Investigating the effect of algal blooms on in passive systems

by Shuang Liang

A thesis submitted to the graduate program in Civil Engineering in conformity with the requirements for the degree of Master of Applied Science

Queen’s University Kingston, Ontario, Canada September 2018

Copyright © Shuang Liang, 2018 ABSTRACT

Wastewater stabilization (WSPs) are low-cost, effective and sustainable passive wastewater treatment technologies. However, algal blooms may occur in WSPs under certain conditions, affecting the complex interactions between water quality parameters. This thesis is an investigation on the effect of algal blooms on water quality and treatment performance of a WSP located in a temperate climate. Studies were conducted at a WSP located in southeastern Ontario, experiencing reoccurring excessive algal growth and pH fluctuations, exceeding government regulatory limits. A full range of water quality parameters were monitored for the system and the biochemical dynamics in the WSPs were assessed through multivariate statistical analysis. One study was conducted to provide an enhanced understanding of the relationships between algal

- blooms, water quality and meteorological factors. Water temperature, pH, DO and NO3 were determined to be the most significant parameters describing the spatial variations in surface algal blooms. A second study examined the seasonal disinfection performance using Escherichia coli, and three novel bacterial indicators, Enterococci, Clostridium perfringens and total coliforms, in the presence and absence of the previously studied algal blooms. Temperature, pH and DO were shown to be significant (p<0.05) factors in disinfection. A long-term assessment of seasonal fluctuations in treatment performance and water quality was also performed. Removal efficiency of nutrients and were observed to have strong seasonal fluctuations and water temperature, pH, DO and TP accounted for majority of the temporal variations observed in water quality. This research provides practical knowledge and uncovering key trends and factors involved in algal bloom dynamics and treatment performance, contributing to the continued improvement of WSPs design and performance in temperate climates.

! ii Acknowledgments

I would like to thank my supervisors, Dr. Pascale Champagne and Dr. Geof Hall, for not only their support in my research but also inspiring me to continuously pursue new academic, extracurricular and professional opportunities.

I would like to acknowledge the staff at Amherstview WPCP for their assistance with sample collection, analysis and site information: Rami Maassarani, Marie-Josée Merritt, Jenna

Campbell, and Lorie McFarland at Loyalist Township.

I am grateful for financial support from NSERC Canadian Graduate Scholarship – Master’s

(CGS-M), NSERC Collaborative Research & Development (CRD), NSERC Collaborative

Research & Training Experience (CREATE), Canada Research Chairs program and Queen’s

Beaty Water Research Centre.

I am grateful all the administrative support and lovely conversations with Debbie Ritchie, Sandra

Martin, Angela Agostino and Susan Palo.

I am also thankful to my colleagues and summer students, in particular, Lei Liu, Meng Jin,

Alexander Rey, David Blair, Meg McKellar, Zoe Zimmerman and Gabrielle Favreau, for their insight and support conducting laboratory and field work.

I am grateful for all the friendships made during my time in Ellis Hall. To my office mates, thank you for your early morning chats and your passion for coffee. To Debrah, I am grateful for our many adventures together over the years and your rivaling passion for snacks. To Rob, thanks for making me play inner tube water polo for the past two years. To the otter circle and many intramural teams, thank you for your friendship and endless laughs. To my parents, thank you for your continued love and support.

! iii Statement of Co-Authorship

Shuang Liang conducted the work presented in this thesis, with the assistance of co-authors who provided technical guidance, reviewed manuscripts, and offered editorial comments and revisions. Chapters 3 to 6 are intended for publication in peer-reviewed journals or conference proceedings and Appendix A is intended for publication as a book. Authorship and publication details are provided as follows:

Chapter 3: Shuang Liang, Geof Hall, and Pascale Champagne

Manuscript to be submitted, in full, for publication in Science of the Total Environment

Chapter 4: Lei Liu, Shuang Liang, Geof Hall, and Pascale Champagne

Manuscript to be submitted, in full, for publication in the Journal of Water Research

Chapter 5: Shuang Liang, Geof Hall, and Pascale Champagne

Published in the 1st International Water Distribution Systems Analysis/Computing and

Control for the Water Industry (WDSA/CCWI) Conference Proceedings.

Chapter 6: Shuang Liang, Geof Hall, and Pascale Champagne

Manuscript to be submitted, in full, for publication in Water Research

Appendix B: Shuang Liang, David Blair, Megan McKellar, Pratik Kumar, Agnieszka Cuprys,

Mehrdad Taheran, Mitra Naghdi, Satinder Kaur Brar, Pascale Champagne, and Geof Hall

Appendix B is to be published in the InterAmerican Network of Academies of Sciences

(IANAS) Water Quality in the Americas textbook.

! iv Table of Contents

ABSTRACT ...... ii

Acknowledgments ...... iii

Statement of Co-Authorship ...... iiv

Table of Contents ...... v

List of Tables ...... iix

List of Figures ...... xi

List of Abbreviations ...... xiii

Chapter 1. Introduction ...... xii 1.1. Background ...... 1 1.2. Research Objective ...... 3 1.3. Scope of Studies ...... 4 1.3.1. Thesis Layout ...... 5 1.4. Works Cited ...... 7

Chapter 2. Literature Review ...... 10 2.1. Introduction ...... 10 2.2. Wastewater Stabilization Ponds ...... 12 2.2.1. Types and System Characteristics of WSPs ...... 13 2.2.1.1. Anaerobic ponds ...... 13 2.2.1.2. Facultative ponds ...... 14 2.2.1.3. Maturation ponds ...... 15 2.2.2. Removal Mechanisms in WSPs ...... 15 2.2.2.1. ...... 15 2.2.2.2. Organic Matter ...... 17 2.2.2.3. Solids ...... 17 2.2.2.4. (N) ...... 18 2.2.2.5. Phosphorous (P) ...... 19

! v 2.2.3. and Macrophyte in WSPs ...... 19 2.2.3.1. Diversity of Algae and Macrophytes ...... 20 2.2.3.2. Algal Growth ...... 22 2.2.3.3. Algae ...... 23 2.2.3.4. Role of Algae in WSP ...... 24 2.3.3.5 Effect of algal blooms on wastewater quality ...... 25 2.2.4. WSP Summary ...... 26 2.3. Statistical Methods ...... 27 2.3.1. Correlations Analysis ...... 27 2.3.2. Multivariate statistical analysis ...... 28 2.3.2.1. Principal Component Analysis (PCA) ...... 28 2.2.3.2. Discriminant Analysis (DA) ...... 31 2.3.3. Time Series Analysis ...... 34 2.3.3.1. Seasonal Kendall ...... 34 2.4. Summary ...... 35

2.5. References ...... 36

Chapter 3. Spatial and temporal variations in water quality in a facultative wastewater stabilization with algal blooms ...... 50 3.1. Introduction ...... 51 3.2. Materials & Methods ...... 52 3.2.1. System overview ...... 52 3.2.2. Water quality ...... 53 3.2.3. Meteorological Parameters ...... 55 3.2.4. Algae Surveying & Image Processing ...... 55 3.2.5. Statistical Analysis ...... 56 3.3. Results ...... 57 3.3.1. Characterization of surface algae cover ...... 57 3.3.2. Effect of meteorological conditions on algal blooms ...... 58 3.3.3. Spatial and temporal changes in water quality ...... 61 3.4. Conclusion ...... 65 3.5. References ...... 67

! vi Chapter 4. Examining disinfection performance and water quality in a wastewater stabilization pond with algal blooms using multivariate statistical analysis ...... 71 4.1. Introduction ...... 72 4.2. Materials ...... 74 4.2.1. System overview ...... 74 4.2.2. Water quality analysis ...... 75 4.2.3. Detection of indicator organisms ...... 75 4.2.4. Statistical Analysis ...... 76 4.3. Results and discussion ...... 77 4.3.1. Summary of WSP system treatment performance ...... 77 4.3.2. Multivariate statistical analysis ...... 79 4.4. Conclusion ...... 86 4.5. References ...... 87

Chapter 5. Monitoring seasonal variations in treatment performance of a wastewater stabilization pond with pH fluctuations ...... 94 5.1. Introduction ...... 96 5.2. Materials & Methods ...... 98 5.2.1. System overview ...... 98 5.2.2 Field Sampling ...... 98 5.2.3. Statistical Analysis ...... 100 5.3. Results & Discussion ...... 100 - 5.3.1. Time Series of temperature, pH, DO, E. coli, TP, and NO3 ...... 100 5.3.2. Seasonal variations in treatment efficiency ...... 102 5.3.3. Correlations between water quality parameters ...... 103 5.3.4. Seasonal trends ...... 104 5.4. Conclusions ...... 105 5.5. References ...... 107

Chapter 6. Investigating long-term seasonal dynamics in a wastewater stabilization pond with algal blooms using multivariate statistical analysis ...... 110 6.1. Introduction ...... 111 6.2. Materials ...... 113

! vii 6.2.1. System overview ...... 113 6.2.2. Field Sampling ...... 114 6.2.3. Statistical Analysis ...... 114 6.2.3.1. Data treatment and calculations ...... 114 6.2.3.2. Factor Analysis/Principal Component Analysis ...... 115 6.2.3.3. Discriminant Analysis ...... 116 6.3. Results & Discussion ...... 116 6.3.1. Time series of wastewater parameters ...... 116 6.3.2. Treatment performance of WSP ...... 120 6.3.3. Analysis of intra-seasonal trends ...... 121 6.3.4. Correlations between water quality parameters ...... 122 6.3.5. Factor Analysis/Principal Component Analysis ...... 123 6.3.6. Discriminant Analysis ...... 126 6.4. Conclusions ...... 129 6.5. References ...... 130

Chapter 7. Summarizing Conclusions ...... 136 7.1. Thesis Summary ...... 136 7.2. Future Work ...... 139 7.2. Contributions to Science and Engineering ...... 140

APPENDIX A ...... 143 A.1. Pre-Screening Data Calculations ...... 143 A.2. Criteria for Spearman’s ρ correlation coefficient ...... 147 A.3. Data Screening for Principal Component Analysis ...... 148 A.4. Variables Entered/Removed in Discriminant Analysis ...... 149 A.5. Numerical Values for Seasonal Kendall Tests ...... 152 A.6. Meteorological Data ...... 153 A.7. Works Cited ...... 154

APPENDIX B ...... 155

B.1. Water Quality in the Americas – Canada ...... 155

! viii List of Tables

TABLE 2.1. VARIOUS WASTEWATER EFFLUENT DISCHARGE LIMITS UNDER WSER AND ONTARIO

PROVINCIAL REGULATIONS. (GOVERNMENT OF CANADA, 2012A; ONTARIO MINISTRY OF

ENVIRONMENT AND CLIMATE CHANGE, 2017)...... 11

TABLE 2.2.OVERVIEW OF ALGAE AND PLANT SPECIES COMMONLY FOUND IN WSPS...... 21

TABLE 3.1. DESCRIPTIVE STATISTICS OF WASTEWATER QUALITY, CHLOROPHYLL AND

PHYCOCYANIN LEVELS FROM MAY TO SEPTEMBER AT WSP 1...... 67

TABLE 3.2. VARIMAX ROTATION PCA LOADING MATRIX...... 74

TABLE 3.3. CLASSIFICATION FUNCTIONS FOR DISCRIMINANT ANALYSIS OF SPATIAL VARIATIONS IN

WATER QUALITY...... 78

TABLE 3.4. CLASSIFICATION MATRIX FOR DISCRIMINANT ANALYSIS (DA) OF TEMPORAL

VARIATION IN WSP 1...... 78

TABLE 4.1. SUMMARY OF WATER QUALITY PARAMETERS, WITH MEAN VALUES AND STANDARD

DEVIATIONS, MEASURED IN BOTH INFLUENT AND EFFLUENT OF THE WSP SYSTEM OVER THE

SAMPLING PERIOD IN BOTH WARM AND COLD SEASONS...... 78

TABLE 4.2. SPEARMAN’S Ρ CORRELATION COEFFICIENTS BETWEEN ALL PARAMETERS...... 81

TABLE 4.3.DESCRIPTIVE STATISTICS OF SELECTED PCS FOR ALL MEASURED WATER QUALITY

PARAMETERS...... 82

TABLE 4.4. ROTATION PCA LOADING MATRIX FOR RETAINED PCS AND LOADINGS FOR RESPECTIVE

WATER QUALITY PARAMETERS...... 84

TABLE 5.1. WEEKLY FINAL EFFLUENT CONCENTRATIONS IN WSP1 FROM APRIL 2011 TO MARCH

2015. OVERALL AND SEASONAL AVERAGES AND REMOVAL EFFICIENCIES FOR FINAL EFFLUENT

PARAMETERS...... 102

TABLE 5.2. WEEKLY FINAL EFFLUENT CONCENTRATIONS IN WSP1 FROM APRIL 2011 TO MARCH 2015...... 103

TABLE 5.3. SEASONAL KENDALL TEST OF TRENDS FOR FINAL EFFLUENT PARAMETERS IN ORDER OF

SPRING, SUMMER, FALL AND WINTER (RESPECTIVELY)...... 104

! ix TABLE 6.1. DESCRIPTIVE STATISTICS OF OVERALL AND SEASONAL WASTEWATER QUALITY

PARAMETERS MONITORED AT WSP 1...... 118

TABLE 6.2. TIME SERIES PLOTS DEPICTING SEASONAL VARIATIONS OF A) TEMPERATURE, B) PH, C)

DO, D) E. COLI; E) NO3- ; F) NH4+; G) TP; G) CBOD; H) TSS; I) ALKALINITY IN THE FINAL

EFFLUENT WATER QUALITY...... 120

TABLE 6.3. REMOVAL EFFICIENCIES FOR MOECC EFFLUENT DISCHARGE PARAMETERS...... 135

TABLE 6.4. SEASONAL KENDALL TEST OF TRENDS FOR FINAL EFFLUENT PARAMETERS IN ORDER OF

SPRING, SUMMER, FALL AND WINTER (RESPECTIVELY)...... 136

TABLE 6.5. FACTOR LOADING MATRIX AND EXPLAINED VARIANCE OF WATER QUALITY

PARAMETERS FOR OVERALL PERIOD AND FOUR SEASONS...... 139

TABLE 6.6. EIGENVALUES FOR THE DISCRIMINANT FACTORS (DFS)...... 140

TABLE 6.7. CLASSIFICATION FUNCTIONS FOR DISCRIMINANT ANALYSIS OF SPATIAL VARIATIONS OF

WATER QUALITY...... 142

TABLE 6.8. CLASSIFICATION MATRIX FOR DISCRIMINANT ANALYSIS (DA) OF TEMPORAL

VARIATION IN WSP 1...... 142

TABLE A.1. SHAPIRO-WILK TEST OF NORMALITY FOR WATER QUALITY DATA FOUND IN CHAPTER 3...... 144

TABLE A.2. SHAPIRO-WILK TEST OF NORMALITY FOR WATER QUALITY DATA FOUND IN CHAPTER 4...... 145

TABLE A.3. SHAPIRO-WILK TEST OF NORMALITY FOR WATER QUALITY DATA FOUND IN CHAPTER 6...... 146

TABLE A.4. VARIABLES ENTERED/REMOVED AT EACH STEP IN THE BACKWARD STEPWISE MODE FOR

WATER QUALITY DATA SET FROM CHAPTER 4...... 150

TABLE A.5. VARIABLES ENTERED/REMOVED AT EACH STEP IN THE BACKWARD STEPWISE MODE FOR

WATER QUALITY DATA SET FROM CHAPTER 6...... 151

TABLE A.6. SEASONAL KENDALL TESTS PERFORMED ON EFFLUENT WATER QUALITY PARAMETERS

IN CHAPTER 6...... 152

! !

