ± 1.4, as shown in Figure 5.20. Although fold changes were small, results of Student’s t-test indicate the differences in NME copy numbers in test fish compared to controls, were significantly different in five out the eight bioassays. This means that exposure to every effluent at 10 % concentration (AB1, AB2, MP1, MP2), as well as MP1 at 1% concentration, led to a significant differences in NME expression compared to the corresponding expression in controls.  Comparatively, significant differences in copy numbers between test fish and controls were only noted from one of the bioassays (MP2 10%) for the other estrogen-responsive genes.  Notably, NME was the only gene to show a significant response to aerobic effluents, with NME induction after exposure to 10 % aerobic effluent with total sterols concentration as low as 7.30 μg/L (in 100 % effluent).  It is important to remember here that no estrone was detected in the samples collected from effluent used in the bioassays; therefore, it may be assumed that the estrogenic effects were caused by sterols, or some other, unmeasured compound(s).  The results for NME show that effluent from each set of treatment conditions was capable of inducing an estrogenic response.  The results also demonstrate that NME was significantly more sensitive than VEPb, VEPg and VTG1 in showing an estrogenic response.  These findings are in agreement with work conducted by Gunnarsson et al., (2007), confirming that NME can be significantly induced at concentrations of xenoestrogens that are too low to induce VTG, and in the case of the present study, also too low to induce VEP.  Results also demonstrated that exposure to effluents generated using a 20 d SRT led to significant repression of NME (negative fold change), whereas exposure to effluents generated using a 5 d SRT led to significant induction of NME (positive fold change).  This may indicate that treatment system  117 operation under a shorter SRT is more likely to produce estrogenic effects than operation under a long SRT.  This finding is supported by the sterols chemistry data discussed previously; this shows that the microaerophilic effluent induced a stronger estrogenic response when generated with a 5 d SRT, than when generated with a 20 d SRT, despite higher sterols concentrations being associated with the latter.  On the other hand, down-regulation of NME from exposure to effluents generated with a 20 d SRT suggests that at longer SRTs, the effluent induced anti-estrogenic effects.  Anti- estrogenic effects have been previously reported in fish from exposure to PMEs, namely via inhibition of vitellogenin production (Kukkonen et al. 1999; Christianson-Heiska & Isomaa 2008).  5.3.2.3.2.2 Androgenic Response Fold change was relatively low for AR gene transcripts between all exposure groups and controls, and ranged from -1.1 from exposure to AB2 (aerobic effluent generated with a 5 d SRT) at 1 % concentration to 1.5 from exposure to MP1 (microaerophilic effluent generated with a 20 d SRT) at 10 % concentration.  However, results of a Student’s t-test showed that no changes in AR copy numbers were significant between test fish and controls.  It is logical that no significant androgenic effects were detected in the test fish, as no androgenic compounds were detected in any of the effluents.  5.3.2.3.2.3 Thyroid Hormone Receptor Response THRa was second in sensitivity to endocrine related effects of all six genes tested, next to NME. There were significant differences in THRa copy numbers between exposure groups and controls for three out of the four bioassays, using microaerophilic effluents.  None of the aerobic effluents induced a THRa response, nor did MP1 at 1 % concentration.  Interestingly, significant differences in copy numbers between test fish and controls were resultant from negative fold changes from exposure to MP1 at 10 % concentration and from positive fold changes from exposure to MP2, at 1 % and 10% concentration.  According to these results, exposure to microaerophilic effluent generated under a 5 d SRT led to up-regulation of THRa, and exposure to microaerophilic effluent generated under a 20 d SRT led to down-regulation of THRa.  This trend is similar to that observed for NME, showing effluents treated with shorter SRTs induced endocrine related genes and those treated with longer SRTs repressed endocrine related genes.    118 5.3.2.4 Summary of Genomic Analysis Significant changes in gene expression were demonstrated between fish exposed to microaerophilic effluents generated with a 5 d SRT (MP2) at 10 % concentration and control fish for every experimental gene except AR.  In comparison, aerobic effluents only induced significant changes in gene expression between test fish and controls for NME.  These results suggest that effluents treated under oxygen-limited conditions are more conducive to producing endocrine disruption effects in fish, relative to effluents treated using fully aerated conditions.  NME was the most sensitive of the six experimental genes in that differences in NME expression were significant between test fish and controls for the majority of the bioassay groups.  Significant changes in NME expression between test fish and controls were apparent from exposure to all aerobic and microaerophilic effluents at 10 % concentration, as well as microaerophilic effluent generated with a 20 d SRT at 1 % concentration.  This compares to the other estrogenic genes, VTG1, VEPb and VEPg, for which significant up-regulation was only induced by exposure to MP2. The sensitivity of NME was demonstrated as it was the only gene to significantly respond to exposure to aerobic effluents, though only at 10 % concentrations.  Comparatively, microaerophilic effluents were able to induce significant changes in gene expression at 1 % concentration for both NME and THRa.  THRa was thus second in sensitivity to endocrine disruption of all the genes tested, with significant changes in THRa expression between test fish and controls from exposure to MP2 10%, MP1 10 % and MP1 1%.  Solids retention time also influenced the endocrine disruption potential of the effluents, as demonstrated by a greater occurrence of significant changes in gene expression from exposure to effluents treated with a 5 d SRT than a 20 d SRT.  Interestingly, it was also the 5 d SRT that led to the generation of a greater amount of sterols biotransformation products.  Also, exposure to microaerophilic effluent generated with a 5 d SRT led to the greatest degree of genomic response, even though this was not the effluent with the highest measured sterols concentrations.  Hence it appears that operating the treatment system under a 5 d SRT increased the endocrine disruption potential of the effluent, independent of the chemistry results obtained in this study.  Finally, down-regulation of endocrine genes was more prevalent from exposure to effluents treated using a 20 d SRT, and up-regulation of endocrine genes was more prevalent from exposure to  119 effluents treated with a 5 d SRT.  Down-regulation was only significant for NME and THRa; however, this may be an important finding, as these were the most sensitive genes to endocrine disruption out of all six experimental genes evaluated in this study.   120 6 CONCLUSIONS AND RECOMMENDATIONS The objectives of this study were to: 1) Assess whether phytosterol biotransformation products can be identified in biologically treated pulp mill effluents, 2) Determine if wastewater treatment conditions affect the levels of phytosterols and their biotransformation products in treated effluents, 3)  Assess the genomic response in fish exposed to biologically treated pulp mill effluents, using a range of wastewater treatment conditions, and 4)  Determine whether a relationship exists between the genomic response in exposed fish and (a) the concentrations of phytosterols and their biotransformation products in test effluents, or (b) the wastewater treatment conditions used to produce the test effluents.  In order to assess these objectives, it was first necessary to evaluate the operation of the BBWTS, in order to have confidence in the results of the chemical and genomic analysis.  The goals of the experimental design were met in operation of the BBWTS, as DO and SRT were substantially different between treatment trains and experimental runs, respectively.  It was difficult to reach a 20 d SRT in the microaerophilic system due to poor settleability and resultant solids losses in final treated effluent; however, an average SRT of 16 d was obtainable in this system, which was considered to be sufficiently higher than the low SRT (5 d) to evaluate changes based on this parameter.  It is recommended that future studies utilizing a BBWTS strive to minimize thermal density currents in the secondary clarifiers by heating them to the same temperature as the reactors, thus maximizing biomass settling.  COD and nutrient concentrations demonstrated that the treatment system was successful in removing organic material from the effluent, with the aerobic system consistently removing more COD than the microaerophilic system.  Nitrification was observed, further confirming biological activity in the BBWTS, and as expected, this process was confined to the aerobic system only. Interestingly, a change in the feed water occurred during the 5 d SRT, which led to a sustained increase in influent and effluent COD and the loss of nitrification half way through this experimental run.  However, these changes did not significantly impact COD removal efficiencies, influent sterols concentrations, or sterols removals or metabolite generation.  121  Overall, the BBWTS was considered to be biologically active and relatively robust.  As such, the study objectives can be evaluated knowing that any changes in sterols chemistry and genomic responses were not likely to be resultant of operational upsets.  Objective 1:  Assess whether phytosterol biotransformation products can be identified in biologically treated pulp mill effluents.  The results of the organic analyses indicated that phytosterol biotransformation products were identified in biologically treated pulp mill effluents.  This was evidenced by net generation of small amounts of cholesterol, dihydrocholesterol, desmosterol and estrone throughout the experiments. Generation of estrone was an important finding, as it is a natural estrogen belonging to the animal kingdom and also acts as a potent EDC.  The fact that estrone was found to be produced in the BBWTS shows that animal steroids can be generated during biological treatment of pulp mill effluents.  This finding warrants investigation into full scale systems. Further work should include testing of a broad range of animal steroids, given the complexity of steroid biosynthesis pathways and the number of intermediaries that may potentially be involved.  Also, future work should maintain testing for androgenic as well as estrogenic compounds, as both of these types of responses have been observed from exposure of fishes to biologically treated PMEs.  