! x List of Figures

FIGURE 2. 1. BIOLOGICAL AND CHEMICAL PROCESSES IN A FACULTATIVE WSP...... 14

FIGURE 2.2. COMMON STOPPING PROCEDURES FOR PC SELECTION DISPLAYED: KAISER CRITERION,

SCREE PLOT AND BROKEN-STICK DISTRIBUTION MODEL...... 30

FIGURE 3.1. SCHEMATIC OF AMHERSTVIEW WPCP AND SIX SURFACE SAMPLING POINTS (INFLUENT

INDICATED BY INF AND EFFLUENT INDICATED BY EFF)...... 53

FIGURE 3.2. SAMPLE OF SURFACE ALGAL BLOOM SPECIES AT 10X MAGNIFICATION: HYDRODICTYON

SP. B) SPIROGYRA SP...... 58

FIGURE 3.3. GREEN ALGAE SURFACE COVER AND WATER TEMPERATURE IN WSP1 WITH

METEOROLOGICAL PARAMETERS (DAILY PRECIPITATION, AIR TEMPERATURE, WIND SPEED AND

WIND DIRECTION) OVER MONITORING PERIOD...... 59

FIGURE 3.4. SURFACE COVER OF ALGAL BLOOMS FROM MAY 24 TO AUGUST 30, 2017 ON WSP 1 ...... 60

FIGURE 3.5. LOASINGS FOR WATER QUALITY PARAMETERS IN WSP1 BASED ON RETAINED PCS ..... 60 ! FIGURE 4.1. OVERALL SCHEMATIC OF AMHERSTVIEW WPCP AND LOCATIONS OF THE SIX

SAMPLING POINTS IN WSP 1...... 74!

FIGURE 5.1. AERIAL VIEW OF AMERSTVIEW CONTROL PLANT...... 98

FIGURE 5.2. SCHEMATIC OF AMHERSTVIEW WPCP AND LOCATIONS OF PLANT SECONDARY

EFFLUENT (SE) AND FINAL EFFLUENT (FE) SAMPLING POINTS...... 99

FIGURE 5.3. WEEKLY FINAL EFFLUENT CONCENTRATIONS IN WSP1 FROM APRIL 2011 TO MARCH 2015...... 101

FIGURE 6.1. GEOGRAPHIC LOCATION AND SCHEMATIC OF AMHERSTVIEW WPCP, INCLUDING

LOCATION OF PLANT SECONDARY EFFLUENT (SE) AND FINAL EFFLUENT (FE) SAMPLING POINTS

(IMAGE ADAPTED FROM NATURAL RESOURCE CANADA)...... 113

!

! xi !

List of Abbreviations

BOD biochemical demand °C degrees Celsius C. perfringens Clostridium perfringens C. vulgaris Chlorella vulgaris cBOD carbonaceous biochemical oxygen demand CFU colony forming units chl chlorophyll chl-a chlorophyll-a cm centimetre CCME Canadian Council of Ministers of the Environment

CO2 carbon dioxide 2- CO3 carbonate ion COD CW DA discriminant analysis DF discriminant factors DL detection limit DO dissolved oxygen EC Environment Canada E. coli Escherichia coli Eff effluent FA factor analysis FC fecal coliforms H+ hydrogen ion HCO3- bicarbonate ion

H2CO3 carbonic acid HRT hydraulic retention time Inf influent ! xii ! ! km kilometre km2 kilometre squared km/hr kilometres per hour L litre LDO luminescent dissolved oxygen m metre m2 metre squared m3 cubic metre m/s metres per second mg milligram mL millilitre µg microgram MOECC Ministry of Environment and Climate Change “MOECP” Ministry of Environment. Conservation and Parks N nitrogen/nitrogenous

NH3 ammonia + NH4 ammonium nm nanometer - NO2 nitrite - NO3 nitrate

O2 oxygen OH- hydroxide P PC principal component PCA principal component analysis pH negative logarithm to base 10 of hydrogen ion concentration phyco phycocyanin 3- PO4 phosphate ion ROS reactive oxygen species SD standard deviation SRT solids retention time ! xiii ! ! TC total coliform UAV unmanned aerial vehicle UV PAR photosynthetically active radiation WPCP water pollution control plant WSER Wastewater systems effluent regulations WSP wastewater stabilization pond WSP#1 Amherstview WPCP WSP#1 WSP#2 Amherstview WPCP WSP#2

! xiv ! !

Chapter 1. Introduction ! 1.1.! Background ! In 2015, the United Nations created an agenda, Transforming our world: the 2030

Agenda for Sustainable Development, which outlines 17 Sustainable Development Goals (SDG).

These 17 Global Goals, comprised of 169 targets, cover a broad range of development issues including from poverty and hunger, health and education, gender equality, water accessibility, and climate change (United Nations General Assembly, 2015). They provide clear guidelines and targets for all countries to adopt, acting as a universal call to action to end poverty and protect the planet. One goal in particular, SDG 6, focuses on “ensuring the availability and sustainable management of water and sanitation for all” (United Nations General Assembly, 2015).

Wastewater treatment plays an important role in the management of water resources, as as sanitation and protection of the natural environment. If wastewater is improperly or entirely untreated, it can have deleterious characteristics, including higher pH, low dissolved oxygen (DO), high heavy metal concentrations, low alkalinity, and high levels of un-ionized ammonia (Metcalf & Eddy, 2003). Untreated wastewater can also contain biological such as pathogens and emerging contaminants (Holeton, Chambers, Grace, & Kidd, 2011).

Improper wastewater management can negatively affect the environment, human health and activities related to water use and access. Environmental consequences of improper wastewater treatment include eutrophication, loss of habitat, reduced aquatic and wildlife populations, thermal enhancement and depletion of dissolved oxygen (Government of Canada,

1994, 2014; Metcalf & Eddy, 2003; United Nations, 2017). Cumulative impacts can cause

! 1 ! ! larger-scale morphological and changes to aquatic and habitats

(Holeton et al., 2011). contamination can also lead to restrictions on recreational water uses, restrictions on fish and shellfish consumption, degradation of aesthetics and in serious cases, water-borne diseases (Environment and Climate Change Canada, 2013; Government of

Canada, 1994; Tchobanoglous & Crites, 1998).

In Canada, while water sanitation and treatment is perceived as universally accessible and of high quality, this is often not the case for many rural and remote regions. In particular, inadequate wastewater infrastructure is prevalent within First Nations and northern communities.

In fact, 21% of the First Nations wastewater systems exceeded design capacities, with 6% and

41% of wastewater treatment systems to be reported at high-risk and medium-risk, respectively

(Neegan Burnside Ltd., 2011). Additionally, in 2016, it was reported that an estimated 150 billion liters of untreated and raw sewage were discharged directly into Canadian waters and bordering oceans (Government of Canada, 2012a). The federal government reported municipal wastewater effluent as the largest single source of pollutant discharge by volume (Government of

Canada, 2014). Thus, unregulated wastewater discharge and proper wastewater treatment is an important concern for Canadian communities.

In 2012, the Canadian Council of Ministers of the Environment (CCME) released the

Wastewater Systems Effluent Regulations (WSER), containing federal requirements for municipalities to achieve minimum national performance standards for discharged wastewater effluent (Government of Canada, 2012b). While WSER will improve the water quality in receiving aquatic environment, the wastewater treatment plant infrastructure upgrades required to meet these regulations is a massive financial undertaking for municipalities, especially for rural, remote and Indigenous communities (Champagne, 2017).

! 2 ! ! As a solution, many communities employ passive, naturalized or eco-engineered technologies, such as wastewater stabilization ponds (WSPs) and constructed wetlands (CWs), to properly treat wastewater (Shrestha, 2008; Sundaravadivel & Vigneswaran, 2001). WSPs are a sustainable alternative to conventional wastewater treatment due to their ease of operation, minimal chemical requirements and low capital requirements (Al-Hashimi & Hussain, 2013;

Metcalf & Eddy, 2003; Steinmann, Weinhart, & Melzer, 2003). These systems can effectively reduce nutrient loads and remove pathogens via naturally occurring biological, chemical and physical treatment mechanisms (Tchobanoglous & Crites, 1998).

However, as WSPs are open systems, they are susceptible to seasonal fluctuations in water quality parameters, resulting in some maintenance and operational issues (Wallace,

Champagne, & Hall, 2016). Developing a comprehensive understanding of the treatment mechanisms, water quality, biological dynamics and their associated challenges is crucial to improving the design and performance of WSP systems. This research will ultimately improve the effectiveness and efficiency of eco-engineered wastewater treatment technologies as wastewater management becomes an increasingly costly and persistent global challenge, with increased pressures associated with climate change, resource scarcity, and population growth

(Muga & Mihelcic, 2008; Sundaravadivel & Vigneswaran, 2001).

1.2.! Research Objective ! Elevated pH and reoccurring algal blooms are important aspects to consider for proper operation and efficient treatment using WSPs. The objectives for this project will be to: 1)

Investigate and discern relations between algal blooms, water quality parameters and meteorological factors of a full-scale, operational WSP; 2) Assess the impact of algal blooms on

! 3 ! ! short-term and long-term water quality and treatment performance; 3) Provide guidance relating to WSP treatment design and operation regarding mitigation of algal blooms. This thesis seeks to understand the relationships and dynamics between algal blooms and wastewater treatment through literature review, field and laboratory work and statistical analysis.

1.3.! Scope of Studies ! Amherstview Water Pollution Control Plant (WPCP) is a medium-sized facility that employs both active and passive wastewater treatment (Gore and Storrie, 1997). The facility is managed by Loyalist Township, operated by the Ontario Clean Water Agency, and serves a population of approximately 11,000. The plant has both active and passive wastewater treatment technologies. The system currently employs secondary and tertiary treatment, including an activated process, and two in-series WSPs and a recently built constructed wetland. The

WSPs, in particular, have experienced elevated pH levels, regularly exceeding the Ontario

Ministry of the Environment and Climate change dischargeable effluent limit. Large green algal blooms have been observed to occur in conjunction with high pH levels.

The treatment performance of the WSP was examined first, particularly with regards to pathogenic disinfection and other water quality parameters over a one-year study. Interactions between algae growth and water quality parameters were subsequently examined, during both the short period of algal bloom and decay, as well as the long-term relationships between parameters over a five-year period. Ultimately, the breadth of this research aims to provide suggestions for improving the design and operation of the system to mitigate the detrimental impacts of excessive algal growth.

! 4 ! ! 1.3.1.!Thesis Layout ! Chapter 2 presents a comprehensive literature review of WSPs, in particular, types of

WSPs, pollutant removal mechanisms, the role of algae and other macrophytes as well as the relationship between algal growth and pH. This chapter also provides a background on statistical methods extensively used throughout this thesis, such as correlation analysis, multivariate statistical methods and time-series analysis.

Chapter 3 focused the investigating the effect of water quality and meteorological parameters on algal growth, from both a spatial and temporal perspective. From May to August

2017, surface algal cover was quantified using image processing techniques in conjunction with collected water quality parameters and climate data. Multivariate statistical methods, namely correlation analysis, PCA and DA, were employed to determine the most significant factors affecting surface algal growth.

The purpose of Chapter 4 was to determine significant water quality parameters affecting disinfection performance of the WSP. E. coli, C. perfringens, Enterococci and Total Coliform

(TC) were selected as bacterial indicators of disinfection. Physical, chemical, and biological water quality parameters were collected from October 2016 to August 2017 in parallel with the microbial indicators. Correlation analyses were conducted to provide information regarding types of relationships between bacterial indicators and water quality parameters. PCA was used to determine the water quality parameters with the most impact on disinfection performance of the system.

In Chapter 5, a study of government-regulatory effluent parameters, temperature, pH,

COD, DO, nitrate, phosphate, total phosphorus and suspended solids, over a long-term period

! 5 ! ! was conducted to reveal seasonal trends and relationships between water quality parameters.

Cross-correlation analysis and seasonal Kendall tests were conducted on a three-year period data set to determine any time lags present and prominent seasonal trends.

In Chapter 6, a statistical analysis was conducted using a comprehensive water quality data set over a five-year period. Time series analysis provided insight into the time-dependent relationships between parameters and multivariate statistics, namely FA/PCA and DA were used to delineate parameters responsible for majority of the seasonal variations in water quality.

Chapter 7 summarizes conclusions and future work from each study and contributions applicable to science and engineering. This research contributes to the field of environmental engineering, in particular, improvement wastewater treatment performance and resolving associated operational issues. Additionally, multivariate statistical methods are valuable tools in monitoring large water quality data sets over varying time periods to extract meaningful results.

Appendix A provides additional details regarding analytical methods and statistical analysis of raw data. Appendix B entails a chapter from the Water Quality in the Americas textbook commissioned by the Inter-American Network of Academies of Sciences (IANAS) organization. This book examines water quality and water treatment for countries worldwide, with this chapter providing insight on environmental, social and policy challenges surrounding water quality in a Canadian context. Personal writing contributions for the section on Canada include “Chapter 5: Coping with Sustainable Development Goals 6 (SDG-6)” and “Chapter 6:

Successful experiences in improving water quality” as well as edits, formatting and coordination within the Queen’s University chapter contributions.

! 6 ! ! 1.4.! Works Cited

Al-Hashimi, M. A. I., & Hussain, H. T. (2013). Stabilization pond for wastewater treatment.

European Scientific Journal May, 9(14), 1857–7881.

Champagne, P. (2017). Wastewater in High Latitudes Wastewater in High Latitudes, (January),

SCIENTIA GLOBAL.

Environment and Climate Change Canada. (2013). Impacts of Municipal Wastewater Effluents.

Retrieved January 27, 2018, from http://ec.gc.ca/eu-

ww/default.asp?lang=En&n=BE6F254A-1

Government of Canada. (1994). Summary and Update of the 1997 Science Assessment of the

Impacts of Municipal Wastewater Effluents ( MWWE ) on Canadian Waters and Human

Health.

Government of Canada. (2012a). Wastewater regulations overview. Retrieved February 10,

2018, from https://www.canada.ca/en/environment-climate-

change/services/wastewater/regulations.html

Government of Canada. (2012b). Wastewater Systems Effluent Regulations. Retrieved from

http://laws-lois.justice.gc.ca/eng/regulations/SOR-2012-139/FullText.html

Government of Canada. (2014). Wastewater pollution. Retrieved February 12, 2018, from

https://www.canada.ca/en/environment-climate-change/services/wastewater/pollution.html

Holeton, C., Chambers, P. A., Grace, L., & Kidd, K. (2011). Wastewater release and its impacts

on Canadian waters. Canadian Journal of Fisheries and Aquatic Sciences, 68(10), 1836–

1859. http://doi.org/10.1139/f2011-096

Metcalf & Eddy. (2003). Wastewater Engineering: Treatment and Resource Recovery (Fourth).

McGraw-Hill Science.

! 7 ! ! Muga, H. E., & Mihelcic, J. R. (2008). Sustainability of wastewater treatment technologies.

Journal of Environmental Management, 88(3), 437–447.

http://doi.org/10.1016/j.jenvman.2007.03.008

Neegan Burnside Ltd. (2011). National Assessment of First Nations Water and Wastewater

Systems - Ontario Regional Roll-Up Report. Retrieved from https://www.aadnc-

aandc.gc.ca/eng/1314634863253/1314634934122#chp3_3_1

Roshan R. Shrestha. (2008). Constructed Wetlands Manual. United Nations Development

Programme. Retrieved from www.unhabitat.org

Steinmann, C. R., Weinhart, S., & Melzer, A. (2003). A combined system of and

constructed wetland for an effective wastewater treatment. Water Research, 37(9), 2035–

2042. http://doi.org/10.1016/S0043-1354(02)00441-4

Sundaravadivel, M., & Vigneswaran, S. (2001). Constructed wetlands for wastewater treatment.

Critical Reviews in Environmental Science and Technology, 31(4), 351–409.

http://doi.org/10.1080/20016491089253

Tchobanoglous, G., & Crites, R. (1998). Small & Decentralized Wastewater Management

Systems. Michigan: WCB/McGraw-Hill.

United Nations. (2017). The United Nations World Water Development Report: Wastewater, The

Untapped Resource. Retrieved from

http://unesdoc.unesco.org/images/0024/002471/247153e.pdf

United Nations General Assembly. (2015). Transforming our world: The 2030 agenda for

sustainable development.

https://sustainabledevelopment.un.org/content/documents/7891Transforming%20Our%20W

orld. Pdf, (1), 1–5. http://doi.org/10.1007/s13398-014-0173-7.2

! 8 ! ! Wallace, J., Champagne, P., & Hall, G. (2016). Multivariate statistical analysis of water

chemistry conditions in three wastewater stabilization ponds with algae blooms and pH

fluctuations. Water Research, 96, 155–165. http://doi.org/10.1016/j.watres.2016.03.046

! 9 ! ! Chapter 2.! Literature Review

2.1.! Introduction ! Effective wastewater treatment is fundamental to safeguarding both human health and the natural environment. In Canada, governance of wastewater management is divided among federal, provincial, and municipal governments. The responsibility of water resources and wastewater management is primarily delegated to provinces and municipalities, as covered under section 109 of the Constitution Act of 1867. The Water Act (1985), Fisheries Act (1985) and

Environmental Protection Act (1999) are the predominant legislations guiding wastewater management in Canada (Government of Canada, 1985a, 1985b, 1999). The Fisheries Act (1985) and Canadian Environmental Protection Act (1999) are the governing legislation for wastewater treatment; outlining effluent limits, required treatment, and operational procedures (Government of Canada, 1985a, 1999).

Despite comprehensive regulations, adequate wastewater treatment continues to be an issue for Canada as wastewater is often improperly treated or discharged in the form of raw sewage into receiving environments. According to the Government of Canada (2012a) , more than 150 billion litres of untreated and raw sewage were discharged directly into Canadian and surrounding oceans annually. Prior to 2012, only recommended guidelines were provided as no enforceable federal standards for wastewater quality parameters existed. In 2012, the

Canadian Council of Ministers of the Environment (CCME) released the new federal Wastewater

Systems Effluent Regulations (WSER), under the Fisheries Act (Federation of Canadian

Munipalities, 2017; Government of Canada, 2012b). WSER outlines mandatory minimum effluent quality standards that must be achieved following secondary wastewater treatment

! 10 ! ! (Table 2.1) (Government of Canada, 2016). The revised federal regulations will require municipalities to undertake an estimated $20 billion in upgrades to existing infrastructure over the next 20 years, in addition to an estimated $80 billion required to replace aging water systems

(Canadian Water Network, 2014). Municipalities have a 30-year timeline to meet the minimum national performance standards (NPS) for effluent discharge parameters (Government of Canada,

2012b).

Provincial guidelines can be more comprehensive and differ slightly from federal regulations, as they are typically directed towards regional environmental issues. In Ontario, the

Ministry of Environment and Climate Change (MOECC) oversees and implements their own criteria for wastewater effluent quality parameters, with additional requirements for total phosphorus (TP), E. coli and pH (Table 2.1)

Table 2.1. Wastewater effluent discharge limits under WSER and Ontario provincial regulations (Government of Canada, 2012a; Ontario Ministry of Environment and Climate Change, 2017).

Wastewater Ontario Ministry of Systems Effluent Environment and Wastewater Quality Parameter Regulations Climate Change (WSER) (MOECC) Carbonaceous Biological Oxygen 25 25 Demand (mg/L) Unionized Ammonia-Nitrogen (mg/L) 1.25 0.02 (mg/L) 25 20 Total Phosphorus (mg/L) - 1.0 E. coli (CFU/100mL) - 200 pH (units) - 6.0-9.5

Many small and remote municipalities employ naturalized wastewater treatment systems, such as wastewater stabilization ponds (WSPs), as simple, accessible and cost-effective technologies to remain compliant with the WSER regulations. WSPs are designed to attenuate nutrients while also providing environmental conditions suitable for disinfection of pathogenic

! 11 ! ! and viral (Tchobanoglous & Crites, 1998). WSPs usually require low capital costs and minimal maintenance, providing accessible and sustainable, low energy wastewater technologies for communities with limited resources.

The effectiveness of WSPs depends on the physical, chemical and biological processes within these systems, as well as the complex interactions between biological communities, wastewater compositions and climatic conditions. Biological communities within WSPs, comprised mainly of algae, plants, and bacteria, heavily influence the efficiency of effluent polishing and nutrient recovery processes (Tchobanoglous & Crites, 1998). In particular, algae play an important role in nutrient removal and disinfection. However, the long hydraulic retention times (HRT) and deep light penetration in WSPs are also conducive to excessive algal growth, referred to as algal blooms (Davis, Berry, Boyer, & Gobler, 2009). Consequently, algal blooms can be responsible for regulatory exceedances in effluent parameters, such as pH, biochemical oxidation demand (BOD), dissolved oxygen (DO), E. coli and chlorophyll counts

(Steinmann et al., 2003; Tchobanoglous & Crites, 1998).

2.2.! Wastewater Stabilization Ponds

WSPs are large, shallow basins constructed for the treatment of raw sewage, primary domestic, secondary domestic or industrial wastewaters in a variety of climates, including polar, temperate and tropical regions (Shammas, Wang, & Wu, 2008). WSPs are used to remove nutrient loads in wastewater, mainly BOD, nitrogen, and phosphorus-based nutrients, as well as pathogenic microorganisms (Kayombo, Mbwette, Katima, & Jorgensen, 2003). WSPs have been reported to be effective in attenuating high nutrient loads and removing pathogens, such as bacteria, , helminths and protozoans (Curtis, Mara, & Silva, 1992). These ponds are typically constructed in a low permeability environment, enclosed with clay-based berms ! 12 ! ! and contain outlet structures, such as a drain and overflow weir to enable water level adjustment.

There are three general types of WSPs based on their intended functions and biochemical environments; anaerobic, facultative, and aerobic ponds (Sah, Rousseau, & Hooijmans, 2012).

They can be applied as a single-cell pond or an in-series, multiple cell system and sometimes in conjunction with a constructed wetland, depending on the intended application (Pearson, Mara,

& Bartone, 1987; Tchobanoglous & Crites, 1998).

2.2.1.! Types and System Characteristics of WSPs

2.2.1.1. Anaerobic ponds Anaerobic ponds, typically ranging from 2-10 m in depth, are used to treat wastewaters with high organic loads (usually exceeding 100 g BOD/m3 day). The low surface area to volume ratio minimizes water-atmosphere gas exchange, causing anaerobic conditions to dominate these ponds. The primary purpose is to treat high organic loads via of settled solids for BOD removal (Kayombo et al., 2003; Tchobanoglous & Crites, 1998). As such, they are most suitable for primary treatment in WSPs when organic loadings are high and are the first cell in a series of WSP systems. These ponds can efficiently treat concentrated wastewater over a relatively short period of time, approximately 1-1.5 days retention. Anaerobic digestion is temperature and pH-dependent, with increased digestion in temperatures above 15°C and decreased activity in conditions with pH<6.2 (Spuhler, 2011). BOD reductions can range between 50-85% in anaerobic ponds, depending on environmental conditions (Spuhler, 2011).

BOD removal is achieved either through the sedimentation of solids and the digestion of settled sludge or suspended organic matter by anaerobic microorganisms via acidogenesis, acetogenesis, and methanogenesis (Sah et al., 2012).

! 13 ! ! 2.2.1.2. Facultative ponds Facultative ponds, normally 1-2 m in depth, contain both aerobic and anaerobic zones for

BOD and pathogen removal. The upper, aerobic layer provides conditions suitable for algal growth due to the high surface area to volume ratio and higher temperatures (Tchobanoglous &

Crites, 1998). BOD and pathogen removal occur in the aerobic layer as a result of mutualistic relationships between algae and bacteria. Algae provide DO, as a by-product of photosynthesis, to bacteria while algae uptake their required carbon dioxide, ammonia and phosphorus nutrients produced by bacteria and readily available in wastewater (Figure 2.1). Facultative bacteria decompose dissolved and suspended organic matters, and solids subsequently settle in the lower layer. Settled organic matter form a sludge layer, decomposed by anaerobic bacteria (Mara et al.,

1992; Metcalf & Eddy, 2003; Tchobanoglous & Crites, 1998).

)!

Figure 2.1. Biological and chemical processes in a facultative WSP. ! 14 ! ! 2.2.1.3. Maturation ponds Maturation ponds, typically 1-1.5 m in depth, are aerobic ponds as light penetration and dissolved oxygen is present throughout the water column (Maynard, Ouki, & Williams, 1999).

The primary purpose of these ponds is pathogen removal, serving as tertiary treatment by improving water quality prior to discharge into the natural environment. The high surface area to volume ratio allows for higher light penetration, enabling higher rates of photosynthesis and oxygen transfer between the air-water interface. These conditions facilitate aerobic degradation of organics and UV disinfection processes (Curtis et al., 1992; Kadir & Nelson, 2014). Removal of bacteria, viruses, and helminths has been observed in maturation ponds, with up to 4-

6 log units for fecal coliforms and 2-3 log units for viruses (Pearson et al., 1987). Maturation ponds are often the last polishing step in a system consisting of multiple cell series, where the size and retention time vary depending on the regional standards and the incoming wastewater.

2.2.2.! Pollutant Removal Mechanisms in WSPs

2.2.2.1. Pathogens

Pathogen removal is a primary concern in wastewater treatment as there are approximately 60 identified types of waterborne pathogens contributing to human diseases (Wu,

Carvalho, Müller, Manoj, & Dong, 2016). Pathogenic organisms can include bacteria, viruses, protozoa, helminths and fungi, with bacterial pathogens being the most commonly found type in wastewater. Disinfection is typically assessed using bacterial indicator organisms such as E. coli, as representatives for pathogens. In WSPs, the process of pathogen removal involves a number of factors such as sunlight, pH, DO level, solids retention time (SRT), temperature, predation, attachment, sedimentation and nutrient availability (Awuah, Anohene, Asante, Lubberding, &

! 15 ! ! Gijzen, 2001; Liu, Macdougall, Hall, & Champagne, 2017; Maynard et al., 1999).

Sunlight, pH and DO are three primary mechanisms for pathogen removal (MacDougall,

Champagne, & Hall, 2017; Tyagi et al., 2011). DO, provided via either diffusion from the atmosphere or photosynthesis, is involved in disinfection processes found in the aerobic zone

(Schueder, Champagne, & Hall, 2016). Photo-oxidation requires the presence of oxygen to form reactive oxygen species (ROS), suggesting that higher DO concentrations lead to increased rates of photo-oxidation. Increased DO concentrations have been observed to decrease in E. coli and

Enterococci populations (Liu, Champagne, & Hall, 2018).

UV irradiation has a lethal impact on coliforms, with different wavelength regions, such as the ultraviolet (UV) spectrum (290 – 400 nm) and photosynthetically active radiation (PAR)

(400 – 700 nm), contributing to disinfection in wastewater (Bolton, 2010). The three UV-related mechanisms are: (1) direct damage to DNA by UV-B wavelengths (280 – 320 nm); (2) indirect endogenous damage involving UV-B; (3) indirect exogenous damage involving photosynthetically active radiation (PAR) (400 – 700 nm), UV-A (320-400nm) and UV-B

(Davis-Colley, Donnison, & Speed, 2000; Kadir & Nelson, 2014).

Additionally, pH affects the effectiveness of pathogen removal in WSPs. pH between 6.5 and 7.5 have been reported as optimal for fecal bacteria growth, while pH levels exceeding 8.5 and frequent fluctuations have been noted to negatively affect E. coli survival (Ansa,

Lubberding, Ampofo, Amegbe, & Gijzen, 2012; Awuah et al., 2001; Davis-Colley et al., 2000;

Liu et al., 2016a). HRT influences other factors collectively since longer HRT provides prolonged exposure for other factors contributing to disinfection to continue (Liu, Hall, &

Champagne, 2016b).

! 16 ! ! 2.2.2.2. Organic Matter

Organic matter in wastewater is comprised of a vast array of compounds, ranging from macromolecules, such as humic and fulvic acids, to smaller compounds, such as carbohydrates and amino acids. The main mechanisms for the removal of organic matter is bacterial oxidation in the aerobic zone, followed by sedimentation or additional anaerobic conversion in the benthic layer (Tchobanoglous & Crites, 1998). Organic matter is quantified using chemical oxygen demand (COD) and biological oxygen demand (BOD). COD is a measurement of the amount of total organic matter that can be chemically oxidized in a water sample, while BOD measures the

DO consumed by microorganisms in biologically oxidizing organic matter. COD accounts for all oxidizable organic matter in the wastewater, whereas BOD only accounts for the organic matter than can be consumed by microorganisms (Metcalf & Eddy, 2003).

2.2.2.3. Solids In wastewater, solids can be found either in suspended or dissolved form, and can be comprised of both organic and inorganic matter. Bacteria attached to surfaces of plants and sediment particles are primarily responsible for aerobic and anaerobic degradation of solids.

Facultative bacteria can decompose solids, which subsequently settle in the lower layer where they are further decomposed by anaerobic bacteria. Additionally, reductions in TSS have been shown to have statistically significant correlation with the reduction of E. coli (Huang,

Truelstrup Hansen, Ragush, & Jamieson, 2017). Rates of sedimentation and within the

WSP are dependent on a number of factors, such as temperature, flow regime and pond depth, and can be enhanced by chemical and biochemical flocculation. Although influent TSS concentrations are often reduced across a WSP, effluent levels may reach unacceptable levels

! 17 ! ! when algal growth rates are high, contributing up to 50-90% of TSS (Mara et al., 1992;

Shammas et al., 2008).

2.2.2.4. Nitrogen (N)

Two of the main forms of nitrogen (N) species important in monitoring WSP

+ - + performance are ammonium (NH4 ) and nitrate (NO3 ). Ammonium (NH4 ), prevalent in from

6.0 to 9.3 in aqueous solution, is produced via hydrolysis of organic nitrogen-containing

- compounds such as proteins and enzymes. Nitrate (NO3 ), a product of nitrification, is not commonly found in municipal wastewater and its presence often indicates the existence of agricultural runoff or prior nitrification (Metcalf & Eddy, 2003).

Nitrogenous (N) compounds are primarily removed via plant uptake, nitrification- denitrification processes, settling of detritus and volatilization. Nitrogen (N) can be incorporated into plant biomass via nitrogen assimilation. Cell growth converts soluble inorganic nitrogen into particulate organic nitrogen (Ferrara & Avci, 1982). At higher pH levels exceeding 9.3, formation of gaseous ammonia (NH3) can occur. The rate of volatilization is governed by

Henry’s Law, depending on the aqueous and atmospheric equilibrium (Tchobanoglous & Crites,

1998). Volatilization is typically higher during the summer when pH levels and temperatures are higher (Shammas et al., 2008).

Remaining ammonia (NH3) is removed via biological nitrification and denitrification processes. Ammonia (NH3) is converted to nitrite by Nitrosomonas bacteria, and subsequently

- into nitrate by Nitrobacter species. Nitrate (NO3 ) is then converted to nitrogen gas (N2) by

Pseudomonas and Achromobacter species through the process of denitrification, which occurs in the anaerobic sludge layer of facultative and maturation WSPs (Tchobanoglous & Crites, 1998).