Further, it may be of use to approach this work from an analytical chemistry perspective, evaluating GC-MS data in full scan mode, in attempt to identify metabolites of interest.  Although sterols biotransformation products were identified in this study, they were detected in very small concentrations, indicating that biotransformation contributed a very small portion to the total sterols removal.  Mass balance analysis indicated 53 and 37 % of the influent sterols were biodegraded or biotransformed in the aerobic and microaerophilic systems respectively.  However, the concentrations of sterols biotransformation products measured in the final effluents and WAS accounted for only 0.1 % of influent sterols under both aerobic and microaerophilic conditions.  In this regard, it is recognized that the recommendation above to analyze GC-MS data in full scan mode would be limited by the combination of low concentrations of metabolites in effluents and the sensitivity of the instrument, when operating in this mode.   122  Objective 2:  Determine if wastewater treatment conditions affect the levels of phytosterols and their biotransformation products in treated effluents.  The results showed that both DO and SRT had an effect on the levels of phytosterols and their biotransformation products in treated effluents.  Sterols removal was much higher in the aerobic system compared to the microaerophilic system, and as a result, microaerophilic effluent sterols concentrations were one to two orders of magnitude higher than those of aerobic effluent.  The aerobic system was able to obtain close to 100 % sterols removal during both the 20 d SRT and 5 d SRT experimental runs, resulting in total sterols concentrations in final effluent of 12.0 and 51.1 μg/L on average for the two runs respectively.  Comparatively, the microaerophilic system produced 50 % sterols removal during the 20 d SRT run and 69 % removal during the 5 d SRT run, resulting in total sterols concentrations in final effluent of 528 and 474 μg/L on average for the two runs respectively.  The differences between the aerobic and microaerophilic systems, during the 20 d SRT run, were significant in terms of β-sitosterol removal.  Differences in β-sitosterol removal were also significant between the experimental runs in the microaerophilic system.  This indicates that fully aerated conditions led to significantly greater sterols removal, and hence lower final effluent sterols concentrations, than did oxygen-limited conditions.  Further, the results show that under oxygen-limited conditions, short SRTs led to greater sterols removal, and hence lower final effluent sterols concentrations, than did long SRTs.  The aerobic system proved to be especially robust in terms of sterols removal, as the change from a long to a short SRT did not significantly impact sterols removal.  In the microaerophilic system on the other hand, the longer SRT was associated with significantly less sterols removal than the short SRT.  In fact, the microaerophilic system operating under a 20 d SRT led to the lowest sterols removal rates and the highest effluent sterols concentrations of all the experimental conditions tested.  The effluent generated under these conditions also exhibited the highest effluent TSS observed during the study.  Levels of suspended solids and sterols are correlated, as the phase- separation analysis indicated that the majority of sterols throughout all process streams were associated with the particulate phase.  Hence, poor settleability and high effluent TSS was associated with the high sterols concentrations observed in microaerophilic effluent treated with a 20 d SRT.   123 In terms of sterols biotransformation, twice as much metabolite production was found to have occurred during the 5 d SRT than the 20 d SRT for both DO conditions, based on the concentrations of biotransformation products measured in the effluents.  This may suggest that more sterols-degrading bacteria were present under short SRT conditions, compared to longer SRTs.  The most ecologically significant biotransformation product identified - estrone - was only produced under aerobic conditions, but was generated at both SRTs.  This is an important point because estrone, like many animal steroids, contain more oxygen molecules than do sterols; therefore, systems with low oxygen availability are not likely to generate these compounds.  Future work on this subject would benefit from spiking experiments in which sterols are added to the influent.  Such experiments would be useful for two key reasons.  First, increased influent sterols may increase the concentrations of biotransformation products in final effluents, making them easier to detect.  This may increase the portion of samples in which biotransformation products are detected, and may lead to the identification of other metabolites not measured in this study. Increasing the concentrations of metabolites would also help identify unknown compounds when operating a GC-MS in full scan mode.  Second, by measuring biotransformation products before and after spiking, the source of the biotransformation products can be confirmed.  For example, if increasing sterols levels in influent leads to increased concentrations of biotransformation products in effluent, a stronger case can be made for the occurrence of the sterols biotransformation process. At present, it can only be assumed that the metabolites identified in this study were generated from sterols.  Objective 3:  Assess the genomic response in fish exposed to biologically treated pulp mill effluents using a range of wastewater treatment conditions.  Both estrogenic and thyroid responses were identified in fish exposed to PME treated using the BBWTS, through the use of genomic analysis.  A common response pattern was witnessed between the estrogenic biomarkers VTG1, VEPB and VEPG.  Significant up-regulation of these three genes was demonstrated in test fish exposed to microaerophilic effluent generated with a 5 d SRT at 10 % concentration, compared to control fish.  The fourth estrogen-responsive gene tested, NME, demonstrated a much more sensitive response from exposure to the test effluents, as fish exposed to each of the four effluents demonstrated significant change in NME expression compared to  124 control fish.  This includes microaerophilic and aerobic effluents generated with both a 5 d SRT and a 20 d SRT at 10 % concentration, as well as the microaerophilic effluent generated with a 20 d SRT at 1 % concentration.  The thyroid response was evidenced by significant changes in THRa expression from exposure to microaerophilic effluents generated under both SRTs at 10 % concentration, as well as the microaerophilic effluent generated with a 5 d SRT at 1 % concentration.  Analysis of the AR gene did not indicate any significant androgenic responses.  However, anti- estrogenic effects were observed through suppression, or down-regulation, of all four of the estrogenic genes.  Although the down-regulation response was witnessed for each of these genes, it was only significant for NME, and only as a result of exposure to effluents generated under 20 d SRT.  A significant down-regulation response was observed in fish exposed to aerobic effluent at 10 % concentration and microaerophilic effluent, at both 10 % and 1 % concentrations.  Future toxicogenomic analysis would benefit from the use of microarray technology to examine the full suite of endocrine related genes.  This would provide much more detail than obtained in the present study, as conclusions drawn from the analysis of a few select genes are limited.  Objective 4:    Determine whether a relationship exists between the genomic response in exposed fish and (a) the concentrations of phytosterols and their biotransformation products in test effluents, or (b) the wastewater treatment conditions used to produce the test effluents.  Some commonalities were found between the genomic response, wastewater treatment conditions and concentrations of phytosterols and their biotransformation products, although some of these relationships were more clear than others.  Overall, the strongest, most consistent genomic response was evidenced in fish exposed to microaerophilic effluent treated using a 5 d SRT.   There were also some significant responses from microaerophilic effluent generated using a 20 d SRT.  Both these effluents were able to generate a significant response at concentrations of both 10 % and 1 %, although significant responses were much more common from exposure to 10 % effluent. Comparatively, exposure to aerobic effluents generated a significant response in only the most sensitive gene evaluated (NME) and only at 10 % effluent concentration.  These results indicate that  125 exposure to effluents generated under oxygen-limited conditions is much more likely to elicit genomic-level endocrine disruption than is exposure to effluents generated under fully aerated conditions.  However, it is also evident that fully aerated effluents are not exempt from causing endocrine disruption; rather, genomic-level endocrine disruption can be detected from exposure to aerobically treated effluents when using the most sensitive biomarkers.  Interestingly, microaerophilic effluents generated using a 5 d SRT elicited more genomic responses than did effluents generated using a 20 d SRT even though sterols concentrations were higher in the effluent generated using a 20 d SRT.  This indicates that sterols concentrations alone are not indicative of the level of endocrine disruption that may be expected from PME exposure.  There were twice as much sterols biotransformation products detected in the effluents generated with a 5 d SRT compared to those generated with a 20 d SRT, however, in the case of the microaerophilic effluents, the total concentrations of metabolites measured was only 1.05 and 0.580 μg/L during the two experimental periods respectively.  Further, the compounds identified as sterols biotransformation products in these effluents were chole, desmo and DHC, and these are not expected to elicit a stronger endocrine disruption effect than β-sito, which was present in concentrations two to three magnitudes higher.  Therefore, the higher level of measured sterols biotransformation products in microaerophilic effluent generated with a 5 d SRT is not considered to be the cause of the greater genomic response observed from exposure to this effluent.  As such, it can be concluded that SRT appeared to be a driver of endocrine disruption, with a stronger genomic response observed as a result of shorter SRT.  There also appeared to be a pattern between up- and down-regulation of endocrine genes and SRT. Up-regulation was observed for all genes tested from exposure to effluents generated using a 5 d SRT, except for AR.  Comparatively, down-regulation was witnessed for the most sensitive genes (NME and THRa) from exposure to effluents generated using a 20 d SRT.  Hence, it appears that short SRTs led to induction of endocrine genes, whereas long SRTs led to suppression of endocrine genes.  Future studies would be required to confirm that the genomic responses were a result of the specific wastewater treatment conditions identified in this work.  