! 18 ! ! 2.2.2.5. Phosphorous (P)

Phosphorous (P) is often a limiting factor in the growth of plants and algae in natural environments (Wetzel, 2001). There are generally three groups of phosphorous (P) species in wastewater: (1) Orthophosphates: simple and biologically available phosphorous compounds,

- 2- 3- including H3PO4, H2PO4 , HPO4 , and PO4 ; (2) Polyphosphates: derivatives of larger molecules which can be hydrolyzed into orthophosphates; (3) Organic phosphorous: larger compounds which are generally a larger concern (Metcalf & Eddy, 2003).

3- Phosphate (PO4 ) can be sequestered by algae and incorporated into their biomass. In certain conditions involving long HRT, warmer temperatures and high nutrient loading, WSP systems are susceptible to excessive algal growth (Schueder et al., 2016; Wallace, Champagne,

Hall, Yin, & Liu, 2015). Once algae decay, they settle and form the benthic layer of sludge

(Dickinson, Whitney, & McGinn, 2013). At higher pH levels (>8), phosphorous can precipitate out into sediments with calcium and magnesium (Larsdotter, La Cour Jansen, & Dalhammar,

3- 2007). The release of both orthophosphate (PO4 ) and polyphosphate originating as microbial and algal biomass, can also occur at low oxygen concentrations, redox potentials, and high temperatures (Powell, Shilton, Pratt, & Chishti, 2011; Vendramelli, Vijay, & Yuan, 2016).

Phosphorous (P) can also be removed through processes such as adsorption, complexation, and precipitation with aluminum, iron and calcium, as well as clay minerals in the sediment layer.

2.2.3.! Algae and Macrophyte in WSPs

WSP systems have large surface areas with exposure to sunlight, high nutrient fluxes, and long HRT, providing suitable conditions for the growth of algae and other aquatic plants. Algae are a broadly defined class of predominantly autotrophic aquatic organisms, ranging from

! 19 ! ! microscopic, unicellular microalgae to macroscopic, multicellular forms. Algae and other plant matter are predominant sources of new organic matter and energy that drives system dynamics in aquatic environments (Wetzel, 2001). In WSPs, they can affect DO and pH levels, nutrient assimilation, adsorption of heavy metals, decay of organic matter, and settling of suspended solids (Abdel-Raouf, Al-Homaidan, & Ibraheem, 2012). Algae can treat wastewater through a combination of processes involving nutrient uptake, elevated pH, and high DO concentrations. In appropriately designed systems, algae-mediated treatment can be considered a less expensive and more effective method of removing nutrients and metals than conventional tertiary treatment

(Hoffmann, 1998; Oswald & Golueke, 1966; Oswald & Gotaas, 1985; Redalje et al., 1989).

2.2.3.1. Diversity of Algae and Macrophytes

The types of algae found in WSPs vary significantly and are dependent on the site- specific environmental conditions, such as regional climate, season, site conditions, wastewater constituents and flow (Hosetti & Frost, 1998; Mara et al., 1992; Shammas et al., 2008). Green algae, Chlorophyta, are typically the most dominant and broad taxa, in terms of variety and quantity, in lagoon systems (Table 2.2). Chlorophyta are a large and morphologically diverse algal group found in both freshwater and municipal wastewater systems (Barsanti & Gualtieri,

2006; Tchobanoglous & Crites, 1998). Spirogyra and Cladophora spp., both types of benthic filamentous green macroalgae, are two naturally abundant Chlorophyta species found in eutrophic wastewaters (Simons & van Beem, 1990). These algae can form free-floating mats found along downwind shores of ponds and are capable of undergoing sudden changes in buoyancy to avoid more extreme weather periods. Another common species, Hydrodictyon reticulatum, aggregate and form free-floating single-layered cell colonies attached to

Cladophora. H. reticulatum frequently form these extensive surface and subsurface mats in ! 20 ! ! shallow regions of ponds during low flow conditions in the summer months. Other Chlorophyta species commonly found in wastewater include Chlorella spp., Scenedesmus spp.,

Chlamydomonas spp. (Schueder et al., 2016; Wallace et al., 2015). Depending on the in ponds, microalgal composition may vary as motile species, such as Euglena and

Chlamydomonas, dominate in more turbid water compared to non-motile algae, such as

Chlorella, which are more prevalent in less turbid waters (Abdel-Raouf et al., 2012).

Table 2.2.Overview of algae and plant species commonly found in WSPs.

Plant Common species Green Algae (Chlorophyta) Spirogyra spp. Cladophora spp. Bacillariophyta. Euglenophyta spp. Hydrodictyon spp. Chlorella spp. Scenedesmus spp. Chlamydomonas spp. Micractinium spp. Golenkinia spp. Submerged &/or free- Lemna spp. floating macrophytes Spirodella spp. Wolffia spp. Myriophyllum spicatum Potamogeton spp. Eichhornia crassipes Pistia stratiotes Ipomoea aquatica Ceratophyllum demersum Elodea spp.

Aquatic plants also play a role in treatment of wastewater in WSP. Similar to algae, macrophytes are also capable of reducing nutrient loads, removing COD and BOD, as well as decreasing heavy metal concentrations in wastewater. Aquatic macrophytes are commonly found in wastewater and have extensive capacity to uptake nutrients, similar to free-floating

! 21 ! ! algae (Wolverton & McDonald, 1979). Examples of common free-floating macrophytes include

Lemna spp., Eichhornia crassipes and Elodea spp. (Table 2.2). The broad surface and root zones of macrophytes act as habitats for epiphytic algae and bacteria. For instance, Lemna spp. and bacteria have a symbiotic relationship, with Lemna spp. providing oxygen required for heterotrophic bacteria and bacteria decomposing organic matter. Organic solids can be directly assimilated into carbohydrates and various amino acids, removing nutrients, BOD and COD

(Hillman, & Culley, 1978). Macrophytes are capable of removing nutrients and contaminants through both direct assimilation into tissues and providing suitable environments for algae and microorganisms.

2.2.3.2. Algal Growth

99.9% of algal biomass is composed of carbon, hydrogen, oxygen, nitrogen, phosphorus and sulphur, while the remaining 0.01% is most comprised of other elements, such as calcium, potassium, sodium, chlorine, magnesium, iron, and silicon (Barsanti & Gualtieri, 2006). The nutrient in the lowest quantity in relation to the requirements of the algae is defined as the limiting nutrient, and limits algal growth rates. Of these elements, nitrogen (N) and phosphorus

(P) are considered to be the most important nutrients (Shammas et al., 2008). For growth of algae, nitrogen (N) is an important nutrient as it is required for many biological constituents such as proteins, chlorophyll, and enzymes. While nitrogen (N) is a key for algae, excessive concentrations will negatively affect algal growth. The optimal range of ammonia (NH3) is

+ between 12-20 mg/L, while ammonium (NH4 ) and nitrate (NO3-) concentrations exceeding 45 mg/L and 35 mg/L, respectively, have been shown to inhibit algal growth (Shammas et al.,

3- 2008). Phosphate (PO4 ) is also an important nutrient for algal growth and is often seen as the rate-limiting nutrient. Nitrogen (N) is often observed as the limiting nutrient in high and ! 22 ! ! low flow marine systems, while phosphorus (P) is typically the limiting nutrient in freshwater

(Barsanti & Gualtieri, 2006).

2.2.3.3. Algae photosynthesis

Algae assimilate CO2 in the presence of sunlight and produce O2 and sugar, in a process referred to as photosynthesis. Photosynthesis is a temperature-dependent process, with algal and plant growth primarily occurring during spring, summer and fall in temperate climates, such as

Canada. From a study examining the effect of light, temperature and organic loading rates on

WSP dynamics, temperature was observed to have the greatest overall influence on phytoplankton growth and population (Ragush & Jamieson, 2016). Chlorophyll, the green pigment found in the cells of green algae and macrophytes, is the primary molecule involved in photosynthesis as it absorbs solar radiation used for glucose conversion. There are numerous types of chlorophyll, with chlorophyll-a (chl-a) being the most common type (Barsanti &

Gualtieri, 2006; Speers, 1997). Between 430 to 670 nm, photosynthetic activity is maximized, corresponding to the absorption maxima for chl-a (Whitmarsh & Govindjee, 1999). Chlorophyll is fluorescent and can be a measured using fluorometric instruments, providing an indication of algae concentrations in WSPs (Jakob, Schreiber, Kirchesch, Langner, & Wilhelm, 2005).

Algal photosynthesis is an autotrophic process whereby carbon dioxide (CO2) is assimilated to produce organic matter for energy, as shown below:

)*+,-,///0,)1213,4)) 6"#$ + 12($# /"5(6$#5 + 6($# + 6#$ (Equation 2.1.)

At the water-atmosphere interface, atmospheric CO2 is dissolved in water to form dissolved carbonic acid. Increased pH levels during high temperatures can be attributed to biological activity as the carbonate – bicarbonate equilibrium is impacted by algae using carbon dioxide

(Ragush et al., 2015). Algal uptake of CO2 causes a shift in the acid-base equilibrium between ! 23 ! ! - 2- dissolved CO2, carbonic acid (H2CO3), bicarbonate ions (HCO3 ), and carbonate ions (CO3 )

(Uusitalo, 1996):

;<.> 7896: ∗ "#$ +/($#/ /($"#? (Equation 2.2.)

7 96:;A.B < C D ($"#? /("#? + ( (Equation 2.3.)

7 96:;

2.2.3.4. Role of Algae in WSP

Algae have been reported to play a significant role in wastewater treatment, ranging from disinfection, BOD/COD removal, to nutrient removal, and sequestration of heavy metals (Abdel-

Raouf et al., 2012). Diurnal variations in CO2, pH and DO, as a result of algal photosynthesis, affect various treatment mechanisms such as the precipitation of metals and phosphorous, volatilization of ammonia, bacterial activity and pathogenic disinfection (Amengual-Morro et al.,

2012).

Algal growth contributes to elevated pH and DO levels, which aids in pathogen inactivation (Liu et al., 2016a). A study conducted with algae-treated domestic sewage reported a reduction of 88.8% in 11.4 days while another study reported a reduction of 99.6% in 10 days

(Meron, Rebhun, & Sless, 1965; Sebastian & Nair, 1984). During periods of high solar radiation, the photosynthetic activity of algae increases, shifting the carbonate system to produce more OH- ions (Mayes et al., 2009). Higher pH levels have been shown to be beneficial for pathogen removal. For instance, pH levels higher than 8.5 have been observed to be effective for the removal of indicator organisms, while fluctuations in pH have been shown to negatively affect the survival of E. coli (Awuah et al., 2001; Davis-Colley et al., 2000).

! 24 ! ! Algae also exhibit a high capacity for the uptake of inorganic nutrients and heavy metals in wastewater (Blier, Laliberté, & de la Noüe, 1995). In a study investigating the use of native

Ecuadorian algae in secondary wastewater treatment, results indicated that microalgal cultures

+ 3- - decreased ammonium (NH4 ), phosphate (PO4 ) and nitrate (NO3 ) levels by 55.6%, 20.4% and

93.1%, respectively, in aerated photobioreactors (Benítez, Champagne, Ramos, Torres, &

Ochoa-Herrera, 2018). In another study examining enhanced nutrient removal using Chlorella vulgaris, all nutrients were exhausted within the 12-day cultivation period with 99.0% removal

+ - of total nitrogen, ammonium (NH4 ) and nitrate (NO3 ), and 100% for removal of phosphate

3- (PO4 ) (Ge et al., 2018). Macroalgae, such as Spirogyra sp., have also been shown to be capable of nitrogen and phosphorous removal in diluted primary, secondary and centrate wastewaters

(Ge, Madill, & Champagne, 2016). Furthermore, algal species are capable of heavy metal sequestration from aqueous solutions through different mechanisms as trace nutrient metals can be accumulated intracellularly via active biological transport (Leusch, Holan, & Volesky, 1995;

Rehman & Shakoori, 2001; Wilde & Benemann, 1993; Yee, Benning, Phoenix, & Ferris, 2004).

2.3.3.5!Effect of algal blooms on wastewater quality

Algal blooms can occur in facultative WSPs, as they are particularly susceptible to external meteorological conditions. Wind within the upper water layer contributes to mixing and a uniform distribution of BOD, DO, bacteria and algae, resulting in a higher degree of waste stabilization (Tchobanoglous & Crites, 1998). Conversely, low winds may lead to insufficient mixing and vertical stratification of the water column (Tchobanoglous & Crites, 1998). In temperate regions, changing temperatures increase mixing, while stagnant, warmer temperatures induce vertical stratification (Wetzel, 2001). Additionally, eutrophication, often characterized by

! 25 ! ! algal blooms, leads to increased pH, organic matter and TSS, DO depletion, and reduced biodiversity in receiving systems (Barsanti & Gualtieri, 2006).

During periods of high photosynthetic activity, pH levels can rise to above 9.0 and have significant impacts on aquatic wildlife (Barsanti & Gualtieri, 2006; Kayombo et al., 2003;

Veeresh, Veeresh, Huddar, & Hosetti, 2010). As the rate of photosynthesis increases, DO levels may rise above normal saturation levels, termed supersaturation. Supersaturation can cause gas bubble trauma, whereby gas bubbles form in the internal organs and tissue of fish and affect metabolic activity (Boyd, Watten, Goubier, & Wu, 1994). As pH and ammonia levels increase, toxicity to aquatic wildlife also becomes more severe. For instance, fish may suffer from changes in gill tissues, loss of equilibrium, hyperexcitability, increased respiratory activity, and increased heart rate (Randall & Tsui, 2002; Smart, 1975, 1978) . High levels of ammonia ions suppress proper excretion of ammonia and subsequently high blood-ammonia levels, potentially resulting in convulsions, coma, or death (Smart, 1978).

2.2.4.! WSP Summary

Wastewater stabilization ponds (WSPs) can effectively facilitate treatment with a wide range of characteristics, facilitated by naturally-occurring biogeochemical processes. Complex aquatic ecosystems, involving algae, plants and bacteria, can attenuate nutrient loads and reduce pathogenic organisms in WSPs in a variety of climactic conditions. As WSPs subject to both seasonal weather patterns and wastewater variations, they can be highly complex open systems, presenting challenges for monitoring and performance prediction. While algae are involved in treatment mechanism for disinfection, nutrient removal and removal of heavy metals, algal blooms can lead to potential issues. Algal blooms can result in pH levels that exceed government effluent limits and also lead to detrimental impacts on aquatic environments receiving the ! 26 ! ! effluent discharge. Thus, a comprehensive understanding of the biological, chemical, and physical conditions that influence wastewater quality within WSPs is crucial to effective wastewater treatment and safeguarding our water resources.

2.3.! Statistical Methods

2.3.1.! Correlations Analysis

Correlation coefficient tests can provide an indication of the relationship between variables and how they change with respect to each other, with correlation coefficients indicating the strength of the statistical relationships. While there are a number of tests, each with unique characteristics, they generally assume coefficient values in the range from −1 to +1, with +1 indicating the strongest positive correlation and −1 the strongest negative correlation. However, correlation tests often only indicate positive and negative correlations, not causation and magnitude of the relationship (Lee & Nicewander, 1988).

Spearman’s ρ correlation, one of the most commonly used correlation coefficient test, is a nonparametric measure of rank correlation, indicating the statistical dependence between the rankings of two variables. This test can be applied when the data is nonparametric, does not fit any particular distribution and is appropriate for both continuous and discrete ordinal variables.

Using less assumptions regarding distribution than parametric tests, the results describe suitability of a monotonic relationship between two variables (Lee & Nicewander, 1988).

Spearman correlation between two variables will be high when observations have a similar rank and low when observations have a low rank between the two variables.