The PME exposure literature shows that there have been extensive challenges in demonstrating reliable results between studies.  This  126 suggests that rigorous testing is required in order to extrapolate findings of any one study to another. Hence, additional work to confirm the effects of wastewater treatment SRT and DO on endocrine disruption effects would be useful in strengthening the conclusions of this study.  127 REFERENCES Agilent Technologies (2006) Agilent RNA 6000 Nano Kit Guide. G2938-90034.  Waldbronn, Germany. Ahmad, S., Garg, S.K. and Johri, B.N. (1992) Biotransformation of sterols selective cleavage of the side-chain. Biotechnology Advances, 10, 1-67. Akihisa, T., Kokke, W.C.M.C. and Tamura, T. (1991) Naturally occuring sterols related comopunds from plants. In: Patterson, G.W. & Nes, W.D. (Eds), Physiology and Biochemistry of Sterols. American Oil Chemists' Society, Champaign, Illinios. American Public Health Association, American Water Works Association, Water Environment Federation (1998) Standard Methods for the Examiniation of Water and Wastewater. American Public Health Association, Washington, DC. Andersson, T., Forlin, L., Hardig, J. and Larsson, A. (1988) Physiological Disturbances in Fish Living in Coastal Water Polluted with Bleached Kraft Pulp-Mill Effluents. Canadian Journal of Fisheries and Aquatic Sciences, 45, 1525-1536. Benninghoff, A.D. and Williams, D.E. (2008) Identification of a transcriptional fingerprint of estrogen exposure in rainbow trout liver. Toxicological Sciences, 101, 65-80. Bulich, A.A. and Isenberg, D.L. (1981) Use of the luminescent bacterial system for the rapid assessment of aquatic toxicity. ISA Transactions, 20, 29-33. Carson, J.D., Jenkins, R.L., Wilson, E.M., Howell, W.M. and Moore, R. (2008) Naturally occurring progesterone in loblolly pine (Pinus taeda L.): A major steroid precursor of environmental androgens. Environmental Toxicology and Chemistry, 27, 1273-1278. Christianson-Heiska, I. and Isomaa, B. (2008) The use of primary hepatocytes from brown trout (Salmo trutta lacustris) and the fish cell lines RTH-149 and ZF-L for in vitro screening of (anti)estrogenic activity of wood extractives. Toxicology in Vitro, 22, 589-597. Christianson-Heiska, I., Smeds, P., Granholm, N., Bergelin, E. and Isomaa, B. (2007) Endocrine modulating actions of a phytosterol mixture and its oxidation products in zebrafish (Danio rerio). Comparative Biochemistry and Physiology C-Toxicology & Pharmacology, 145, 518-527. Cody, R.P. and Bortone, S.A. (1997) Masculinization of mosquitofish as an indicator of exposure to kraft mill effluent. Bulletin of environmental contamination and toxicology, 58, 429-436. Cook, D., LaFleur, L., Parrish, A., Jones, J. and Hoy, D. (1997) Characterization of plant sterols from 22 US pulp and paper mills. Water, Science and Technology, 35, 297-303. Conner, A.H., Nagaoka, M., Rowe, J.W. and Perlman, D. (1976) Microbial Conversion of Tall Oil Sterols to C19 Steroids. Applied and Environmental Microbiology, 32, 310-311.  128 Denslow, N.D., Bowman, C.J., Ferguson, R.J., Lee, H.S., Hemmer, M.J. and Folmar, L.C. (2001) Induction of gene expression in sheepshead minnows (Cyprinodon variegatus) treated with 17 beta-estradiol, diethylstilbestrol, or ethinylestradiol: The use of mRNA fingerprints as an indicator of gene regulation. General and Comparative Endocrinology, 121, 250-260. Denton, T.E., Howell, W.M., Allison, J.J., Mccollum, J. and Marks, B. (1985) Masculinization of female mosquitofish by exposure to plant sterols and Mycobacterium-smegmatis. Bulletin of Environmental Contamination and Toxicology, 35, 627-632. De Silva, S.S. and Anderson, T.A. (1995) Fish Nutrition in Aquaculture, Chapman & Hall Aquaculture Series 1. Chapman & Hall, London. Dias, A.C.P., Fernandes, P., Cabral, J.M.S. and Pinheiro, H.M. (2002) Isolation of a biodegradable sterol-rich fraction from industrial wastes. Bioresource Technology, 82, 253-260. Durhan, E.J., Lambright, C., Wilson, V., Butterworth, B.C., Kuehl, D.W., Orlando, E.F., Guillette, L.J.,Jr, Gray, L.E. and Ankley, G.T. (2002) Evaluation of androstenedione as an androgenic component of river water downstream of a pulp and paper mill effluent. Environmental Toxicology and Chemistry, 21, 1973-1976. Ellis, R.J., Van Den Heuvel, M.R., Bandelj, E., Smith, M.A., McCarthy, L.H., Stuthridge, T.R. and Dietrich, D.R. (2003) In vivo and in vitro assessment of the androgenic potential of a pulp and paper mill effluent. Environmental Toxicology and Chemistry, 22, 1448-1456. Environment Canada (2007) Biological Test Method: Acute Lethality Test Using Rainbow Trout. EPS 1/RM/9. Environment Canada (2006a) Standard Operating Procedure for the Extraction of Total RNA from Tissues Stored in RNALaterTM. RNAEXTRACT2.0SOP.  Pacific Environmental Science Center, North Vancouver, BC. Environment Canada (2006b) Protocol for Cloning from Excised Bands.  Pacific Environmental Science Center, North Vancouver, BC. Environment Canada. (2005a) Study Summary (2002-2004): Effect of effluent from nine pulp mills in British Columbia on underyearling rainbow trout.  Pacific Environmental Science Center, North Vancouver. Environment Canada (2005b) Sterols: Gas Chromatography / Mass Spectrometry V1.1.  Pacific Environmental Science Center, North Vancouver, BC. Environment Canada (1992) Biological Test Method: Toxicity Test Using Luminescent Bacteria. EPS 1/RM/24.  Pacific Environmental Science Center, North Vancouver, BC. Fardim, P. and Duran, N. (2003) Modification of fibre surfaces during pulping and refining as analysed by SEM, XPS and ToF-SIMS. Colloids and Surfaces A-Physicochemical and Engineering Aspects, 223, 263-276.  129 Fengel, D. and Wegener, G. (1983) Wood: Chemistry, Ultrastructure, Reactions. Walter de Gruyter, Berlin. Fentress, J.A., Steele, S.L., Bart, H.L.,Jr and Cheek, A.O. (2006) Reproductive Disruption in Wild Longear Sunfish (Lepomis megalotis) Exposed to Kraft Mill Effluent. Environmental Health Perspectives, 114, 40-45. Fernandes, P., Cruz, A., Angelova, B., Pinheiro, H.M. and Cabral, J.M.S. (2003) Microbial conversion of steroid compounds: recent developments. Enzyme and microbial technology, 32, 688- 705. Fernandez, M.P., Ikonomou, M.G. and Buchanan, I. (2007) An assessment of estrogenic organic contaminants in Canadian wastewaters. Science of the Total Environment, 373, 250-269. Filby, A.L., Thorpe, K.L., Maack, G. and Tyler, C.R. (2007) Gene expression profiles revealing the mechanisms of anti-androgen-and estrogen-induced feminization in fish. Aquatic Toxicology, 81, 219-231. Flouriot, G., Pakdel, F., Ducouret, B. and Valotaire, Y. (1995) Influence of Xenobiotics on Rainbow-Trout Liver Estrogen-Receptor and Vitellogenin Gene-Expression. Journal of Molecular Endocrinology, 15, 143-151. Francois, E., Wang, D.Y., Fulthorpe, R., Liss, S.N. and Edwards, E.A. (2003) DNA microarrays for detecting endocrine-disrupting compounds. Biotechnology Advances, 22, 17-26. Gilman, C.I., Leusch, F.D., Carl Breckenridge, W. and MacLatchy, D.L. (2003) Effects of a phytosterol mixture on male fish plasma lipoprotein fractions and testis P450scc activity. General and Comparative Endocrinology, 130, 172-184. Gunnarsson, L., Kristiansson, E., Forlin, L., Nerman, O. and Larsson, D.G.J. (2007) Sensitive and robust gene expression changes in fish exposed to estrogen - a microarray approach. BMC Genomics, 8, 149. Hewitt, L.M., Kovacs, T.G., Dube, M.G., MacLatchy, D.L., Martel, P.H., McMaster, M.E., Paice, M.G., Parrott, J.L., Van den Heuvel, M.R. and Van der Kraak, G.J. (2008) Altered reproduction in fish exposed to pulp and paper mill effluents: Roles of individual compounds and mill operating conditions. Environmental Toxicology and Chemistry, 27, 682-697. Hewitt, L.M., Parrott, J.L. and McMaster, M.E. (2006) A decade of research on the environmental impacts of pulp and paper mill effluents in Canada: Sources and characteristics of bioactive substances. Journal of Toxicology and Environmental Health-Part B-Critical Reviews, 9, 341-356. Hewitt, L.M. and Marvin, C.H. (2005) Analytical methods in environmental effects-directed investigations of effluents. Mutation Research, 589, 208-232. Hocking, M.B. (1998) Handbook of Chemical Technology and Pollution Control. Academic Press, Toronto.  130 Honkanen, J.O., Kostamo, A. and Kukkonen, J.V.K. (2005) Toxicity of a phytosterol mixture to grayling (Thymallus thymallus) during early developmental stages. Archives of Environmental Contamination and Toxicology, 48, 391-396. Howell, W.M., Black, D.A. and Bortone, S.A. (1980) Abnormal Expression of Secondary Sex Characters in a Population of Mosquitofish, Gambusia-Affinis-Holbrooki - Evidence for Environmentally-Induced Masculinization. Copeia, 676-681. Howell, W.M. and Denton, T.E. (1989) Gonopodial morphogenesis in female mosquitofish, Gambusia affinis affinis , masculinized by exposure to degradation products from plant sterols. Environmental Biology of Fishes, 24, 43-51. Iguchi, T., Katsu, Y., Urushitani, H., Lange, A. and Tyler, C.R. (2007) Developmental reproductive effects of exposure to pharmaceutical steroids in the aquatic environment: Studies on mosquitofish (Gambusia affinis affinis), roach (Rutilus rutilus) and medaka (Oryzias latipes). Journal of Marine Science and Technology-Taiwan, 15, 29-36. Imai, S., Koyama, J. and Fujii, K. (2007) Effects of estrone on full life cycle of Java medaka (Oryzias javanicus), a new marine test fish. Environmental toxicology and chemistry / SETAC, 26, 726-731. Invitrogen Life Technologies (2004) SuperScriptTM III First-Strand Synthesis SuperMix. 18057N. Jenkins, R., Angus, R.A., McNatt, H., Howell, W.M., Kemppainen, J.A., Kirk, M. and Wilson, E.M. (2001) Identification of androstenedione in a river containing paper mill effluent. Environmental Toxicology and Chemistry, 20, 1325-1331. Jenkins, R.L., Wilson, E.M., Angus, R.A., Howell, W.M. and Kirk, M. (2003) Androstenedione and progesterone in the sediment of a river receiving paper mill effluent. Toxicological Sciences, 73, 53-59. Kampfer, P., Erhart, R., Beimfohr, C., Bohringer, J., Wagner, M. and Amann, R. (1996) Characterization of bacterial communities from activated sludge: Culture-dependent numerical identification versus in situ identification using group- and genus-specific rRNA-targeted oligonucleotide probes. Microbial Ecology, 32, 101-121. Karels, A., Markkula, E. and Oikari, A. (2001) Reproductive, biochemical, physiological, and population responses in perch (Perca fluviatilis L.) and roach (Rutilus rutilus L.) downstream of two elemental chlorine-free pulp and paper mills. Environmental Toxicology and Chemistry, 20, 1517-1527. Karels, A.E., Soimasuo, M., Lappivaara, J., Leppaenen, H., Aaltonen, T., Mellanen, P. and Oikari, A.O.J. (1998) Effects of ECF-bleached kraft mill effluent on reproductive steroids and liver MFO activity in populations of perch and roach. Ecotoxicology, 7, 123-132. Kieslich, K. (1980) Industrial aspects of biotechnological production of steroids.  Biotechnology Letters, 2(5), 211-217.  131 Kime, D.E. (1998) Endocrine Disruption in Fish. Kluwer Academic Publishers, Boston. Kime, D.E. (1999)  A stragegy for assessing the effects of xenobiotics on fish reproduction.  