! 27 ! ! The Spearman correlation coefficient is defined and calculated as the Pearson correlation coefficient (r) between the ranked variables (Myers & Well, 2003). The Pearson’s correlation coefficient is a measure of the bivariate linear relationship of raw values, calculated as follows:

J/( LM)C( L)( M) GH = / (Equation 2.5.) (J LOC LO)(J MOC MO)

Where rp represents the dimensionless Pearson correlation coefficient x represents one variable, y represents the second variable n represents the number of observations

Subsequently, the equation for Spearman correlation coefficient test is shown as follows, assuming all n ranks are distinct integers:

S TO G = Q −/ U (Equation 2.6.) P J/(JOCQ)

Where VW represents the dimensionless Spearman correlation coefficient X* represents the difference between the two ranks of each observation X* represents the rank difference of sample values, shown as VY Z* − VY [* / n represents the number of observations

As Spearman’s correlation coefficient calculates the linear relationship between ranked variables of individual samples (xi, yi), instead of the samples themselves, the weight of the correlation coefficient indicates the strength of the relationship (Rodgers & Wander, 1988).

2.3.2.! Multivariate statistical analysis

2.3.2.1.Principal Component Analysis (PCA) Principal Components Analysis (PCA), a multivariate statistical method, is a type of unconstrained ordination analysis used to understand the variables that contribute most to the variance observed in samples (Abdi & Williams, 2010; Jolliffe, 1986). PCA compresses highly dimensional and correlated data into a few uncorrelated latent variables (combinations of original variables), by reducing the number of original variables.

! 28 ! ! By reducing the dimensionality of the data matrix into bilinear terms, latent patterns and relationships that drive processes and phenomena in the data are more easily uncovered. To obtain this, an orthogonal transformation is performed to convert the set of variables into a set of linearly uncorrelated values, referred to as Principal Components (PCs) (Wold, Esbensen, &

Geladi, 1987). The number of PCs produced is equal to number of variables minus one. PCs have eigenvectors and eigenvalues (λ) in the new covariance matrix, which describes the variance of the data and covariance among variables. Eigenvectors are used to reorient the data from the original xy-axis to new PC axes. Eigenvalues, the coefficients of eigenvectors, provide the axes magnitude and serve as a measure of the data’s covariance. Eigenvectors are ranked in terms of the eigenvalues in descending order, listing the PCs in order of significance. The first

PC is the linear combination of x-variables that accounts for the maximum variance within all linear combinations. Each succeeding PC will account for the next highest variance possible, subject to the constraint that each new component will be orthogonal to previously defined PCs.

The resulting PCs are subject to various criteria that determine retained components and, consequently, the retained variables (Jackson, 1993; J R King & Jackson, 1999). Various heuristic procedures are used as stopping rules in PCA to determine PC retention, with the most popular methods described below:

(1)!Kaiser criterion: PCs with eigenvalues exceeding 1.0 are retained based on the rationale

these PCs summarize more information than any single original variable (Guttman, 1954;

Kaiser, 1960). This is the most commonly used stopping rule in ecology.

(2)!Scree plot: A graphical representation of PCs with PCs on the x-axis and their

corresponding eigenvalues on the y-axis in a successive rank order (Figure 2.2). The point

or bend at which the eigenvalues deviate from the steeper section into a flattened line

! 29 ! ! determines which PCs are retained. Depending on the interpretation method, the

components directly to the left of the straight line segment or including the first PC to the

right are retained (Cattell, 1966; Cattell & Vogelmann, 1977; Ledesma, Valero-Mora, &

Macbeth, 2015).

(3)!Broken-Stick distribution model: A theoretical stick of unit length is divided into the same

number of pieces as there are PCA axes and then sorts these pieces from the longest to the

shortest. This procedure is repeated with all possible permutations computed, producing an

average of all computed eigenvalues. With this model, total variance is assumed to be

proportional among components. Observed components whose eigenvalues are larger than

the corresponding eigenvalues generated by the Broken-Stick model serve as the stopping

point (King & Jackson, 1999).

4 (1) Kaiser Criterion: λ>1 (2) Scree Plot (3) Broken Stick Distribution Model

3

2 Eigenvalue Eigenvalue

1

0 0 2 4 6 8 10 12 PC Figure 2.2. Common stopping procedures for PC selection displayed: Kaiser criterion, Scree plot and Broken-stick distribution model.

! 30 ! ! PCs are then extracted and rotated, using the Varimax normalization, to determine the rotated factor loadings for each original variable (Kaiser, 1960). PC loadings represent the correlation between the original variables and PCs, indicating the weight by which each standardized original variable should be multiplied to get the component score. Interpretation of the PCs is heavily reliant on loadings, indicating which of the original variables contribute most, or are most heavily weighted, towards a certain component (Peres-Neto, Jackson, & Somers,

2017).

However, PCA is sensitive to missing data and data sets with missing entries and generally require some form of pre-processing to remove outliers or to fill in using estimated values (Pop, Einax, & Sârbu, 2009). For a few missing values, removal of the sample row(s) is acceptable. However, with more incomplete data sets, removal of row(s) with incomplete entries can exclude too much information and lead to incomplete analysis. Thus, imputation is commonly used and produces satisfactory results (Abdi & Williams, 2010). Despite these shortcomings, PCA is widely used in ecological studies, providing an objective and reliable statistical method for managing large, multi-parameter data sets (Arhonditsis, Brett, & Frodge,

2003; Reisenhofer, Picciotto, & Li, 1995; Zitko, 1994). PCA is also becoming an increasingly more prevalent statistical method used in monitoring wastewater treatment and assessing water quality (Barbieri et al., 1999; Bengraïne & Marhaba, 2003; Iordache, & Dunea, 2013; Perkins &

Underwood, 2000; Singh et al., 2004; Wallace, 2013; Yidana et al., 2008).

2.2.3.2. Discriminant Analysis (DA)

DA is used to determine whether a variable known a priori is effective in predicting category membership for a set of observations. This method can be used as either a linear classifier or for dimensionality reduction before classification. DA creates one or more linear combinations of

! 31 ! ! predictors, referred to as Discriminant Functions (DFs), creating a new latent variable for each

DF that will discriminate between categories of the dependent variable. The first DF created maximizes the differences between groups on that function. The second DF maximizes differences on that function, but also must not be correlated with the previous DF. This continues with subsequent functions with the requirement that the new function not be correlated with any of the previous functions (Wolfgang & Simar, 2012). Using original data, a discriminant function

(DF) is created for each group using the following equation:

b \ ]* = / ^* +/ `96 _*`a*` (Equation 2.7) Where j represents number of variables i represents the component represented as kth component bk represents the size of λ for i i is the number of groups (G) ki is the constant inherent to each group n is the number of parameters used to classify dataset into a group wj is the weight coefficient assigned by DA to a given parameter (pj).

This procedure continues until the maximum number of DFs produced, determined by the number of predictors and categories in the dependent variable (Alberto et al., 2001; Singh et al.,

2004).

Forward stepwise and backward stepwise modes are two different DA methods, where a model of discrimination is built step-by-step either forward or backward. All variables are evaluated to determine which ones contribute most to the discrimination between groups at each step. In forward stepwise DA mode, variables are included step-by-step starting with the highest significance and moves downward. According to this approach, all variables are retained.

Alternatively, backward stepwise mode uses the opposite approach and variables are removed beginning with the least significant until no significant variables remain. In this case, all variables are retained in the model and then, at each step, the variable that contributes least to the

! 32 ! ! prediction of group membership is eliminated (McGarigal, Cushman, & Stafford, 2000).

According to this approach, only the variables that contribute the most to the discrimination between groups are kept.

The classification table, also known as a confusion, assignment or prediction matrix, compares the predicted results by the logistic regression model compared to the actual group membership. By comparing these two values, the suitability the derived function in predicting group membership can be quantitatively assessed (Wolfgang & Simar, 2012).

Linear DA (DA) is related to analysis of variance (ANOVA) and regression analysis, which all express one dependent variable as a linear combination of other variables. However,

ANOVA uses categorical independent variables and a continuous dependent variable, whereas

DA uses continuous independent variables and a categorical dependent variable. DA is also closely related to principal component analysis (PCA) as they both attempt to find linear combinations of variables which best explain the variance in data. However, for DA, a distinction between independent variables and dependent variables is made.

DA can guide the interpretation of complex data sets to better understand spatial and temporal changes in large systems. In particular, DA can facilitate the identification of possible factors that could influence water systems, providing valuable information for water resources management. This method has been especially popular in in freshwater and groundwater research, in particular, exploring and evaluating the impacts of polluting and natural sources on water quality (Shrestha & Kazama, 2007; Sangam Shrestha, Kazama, & Nakamura, 2008; Singh et al., 2004). In recent years, these methods have also been applied to wastewater treatment and issues associated with pathogen detection, microbial populations and pollutants in wastewater effluent (Amaral et al., 2004; Harwood, Whitlock, & Withington, 2000; Singh et al., 2005). ! 33 ! ! 2.3.3.! Time Series Analysis Time series analysis is used to extract meaningful statistics and other characteristics from time-dependent data. The analysis requires for the data to have a natural temporal ordering and assumes a non-parametric approach (Brockwell & Davis, 1997).

2.3.3.1. Seasonal Kendall

The Seasonal Kendall test (SK test) is a nonparametric test that analyzes data for monotonic (consistently increasing or decreasing) trends in seasonal data. The SK test is a special case of the Mann Kendall Trend Test, which accounts for seasons within a data set. The

Mann Kendall test is nonparametric test used to identify a trend in a series over time, where no assumption of normality is required. The overall Sk statistic is calculated from summing each season’s Kendall S statistic:

e cd = *96 c* (Equation 2.8) Where m is the number of seasons S is the seasonal data set

The seasonal Kendall test runs a separate Mann Kendall trend test on each of m seasons separately so that data is compared to the same season. For example, Spring would be only be compared with Spring and Summer would only be compared with Summer.

For passive wastewater treatment systems, multivariate statistics are useful exploratory data analysis techniques that examine spatial and temporal trends and relationships between variables in multi-parameter datasets. As WSP are highly dynamic systems affected by their natural environment, it can be difficult to fully characterize a system, assess treatment performance and monitor biogeochemical interactions. Thus, correlations analysis, PCA and DA are multivariate techniques that can be used to improve understanding and interpretation of large environmental datasets containing many interdependent variables. Time series analysis, another ! 34 ! ! statistical tool, is particularly helpful for understanding temporal relationships between variables and revealing the presence of dynamic time lags between variables.

2.4.! Summary

WSP systems are effective low-cost and low-energy passive wastewater treatment technologies extensively used in Canadian small, rural, remote and Indigenous communities.

There are three primary types of WSPs, anaerobic, facultative and aerobic, which are employed for different treatment objectives. WSPs use complex biochemical and physical interactions to remove pollutants such as organic matter, pathogens, nitrogen, phosphorus and suspended solids.

WSPs rely heavily on algae and macrophytes for wastewater treatment, as they are capable of directly assimilating nitrogen and phosphorus-based compounds, reducing BOD/COD as well as creating conditions favourable for other treatment mechanisms by influencing water quality parameters. Algae and aquatic macrophytes play a particularly essential role in the treatment of wastewater. However, given specific environmental conditions, excessive algal growth can lead to algal blooms and high pH levels, which can exceed regulatory limits and have detrimental impacts to aquatic ecosystems.

Evidently, WSPs are highly complex systems whose performance is affected by many interacting biochemical and physical processes making difficult to understand and predict their performance at times. Analytical tools are crucial to developing a better understanding of the interacting factors and are used in modelling and performance prediction in wastewater management. Multivariate statistical techniques are useful for determining the type, direction and strength of relationships between parameters in large datasets, revealing meaningful, time- dependent relationships between variables. This research will contribute to the continued improvement and design of naturalized wastewater treatment technologies in a changing climate. ! 35 ! ! 2.5.!References

Abdel-Raouf, N., Al-Homaidan, A. A., & Ibraheem, I. B. M. (2012). Microalgae and wastewater

treatment. Saudi Journal of Biological Sciences, 19(3), 257–275.

http://doi.org/10.1016/j.sjbs.2012.04.005

Abdelaziz, A. E. M., Leite, G. B., Belhaj, M. A., & Hallenbeck, P. C. (2014). Screening

microalgae native to Quebec for wastewater treatment and biodiesel production.

Bioresource Technology, 157, 140–148. http://doi.org/10.1016/j.biortech.2014.01.114

Abdi, H., & Williams, L. J. (2010). Principal Component Analysis. Wiley Interdisciplinary

Reviews: Computational Statistics, 2(4), 433–470. http://doi.org/10.1002/wics.101

Alberto, W. D., Marı́a del Pilar, D., Marı́a Valeria, A., Fabiana, P. S., Cecilia, H. A., & Marı́a de

los Ángeles, B. (2001). Pattern Recognition Techniques for the Evaluation of Spatial and

Temporal Variations in Water Quality. A Case Study:: Suquı́a Basin (Córdoba–

Argentina). Water Research, 35(12), 2881–2894. http://doi.org/10.1016/S0043-

1354(00)00592-3

Amaral, A. L., Da Motta, M., Pons, M. N., Vivier, H., Roche, N., Mota, M., & Ferreira, E. C.

(2004). Survey of Protozoa and Metazoa populations in wastewater treatment plants by

image analysis and discriminant analysis. Environmetrics, 15(4), 381–390.

http://doi.org/10.1002/env.652

Amengual-Morro, C., Moyà Niell, G., & Martínez-Taberner, A. (2012). Phytoplankton as

bioindicator for waste stabilization ponds. Journal of Environmental Management,

95(SUPPL.), S71–S76. http://doi.org/10.1016/j.jenvman.2011.07.008

Ansa, E. D. O., Lubberding, H. J., Ampofo, J. A., Amegbe, G. B., & Gijzen, H. J. (2012).

Attachment of faecal coliform and macro-invertebrate activity in the removal of faecal

! 36 ! ! coliform in domestic wastewater systems. Ecological Engineering, 42, 35–

41. http://doi.org/10.1016/j.ecoleng.2012.01.018

Arhonditsis, G. B., Brett, M. T., & Frodge, J. (2003). Environmental control and limnological

impacts of a large recurrent spring bloom in Washington, USA. Environmental

Management. http://doi.org/10.1007/s00267-002-2891-4

Awuah, E., Anohene, F., Asante, K., Lubberding, H., & Gijzen, H. (2001). Environmental

conditions and pathogen removal in macrophyte- and algal-based domestic wastewater

treatment systems. Water Science and Technology, 44(6), 11 LP-18. Retrieved from

http://wst.iwaponline.com/content/44/6/11.abstract

Barbieri, P., Andersson, C. A., Massart, D. L., Predonzani, S., Adami, G., & Reisenhofer, E.

(1999). Modeling bio-geochemical interactions in the surface waters of the Gulf of Trieste

by three-way principal component analysis (PCA). Analytica Chimica Acta, 398(2), 227–

235. http://doi.org/10.1016/S0003-2670(99)00418-3

Barsanti, L., & Gualtieri, P. (2006). Algae: Anatomy, Biochemistry, and Biotechnology (Second).

Boca Raton, USA: CRC Press.

Bengraïne, K., & Marhaba, T. F. (2003). Using principal component analysis to monitor spatial

and temporal changes in water quality. Journal of Hazardous Materials, 100(1–3), 179–

195. http://doi.org/10.1016/S0304-3894(03)00104-3

Benítez, M. B., Champagne, P., Ramos, A., Torres, A. F., & Ochoa-Herrera, V. (2018).

Wastewater treatment for nutrient removal with Ecuadorian native microalgae.

Environmental Technology (United Kingdom), 0(0), 1–9.

http://doi.org/10.1080/09593330.2018.1459874

Blier, R., Laliberté, G., & de la Noüe, J. (1995). Tertiary treatment of cheese factory anaerobic

! 37 ! ! effluent with Phormidium bohneri and Micractinum pusillum. Bioresource Technology,

52(2), 151–155. http://doi.org/https://doi.org/10.1016/0960-8524(95)00014-6

Boyd, C. E., Watten, B. J., Goubier, V., & Wu, R. (1994). Gas supersaturation in surface waters

of ponds. Aquacultural Engineering, 13(1), 31–39. http://doi.org/10.1016/0144-

8609(94)90023-X

Brockwell, P. J., & Davis, R. A. (1997). Springer Series in Statistics. Springer Series in

Statistics. http://doi.org/10.1007/978-0-387-98135-2

Canadian Water Network. (2014). 2014 Canadian Municipal Water Priorities Report. Retrieved

from http://www.cwn-rce.ca/initiatives/municipal-consortium/2014-canadian-municipal-

water-priorities-report/

Cattell, R. (1966). The Scree Test For The Number Of Factors: Multivariate': Vol 1, No 2.

Behavioral Research, 1(2), 245–276.

http://doi.org/http://dx.doi.org/10.1207/s15327906mbr0102_10

Cattell, R. B., & Vogelmann, S. (1977). A comprehensive trial of the scree and kg criteria for

determining the number of factors. Multivariate Behavioral Research, 12(3), 289–325.

http://doi.org/10.1207/s15327906mbr1203_2

Curtis, T. P., Mara, D. D., & Silva, S. a. (1992). Influence of pH, Oxygen, and Humic substances

on ability of sunlight to dammage fecal-Coliforms in waste stabilisation pond water.

Applied Environmental Microbiology, 58(4), 1335–1343.

Davis-Colley, R. J., Donnison, A. M., & Speed, D. J. (2000). Towards a mechanistic

understanding of pond disinfection. Water Science and Technology, 42(10–11), 149 LP-

158. Retrieved from http://wst.iwaponline.com/content/42/10-11/149.abstract

Davis, T. W., Berry, D. L., Boyer, G. L., & Gobler, C. J. (2009). The effects of temperature and

! 38 ! ! nutrients on the growth and dynamics of toxic and non-toxic strains of Microcystis during

cyanobacteria blooms. Harmful Algae, 8(5), 715–725.

http://doi.org/10.1016/j.hal.2009.02.004

Dickinson, K. E., Whitney, C. G., & McGinn, P. J. (2013). Nutrient remediation rates in

municipal wastewater and their effect on biochemical composition of the microalga

Scenedesmus sp. AMDD. Algal Research, 2(2), 127–134.

http://doi.org/10.1016/j.algal.2013.01.009

Federation of Canadian Munipalities. (2017). Federal Wastewater Systems Effluent Regulations.

Ferrara, R. A., & Avci, C. B. (1982). Nitrogen Dynamics in Waste Stabilization Ponds. Journal

(Water Pollution Control Federation), 54(4), 361–369.

Ge, S., Madill, M., & Champagne, P. (2016). Use of freshwater macroalgae Spirogyra sp. for the

treatment of municipal wastewaters and biomass production for biofuel applications.

Biomass and Bioenergy, 111, 213–223. http://doi.org/10.1016/j.biombioe.2017.03.014

Ge, S., Qiu, S., Tremblay, D., Viner, K., Champagne, P., & Jessop, P. G. (2018). Centrate

wastewater treatment with Chlorella vulgaris: Simultaneous enhancement of nutrient

removal, biomass and lipid production. Chemical Engineering Journal, 342(December

2017), 310–320. http://doi.org/10.1016/j.cej.2018.02.058

Government of Canada. Canada Fisheries Act (1985). Retrieved from http://laws-

lois.justice.gc.ca/PDF/F-14.pdf

Government of Canada. Canada Water Act (1985).

Government of Canada. Canadian Environmental Protection Act (1999).

Government of Canada. (2012a). Wastewater regulations overview. Retrieved February 10,

2018, from https://www.canada.ca/en/environment-climate-

! 39 ! ! change/services/wastewater/regulations.html

Government of Canada. (2012b). Wastewater Systems Effluent Regulations. Retrieved from

http://laws-lois.justice.gc.ca/eng/regulations/SOR-2012-139/FullText.html

Government of Canada. (2016). Wastewater Systems Effluent Regulations: SOR/2012-139.

Règlement sur les effluents des systèmes d ’ assainissement des eaux usées: DORS/2012-

139. Government of Canada Justice Laws Website. Retrieved from http://laws-

lois.justice.gc.ca/eng/regulations/SOR-2012-139/

Guttman, L. (1954). Some necessary conditions for common-factor analysis. Psychometrika,

19(2), 149–161. http://doi.org/10.1007/BF02289162

Harwood, V. J., Whitlock, J., & Withington, V. (2000). Classification of Antibiotic Resistance

Patterns of Indicator Bacteria by Discriminant Analysis': Use in Predicting the Source of

Fecal Contamination in Subtropical Waters Classification of Antibiotic Resistance Patterns

of Indicator Bacteria by Discrimin. Applied and Environmental Microbiology, 66(9), 3698–

3704. http://doi.org/10.1128/AEM.66.9.3698-3704.2000.Updated

Hillman, W. S., Jr, D. D. C., & Culley, D. D. (1978). The Uses of Duckweed. American

Scientist, 66(May), 442–451. Retrieved from

http://adsabs.harvard.edu.ezproxy.lib.utexas.edu/abs/1978AmSci..66..442H

Hoffmann, J. P. (1998). Wastewater treatment with suspended and nonsuspended algae. Journal

of Phycology, 34(5), 757–763. http://doi.org/10.1046/j.1529-8817.1998.340757.x

Hosetti, B., & Frost, S. (1998). A review of the control of biological waste treatment in

stabilization ponds. Critical Reviews in Environmental Science and Technology, 28(2),

193–218. http://doi.org/10.1080/10643389891254205

Huang, Y., Truelstrup Hansen, L., Ragush, C. M., & Jamieson, R. C. (2017). Disinfection and

! 40 ! ! removal of human pathogenic bacteria in arctic waste stabilization ponds. Environmental

Science and Pollution Research, 1–13. http://doi.org/10.1007/s11356-017-8816-9

Iordache, Ş., & Dunea, D. (2013). Multivariate analysis applied to physicochemical parameters

of treated wastewater effluents discharged in Teleajen River, 58, 717–723.

Jackson, D. A. (1993). Stopping Rules in Principal Components Analysis: A Comparison of

Heuristical and Statistical Approaches. Ecology, 74(8), 2204–2214.

http://doi.org/10.2307/1939574

Jakob, T., Schreiber, U., Kirchesch, V., Langner, U., & Wilhelm, C. (2005). Estimation of

chlorophyll content and daily primary production of the major algal groups by means of

multiwavelength-excitation PAM chlorophyll fluorometry: Performance and

methodological limits. Photosynthesis Research, 83(3), 343–361.

http://doi.org/10.1007/s11120-005-1329-2

Jolliffe, I. . T. (1986). Principal Component Analysis. Journel, A.G., 1989., 487. Retrieved from

https://books.google.co.id/books?id=_olByCrhjwIC

Kadir, K., & Nelson, K. L. (2014). Sunlight mediated inactivation mechanisms of Enterococcus

faecalis and Escherichia coli in clear water versus waste stabilization pond water. Water

Research, 50, 307–317. http://doi.org/10.1016/j.watres.2013.10.046

Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educational and

Psychological Measurement, 20(1), 141–151.

Kayombo, S., Mbwette, T. S. A., Katima, J. H. Y., & Jorgensen, S. E. (2003). Effects of

substrate concentrations on the growth of heterotrophic bacteria and algae in secondary

facultative ponds. Water Research, 37(12), 2937–2943. http://doi.org/10.1016/S0043-

1354(03)00014-9

! 41 ! ! King, J. R., & Jackson, D. a. (1999). Variable selection in large environmental data sets using

principal components analysis. Environmetrics. http://doi.org/10.1002/(SICI)1099-

095X(199901/02)10:1<67::AID-ENV336>3.0.CO;2-0

Larsdotter, K., La Cour Jansen, J., & Dalhammar, G. (2007). Biologically mediated phosphorus

precipitation in wastewater treatment with microalgae. Environmental Technology, 28(9),

953–960. http://doi.org/10.1080/09593332808618855

Ledesma, R. D., Valero-Mora, P., & Macbeth, G. (2015). The Scree Test and the Number of

Factors: a Dynamic Graphics Approach. The Spanish Journal of Psychology, 18, E11.

http://doi.org/10.1017/sjp.2015.13

Lee, J., & Nicewander, W. A. (2016). Thirteen Ways to Look at the Correlation Coefficient

Published by': Taylor & Francis , Ltd . on behalf of the American Statistical Association

Stable URL': http://www.jstor.org/stable/2685263 REFERENCES Linked references are

available on JSTOR for this ar, 42(1), 59–66.

Lee Rodgers, J., & Alan Nice Wander, W. (1988). Thirteen ways to look at the correlation

coefficient. American Statistician, 42(1), 59–66.

http://doi.org/10.1080/00031305.1988.10475524

Leusch, A., Holan, Z. R., & Volesky, B. (1995). Biosorption of Heavy Metals ( Cd , Cu , Ni , Pb

, Zn ) by Chemically-Reinforced Biomass of Marine Algae, 279–288.

Liu, L., Champagne, P., & Hall, G. (2018). The role of algae in disinfection processes in

wastewater stabilization ponds. Queen’s University.

Liu, L., Hall, G., & Champagne, P. (2016a). Effects of environmental factors on the disinfection

performance of a wastewater stabilization pond operated in a temperate climate. Water

(Switzerland), 8(1). http://doi.org/10.3390/w8010005

! 42 ! ! Liu, L., Hall, G., & Champagne, P. (2016b). Effects of environmental factors on the disinfection

performance of a wastewater stabilization pond operated in a temperate climate. Water

(Switzerland), 8(1). http://doi.org/10.3390/w8010005

Liu, L., Macdougall, A., Hall, G., & Champagne, P. (2017). Disinfection Performance in

Wastewater Stabilization Ponds in Cold Climate Conditions: A Case Study in Nunavut,

Canada. Environments, 4(4), 93. http://doi.org/10.3390/environments4040093

MacDougall, A., Champagne, P., & Hall, G. (2017). The effects of climate and humic substances

on disinfection performance in Arctic wastewater stabilization ponds. Queen’s University.

Retrieved from

https://search.proquest.com/docview/1886374899?accountid=6180%5Cnhttp://dw2zn6fm9z

.search.serialssolution.com?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-

8&rfr_id=info:sid/Dissertations+%26+Theses+%40+Queen%27s+University+at+Kingston

%2C+Ontario&rft_val_f

Mara, D., Mills, S., Pearson, H., & Alabaster, G. (1992). Waste Stabilization Ponds - a Viable

Alternative for Small Community Treatment Systems. Journal of the Institution of Water

and Environmental Management, 6(1), 72–78. http://doi.org/10.1111/j.1747-

6593.1992.tb00740.x

Mayes, W. M., Batty, L. C., Younger, P. L., Jarvis, A. P., Kõiv, M., Vohla, C., & Mander, U.

(2009). Wetland treatment at extremes of pH: A review. Science of The Total Environment,

407(13), 3944–3957. http://doi.org/https://doi.org/10.1016/j.scitotenv.2008.06.045

Maynard, H. E., Ouki, S. K., & Williams, S. C. (1999). Tertiary lagoons: A review of removal

mechanisms and performance. Water Research, 33(1), 1–13. http://doi.org/10.1016/S0043-

1354(98)00198-5

! 43 ! ! McGarigal, K., Cushman, S., & Stafford, S. G. (2000). Multivariate statistics for wildlife and

ecology research. New York, USA: Springer.

Meron, A., Rebhun, M., & Sless, B. (1965). Quality Changes as a Function of Detention Time in

Wastewater Stabilization Ponds. Journal (Water Pollution Control Federation), 37(12),

1657–1670. Retrieved from http://www.jstor.org/stable/25035436

Metcalf & Eddy. (2003). Wastewater Engineering: Treatment and Resource Recovery (Fourth).

McGraw-Hill Science.

Myers, J., & Well, A. (2003). Research design and statistical analysis (Second). New Jersey,

USA: Lawrence Erlbaum Associates.

Oswald, W. J., & Golueke, C. G. (1966). Eutrophication Trends in the USA - A problem?

Journal Water Pollution Control Federation, 38(6), 964–975.

Oswald, W. J., & Gotaas, H. B. (1985). Photosynthesis in . American Society of

Civil Engineers (Vol. 2849, p. 33).

Pearson, H. W., Mara, D. D., & Bartone, C. R. (1987). Guidelines for the minimum evaluation of

the performance of full-scale waste stabilization pond systems. Water Research, 21(9),

1067–1075. http://doi.org/10.1016/0043-1354(87)90028-5

Peres-Neto, P. R. ., Jackson, D. A., & Somers, K. M. (2017). Giving Meaningful Interpretation to

Ordination Axes': Assessing Loading Significance in Principal Component Analysis.

Ecology, 84(9), 2347–2363.

Perkins, R. G., & Underwood, G. J. C. (2000). Gradients of chlorophyll a and water chemistry

along an eutrophic with determination of the limiting nutrient by in situ nutrient

addition. Water Research, 34(3), 713–724. http://doi.org/10.1016/S0043-1354(99)00228-6

Pop, H. F., Einax, J. W., & Sârbu, C. (2009). Classical and fuzzy principal component analysis

! 44 ! ! of some environmental samples concerning the pollution with heavy metals. Chemometrics

and Intelligent Laboratory Systems, 97(1), 25–32.

http://doi.org/10.1016/j.chemolab.2008.06.006

Powell, N., Shilton, A., Pratt, S., & Chishti, Y. (2011). Phosphate release from waste

stabilisation pond sludge: Significance and fate of polyphosphate. Water Science and

Technology, 63(8), 1689.

Ragush, C. M., & Jamieson, R. C. (2016). Oxygen Dynamics and Carbon Removal in Municipal

Waste Stabilization Ponds in Arctic Regions. Dalhousie University.

Ragush, C. M., Schmidt, J. J., Krkosek, W. H., Gagnon, G. A., Truelstrup-Hansen, L., &

Jamieson, R. C. (2015). Performance of municipal waste stabilization ponds in the Canadian

Arctic. Ecological Engineering, 83, 413–421. http://doi.org/10.1016/j.ecoleng.2015.07.008

Randall, J., & Tsui, N. (2002). Ammonia toxicity in fish. Marine Pollution Bulletin, 45, 17–23.

Redalje, D. G., Duerr, E. O., De La Noue, J., Mayzaud, P., Nomomura, A. H., & Cassin, R. C.

(1989). Algae as ideal waste removers: biochemical pathways (Biotreatme). Boca Raton,

USA: CRC Press.

Rehman, A., & Shakoori, A. R. (2001). Heavy metal resistance Chlorella spp., isolated from

tannery effluents, and their role in remediation of hexavalent chromium in industrial waste

water. Bulletin of Environmental Contamination and Toxicology, 66(4), 542–547.

http://doi.org/10.1007/s00128010041y

Reisenhofer, E., Picciotto, A., & Li, D. (1995). A factor analysis approach to the study of the

eutrophication of a shallow, temperate lake (San Daniele, North Eastern Italy). Analytica

Chimica Acta, 306(1), 99–106. http://doi.org/10.1016/0003-2670(94)00584-9

Sah, L., Rousseau, D. P. L., & Hooijmans, C. M. (2012). Numerical Modelling of Waste

! 45 ! ! Stabilization Ponds: Where Do We Stand? Water, Air, & Pollution, 223(6), 3155–3171.

http://doi.org/10.1007/s11270-012-1098-4

Schueder, M. R., Champagne, P., & Hall, G. (2016). Water quality spatial heterogeneity in

wastewater stabilizations ponds at elevated pH. Queen’s University.