The Science of the Total Environment, 225, 3-11. Kostamo, A., Holmbom, B. and Kukkonen, J.V.K. (2004) Fate of wood extractives in wastewater treatment plants at kraft pulp mills and mechanical pulp mills. Water Research, 38, 972-982. Kovacs, T., Hewitt, M., Dube, M., MacLatchy, D., Parrott, J., McMaster, M., Martel, P. and Paice, M. (2006) Towards the elimination of the changes in reproductive indicators of fish associated with exposure to pulp and paper mill effluents. In: Proceedings of the 6th International Conference on the Environmental Fate and Effects of Pulp and Paper Effluents, Vitoria, Brazil, submitted. Kovacs, T., Martel, P., Ricci, M., Michaud, J. and Voss, R. (2005) Further insights into the potential of pulp and paper mill effluents to affect fish reproduction. Journal of Toxicology and Environmental Health, Part A: Current Issues, 68, 1621-1641. Kukkonen, J., Punta, E., Koponen, P., Paranko, J., Leppaenen, H., Holopainen, I. and Hyvaerinen, H. (1999) Biomarker responses by crucian carp (Carassius carassius) living in a pond of secondary treated pulp mill effluent. Elsevier Science Ltd., Oxford. Lahdelma, I. and Oikari, A. (2006) Stratigraphy of wood-derived sterols in sediments historically contaminated by pulp and paper mill effluents. Journal of Paleolimnology, 35, 323-334. Lappivaara, J. (2001) Effects of acute handling stress on whitefish Coregonus lavaretus after prolonged exposure to biologically treated and untreated bleached kraft mill effluent. Archives of Environmental Contamination and Toxicology, 41, 55-64. Larkin, P., Folmar, L.C., Hemmer, M.J., Poston, A.J., Lee, H.S. and Denslow, N.D. (2002) Array technology as a tool to monitor exposure of fish to xenoestrogens. Marine Environmental Research, 54, 395-399. Larsson, D.G.J., Hyllner, S.J. and Haux, C. (1994) Induction of Vitelline Envelope Proteins by Estradiol-17-Beta in 10 Teleost Species. General and Comparative Endocrinology, 96, 445-450. Leanos-Castaneda, O. and Van der Kraak, G. (2007) Functional characterization of estrogen receptor subtypes, ER alpha and ER beta, mediating vitellogenin production in the liver of rainbow trout. Toxicology and Applied Pharmacology, 224, 116-125. Lehtinen, K.J., Mattsson, K., Tana, J., Engstrom, C., Lerche, O. and Hemming, J. (1999) Effects of wood-related sterols on the reproduction, egg survival, and offspring of brown trout (Salmo trutta lacustris L.). Ecotoxicology and Environmental Safety, 42, 40-49. Leusch, F.D.L. and MacLatchy, D.L. (2003) In vivo implants of beta-sitosterol cause reductions of reactive cholesterol pools in mitochondria isolated from gonads of male goldfish (Carassius auratus). General and Comparative Endocrinology, 134, 255-263.  132 Lin, K.H., Wang, W.J., Wu, Y.H. and Cheng, S.Y. (2002) Activation of antimetastatic Nm23-H1 gene expression by estrogen and its alpha-receptor. Endocrinology, 143, 467-475. Lister, A. and Van der Kraak, GJ. (2001) Endocrine Disruption: Why Is It So Complicated? Water Quality Research Journal of Canada, 36, 175-190. Lo, C., Pan, C. and Liu, W. (2002) Production of testosterone from phytosterol using a single-step microbial transformation by a mutant of Mycobacterium sp. Journal of Industrial Microbiology & Biotechnology, 28, 280-283. Lowell, R.B., Ring, B., Pastershank, G., Walker, S., Trudel, L. and Hedley, K. (2005) National assessment of pulp and paper environmental effects monitoring data: Findings from cycles 1 through 3.  National Water Research Institute Scientific Assessment Report Series No. 5. MacLatchy, D.L., Shaughnessy, K.S., Belknap, A.M., Dube, M.G. and Hewitt, M.L. (2006) Reproductive steroid responses in fish exposed to pulp mill condensates. In: Proceedings of the 6th International Conference on the Environmental Fate and Effects of Pulp and Paper Effluents, Vitoria, Brazil, submitted. Maclatchy, D.L. and Van der Kraak, G.J. (1995) The phytoestrogen beta-sitosterol alters the reproductive endocrine status of goldfish. Toxicology and Applied Pharmacology, 134, 305-312. Madyastha, K.M. and Shankar, V.N. (1994) Role of Neutral Metabolites in Microbial Conversion of 3-Beta-Acetoxy-19-Hydroxycholest-5-Ene into Estrone. Applied and Environmental Microbiology, 60, 1512-1518. Magnus, E., Hoel, H. and Norske, G.E.C. (2000) TMP wastewater treatment, including a biological high-efficiency compact reactor - Toxicity reduction and removal of extractives. Nordic Pulp & Paper Research Journal, 15, 37-45. Mahato, S.B. and Garai, S. (1997) Advances in microbial steroid biotransformation. Steroids, 62, 332- 345. Mahmood, T. and Paice, M. (2006) Aerated stabilization basin design and operating practices in the Canadian pulp and paper industry. Journal of Environmental Engineering and Science, 5, 383-395. Mahmood-Khan, Z. (2005) Occurrence and Treatment of Plant Sterols in Pulp and Paper Mill Effluents.  PhD dissertation, University of  British Columbia, Vancouver, BC. Mahmood-Khan, Z. and Hall, E.R. (2003) Occurrence and removal of plant sterols in pulp and paper mill effluents. Journal of Environmental Engineering and Science, 2, 17-26. Malaviya, A. and Gomes, J. (2008) Androstenedione production by biotransformation of phytosterols. Bioresource Technology, 99, 6725-6737. Marsheck, W.J., Kraychy, S. and Muir, R.D. (1972) Microbial Degradation of Sterols. Applied Microbiology, 23, 72-77.  133 Martel, P. Kovacs, T., and Berbue, V. (2008) The benefits of biotreatment for reducing the effects of pulp and paper mill effluents on fish reproduction in laboratory tests.  Water Quality Research Journal of Canada,43(2/3), 93-103. Martinovic, D., Blake, L.S., Durhan, E.J., Greene, K.J. and Kahl, M.D. (2008) Reproductive toxicity of vinclozolin in the fathead minnow: Confirming an anti-androgenic mode of action. Environmental Toxicology and Chemistry, 27, 478-488. Mattsson, K., Tana, J., Engstrom, C., Hemming, J. and Lehtinen, K.J. (2001) Effects of wood- related sterols on the offspring of the viviparous blenny, Zoarces viviparus L. Ecotoxicology and Environmental Safety, 49, 122-130. McLeay, D. (1987) Aquatic toxicity of pulp and paper mill effluent: A review.  Environment Canada, Ottawa. McMaster, M.E., Hewitt, L.M. and Parrott, J.L. (2006) A decade of research on the environmental impacts of pulp and paper mill effluents in Canada: Field studies and mechanistic research. Journal of Toxicology and Environmental Health-Part B-Critical Reviews, 9, 319-339. McMaster, M., Munkittrick, K., Van der Kraak, G., Flett, P. and Servos, M. (1996) Detection of steroid hormone disruptions associated with pulp mill effluent using artificial exposures of goldfish. St. Lucie Press, Delray Beach, Florida. Mellanen, P., Soimasuo, M., Holmbom, B., Oikari, A. and Santti, R. (1999) Expression of the vitellogenin gene in the liver of juvenile whitefish (Coregonus lavaretus L-s.l.) exposed to effluents from pulp and paper mills. Ecotoxicology and Environmental Safety, 43, 133-137. Mellanen, P., Petaenen, T., Lehtimaeki, J., Maekelae, S., Bylund, G., Holmbom, B., Mannila, E., Oikari, A. and Santti, R. (1996) Wood-derived estrogens: Studies in vitro with breast cancer cell lines and in vivo in trout. Toxicology and Applied Pharmacology, 136, 381-388. Miracle, A.L. and Ankley, G.T. (2005) Ecotoxicogenomics: linkages between exposure and effects in assessing risks of aquatic contaminants to fish. Reproductive Toxicology, 19, 321-326. Moncecchi, D., Nemecz, G., Schroeder, F. and Scallen, T.J. (1991) The participation of sterol carrier protein-2 (SPC-2) in cholesterol metabolism. In: Patterson, G.W. and Nes, W.D. (Eds), Physiology and Biochemistry of Sterols. American Oil Chemists' Society, Champaign, Illinios. Munkittrick, K.R., McMaster, M.E., McCarthy, L.H., Servos, M.R. and Van Der Kraak, G.J. (1998) An overview of recent studies on the potential of pulp-mill effluents to alter reproductive parameters in fish. Journal of Toxicology and Environmental Health-Part B-Critical Reviews, 1, 347- 371. Munkittrick, K.R., Vanderkraak, G.J., Mcmaster, M.E. and Portt, C.B. (1992) Response of hepatic MFO activity and plasma sex steroids to secondary-treatment of bleached kraft pulp-mill effluent and mill shutdown. Environmental Toxicology and Chemistry, 11, 1427-1439.  134 Munkittrick, K., Servos, M., Carey, J. and van der Kraak, G. (1997) Environmental impacts of pulp and paper wastewater: Evidence for a reduction in environmental effects at North American pulp mills since 1992. Forest Industry Wastewater 5:  Selected proceedings of the 5th IAWQ International Symposium on Forest Industry Wastewaters, Vancouver, BC. Nakajima, N., Fujioka, S., Tanaka, T., Takatsuto, S. and Yoshida, S. (2002) Biosynthesis of cholestanol in higher plants. Phytochemistry, 60, 275-279. Nakari, T. and Erkomaa, K. (2003) Effects of phytosterols on zebrafish reproduction in multigeneration test. Environmental Pollution, 123, 267-273. National Center for Biotechnology Information (2008) GenBank, National Library of Medicine and National Institutes of Health. Available from: http://www.ncbi.nlm.nih.gov/Genbank/index.html.  Accessed April 2008. National Institute of Standards and Technology and SEMATECH. (2006) e-Handbook of Statistical Methods, US Commerce Department, Technology Administration. Available from: http://www.itl.nist.gov/div898/handbook/.  Accessed August 2008). Oakes, K.D., Tremblay, L.A. and Van der Kraak, G.J. (2005) Short-term lab exposures of immature rainbow trout (Oncorhynchus mykiss) to sulfite and kraft pulp-mill effluents: Effects on oxidative stress and circulating sex steroids. Environmental Toxicology and Chemistry, 24, 1451- 1461. Oern, S., Svenson, A., Viktor, T., Holbech, H. and Norrgren, L. (2006) Male-biased sex ratios and vitellogenin iInduction in zebrafish exposed to effluent water from a Swedish pulp mill. Archives of Environmental Contamination and Toxicology, 51, 445-451. Orlando, E.F., Bass, D.E., Caltabiano, L.M., Davis, W.P., Gray, L.E. and Guillette, L.J. (2007) Altered development and reproduction in mosquitofish exposed to pulp and paper mill effluent in the Fenholloway River, Florida USA. Aquatic Toxicology, 84, 399-405. Orrego, R., Moraga, G., Gonzalez, M., Gavilan, J.F., Valenzuela, A., Burgos, A. and Barra, R. (2006) Pulp mill effluent effects on rainbow trout in the Biobio River, Chile. In: Proceedings of the 6th International Conference on the Environmental Fate and Effects of Pulp and Paper Effluents, Vitoria, Brazil, submitted. Osachoff, H. (2008) Toxicogenomics effects of sewage on juvenile chinook salmon (Oncorhynchus tshawytscha).  PhD dissertation, Simon Fraser University, Burnaby, BC. Parks, L.G., Lambright, C.S., Orlando, E.F., Guillette, L.J., Ankley, G.T. and Gray, L.E. (2001) Masculinization of female mosquitofish in kraft mill effluent-contaminated Fenholloway River water is associated with androgen receptor agonist activity. Toxicological Sciences, 62, 257-267. Paice, M.G., Kovacs, T., Bergernon, J. and O'Connor, B. (2003) Current status of the pure oxygen activated sludge process in Canadian mills. Tappi Journal, 2, 17-22.  