Sebastian, S., & Nair, K. V. K. (1984). Total removal of coliforms and E. coli from domestic

sewage by high-rate pond mass culture of Scenedesmus obliquus. Environmental Pollution.

Series A, Ecological and Biological, 34(3), 197–206. http://doi.org/10.1016/0143-

1471(84)90116-8

Shammas, N., Wang, L. K., & Wu, Z. (2008). Waste Stabilization Ponds and Lagoons. In

Biological Treatment Processes - Volume 8. New Delhi, India: Humana Press.

Shrestha, S., & Kazama, F. (2007). Assessment of quality using multivariate

statistical techniques: A case study of the Fuji river basin, Japan. Environmental Modelling

and Software, 22(4), 464–475. http://doi.org/10.1016/j.envsoft.2006.02.001

Shrestha, S., Kazama, F., & Nakamura, T. (2008). Use of principal component analysis, factor

analysis and discriminant analysis to evaluate spatial and temporal variations in water

quality of the Mekong River. Journal of Hydroinformatics, 10(1), 43.

http://doi.org/10.2166/hydro.2008.008

Simons, J., & van Beem, A. P. (1990). Spirogyra species and accompanying algae from pools

and ditches in The Netherlands. Aquatic Botany, 37(3), 247–269.

http://doi.org/10.1016/0304-3770(90)90073-T

Singh, K. P., Malik, A., Mohan, D., & Sinha, S. (2004). Multivariate statistical techniques for the

evaluation of spatial and temporal variations in water quality of Gomti River (India) - A

case study. Water Research, 38(18), 3980–3992.

! 46 ! ! http://doi.org/10.1016/j.watres.2004.06.011

Singh, K. P., Malik, A., Mohan, D., Sinha, S., & Singh, V. K. (2005). Chemometric data analysis

of pollutants in wastewater - A case study. Analytica Chimica Acta, 532(1), 15–25.

http://doi.org/10.1016/j.aca.2004.10.043

Smart, G. R. (1975). The effect of ammonia on gill structure of the rainbow trout (Salmo

gairdneri), 85(1), 1386–1395.

Smart, G. R. (1978). Investigations of the toxic mechanisms of ammonia to fish–gas exchange in

rainbow trout (Salmo gairdneri) exposed to acutely lethal concentrations. Journal of Fish

Biology, 12(1), 93–104. http://doi.org/10.1111/j.1095-8649.1978.tb04155.x

Speers, B. R. (1997). Photosynthetic pigments. Retrieved January 21, 2018, from

http://www.ucmp.berkeley.edu/glossary/gloss3/pigments.html

Spuhler, D. (2011). Waste Stabilization Ponds (WSP). Retrieved from

https://www.sswm.info/water-nutrient-cycle/wastewater-treatment/hardwares/semi-

centralised-wastewater-treatments/waste-stabilization-ponds-%28wsp%29

Steinmann, C. R., Weinhart, S., & Melzer, A. (2003). A combined system of lagoon and

constructed wetland for an effective wastewater treatment. Water Research, 37(9), 2035–

2042. http://doi.org/10.1016/S0043-1354(02)00441-4

Tchobanoglous, G., & Crites, R. (1998). Small & Decentralized Wastewater Management

Systems. Michigan: WCB/McGraw-Hill.

Tyagi, V. K., Sahoo, B. K., Khursheed, A., Kazmi, A. A., Ahmad, Z., & Chopra, A. K. (2011).

Fate of coliforms and pathogenic parasite in four full-scale sewage treatment systems in

India. Environmental Monitoring and Assessment, 181(1–4), 123–135.

http://doi.org/10.1007/s10661-010-1818-4

! 47 ! ! Uusitalo, J. (1996). Algal carbon uptake and the difference between alkalinity and high pH

(“alkalization”), exemplified with a pH drift experiment. Scientia Marina (Vol. 60).

Veeresh, M., Veeresh, A. V., Huddar, B. D., & Hosetti, B. B. (2010). Dynamics of industrial

waste stabilization pond treatment process. Environmental Monitoring and Assessment,

169(1–4), 55–65. http://doi.org/10.1007/s10661-009-1150-z

Vendramelli, R. A., Vijay, S., & Yuan, Q. (2016). Phosphorus removal mechanisms in a

facultative wastewater stabilization pond. Water, Air and Soil Pollution, 11, 1–10.

Wallace, J. (2013). Evaluating the performance and water chemistry dynamics of passive

systems treating municipal wastewater and landfill . Queen's University.

Wallace, J., Champagne, P., Hall, G., Yin, Z., & Liu, X. (2015). Determination of algae and

macrophyte species distribution in three wastewater stabilization ponds using metagenomics

analysis. Water (Switzerland), 7(7), 3225–3242. http://doi.org/10.3390/w7073225

Wetzel, R. (2001). Limnology: Lake and River Ecosystems (Third). London, UK: Academic

Press.

Whitmarsh, J., & Govindjee. (1999). The Photosynthetic Process BT - Concepts in

Photobiology: Photosynthesis and Photomorphogenesis. In G. S. Singhal, G. Renger, S. K.

Sopory, K.-D. Irrgang, & Govindjee (Eds.) (pp. 11–51). Dordrecht: Springer Netherlands.

http://doi.org/10.1007/978-94-011-4832-0_2

Wilde, E. W., & Benemann, J. R. (1993). Bioremoval of heavy metals by the use of microalgae.

Biotechnology Advances, 11(4), 781–812. http://doi.org/10.1016/0734-9750(93)90003-6

Wold, S., Esbensen, K. I. M., & Geladi, P. (1987). Principal component analysis. Chemometrics

and Intelligent Laboratory Systems, 2, 37–52. http://doi.org/Doi 10.1016/0169-

7439(87)80084-9

! 48 ! ! Wolfgang, K. H., & Simar, L. (2012). Applied Multivariate Statistical Analysis. Use R!

http://doi.org/10.1007/978-3-642-17229-8

Wolverton, B. C., & McDonald, R. C. (1979). Upgrading facultative wastewater lagoons with

vascular aquatic plants. Journal of the Water Pollution Control Federation, 51(2), 305–313.

Retrieved from

http://www.jstor.org/stable/10.2307/25039826%5Cnhttp://www.scopus.com/inward/record.

url?eid=2-s2.0-0018438102&partnerID=40&md5=f335bd2a0c5b34ffcba7a27896954ff8

Wu, S., Carvalho, P. N., Müller, J. A., Manoj, V. R., & Dong, R. (2016). Sanitation in

constructed wetlands: A review on the removal of human pathogens and fecal indicators.

Science of The Total Environment, 541(Complete), 8–22.

http://doi.org/10.1016/j.scitotenv.2015.09.047

Yee, N., Benning, L. G., Phoenix, V. R., & Ferris, F. G. (2004). Characterization of

Metal−Cyanobacteria Sorption Reactions:' A Combined Macroscopic and Infrared

Spectroscopic Investigation. Environmental Science & Technology, 38(3), 775–782.

http://doi.org/10.1021/es0346680

Yidana, S. M., Ophori, D., & Banoeng-Yakubo, B. (2008). A multivariate statistical analysis of

surface water chemistry data-The Ankobra Basin, Ghana. Journal of Environmental

Management. http://doi.org/10.1016/j.jenvman.2006.11.023

Zitko, V. (1994). Principal component analysis in the evaluation of environmental data. Marine

Pollution Bulletin, 28(12), 718–722. http://doi.org/10.1016/0025-326X(94)90329-8

! 49 ! ! Chapter 3. Spatial and temporal variations in water quality in a facultative wastewater stabilization pond with algal blooms

ABSTRACT

A WSP located in southeastern Ontario, Canada was selected as a model WSP system with excessive algal growth and high pH fluctuations in a temperate climate. Water quality and climatic parameters were collected weekly over a 4-month period. In order to elucidate the relationships between algal blooms, water quality and meteorological parameters, surveying imaging techniques and multivariate statistics methods were employed. The epilimnetic algal blooms were a dominant contributing factor of the observed spatial heterogeneity in the water quality of the WSP, with algal blooms strongly influenced by wind and precipitation. From principal component analysis, the first three Principal Components accounted for 70.5% of the total variance. Although this analysis did not result in a significant data reduction, it helped to identify the factors/sources responsible for water quality variations within the WSP, mainly related to temperature, pH and nutrients. Using discriminant analysis, water temperature, pH, DO

- and NO3 were determined to be the most significant parameters discriminating spatial variation within the WSP. The relationships between algae, climate and water quality parameters provides practical knowledge for passive wastewater treatment system operators experiencing similar eutrophication issues in temperate climates.

! 50 ! ! 3.1.! Introduction ! Algae are important in passive wastewater treatment processes as they: 1) provide conditions to reduce the biochemical oxygen demand (BOD) and chemical oxygen demand

(COD) (Steinmann et al., 2003), 2) assimilate inorganic nutrients, such as nitrogen (N) and phosphorous (P) (Tchobanoglous & Crites, 1998), and 3) maintain the aerobic phase of facultative ponds, promoting disinfection of pathogens (Reinoso, Torres, & Bécares, 2008).

However, algae can also pose operational and treatment issues under certain environmental conditions. For instance, high rates of photosynthetic activity may result in excessive algal blooms, commonly seen in lacustrine and coastal systems (Arhonditsis et al., 2003). Algal blooms are often attributed to high nutrient loadings, warmer temperatures and minimal hydraulic mixing and can subsequently lead to elevated pH, total suspended solids (TSS) and chemical oxygen demand (COD) (Gschlößl, Steinmann, Schleypen, & Melzer, 1998).

Within the context of passive wastewater treatment, factors and mechanisms that regulate algal blooms are not fully understood and could represent an important tool for water resource management. Multivariate statistical techniques, such as principal component analysis (PCA), factor analysis (FA) and discriminant analysis (DA), can aid in the interpretation of multivariable, complex data sets to better understand spatial and temporal processes in ecological systems (Shrestha et al., 2008). Multivariate statistical methods can be used as an important tool for wastewater management and have been applied to wastewater monitoring, control and performance assessment, including passive treatment systems, in a number of studies (Dong &

Reddy, 2010; Ebrahimi, Gerber, & Rockaway, 2017; Singh et al., 2005).

A previous study was conducted by Wallace et al. examining algal growth and wastewater chemistry at the same water pollution control plant (WPCP) for three WSPs in-series ! 51 ! ! (Wallace et al., 2016). This research focused on temporal changes in the three WSPs and tracked changes in chlorophyll-a measurements with respect to water quality parameters from July to

October. While chlorophyll is commonly used as a green algae indicator, it is not always consistent or representative (Schueder et al., 2016). As such, spatial imaging of surface algal blooms will be used in conjunction with photosynthetic pigment measurements, taking into consideration the effects of algal surface cover on water quality. This paper will build on this previous work and include interacting dynamics between surface macroalgal blooms, water quality and meteorological parameters to discern key factors involved in algal blooms in WSPs.

The study focuses on an exploratory analysis of spatial and temporal changes in algal development and decline within a single facultative WSP at a municipal wastewater treatment plant. We investigated the relationships between the algal communities present, wastewater quality parameters and external climactic factors over a four-month period. Multivariate statistical methods were used, primarily principal component analysis and discriminant analysis, to determine wastewater quality parameters affecting algal growth. An understanding of algal dynamics and wastewater quality in WSPs will be beneficial for long-term management of passive wastewater treatment systems.

3.2.! Materials & Methods

3.2.1.! System overview ! Amherstview water pollution control plant (WPCP) is a municipal secondary wastewater treatment facility, treating an average of 3500 m3/day of municipal wastewater received from the community of Amherstview. The studied system consists of two facultative WSPs in-series and a surface-flow constructed wetland, for a total volume of 170,000 m3 (CH2M Hill, 2011). The

! 52 ! ! three passive treatment systems have average operating depths of approximately 1.61 m, 1.42 m and 1.17 m, and surface areas of 48 600 m2, 28 200 m2 and 31 200 m2, respectively (Figure 3.1).

WSP 1 receives highly nitrified wastewater with low carbonaceous biochemical oxygen demand

(cBOD5 < 5 mg/L). In the last decade, excessive algal bloom has been observed at the surface of

WSP 1, causing concerns surrounding water quality and toxicity for the final effluent discharged to the Bayview Bog wetland. As a result, the third WSP was converted into a constructed wetland (CW), with three individually operated treatment trains, for pH attenuation.

Figure 3.1. Schematic of Amherstview WPCP and six surface sampling points (Influent indicated by Inf and Effluent indicated by Eff). Black arrows indicate flow of WSP. 3.2.2.! Water quality ! As shown in Figure 4.1, weekly samples were collected at six surface points within WSP

1, from influent to effluent, from May to September 2017. The water quality parameters were characterized using field probes or according to Standard Methods for the Examination of Water and Wastewater (APHA/AWWA/WEF, 2012). An overview of the monitored parameters can be ! 53 ! ! seen in Table 3.1. The water quality parameters, measured using Hydrolab DS5 –

Multiparameter Data Sonde (Sutron, Sterling, VA, USA) included pH, water temperature (WT),

DO, chlorophyll and phycocyanin. Chlorophyll (chl) served as an indicator of green algae, and phycocyanin (phyco), an indicator of cyanobacteria. Standard methods and Hach kits (Colorado,

+ USA) were used to determine nitrogen species including ammonium (NH4 ) (APHA Standard

- Method 4500) and nitrate (NO3 ) (APHA Standard Method 4500C). Phosphorous (P) species

3- included phosphate (PO4 ) (APHA Standard Method 4500) and total phosphorous (TP) (U.S.

EPA Reactor Digestor Method, Hach Method 8190). Organic matter was reported in terms of total chemical oxygen demand (COD) (U.S. EPA Reactor Digestion Method, Hach Method

8000). Procedural blank measurements and duplicate samples were taken for QA/QC.

Table 3.1. Descriptive statistics of wastewater quality, chlorophyll and phycocyanin levels from May to September at WSP 1. Min Max Mean Std Dev Variance Skewness Kurtosis WT Inf 9.76 24.86 19.22 4.18 17.45 -1.08 0.70 (°C) Eff 9.88 23.34 19.12 3.90 15.23 -1.27 1.11 pH Inf 6.97 9.13 7.69* 2.13 4.53 -3.53 13.32 (units) Eff 7.25 9.73 8.35* 0.58 0.34 -0.86 -0.29 DO Inf 6.29 15.93 10.25* 2.77 7.66 0.48 -0.33 (mg/L) Eff 4.47 14.90 9.09* 3.02 9.11 0.44 -0.21 3- PO4 Inf 0.20 0.70 0.39 0.13 0.02 0.64 1.16 (mg/L) Eff 0.03 0.58 0.35 0.15 0.02 -0.37 0.38 TP Inf 1.19 2.68 1.61 0.39 0.15 1.49 2.81 (mg/L) Eff 0.98 3.1 1.57 0.60 0.36 1.56 1.88 + NH4 Inf 0.02 2.53 0.79 0.65 0.42 1.55 2.49 (mg/L) Eff 0.02 5.02 0.84 1.14 1.30 3.73 14.54 - NO3 Inf 1.34 18.73 9.95* 4.91 24.13 0.02 -0.79 (mg/L) Eff 1.13 16.97 8.33* 4.02 16.15 0.38 0.35 COD Inf 10.00 38.00 18.25* 6.97 48.53 1.56 0.60 (mg/L) Eff 11.00 34.00 19.18* 7.129 50.82 0.944 0.11 chl Inf 0.36 262.52 24.21* 64.17 4118.21 3.87 0.56 (µg/L) Eff 0.41 132.53 13.35* 32.43 1052.00 3.74 14.47 phyco Inf 20.39 17530.79 2637.26 4239.34 17972006.14 3.27 11.56 (mg/L) Eff 0.96 10410.95 2690.76 3305.44 10925924.16 1.74 2.17 Note: * indicates a significant (p<0.05) difference between the data set for influent and effluent water quality parameters, based on t-tests conducted.