135 Parrott, J.L., Wood, C.S., Boutot, P. and Dunn, S. (2004) Changes in growth, secondary sex characteristics, and reproduction of fathead minnows exposed for a life cycle to bleached sulfite mill effluent. Journal of Toxicology and Environmental Health, Part A: Current Issues, 67, 1755- 1764. Parrott, J. (2006) Overview of laboratory testing for reproductive effects of pulp mill effluents in fish. In: Proceedings of the 6th International Conference on the Environmental Fate and Effects of Pulp and Paper Effluents, Vitoria, Brazil, submitted. Parrott, J.L., McMaster, M.E. and Hewitt, L.M. (2006) A Decade of Research on the Environmental Impacts of Pulp and Paper Mill Effluents in Canada: Development and Application of Fish Bioassays. Perez, C., Falero, A., Duc, H.L., Balcinde, Y. and Hung, B.R. (2006) A very efficient bioconversion of soybean phytosterols mixtures to androstanes by mycobacteria. Journal of Industrial Microbiology & Biotechnology, 33, 719-723. Postel, E.H. (1998) NM23-NDP kinase. The International Journal of Biochemistry & Cell Biology, 30, 1291-1295. Rayner, K., Chen, Y.X., Hibbert, B., White, D., Miller, H., Postel, E.H. and O'Brien, E.R. (2008) Discovery of NM23-H2 as an estrogen receptor beta-associated protein: role in estrogen- induced gene transcription and cell migration. The Journal of Steroid Biochemistry and Molecular Biology, 108, 72-81. Rickwood, C.J., Dube, M.G., Hewitt, L.M., Kovacs, T.G. and MacLatchy, D.L. (2006a) Use of paired fathead minnow (Pimephales promelas) reproductive test. Part 2: Source identification of biological effects at a bleached kraft pulp mill. Environmental Toxicology and Chemistry, 25, 1847-1856. Rickwood, C.J., Dube, M.G., Hewitt, L.M., Parrott, J., Kovacs, T.G. and MacLatchy, D.L. (2006b) Endocrine disruption source evaluation at a bleached kraft pulp mill using pair-breeding fathead minnow (Pimephales promelas). In: Proceedings of the 6th International Conference on the Environmental Fate and Effects of Pulp and Paper Effluents, Vitoria, Brazil, submitted. Rozman, D. and Waterman, M.R. (2002) Sterol biosynthesis. In: Mason, J.I. (Ed), Genetics of Steroid Biosynthesis and Function. Taylor & Francis, London. Schefe, J.H., Lehmann, K.E., Buschmann, I.R., Unger, T. and Funke-Kaiser, H. (2006) Quantitative real-time RT-PCR data analysis: current concepts and the novel "gene expression's CT difference" formula. Journal of Molecular Medicine, 84, 901-910. Schuppler, M., Mertens, F., Schon, G. and Gobel, U.B. (1995) Molecular Characterization of Nocardioform Actinomycetes in Activated-Sludge by 16s Ribosomal-Rna Analysis. Microbiology-UK, 141, 513-521.  136 Sepulveda, M.S., Ruessler, D.S., Denslow, N.D., Holm, S.E., Schoeb, T.R. and Gross, T.S. (2001) Assessment of reproductive effects in largemouth bass (Micropterus salmoides) exposed to bleached/unbleached kraft mill effluents. Archives of Environmental Contamination and Toxicology, 41, 475-482. Servos, M.R., Bennie, D.T., Burnison, B.K., Jurkovic, A., McInnis, R., Neheli, T., Schnell, A., Seto, P., Smyth, S.A. and Ternes, T.A. (2005) Distribution of estrogens, 17 beta-estradiol and estrone, in Canadian municipal wastewater treatment plants. Science of the Total Environment, 336, 155-170. Shaughnessy, K.S., Belknap, A.M., Hewitt, L.M., Dube, M.G. and MacLatchy, D.L. (2007) Effects of kraft pulp mill condensates on plasma testosterone levels in mummichog (Fundulus heteroclitus). Ecotoxicology and Environmental Ssafety, 67, 140-148. Sih, C.J., Lee, S.S., Tsong, Y.Y., Wang, K.C. and Chang, F.N. (1965) An Efficient Synthesis of Estrone and 19-Norsteroids from Cholesterol. Journal of the American Chemical Society, 87, 2765- 2766. Soimasuo, M.R., Karels, A.E., Leppanen, H., Santti, R. and Oikari, A.O.J. (1998) Biomarker responses in whitefish (Coregonus lavaretus L. sl) experimentally exposed in a large lake receiving effluents from pulp and paper industry. Archives of Environmental Contamination and Toxicology, 34, 69-80. Stainsby, F.M., Soddell, J., Seviour, R., Upton, J. and Goodfellow, M. (2002) Dispelling the "Nocardia amarae" myth: a phylogenetic and phenotypic study of mycolic acid-containing actinomycetes isolated from activated sludge foam. Water Science and Technology, 46, 81-90. Stefanov, S., Yankov, D. and Beschkov, V. (2006) Biotransformation of phytosterols to androstenedione in two phase water-oil systems. Chemical and Biochemical Engineering Quarterly, 20, 421-427. Svoboda, J.A. (1997) Variability of metabolism and function of sterols in insects. In: Parish, J. and Nes, W.D. (Eds), Biochemistry and Function of Sterols. CRC Press, Boca Raton, FL. Szykula, J., Hebda, C., Orpiszewski, J. and Saganska, K. (1991) Microbial transformation of neutral fraction and upgraded neutral fraction of Polish tall oil. Biotechnology Letters, 13, 917-922. Tchobanoglous, G., Burton, F.L. and Stensel, H.D. (2003) Wastewater Engineering Treatment and Reuse / Metcalf and Eddy, Inc. McGraw-Hill, New York, NY. Temple, S.E., Ramsden, S.D., Haimberger, T.J., Veldhoen, K.M., Veldhoen, N.J., Carter, N.L., Roth, W.M. and Hawryshyn, C.W. (2008) Effects of exogenous thyroid hormones on visual pigment composition in coho salmon (Oncorhynchus kisutch). Journal of Experimental Biology, 211, 2134- 2143. Thain, M. and Hickam, M. (1996) Dictionary of Biology. Penguin Books, Toronto.  137 Thomas, P. (2008) The Endocrine System.  In: Di Giulio, R.T., and Hinton, D.E. (Eds), The Toxicology of Fishes. CRC Press, New York NY. Tremblay, L. and Van der Kraak, G. (1998) Use of a series of homologous in vitro and in vivo assays to evaluate the endocrine modulating actions of beta -sitosterol in rainbow trout. Aquatic Toxicology, 43, 149-162. Tremblay, L. and Van der Kraak, G. (1999) Comparison between the effects of the phytosterol beta-sitosterol and pulp and paper mill effluents on sexually immature rainbow trout. Environmental Toxicology and Chemistry, 18, 329-336. US Environmental Protection Agency. (1975) Process Design Manual: Nitrogen Control. Office of Technology Transfer, Washington, DC. US Environmental Protection Agency. (2006) Endocrine Disruptro Screening Program: Endocrine Primer, US EPA Office of Prevention, Pesticides and Toxic Substances. Available from: http://www.epa.gov/scipoly/oscpendo/edspoverview/primer.htm#2, Accessed April 2007. van den Heuvel, M.R., Ellis, R.J., Tremblay, L.A. and Stuthridge, T.R. (2002) Exposure of reproductively maturing rainbow trout to a New Zealand pulp and paper mill effluent. Ecotoxicology and Environmental Safety, 51, 65-75. Watson, S. and Miller, K. (2004) The Endocrine System. Greenwood Press, Westport, Connecticut.  Westerlund, L., Hyllner, S.J., Schopen, A. and Olsson, P.E. (2001) Expression of three vitelline envelope protein genes in Arctic char. General and Comparative Endocrinology, 122, 78-87. Wojciechowski, Z.A. (1991) Biochemistry of phytosterol conjugates. In: Patterson, G.W. & Nes, W.D. (Eds), Physiology and Biochemistry of Sterols. American Oil Chemists' Society, Champaign, Illinios. Xavier, C.R., Mosquera-Corral, A., Becerra, J., Hernandez, J.V. and Vidal, G. (2009) Activated sludge versus aerated lagoon treatment of kraft mill effluents containing beta-sitosterol and stigmasterol.  Journal of Environmental Science and Health, Part A, A44(4), in press. Young, G., Kuskabe, M., Nakamura, I., Lokman, P.M., and Goetz, F.W. (2005) Gonadal Steroidogenesis in Teleost Fish.  In: Melamed, P. and Sherwood, N. (Eds), Hormones and their Receptors in Fish Reproduction, Molecular Aspects of Fish and Marine Biology, Vol. 4.  World Scientific, Tuck Ling, Singapore.  APPENDIX A:  Distribution Plots for Sterols Chemistry A1: β-Sitosterol Q-Q Plot -2 -1 0 1 2 -1 50 0 -1 00 0 -5 00 0 Normal Q-Q Plot Theoretical Quantiles S am pl e Q ua nt ile s β-Sitosterol -1 50 0 -1 00 0 -5 00 0 S am pl e Q ua nt ile s -1 50 0 -1 00 0 -5 00 0 S am pl e Q ua nt ile s  APPENDIX A 139 A2: Campesterol Q-Q Plot -2 -1 0 1 2 -3 00 -2 50 -2 00 -1 50 -1 00 -5 0 0 Normal Q-Q Plot Theoretical Quantiles S am pl e Q ua nt ile s Campesterol -3 00 -2 50 -2 00 -1 50 -1 00 -5 0 0 S am pl e Q ua nt ile s -3 00 -2 50 -2 00 -1 50 -1 00 -5 0 0 S am pl e Q ua nt ile s  APPENDIX A 140 A3: Stigmasterol Q-Q Plot -2 -1 0 1 2 -6 0 -5 0 -4 0 -3 0 -2 0 -1 0 0 Normal Q-Q Plot Theoretical Quantiles S am pl e Q ua nt ile s Stigmasterol -6 0 -5 0 -4 0 -3 0 -2 0 -1 0 0 S am pl e Q ua nt ile s -6 0 -5 0 -4 0 -3 0 -2 0 -1 0 0 S am pl e Q ua nt ile s  APPENDIX A 141 A4: Cholesterol Q-Q Plot -2 -1 0 1 2 -6 0 -5 0 -4 0 -3 0 -2 0 -1 0 0 Normal Q-Q Plot Theoretical Quantiles S am pl e Q ua nt ile s -6 0 -5 0 -4 0 -3 0 -2 0 -1 0 0 S am pl e Q ua nt ile s Cholesterol  APPENDIX A 142 A5: Dihydrocholesterol Q-Q Plot -2 -1 0 1 2 -5 -4 -3 -2 -1 0 Normal Q-Q Plot Theoretical Quantiles S am pl e Q ua nt ile s Dihydrocholesterol -5 -4 -3 -2 -1 0 S am pl e Q ua nt ile s -5 -4 -3 -2 -1 0 S am pl e Q ua nt ile s  APPENDIX A 143 A6: Estrone Q-Q Plot -2 -1 0 1 2 0. 0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 Normal Q-Q Plot Theoretical Quantiles S am pl e Q ua nt ile s Estrone 0. 0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 S am pl e Q ua nt ile s 0. 0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 S am pl e Q ua nt ile s  APPENDIX A 144 A7: Coprostanol Q-Q Plot -2 -1 0 1 2 -1 .5 -1 .0 -0 .5 0. 0 Normal Q-Q Plot Theoretical Quantiles S am pl e Q ua nt ile s Coprostanol -1 .5 -1 .0 -0 .5 0. 0 S am pl e Q ua nt ile s -1 .5 -1 .0 -0 .5 0. 0 S am pl e Q ua nt ile s  APPENDIX A 145 A8: Desmosterol Q-Q Plot -2 -1 0 1 2 0. 0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 Normal Q-Q Plot Theoretical Quantiles S am pl e Q ua nt ile s Desmosterol 0. 0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 S am pl e Q ua nt ile s 0. 0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 S am pl e Q ua nt ile s   APPENDIX B:  Animal Care Approval     Signature removed as per privacy requirements  APPENDIX C:  Data Tables for BBWTS Operation and Sterols Chemistry C1: Temperature and Dissolved Oxygen Day Temperature ( oC) Dissolved Oxygen (mg/L) AB MP AB MP SET UP - 11 d SRT 2 35.5 33 5.0 0.1 3 36 34 6.0 0.1 6 30 35 6.0 0.8 7 35 32 5.6 0.2 10 34 34 3.6 0.1 12 35 35 3.2 0.2 13 35 35 3.4 0.1 15 35 35 2.5 0.1 16 34 34.5 5.5 0.2 22 34 34 4.8 0.5 23 35 33 4.7 0.2 24 32 35 4.9 0.2 27 35 21 4.2 0.4 RUN 1 - 20 d SRT 28 37 27 4.8 0.4 29 35 36 5.1 0.4 30 35 35 4.2 0.2 31 37 31 4.8 0.3 33 35 34 4.5 0.2 34 37 34.5 5.8 0.3 35 35.5 35 5.3 0.2 36 35 36 4.0 0.1 37 36 35 3.8 0.2 38 36.5 36 2.3 0.2 39 37 36 5.0 0.1 40 37 36 5.5 0.2 41 36 35 3.1 0.1 42 37 35 4.4 0.1 43 34.5 36 5.1 0.3 44 34 36.5 4.6 0.3 45 34 36 4.9 0.3 48 36.5 33 2.3 0.2 49 32 35 5.8 0.2 50 34 35 6.1 0.3 51 33 34 4.3 0.2 52 34 35 4.2 0.5 53 33 34 5.7 0.2 55 29 34.5 5.5 0.2 56 29 35 4.6 0.2 57 35 31 4.2 0.2 58 29 34 5.7 0.4 59 32 33 5.3 0.5 APPENDIX C 148 C1: Temperature and Dissolved Oxygen, continued  Day Temperature ( oC) Dissolved Oxygen (mg/L) AB MP AB MP RUN 1 - 20 d SRT (continued) 60 32.5 32 6 0.3 61 28 34 5.7 0.5 62 36.5 25 3.8 0.3 63 33 29 4.8 0.2 64 24 35 6.7 0.1 65 29 32.5 5.3 0.2 66 34 28 5.1 0.2 68 22 34 3.8 0.2 69 32.5 38.5 4.6 0.1 70 32 30 6.1 0.1 72 31 30 4.5 0.3 73 30 29.5 5.0 0.2 74 35 35 3.5 0.1 75 36 35 4.6 0.2 76 35 35 3.4 0.1 77 36 35 4.8 0.2 78 36 35 5.0 0.1 79 35 35 4.5 0.2 80 36 35 4.9 0.2 81 36 35 4.8 0.2 82 36 35 5.0 0.2 83 36 35 4.5 0.2 84 36 34 5.1 0.1 85 35 35 4.8 0.2 86 36 35 4.9 0.3 87 36 35 3.6 0.2 88 36 35 1.5 0.6 89 36 35 5.3 0.2 90 35.5 35 4.5 0.3 91 36 35 4.8 0.1 92 36 35 5.2 0.1 93 36 35 5.7 0.2 94 36 35 5.6 0.2 95 36 35 5.5 0.2 96 36 35 5.2 0.1 97 35.5 35 5.4 0.1 98 35.5 35 4.5 0.1 99 35.5 35 4.6 0.1 100 35 35 0.9 0.2 101 35.5 35 5.5 0.2 102 35 35 4.8 0.2 103 35 34 5.5 0.1 104 37 36 5.1 0.2 105 37 36 5.1 0.1 APPENDIX C 149 C1: Temperature and Dissolved Oxygen, continued  Day Temperature ( oC) Dissolved Oxygen (mg/L) AB MP AB MP RUN 1 - 20 d SRT (continued) 106 36 35 5.4 0.1 RUN 2 - 5 d SRT 108 34 31 5.1 0.2 109 34.5 34 4 0.2 110 35 34 3.9 0.1 111 34 34 4.7 0.1 112 34.5 34 4.4 0.1 113 32 34 5 0.1 114 35 34 4.5 0.2 115 34.5 34 4 0.1 116 35 34.5 4.3 0.1 117 35 34 3.9 0.1 118 35 34.5 4.5 0.2 119 35 34 4.7 0.1 120 34.5 34 4.5 0.2 121 35 34 4.3 0.2 122 35 34 4.6 0.2 123 35 34 4.6 0.1 124 35 34.5 4.8 0.2 125 35 34.5 4.4 0.2 126 35 34 4.0 0.2 127 35 34.5 3.2 0.1 128 35 34.5 3.6 0.2 129 35 34.5 3.3 0.2 130 35 34 4.6 0.1 131 35.5 35 4.8 0.1 132 35.5 35 4.5 0.1 133 35 34.5 4.1 0.1 134 35 34.5 4 0.1 135 35 34.5 3.8 0.2 136 35.5 34.5 3.4 0.2 137 35.5 34.5 3.1 0.1 138 36 35 3.8 0.1 139 35.5 35 3.4 0.2 140 35.5 35 4.5 0.2 141 35.5 35 4.5 0.1 142 35.5 34.5 4.8 0.1 143 35.5 35 3.6 0.1 144 35.5 35 4.0 0.1 145 35.5 35 2.8 0.1 146 35 34.5 3.3 0.2 147 35.5 35 4.1 0.6 148 36 35 4.5 0.1 149 36 35 3.8 0.2 APPENDIX C 150 C1: Temperature and Dissolved Oxygen, continued  Day Temperature ( oC) Dissolved Oxygen (mg/L) AB MP AB MP RUN 2 - 5 d SRT (continued) 150 36 35 4.6 0.2 151 36 35 4.3 0.3 152 35.5 34.5 4.4 0.2 153 35.5 34.5 4.3 0.2 154 36 35 3.5 0.2 155 35.5 34.5 3.6 0.3 156 35.5 34.5 4.1 0.3 157 36 35 4.1 0.2 158 35.5 34.5 4.8 0.2 159 36 35 4.6 0.1 160 35.5 34.5 5.0 0.1 161 36 34.5 5.5 0.1 162 36 34.5 5.7 0.1 163 35 34 4.0 0.1 164 35.5 34.5 4.4 0.1 165 36 35 4.7 0.1   APPENDIX C 151 C2: HRT and SRT Day HRT (hr) SRT (d) AB MP AB MP SET UP - 11 d SRT 2 - - 11.0 11.0 6 - - 11.0 11.0 10 - - 11.0 11.0 12 - - 11.0 7.1 15 - - 11.0 2.1 17 - - 11.0 10.0 20 - - 11.0 7.4 22 - - 11.0 10.0 24 - - 11.0 - 27 - - 11.0 11.0 RUN 1 - 20 d SRT 30 12.0 11.1 - - 31 - - 37.9 9.7 34 11.6 10.5 30.8 21.6 35 12.7 11.6 - - 36 - 11.5 - - 38 12.1 12.1 26.0 15.5 41 12.1 12.0 21.9 16.8 44 11.8 11.3 - 16.4 48 12.6 12.3 8.9 17.6 51 12.9 12.4 20.0 14.7 56 - 12.3 20.0 18.6 58 - - 20.0 20.1 62 11.7 11.6 20.0 14.3 63 13.4 12.6 - - 64 14.5 13.8 - - 65 11.5 10.7 20.0 12.2 69 12.1 11.2 20.0 10.2 70 - 11.0 - - 72 13.4 11.5 20.0 20.0 74 11.8 14.4 - - 75 12.0 13.5 - - 76 12.8 12.2 20.0 14.7 77 10.3 11.8 - - 79 - - 20.0 19.9 80 11.6 12.1 - - 81 12.0 11.5 - - 82 12.0 11.3 - - 83 - - 20.0 20.0 84 11.4 12.0 - - 85 11.5 11.5 - - 86 11.5 11.9 20.0 20.0 87 12.0 12.0 - - 90 - 11.8 20.0 13.2 APPENDIX C 152 C2: HRT and SRT, continued  Day HRT (hr) SRT (d) AB MP AB MP RUN 1 - 20 d SRT (continued) 91 12.3 - - - 93 11.9 11.4 20.0 11.9 94 12.1 11.6 - - 96 12.7 11.7 20.0 13.7 97 10.2 11.0 - - 99 10.4 - - - 100 10.6 11.7 20.0 11.2 101 11.9 11.3 - - 102 11.4 11.6 - - 103 12.3 - - - 104 12.6 11.4 20.0 12.3 105 12.0 11.4 - - 106 11.6 11.5 20.0 20.0 RUN 2 - 5 d SRT 109 11.8 10.4 15.6 16.5 110 11.6 11.8 - - 111 - 11.6 5.0 5.0 112 11.7 12.3 - - 113 11.9 12.2 5.0 5.0 114 10.4 11.8 - - 116 12.2 11.5 - - 117 11.6 11.5 5.0 5.0 119 11.4 11.7 - - 120 10.9 11.4 5.0 5.0 121 11.1 11.3 - - 122 10.8 11.3 - - 123 11.6 11.6 - - 124 11.4 11.3 5.0 5.0 126 11.9 11.6 - - 127 11.0 11.3 5.0 5.0 128 11.4 11.4 - - 129 11.7 11.6 - - 130 - - 5.0 5.1 131 11.5 11.8 - - 132 11.2 11.6 - - 133 11.1 11.5 5.0 5.0 138 12.5 12.1 5.0 5.0 139 11.2 11.2 - - 141 11.4 11.2 - - 142 11.2 11.2 5.0 5.0 143 11.4 11.1 - - 144 - 11.3 - - 145 11.2 11.6 5.0 5.0  APPENDIX C 153 C2: HRT and SRT, continued  Day HRT (hr) SRT (d) AB MP AB MP RUN 2 - 5 d SRT (continued) 147 11.8 11.8 - - 148 11.8 11.9 5.0 5.0 149 12.0 11.8 - - 150 12.1 12.0 - - 151 10.9 11.3 - - 152 11.3 11.7 5.0 5.0 154 11.7 11.1 - - 155 11.7 11.0 5.0 5.0 156 12.1 11.3 - - 157 11.7 11.6 - - 158 11.4 11.2 - - 161 12.6 12.2 5.0 5.0 163 13.1 13.0 5.0 5.0 164 11.8 12.2 - - 165 12.0 11.7 5.0 5.0     APPENDIX C 154 C3: Suspended Solids Day MLSS (mg/L) TSS (mg/L) Aerobic Reactor Microaerophilic Reactor Influent Aerobic Effluent Microaerophilic Effluent SET UP - 11 d SRT 2 5440 5900 112 104 158 6 7700 5920 1991 110 118 10 6180 3300 438 107 142 12 6260 3000 324 73 203 15 4220 2360 1145 56 550 17 5700 2730 152 129 124 20 4580 2250 587 135 141 22 5320 2720 202 132 120 24 5260 - 210 91 64 27 4880 3300 93 100 83 RUN 1 - 20 d SRT 31 4400 2020 369 58 104 35 5980 3420 162 97 79 38 6600 4110 2269 127 133 41 8720 3890 1071 199 116 44 - 3600 80 - 110 48 7120 3210 4022 400 91 51 7280 3780 302 60 129 56 3600 3350 156 38 90 58 5990 3810 192 28 95 62 5040 4040 1150 30 141 65 5100 3600 269 32 148 69 4280 2380 2079 31 117 72 6460 2700 130 26 46 76 7580 2320 94 30 79 79 4380 2940 479 14 86 83 4740 2360 78 30 54 86 4980 2300 112 29 62 90 5660 2140 79 27 94 93 5980 2060 195 23 123 96 5300 1460 76 21 77 100 5690 1330 173 28 89 104 6580 1640 125 25 96 106 5560 1420 82 21 59  APPENDIX C 155 C3: Suspended Solids, continued  Day MLSS (mg/L) TSS (mg/L) Aerobic Reactor Microaerophilic Reactor Influent Aerobic Effluent Microaerophilic Effluent RUN 2 - 5 d SRT 109 3280 2080 196 105 69 111 4720 2280 330 62 57 113 5040 1220 632 18 36 117 3440 1680 931 25 46 120 2600 1920 352 26 55 124 2200 1500 467 23 81 127 2720 1440 171 25 91 130 2460 1520 406 50 111 133 2180 1240 202 32 64 138 1620 1340 1014 39 94 142 1540 1480 373 41 128 145 2060 1280 258 111 114 148 2070 1067 417 50 74 152 1746 880 716 55 82 155 2060 1013 375 48 79 161 1560 1070 200 43 74 163 1280 1110 364 51 56 165 1040 880 596 52 73  APPENDIX C 156 C4: Chemical Oxygen Demand Day Total COD Dissolved COD Influent Aerobic  Effluent Micro- aerophilic Effluent Influent Aerobic  Effluent Micro- aerophilic Effluent Set Up - 11 d SRT 2 1462 690 893 1201 560 657 6 2267 690 700 1086 579 541 10 1736 775 1004 1134 594 811 12 1664 705 1028 1110 608 743 15 2035 700 1418 1095 633 869 17 1476 806 1110 1312 657 840 22 1476 681 1030 1192 546 849 24 1539 724 960 1307 536 820 27 1037 647 1037 854 546 1442 Run 1 - 20 d SRT 31 1666 507 695 965 420 565 35 1635 661 980 1375 628 845 38 2271 710 1163 1134 589 1033 41 2011 1018 1100 1264 606 878 44 1486  1225 1307  984 48   1158 1394 1331 666 1194 51 1997 767 1322 1592 748 1076 56 1785 734 1177 1592 695 946 58 1741 757 2028 1335 652 1125 62 2118 743 1311 1197 676 1029 65 1736 815 1330 1321 1006 1125 69 2433 776 1242 1173 767 901 72 1792 753 1092  576 1020 76 1925 633 1139 1677 585 915 79 1878 514 991 1486 504 1459 83 1763 509 939 1467 478 867 86 1328 704 906 1249 521 822 90 1526 541 1035 1368 476 864 93 1724 630 1109 1635 541 867 96 1467 521 971 1368 471 827 100 1852 605 1119 1427 516 921 104 3028 578 1040 2998 516 985 106 1714 551 2087 1536 506 946  APPENDIX C 157 C4: Chemical Oxygen Demand, continued   Day Total COD Dissolved COD Influent Aerobic  Effluent Micro- aerophilic Effluent Influent Aerobic  Effluent Micro- aerophilic Effluent Run 2 - 5 d SRT 109 1983 736 1286 1613 555 992 111 2106 641 1108 1519 612 963 113 2694 631 1063 1537 664 963 117 3173 603 1096 1632 593 916 120 2135 1015 1048 1462 589 892 124 2154 598 1053 1433 634 920 127 1775 683 1134 1388 845 904 130 2083 618 1042 1331 551 870 133 1874 618 1004 1482 584 837 138 3255 827 1389 1650 780 1194 142 2436 856 1518 1750 789 1213 145 2322 1004 1561 2226 804 1266 148 2617 875 1504 1922 794 1304 152 2779 899 1542 1807 770 1280 155 2417 823 1480 1874 765 1251 161 2207 889 1418 1931 746 1227 163 2388 813 1423 1855 732 1232 165 2750 856 1485 1826 746 1232  APPENDIX C 158 C5: Nutrients Day PO4 (mg P / L) NOx (mg N / L) NH3 (mg N / L) Influent AB  Effluent MP Effluent Influent AB Effluent MP Effluent Influent AB Effluent MP Effluent Set Up - 11 d SRT 2 17 15 12 0 0 0 54 26 40 6 15 15 13 0 16 0 41 2.5 31 10 14 13 14 0 6.3 0 38 4.4 35 12 14 10 13 0 10 0 39 6.9 32 15 14 13 14 0 19 0 40 4.5 34 17 14 13 13 0 11 0 49 1.1 34 22 14 14 14 0 16 0 47 0.8 36 24 15 14 14 0 23 0 49 0.6 37 27 8.5 14 13 0 29 0 26 0.7 37 Run 1 - 20 d SRT 31 12 10 11 0 10 0 34 1.6 27 35 15 12 14 0 9 0 42 2.6 29 38 14 12 14 0 15 0 34 2.8 33 41 15 13 14 0 16 0 36 2.0 34 44 15 - 15 0 - 0 36 - 39 48 16 15 15 0 23 0 40 3.5 36 51 15 16 14 0 23 0 41 7.3 32 56 15 13 14 0 18 0 44 1.4 36 58 14 13 14 0 19 0 40 0.9 37 62 15 14 14 0 24 0 36 3.5 38 65 16 13 15 0 21 0 45 1.9 40 69 16 14 16 0 0.6 0 41 25 40 72 14 13 14 0 9.4 0 40 15 40 76 18 14 14 0 23 0 54 2.5 36 79 15 14 15 0 21 0 40 2.7 39 83 15 17 15 0 27 0 47 2.3 43 86 14 14 14 0 22 0 40 2.0 40 90 16 14 15 0 22 0 48 2.0 41 93 15 14 15 0 25 0 44 1.9 41 96 15 14 14 0 28 0 45 0.8 40 100 16 14 15 0 21 0 50 4.9 44 104 16 14 15 0 28 0 51 0.8 37 106 16 14 15 0 27 0 51 1.0 40  APPENDIX C 159 C5: Nutrients, continued  Day PO4 (mg P / L) NOx (mg N / L) NH3 (mg N / L) Influent AB  Effluent MP Effluent Influent AB Effluent MP Effluent Influent AB Effluent MP Effluent Run 2 - 5 d SRT 109 13 12 13 0 10 0 39 1.6 34 111 13 11 12 0 11 0 38 1.2 30 113 13 12 12 0 17 0 34 0.6 29 117 12 12 13 0 12 0 36 3.1 30 120 13 13 13 0 20 0 40 1.6 33 124 - - - 0 15 0 39 1.7 32 127 - - - 0 13 0 43 2.2 32 130 - - - 0 19 0 40 1.2 31 133 15 13 13 0 17 0 45 1.1 32 138 13 13 12 0 1.9 0 33 7.4 26 142 13 12 12 0 0.8 0 33 6.2 23 145 14 12 12 0 0.2 0 41 13 28 148 14 13 13 0 3.6 0 39 5.5 30 152 13 12 12 0 0.3 0 36 7.5 29 155 14 13 13 0 0.5 0 - - - 161 15 13 12 0 0.4 0 44 5.5 31 163 - - - - - - 40 4.7 32 165 - - - - - - 40 9.3 33  APPENDIX C 160 C6: Sterols and Potential Biotransformation Products Parameter Run 1 - 20 d SRT Run 2a - 5 d SRT Day 86 Day 104 Day 107 Day 107 Day 107 Day 130 Day 133 Day 134 Day 134 Day 134   Influent Campesterol  249 129 107 120 176 268 141 172 158 227 Cholesterol  67.3 0.881 0.916 0.899 1.11 3.66 2.03 2.20 2.55 3.45 Coprostanol  0.689 LTDL LTDL LTDL LTDL 1.61 LTDL 1.33 1.46 1.29 Desmosterol LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL Dihydrocholesterol  5.44 LTDL LTDL LTDL LTDL 0.93 LTDL 1.93 LTDL LTDL Estrone LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL β-Sitosterol  1010 555 454 512 763 1540 826 995 917 1290 Stigmasterol 10.1 4.05 3.13 3.49 4.76 39.6 20.3 26.3 24.1 37.8 Progesterone - - - - - - - - - -   Aerobic Effluent Campesterol  4.46 0.833 1.12 1.05 1.08 5.15 2.27 2.58 2.14 1.88 Cholesterol  1.26 0.557 0.395 0.35 0.362 0.96 0.42 0.44 0.52 0.71 Coprostanol LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL Desmosterol LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL Dihydrocholesterol  0.402 0.286 0.397 0.361 0.431 LTDL LTDL LTDL LTDL LTDL Estrone  0.334 LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL β-Sitosterol  23.9 4.19 5.88 5.4 5.63 46.8 16.0 17.9 14.5 12.2 Stigmasterol  0.656 0.194 0.219 0.219 0.218 1.00 0.81 0.62 0.52 0.45 Progesterone LTDL - LTDL - LTDL - - - - - Androstedione LTDL - LTDL - LTDL - - - - - Androstadienedione LTDL - LTDL - LTDL - - - - - Testosterone LTDL - LTDL - LTDL - - - - -   Microaerophilic Effluent Campesterol  114 65.1 78.4 112 95.7 87.6 59.2 71.4 63.6 52.0 Cholesterol  5.88 0.919 1.04 1.36 1.26 0.96 0.98 1.37 1.53 0.93 Coprostanol LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL Desmosterol LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL Dihydrocholesterol  1.08 0.336 0.343 0.447 0.422 0.42 LTDL LTDL 0.50 LTDL Estrone LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL β-Sitosterol  549 305 366 503 426 489 327 365 364 300 Stigmasterol  3.72 2.12 2.57 3.48 3.15 14.2 9.15 11.2 8.57 8.86 Progesterone - - LTDL LTDL LTDL - - - LTDL - Androstedione - - LTDL LTDL LTDL - - - LTDL - Androstadienedione - - LTDL LTDL LTDL - - - LTDL - Testosterone - - LTDL LTDL LTDL - - - LTDL -   APPENDIX C 161 C6: Sterols and Potential Biotransformation Products, continued Run 2b - 5 d SRT   Soluble Phase Whole Water Soluble Phase Whole Water Parameter Day 161 Day 163 Day 165 Day 161 Day 163 Day 165 Day 161 Day 165 Day 161 Day 165   Influent Campesterol 104 207 320 41.8 86.6 92.7 Cholesterol 0.69 1.39 1.89 0.47 0.53 0.65 Coprostanol LTDL 0.77 1.08 LTDL LTDL 0.34 Desmosterol LTDL LTDL LTDL LTDL LTDL LTDL Dihydrocholesterol LTDL LTDL LTDL LTDL LTDL LTDL Estrone LTDL LTDL LTDL LTDL LTDL LTDL β-Sitosterol 508 1019 1557 207 384 451 Stigmasterol 17.9 61.3 67.8 7.60 15.7 17.9 Progesterone - - - - - -   Aerobic Effluent Aerobic Mixed Liquor Campesterol 15.1 13.3 11.6 5.56 2.94 1.82 584 398 5.58 3.34 Cholesterol 1.38 0.62 0.76 0.90 0.41 0.34 18.7 14.2 0.56 0.37 Coprostanol LTDL LTDL LTDL LTDL LTDL LTDL 2.04 LTDL LTDL LTDL Desmosterol LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL Dihydrocholesterol LTDL 0.24 0.24 LTDL LTDL LTDL 4.30 4.19 LTDL LTDL Estrone LTDL 0.44 0.62 LTDL LTDL LTDL LTDL LTDL LTDL LTDL β-Sitosterol 85.3 75.2 65.2 29.8 14.4 8.29 3387 2333 27.0 15.5 Stigmasterol 3.89 3.68 3.80 1.44 0.78 0.70 134 122 1.68 1.17 Progesterone LTDL LTDL LTDL - - - - - - - Androstedione LTDL LTDL LTDL - - - - - - - Androstadienedione LTDL LTDL LTDL - - - - - - - Testosterone LTDL LTDL LTDL - - - - - - -   Microaerophilic Effluent Microaerophilic Mixed Liquor Campesterol 66.6 102 96.3 46.6 35.1 55.2 707 652 82.9 172 Cholesterol 0.54 0.86 0.75 0.56 0.41 0.60 6.36 6.12 0.60 0.81 Coprostanol LTDL LTDL 0.30 LTDL LTDL LTDL 2.40 2.89 0.28 0.39 Desmosterol LTDL LTDL 0.60 LTDL LTDL LTDL LTDL 6.49 LTDL LTDL Dihydrocholesterol LTDL LTDL 0.39 LTDL LTDL LTDL LTDL LTDL LTDL 0.45 Estrone LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL LTDL β-Sitosterol 339 487 405 228 176 279 3459 3446 423 769 Stigmasterol 12.6 22.3 19.0 10.5 8.9 13.3 156 161 17.2 36.4 Progesterone - LTDL LTDL - - - - - - - Androstedione - LTDL LTDL - - - - - - - Androstadienedione - LTDL LTDL - - - - - - - Testosterone - LTDL LTDL - - - - - - -   162 APPENDIX D:  QA/QC D1: Benchscale Biological Wastewater Treatment System Variation in the measured parameters was generally low for BBWTS operation, as described in the results portion of the text.  A few exceptions were noted for which variation was high, most notably in influent and effluent TSS and aerobic effluent nitrogen levels.  In regards to influent TSS, variation was high due to change of feed batches collected from the study mill, and this did not detract from the overall stability of the system.  Variations in nitrogen levels were indicative of changes in nitrification in the aerobic system.  The loss of nitrification in this system during the 5 d SRT run was significant, as it was otherwise sustained throughout the experimental program.  Method error was assessed by duplicate analysis of COD and nutrient samples, the results of which are provided in Tables D1 and D2.   Replicates used for total and dissolved COD analysis displayed average CV values of 6 % and 5 % respectively, with a combined range of 0 to 23 %.  Samples duplicated for analysis of NOx-N, NH3-N and PO4-P showed very low CV values, with averages of 1 % across the board.  These data indicate low variability during analysis of COD and nutrients. Overall, the quality of the measured data was judged to be satisfactory for routine monitoring of the laboratory scale secondary treatment systems. APPENDIX D 163 Table D1:  COD duplicate summary.  All concentrations in mg/L. Total COD       Dissolved COD Sample ID Conc X CV Sample ID Conc X CV 20 d SRT AB23(1) 767 775 1% AB34(1) 613 606 2% AB23(dup) 782   AB34(dup) 599 AB40(1) 767 743 5% AB46(1) 456 478 6% AB40(dup) 719     AB46(dup) 499 AB52(1) 546 578 8% MP36(1) 999 1194 23% AB52(dup) 610   MP36(dup) 1389 MP28(1) 1013 1030 2% MP48(1) 862 864 0% MP28(dup) 1047     MP48(dup) 867 MP42(1) 1211 1242 4% FD38(1) 1553 1592 3% MP42(dup) 1273   FD38(dup) 1630 FD31(1) 1399 1666 23% FD50(1) 1447 1368 8% FD31(dup) 1934     FD50(dup) 1289 FD44(1) 1983 1925 4% FD44(dup) 1868 5 d SRT AB71 (1) 1008 1004 1% AB62(1) 603 634 7% AB71(dup) 999   AB62(dup) 664 MP58(1) 1120 1108 2% AB77(1) 742 746 1% MP58(dup) 1096     AB77(dup) 751 MP73(1) 1575 1542 3% MP64(1) 851 870 3% MP73(dup) 1509   MP64(dup) 889 FD60(1) 3510 3173 15% FD69(1) 1655 1650 0% FD60(dup) 2837   FD69(dup) 1645 FD75(1) 2179 2207 2% FD75(dup) 2236 AVERAGE CV   6%       5%  APPENDIX D 164 Table D2:  Nutrient duplicate summary.  All concentrations in mg/L. PO4-P NOx-N NH3-N Sample ID Conc X CV Sample ID Conc X CV Sample ID Conc X CV AB24(1) 9.9 10.1 3% AB24(1) 9.85 10.1 3% AB24(1) 3.5 3.5 2% AB24(dup) 10.3 AB24(dup) 10.3 AB24(dup) 3.4 AB40(1) 14.1 14.3 1% AB40(1) 24.2 24.3 0% AB40(1) 0.79 0.8 1% AB40(dup) 14.4 AB40(dup) 24.3 AB40(dup) 0.8 AB50(1) 14.2 14.2 0% AB50(1) 28.4 28.4 0% MP31(1) 26.3 27.1 4% AB50(dup) 14.1 AB50(dup) 28.3 MP31(dup) 27.8 MP31(1) 10.7 10.7 0% MP31(1) 0 0.0 n.a. MP36(1) 34.8 35.9 4% MP31(dup) 10.7 MP31(dup) 0 MP36(dup) 37 MP36(1) 14.5 14.5 0% MP36(1) 0 0.0 n.a. MP43(1) 39.4 39.6 1% MP36(dup) 14.5 MP36(dup) 0 MP43(dup) 39.7 MP43(1) 14.4 14.4 0% MP43(1) 0 0.0 n.a. MP53(1) 40.1 40.0 0% MP43(dup) 14.3 MP43(dup) 0 MP53(dup) 39.9 MP53(1) 15.1 15.1 0% MP53(1) 0 0.0 n.a. FD34(1) 36.1 36.2 0% MP53(dup) 15.1 MP53(dup) 0 FD34(dup) 36.2 FD34(1) 14.7 14.8 1% FD34(1) 0 0.0 n.a. FD46(1) 46.6 46.7 0% FD34(dup) 14.9 FD34(dup) 0 FD46(dup) 46.8 FD46(1) 15.1 15.1 0% FD46(1) 0 0.0 n.a. FD46(dup) 15.1 FD46(dup) 0 AB63(1) 22.6 22.6 0% AB63(1) 13.0 13.0 0% AB63(1) 2.2 2.2 4% AB63(dup) 22.6 AB63(dup) 13.0 AB63(dup) 2.1 AB70(1) 11.9 11.9 0% AB70(1) 0.9 0.8 2% AB70(1) 6.2 6.2 0% AB70(dup) 11.9 AB70(dup) 0.8 AB70(dup) 6.2 MP73(1) 12.2 12.2 1% MP73(1) 0.0 0.0 n.a. MP73(1) 29.3 29.2 0% MP73(dup) 12.1 MP73(dup) 0.0 MP73(dup) 29.1 FD59(1) 13.0 13.0 0% FD59(1) 0.0 0.0 n.a. FD59(1) 33.6 33.8 1% FD59(dup) 13.0 FD59(dup) 0.0 FD59(dup) 34.0 FD76(1) 39.4 39.7 1% FD76(dup) 39.9 AVERAGE CV 1% 1% 1% 5 d SRT 20 d SRT   APPENDIX D 165 D2: Quantification of Plant Sterols and Potential Biotransformation Products D2.1: Surrogates Surrogate recoveries for experimental samples ranged from 53 % to 134 %. Average surrogate recoveries across the entire experimental program are provided in Table D3 and these ranged from 63 % to 107 %.  These ranges were considered satisfactory given the extensive sample preparation used in the analytical method and the potential matrix effects associated with pulp mill effluents.  Table D3:  Average surrogate recoveries.  “n.m.” indicates not measured. Sample Type 20 day SRT (run 1) 5 day SRT (run 2a) 5 day SRT (run 2b) Influent 95% 66% 68% Aerobic Effluent 97% 63% 91% Microaerophilic  Effluent 107% 72% 107% Aerobic MLSS n.m. n.m. 60% Microaerophilic  MLSS n.m. n.m. 82%    D2.2: Replicates Experimental replicates were collected on every other day during the sampling periods of each experimental run.  This includes the sampling periods for wastewater collection for bioassays (run 1 - 20 d SRT, and run 2a - 5 d SRT) and for phase separation analysis (run 2b - 5 d SRT). Results for experimental replicates are provided in Table D4.  Across experimental runs, CV values averaged by analytes were lowest for campe, β-sito, stigma with a range of 19 to 60 %, and were highest for copro, desmo, DHC and estrone, with a range of 80 to 157 %.  CV values were also averaged across process streams, accounting for all analytes detected in the process stream for the given experimental run.  Using this analysis, CV values ranged from 21 % for the soluble phase replicates collected from microaerophilic effluent during the 5 d SRT experimental run (run 2b) to 95 % for the replicates collected from influent during the 20 d SRT experimental run (run 1).  Mixed liquor replicates generally demonstrated lower CVs than influent and effluent samples with a range of 32 to 54 %.  From these data it can be concluded that operation of the treatment system led to moderate variation in concentrations of organic parameters.  APPENDIX D 166 Method replicates were collected on one day during the sampling periods of each experimental run, during the periods of wastewater collection for bioassays (run 1 - 20 d SRT and run 2a - 5 d SRT). Each method replicate was collected separately, in succession.  The results for the method replicates are provided in Table D5.  Across experimental runs, CV values averaged by analytes were generally low, ranging from 11 to 24 % for all parameters except DHC, for which the data display a CV of 173 % over the 5 d SRT experimental run.  This value was high compared to the other CV values averaged by analyte as DHC was only present in one of the three replicates for both the influent and microaerophilic effluent.  In all other cases, the individual analytes were either present or absent in all three method replicates.  As with the experimental replicates, CV values for method replicates were also averaged across process streams, accounting for all analytes detected in the process stream for the given experimental run.  Using this analysis, CV values ranged from 5 % for the replicates collected from aerobic effluent during the 20 d SRT experimental run to 52 % for the replicates collected from influent during the 5 d SRT experimental run.  Overall, CV data for method replicates were considered to be satisfactory, given the complexity of sample preparation.  It should be noted that in all cases for replicate analysis, values of less than detection limit (LTDL) were replaced with zeros.  In essence, this serves to make CV calculations conservative, by biasing LTDL values low.  APPENDIX D 167 Table D4.  Summary of experimental replicates for analysis of organic parameters.  All concentrations in μg/L.  Results indicating less than detection limit were replaced with 0.00.  Pr oc es s St re am R ep lic at e ID C am pe st er ol C on ce nt ra tio n C V C ho le st er ol C on ce nt ra tio n C V C op ro st an ol  C on ce nt ra tio n C V D es m os te ro l C on ce nt ra tio n C V D ih yd ro ch ol es te ro l C on ce nt ra tio n C V Es tro ne C on ce nt ra tio n C V β- S ito st er ol C on ce nt ra tio n C V St ig m as te ro l C on ce nt ra tio n C V A ve ra ge  C V by  Pr oc es s St re am Aerobic AB47 4.46 97% 1.26 55% 0.00 n.a. 0.00 n.a. 0.402 24% 0.334 141% 23.9 99% 0.656 77% 82% Effluent AB52 0.833 0.557 0.00 0.00 0.286 0.00 4.19 0.194 Microaer. MP47 114 39% 5.88 103% 0.00 n.a. 0.00 n.a. 1.08 74% 0.00 n.a. 549 40% 3.72 39% 59% Effluent MP52 65.1 0.919 0.00 0.00 0.336 0.00 305 2.12 Influent FD47 249 45% 67.3 138% 0.689 141% 0.00 n.a. 5.44 141% 0.00 n.a. 1010 41% 10.1 60% 95% FD52 129 0.881 0.00 0.00 0.00 0.00 555 4.05 Average CV by Analyte 60% 99% 141% n.a. 80% 141% 60% 59% Aerobic AB64 5.15 55% 0.960 55% 0.00 n.a. 0.00 n.a. 0.00 n.a. 0.00 n.a. 46.8 70% 1.00 15% 49% Effluent AB65 2.27 0.422 0.00 0.00 0.00 0.00 16.0 0.81 Microaer. MP64 87.6 27% 0.963 1% 0.00 n.a. 0.00 n.a. 0.42 141% 0.00 n.a. 489 28% 14.2 30% 46% Effluent MP65 59.2 0.979 0.00 0.00 0.00 0.00 327 9.15 Influent FD64 268 44% 3.66 40% 1.61 141% 0.00 n.a. 0.93 141% 0.00 n.a. 1540 43% 39.6 46% 76% FD65 141 2.03 0.00 0.00 0.00 0.00 826 20.3 Average CV by Analyte 42% 32% 141% n.a. 141% n.a. 47% 30% 20 d SRT, Experimental Run 1 5 d SRT - Experimental Run 2a    APPENDIX D 168 Table D4.  Summary of experimental replicates for analysis of organic parameters, continued. Pr oc es s St re am R ep lic at e ID C am pe st er ol C on ce nt ra tio n C V C ho le st er ol C on ce nt ra tio n C V C op ro st an ol  C on ce nt ra tio n C V D es m os te ro l C on ce nt ra tio n C V Es tro ne C on ce nt ra tio n C V β- S ito st er ol C on ce nt ra tio n C V St ig m as te ro l C on ce nt ra tio n C V A ve ra ge C V by  Pr oc es s St re am Aerobic AB75-W 15.1 13% 1.38 44% 0.00 n.a. 0.00 n.a. 0.00 87% 0.00 91% 85.3 13% 3.89 3% 42% Effluent AB76-W 13.3 0.617 0.00 0.00 0.240 0.437 75.2 3.68 AB77-W 11.6 0.762 0.00 0.00 0.241 0.623 65.2 3.80 Microaer. MP75-W 66.6 22% 0.540 22% 0.00 173% 0.00 1.732 0.00 173% 0.00 n.a. 339 18% 12.6 27% 87% Effluent MP76-W 102 0.856 0.00 0.00 0.00 0.00 487 22.3 MP77-W 96.3 0.748 0.298 0.60 0.392 0.00 405 19.0 Influent FD75-W 104 51% 0.685 46% 0.00 90% 0.00 n.a. 0.00 n.a. 0.00 n.a. 508 51% 17.9 55% 59% FD76-W 207 1.39 0.774 0.00 0.00 0.00 1019 61.3 FD77-W 320 1.89 1.08 0.00 0.00 0.00 1557 67.8 Aerobic ABR1-W 584 27% 18.7 19% 2.04 141% 0.00 n.a. 4.30 2% 0.00 n.a. 3387 26% 134 7% 37% MLSS ABR2-W 398 14.2 0.00 0.00 4.19 0.00 2333 122 Microaer. MPR1-W 707 6% 6.36 3% 2.40 13% 0.00 1.414 0.00 n.a. 0.00 n.a. 3459 0% 156 2% 28% MLSS MPR2-W 652 6.12 2.89 6.49 0.00 0.00 3446 161 Average CV by Analyte 24% 27% 104% 157% 87% 91% 22% 19% Aerobic AB75-S 5.56 56% 0.90 56% 0.00 n.a. 0.00 n.a. 0.00 n.a. 0.00 n.a. 29.8 63% 1.44 42% 54% Effluent AB76-S 2.94 0.41 0.00 0.00 0.00 0.00 14.4 0.782 AB77-S 1.82 0.34 0.00 0.00 0.00 0.00 8.29 0.698 Micro. MP75-S 46.6 22% 0.56 18% 0.00 n.a. 0.00 n.a. 0.00 n.a. 0.00 n.a. 228 23% 10.5 20% 21% Effluent MP76-S 35.1 0.41 0.00 0.00 0.00 0.00 176 8.91 MP77-S 55.2 0.60 0.00 0.00 0.00 0.00 279 13.3 Influent FD75-S 41.8 38% 0.47 17% 0.00 173% 0.00 n.a. 0.00 n.a. 0.00 n.a. 207 36% 7.60 39% 61% FD76-S 86.6 0.53 0.00 0.00 0.00 0.00 384 15.7 FD77-S 92.7 0.65 0.342 0.00 0.00 0.00 451 17.9 Aerobic ABR1-S 82.9 49% 0.60 21% 0.282 23% 0.00 n.a. 0.00 141% 0.00 n.a. 423 41% 17.2 51% 54% MLSS ABR2-S 172 0.81 0.394 0.00 0.453 0.00 769 36.4 Micro. MPR1-S 5.58 36% 0.56 28% 0.00 n.a. 0.00 n.a. 0.00 n.a. 0.00 n.a. 27.0 38% 1.68 25% 32% MLSS MPR2-S 3.34 0.37 0.00 0.00 0.00 0.00 15.5 1.17 Average CV by Analyte 40% 28% 98% n.a. 141% n.a. 40% 35% 5 d SRT - Run 2b, Whole Water Samples 5 d SRT - Run 2b, Soluble Fraction Samples D ih yd ro - ch ol es te ro l C on ce nt ra tio n C V  APPENDIX D 169 Table D5.  Summary of method replicates for analysis of organic parameters.  All concentrations in μg/L.  Results indicating less than detection limit were replaced with 0.00.  Pr oc es s St re am R ep lic at e ID C am pe st er ol C on ce nt ra tio n C V C ho le st er ol C on ce nt ra tio n C V C op ro st an ol  C on ce nt ra tio n C V D es m os te ro l C on ce nt ra tio n C V D ih yd ro ch ol es te ro l C on ce nt ra tio n C V Es tro ne C on ce nt ra tio n C V β- S ito st er ol C on ce nt ra tio n C V St ig m as te ro l C on ce nt ra tio n C V A ve ra ge  C V by  Pr oc es s St re am Aerobic AB54 1.12 3% 0.395 6% 0.00 n.a. 0.00 n.a. 0.397 9% 0.00 n.a. 5.88 4% 0.219 0% 5% Effluent AB55 1.05 0.350 0.00 0.00 0.361 0.00 5.40 0.219 AB56 1.08 0.362 0.00 0.00 0.431 0.00 5.63 0.218 Micro. MP54 78.4 18% 1.04 13% 0.00 n.a. 0.00 n.a. 0.343 13% 0.00 n.a. 366 16% 2.57 15% 15% Effluent MP55 112 1.36 0.00 0.00 0.447 0.00 503 3.48 MP56 95.7 1.26 0.00 0.00 0.422 0.00 426 3.15 Influent FD54 107 27% 0.916 12% 0.00 n.a. 0.00 n.a. 0.00 n.a. 0.00 n.a. 454 28% 3.13 23% 23% FD55 120 0.899 0.00 0.00 0.00 0.00 512 3.49 FD56 176 1.11 0.00 0.00 0.00 0.00 763 4.76 Average CV by Analyte 16% 11% n.a. n.a. 11% n.a. 16% 13% Aerobic AB66 2.58 16% 0.44 24% 0.00 n.a. 0.00 n.a. 0.00 n.a. 0.00 n.a. 17.9 19% 0.62 16% 19% Effluent AB67 2.14 0.52 0.00 0.00 0.00 0.00 14.5 0.52 AB68 1.88 0.71 0.00 0.00 0.00 0.00 12.2 0.45 Micro. MP66 71.4 16% 1.37 25% 0.00 n.a. 0.00 n.a. 0.00 173% 0.00 n.a. 365 11% 11.2 15% 48% Effluent MP67 63.6 1.53 0.00 0.00 0.50 0.00 364 8.57 MP68 52.0 0.93 0.00 0.00 0.00 0.00 300 8.86 Influent FD66 172 20% 2.20 24% 1.33 n.a. 0.00 n.a. 1.93 173% 0.00 n.a. 995 18% 26.3 25% 52% FD67 158 2.55 1.46 0.00 0.00 0.00 917 24.1 FD68 227 3.45 1.29 0.00 0.00 0.00 1290 37.8 Average CV by Analyte 17% 24% n.a. n.a. 173% n.a. 16% 19% 5 d SRT - Experimental Run 2a 20 d SRT, Experimental Run 1  APPENDIX D 170 D2.3: Blanks Traces of cholesterol were detected in the blanks collected during the 20 d SRT experimental run and for the first half of the 5 d SRT experimental run, as shown in Table D6.  All other analyte concentrations were below detection limits, or in the case of progesterone, for which no detection limit was determined, the compound was not detected.  The blanks were not tested for AD, ADD or T.  Blank concentrations were not subtracted from experimental results.  Table D6.  Analytical results for blanks.  “n.d.” indicates not detected. Parameter BLANK 1 BLANK 2 BLANK 3 20 d SRT 5 d SRT 5 d SRT Run 1 Run 2a Run 2b 17α-Estradiol < 0.005 <0.05 <0.2 17α-Ethynylestradiol < 0.02 <0.2 <0.2 17β-Estradiol < 0.005 <0.05 <0.2 β-Sitosterol < 0.005 <0.05 < 0.05 Campesterol < 0.005 <0.05 < 0.05 Cholesterol 0.262 0.13 < 0.05 Coprostanol < 0.005 <0.05 <0.05 Desmosterol < 0.005 <0.05 <0.05 Dihydrocholesterol < 0.005 <0.05 <0.05 Epicoprostanol < 0.005 <0.05 <0.05 Equilin < 0.005 <0.05 <0.05 Equol < 0.2 <2 <2 Estriol < 0.02 <0.2 <0.2 Estrone < 0.005 <0.05 <0.05 Mestranol < 0.005 <0.05 <0.05 Norethindrone < 0.2 <2 <2 Norgestrel < 0.2 <2 <2 Stigmasterol < 0.005 <0.05 < 0.05 Progesterone n.d. n.d. n.d.  This analysis shows the experimental samples may have contained small amounts of cholesterol (chole) from external sources.  Possibly, samples were easily contaminated with cholesterol because it is extremely ubiquitous in nature (Moncecchi et al. 1991). APPENDIX D 171 D2.4: Direct spike One direct spike was analyzed during the study by aliquoting 25 μL of solvent containing all compounds in the Environment Canada sterols method in known concentrations from 20 to 100 μg/mL.  Analysis of the spike began at the derivitization stage, from which time on the spike was treated identically to the samples.  Results of analysis for the spike is provided in Table D7. These data show recoveries in the range of 45 % to 97 %, indicating that derivitization did have an effect on compound recoveries.  Table D7:  Spike recoveries for target analytes.  Desmosterol was not included in the spike. Parameter Measured Concentration Spiked Concentration Recovery 17α-Estradiol 0.333 0.5 67% 17α-Ethynylestradiol 1.79 2.5 72% 17β-Estradiol 0.313 0.5 63% β-Sitosterol 0.407 0.5 81% Campesterol 0.383 0.5 77% Cholesterol 0.390 0.5 78% Coprostanol 0.246 0.5 49% Desmosterol n.a. n.a. n.a. Dihydrocholesterol 0.373 0.5 75% Epicoprostanol 0.359 0.5 72% Equilin 0.443 1 44% Equol 1.11 2 55% Estriol 1.45 2.5 58% Estrone 0.331 0.5 66% Mestranol 0.486 0.5 97% Norethindrone 1.95 2.5 78% Norgestrel 2.00 2.5 80% Stigmasterol 0.378 0.5 76%   APPENDIX D 172 D3: Genomic Response of Rainbow Trout to Exposure to Treated Pulp Mill Effluents D3.1: Bioassays Aerobic and microaerophilic effluent used in the bioassays were first tested for acute toxicity using Microtox® to provide evidence that the effluent would not be acutely toxic to the fish. Results of the Microtox® toxicity test indicated that the aerobic effluent was not acutely toxic at 100 % concentration using a 5 min or 15 min exposure period.  The microaerophilic effluent was acutely toxic at 30 % concentration using a 5 min exposure period and at 22.4 % concentration using a 15 min exposure period.  These results indicate that the test concentrations of 1 % and 10 % would not be acutely toxic to the fish for either the aerobic or microaerophilic effluents.  A sample of the fish cohort used in the bioassays was tested using a standard reference toxicant (phenol) to ensure the sensitivity of the test population was within the charted value for the species. The fish used in this test demonstrated an LC50 of 9.29 mg/L which is near to PESC’s historic geometric mean of 9.91 mg/L and well within the two standard deviation warning limits of 7.41 and 13.27 mg/L.  These results indicate the test fish used in the bioassays displayed a normal response to stress.  D3.2: Genomic Analysis Non-threshold controls, dissociation curves and reference samples were within the expected ranges for all QPCR plates run for final analysis.  In some instances during analysis, the QPCR instrument would stall after initial ramp up to 95 oC, and would need to be restarted in order to continue to amplification cycles.  Although NTCs, dissociation curves and reference samples were adequate for these runs, outlier testing of the technical replicates showed that the majority of samples on these plates were outside of the expected range.  For this reason, results for all samples from plates that were re-run on the QPCR instrument after warm up were discarded.  This reduced the number of technical replicates from four to three for analysis of L8, AR, and VTG1.  173 APPENDIX E:  Standard Curves for Genomic Analysis E1: L8 Standard Curve 14 16 18 20 22 24 26 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 Copy Number C t ( dR n) R2 = 0.955 Efficiency = 114.9 % Threshold = 0.009  APPENDIX E 174 E2: AR Standard Curve 19 21 23 25 27 29 31 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 Copy Number C t ( dR n) R2 = 0.977 Efficiency = 98.4 % Threshold = 0.010  APPENDIX E 175 E3: VTG1 Standard Curve 15 17 19 21 23 25 27 29 31 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 Copy Number C t ( dR n) R2 = 0.951 Efficiency = 62.2 % Threshold = 0.013  APPENDIX E 176 E4: VEPb Standard Curve 18 20 22 24 26 28 30 32 34 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 Copy Number C t ( dR n) R2 = 0.973 Efficiency = 65.6 % Threshold = 0.014  APPENDIX E 177 E5: VEPg Standard Curve 16 18 20 22 24 26 28 30 32 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 Copy Number C t ( dR n) R2 = 0.981 Efficiency = 60.8 % Threshold = 0.010  APPENDIX E 178 E6: NME Standard Curve 15 17 19 21 23 25 27 29 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 Copy Number C t ( dR n) R2 = 0.962 Efficiency = 82.7 % Threshold = 0.012  APPENDIX E 179 E7: THRa Standard Curve 14 16 18 20 22 24 26 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 Copy Number C t ( dR n) R2 = 0.967 Efficiency = 87.3 % Threshold = 0.013  `

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0063112/manifest

Comment

Related Items