! 54 ! ! 3.2.3.! Meteorological Parameters

Weather data, air temperature, precipitation ( and snow) and wind speed and direction, was retrieved from Environment Canada for the Kingston Airport weather station

(44.23° N 76.6° W), located 4.5km east of the field site, from May to September 2017. Daily averages were taken from 12:00 am to 10:00 am on the day of weekly sampling as field samples were consistently conducted from 8:00am to 10:00am.

3.2.4.! Algae Surveying & Image Processing

Images were taken weekly from 8:00-10:00am at an elevation of 80m using a DJI

Phantom 3 Standard UAV. A consistent weekly flight path was pre-set, starting from the influent pipe, along the influent side, crossing over the baffle and heading down to the effluent side towards the effluent weir. Images were taken at an elevation of 80m for the duration of the flight, with the camera facing straight down and stabilized using a 3-axis gimbal gyroscope.

Image processing and analysis were performed using Adobe Photoshop and MATLAB

R2018a. Adobe Photoshop was used to remove geometric distortion and compile images to create the WSP surface area. Using MATLAB, all compiled images were registered according to a same reference WSP outline, using ten reference points for shape segmentation. Boundary registration was using the impoly function, which create a polygon outline. A binary mask was created such that the pond area was encoded in white pixels and the surrounding landscape in black pixels. The pond area in pixel was obtained by loading the binary mask created into the image region analyser application. The YCbCr color space was used for algae colour segmentation, with values adjusted based on weather conditions (sunny, overcast, etc.). Each binary mask was then loaded into the image region analyser application to determine the algal surface cover area in pixels for each compiled weekly image. ! 55 ! ! 3.2.5.! Statistical Analysis Pre-screening tests were conducted on the raw data set to determine the suitability of multivariate analysis. Kaiser-Meyer-Olkin (KMO) were conducted to examine the suitability of the data for PCA, with KMO values close to 1 indicating PCA would be an appropriate statistical tool. In this study, the KMO value was 0.71, indicating that PCA was appropriate for the data set.

Principal Component Analysis (PCA) is commonly used to reduce the number of variables in order to determine which variables are responsible for the majority of the variance in the data. PCA is a bilinear projection technique used to examine data structures to interpret complex relationships between variables (Parinet, L'hote, & Legube, 2004). The original m- dimensional measurement space described by a data matrix, X, which consists of n observations

(rows) on p variables (columns) is projected into a lower, A-dimensional space (Singh et al.,

2005). New variables are constructed as weighted averages of the original variables, referred to as principal components (PC). The weights, W, are constructed so that the variance of y1,

Var(y1), is maximized. Also, so that Var(y2) is maximized and that the correlation between y1 and y2 is zero. The remaining are calculated so that their variances are maximized, subject to the constraint that the covariance between yi and yj, for all i and j parameters, as seen in Eq. 4.1:

[*` = / _6*Z6` +/_$*Z$` + ⋯ +/_3*Z3` (Equation 3.1.) PC provides information on the most meaningful parameters, which describe the majority of data set, affording data reduction with minimum loss of original information. Data were standardized to unit variance in order to exclude bias due to the different measurement scales of the various variables (Legendre, 1998).

Discriminant Analysis (DA) uses the variable(s) known a priori to discriminate between naturally occurring groups for a set of observations and determine which variables account for

! 56 ! ! the variations in the observations (Johnson & Wichern, 2007). Using original data, a discriminant function (DF) is created, comprised of the original parameters, using Equation 3.2:

b \(]*) = / ^* +/ `96 _*`a*` (Equation 3.2.) where i is the number of groups (G), ki is the constant inherent to each group, n is the number of parameters used to classify set of data into a given group and wj is the weight coefficient assigned by DA to a given selected parameter (pj) (Singh et al., 2004). Standard/forward and backward stepwise DA mode were employed to analyze the data set (Wolfgang & Simar, 2012).

The classification table reveals the predicted versus actual group membership, and reveals how well the derived function predicts group membership, using both modes of DA.

3.3.! Results

3.3.1.! Characterization of surface algae cover ! The WSP was observed to have large mats of filamentous surface green algae, generally from June to September (Schueder et al., 2016). A previous study conducted at the site identified the filamentous algae Chlorophyta chlorophyceae, belonging to the genus Oedogonium. Species under this genus include Spirogyra sp. and Hydrodictyon sp. (Figure 3.2), and were visually identified from samples at the site (Wallace et al., 2015). This epiphytic genus exhibits cyclical growth patterns, originating from the benthic layer growing on the submerged macrophytes, and during warmer periods, undergo rapid growth and entangling other submerged plants. As the submerged macrophytes reach the surface of the water, Oedogonium also emerge, forming dense canopies, stemming from macrophytes, to increase their exposure to sunlight. The higher levels of phytoplankton activity on the surface shades the epiphytic macrophytes, resulting irradiation attenuation at the surface of the water column and minimal irradiation in the benthic layer.

! 57 ! ! a) b)

Figure 3.2. Surface algal species at 10x magnification: a) Hydrodictyon sp. b) Spirogyra sp.

3.3.2.! Effect of meteorological conditions on algal blooms

Meteorological conditions, in particular ambient temperature and wind effects, had a notable effect on the WSP, due to its high surface area to volume ratio. During periods of higher temperature in the water column, warmer surface water and minimal mind can prevent water from mixing, allowing for vertical stratification to occur and in turn, allowing algae to grow thicker and faster (Mahyari et al., 2018). Additionally, heavy precipitation and intense rain events allow for more mixing within the water column and can influence water levels and reduce surface algal bloom activity (Cattaneo, Hudon, Vis, & Gagnon, 2013).

In early May, the water temperature was still relative low, around 15-16°C (Figure 3.4).

Colder temperatures is typically associated with less surface floating algae, as previous studies have shown that the buoyancy of algal mats depend high levels of photosynthetic oxygen production, which primarily occur in warmer temperatures (Graham, James, Lemhi, Adrian, &

Spencer, 1995). The period from May 17th to July 5th was characterized as a period of relatively low precipitation, increasing ambient temperatures and relatively constant winds (Figure 3.4). As light penetration increases, and current velocities fall due to lower wind speeds, these conditions facilitate macroalgal growth, allowing for blooms to occur. As a result, as seen in Figure 3.4, the surface algal bloom was shown to increasing extensively from 3.5% to 88.6% surface area cover.

! 58 ! !

June 21 June 28

30 100 Total Precipitation Air Temperature 90 25 WSP 1 Influent Temperature 80 WSP 1 Effluent Temperature

70 20

C) C) 60 ° ( 15 ! 50

40

(mm) Precipitation Temperature 10 30

20 5 10

0 0 15-05-17 31-05-17 16-06-17 02-07-17 18-07-17 03-08-17 19-08-17 04-09-17

July 19 August 9

Figure 3.3. Green algae surface cover and water temperature in WSP1 with meteorological parameters (daily precipitation, air temperature, wind speed and wind direction) over monitoring period. Green bars indicate the meteorological conditions of the highlighted day. ! 59 ! ! However, following a heavy rainfall event (90mm) on July 31, the surface algae was observed to decrease to from 91.2% surface cover on July 26th to 58.9% surface cover on August 9th (Figure

3.4). Heavy precipitation has been shown to cause a partial overturn in the water column, causing algal death and submergence in the bottom (Cattaneo et al., 2013). However, another period of algal blooms following a period of low precipitation, constant wind speeds and higher ambient temperatures were observed a few weeks following the large rainfall event on July 31. By

August 14, re-emergence of epilimnetic algal blooms were seen, covering 60.8% of the WSP surface (Figure 3.4).

100 91.2 88.6 90 82.7 80 76.9

70 66.5 58.9 60.8 60 50.5 50 45.4

40 Surface Surface (%) Cover 29.8 30

20 12.3 6.4 7.1 10 3.5 0 05/24 05/31 06/07 06/14 06/21 06/28 07/05 07/12 07/19 07/26 08/09 08/14 08/23 08/31

Figure 3.4. Surface cover of algal blooms from May 24 to August 30, 2017 on WSP 1. Furthermore, spatial heterogeneity in algal blooms were observed to move in accordance with the wind conditions. As seen from the images taken on June 21 and August 9, the algal blooms were predominantly in the southeast side of the WSP, near the influent and effluent side transported by intense prevailing winds originating from the northwest (Figure 3.3). Similar results were seen in a eutrophic lake in Florida where a strong relationship between wind and the

! 60 formation of epilimnetic algal blooms were observed (Havens, Aumen, James, & Smith, 1996).

However, during periods of low wind speeds, surface algal cover was more homogeneous in nature, shown on June 28th (Figure 3.3) Previous studies conducted in eutrophic , wind mixing near the water surface were reported to control the occurrence and disappearance of algal blooms, as they provided energetic controls for growth (Havens et al., 2001; Valiela et al., 1997).

3.3.3.! Spatial and temporal changes in water quality

PCA were performed on the overall data set to explore the relationships between water quality parameters and identify the important parameters that affected algal blooms over the monitoring period (Table 3.2) PCs were retained if they accounted for a large proportion of the variation, based on the Kaiser criterion (Kaiser, 1960). The results of the PCA analysis show that the first three PC accounted for 70.46% of the total variance, within the first component (PC1) accounted for 33.87%, the second component (PC2) accounted for 19.20%, and the third component (PC3) accounted for 17.40% of the total variance in the dataset.

Table 3.2. Varimax rotation PCA loading matrix.

Extraction Sums of Squared Rotation Sums of Squared PC Initial Eigenvalues Loadings Loadings Variance Cumulative Variance Cumulative Variance Cumulative Total Total Total (%) Variance (%) (%) Variance (%) (%) Variance (%) 1 2.67 33.87 33.87 2.67 33.87 33.87 2.47 33.87 33.87

2 1.28 19.20 53.06 1.28 19.20 53.06 1.36 19.20 53.06

3 1.21 17.40 70.46 1.21 17.40 70.46 1.32 17.40 70.46 4 0.89 11.29 81.75

5 0.88 9.14 90.89

6 0.48 5.11 96.00

7 0.27 2.42 98.42

8 0.20 1.58 100.00

! 61 The variable loadings from the retained PCs were examined to determine the relative importance of a variable compared to other variables within the PC (Figure 3.5). Within PC1, water temperature, phosphate and nitrate had significant (>|0.5|) positive loadings while ammonium had strong negative loadings, representative of the various nutrient cycles. As water temperatures increased, algal photosynthetic activity increased and subsequent uptake of nitrogen and phosphorous species. PC2 had significant high loadings (>|0.5|) for pH and DO, supporting the observed extensive algal blooms during the monitoring period. While algae produce oxygen during photosynthesis, oxygen is consumed during respiration, leading to oxygen depletion. This observation was supported by low summer DO concentrations

(5.38mg/L) in another study conducted in WSP 1 (Liang, Hall, & Champagne, 2018). High pH levels during the summer can be attributed to algal blooms as algal consumption of CO2 during photosynthesis leads to an increase in OH- production. Similar strong correlations were found in a disinfection performance study conducted at the same site (Liu et al., 2016b). High loadings for

TP and ammonium were observed in PC3, suggesting algae occurred.

Decomposition of algae in the benthic layer causes a release of phosphate and ammonia, which can then form ammonium, which are released into the water column.

! 62

Figure 3.5. Loadings for water quality parameters in WSP 1 based on retained PCs.

DA was conducted to determine the most significant parameters associated with the spatial variation in the WSP. The influent and effluent samples on either side of the baffle structure were used as spatial grouping variables, while the water quality parameters were considered independent variables. DA was performed on each raw data matrix using standard, forward stepwise and backward stepwise modes (Table 3.3). In forward stepwise mode, all variables are retained. Alternatively, in the backward stepwise mode, all variables are initially retained in the model and at each step, the variable contributing least to the prediction of group membership is eliminated until only the significant variables remain (McGarigal et al., 2000).

! 63 Table 3.3. Classification functions for discriminant analysis of spatial variations in water quality.

Standard Mode Forward Stepwise Mode Backward Stepwise Mode Parameters Influent Effluent Influent Effluent Influent Effluent WT 0.871 0.476 0.871 0.476 -0.142 -0.452 pH 34.756 38.46 34.756 38.46 28.06 31.873 DO -1.105 -1.479 -1.105 -1.479 -1.449 -1.773 3- PO4 15.096 14.494 15.096 14.494 TP 10.645 10.205 10.645 10.205 + NH4 14.953 14.117 14.953 14.117 - NO3 0.682 0.239 0.682 0.239 -0.194 -0.537 COD -0.348 -0.317 -0.348 -0.317 Constant -163.402 -176.739 -163.402 -176.739 -103.58 -122.107

The classification matrices produced for each DA mode compare the number or percentage of correctly predicted cases versus actual group membership. In this case, the standard and forward stepwise mode correctly assigned 82.5% of cases using all eight parameters while the backward stepwise DA mode correctly assigned 79.7% of cases using only four

- discriminant parameters, temperature, pH, DO, and NO3 (Table 3.4). Only four variables were used to account for the correct assignations, with only a 2.8% difference of correctly assigned cases compared to the standard and forward stepwise modes using all parameters. Temperature,

- pH, DO and NO3 were the most significant parameters discriminating between spatial variations within the WSP.

Table 3.4. Classification matrix for Discriminant Analysis (DA) of temporal variation in WSP 1.

Standard/Forward Stepwise Mode Backward Stepwise Mode Influent Effluent Correct Influent Effluent Correct Influent 32 8 32 31 9 31 Count Effluent 6 34 34 7 32 39 Total 66 70 Influent 80 20 80 77.5 22.5 77.5 % Effluent 15 85 85 17.9 82.1 82.1 Total 82.5 % 79.7%

! 64

From Table 3.1, it can be seen that water temperature at the influent and effluent did not

- differ significantly while influent and effluent values for pH, DO and NO3 did differ significantly, based on t-tests conducted at the p<0.05 level. pH levels were higher on the

- effluent side while DO and NO3 levels were higher on the influent side. These spatial differences could be attributed to the occurrence of the surface algal blooms. On average, algal biomass was seen to be statistically higher on the effluent side where algae were observed to typically cover more surface area, suggesting that the carbonate–bicarbonate equilibrium was

- - impacted by algal photosynthetic activity. OH and HCO3 production increased and

- subsequently increased the pH on the effluent side (Ragush et al., 2015). Similarly, NO3 concentrations different significantly (p<0.05) by 1.65mg/L in the effluent compared to the influent side, suggesting that the increased algal photosynthetic activity also led to increased nitrogen assimilation (Picota, Andrianarisona, & Olijnyka, 2009). Slightly higher average DO levels (10.25mg/L) in the influent side were observed in comparison to lower DO levels

(9.09mg/L) in the effluent. During periods of decreased epilimnetic algal blooms, algal decay can lead to increased DO concentrations in the water column (Liang et al., 2018). Bacterial decomposition of algae involves larger O2 consumption, supported by the higher COD in the effluent side (19.28mg/L) compared to the influent COD (18.25mg/L) levels. It can be inferred that the algal blooms were a dominant contributing factor in the observed spatial heterogeneity in

- the water in addition to temperature, pH, DO and NO3 .

3.4.! Conclusion

In this study, surveying imaging techniques and multivariate statistics methods were employed to gain a deeper understanding of the spatial and temporal variations within a

! 65 facultative WSP. It can be inferred that the algal blooms were a dominant contributing factor in the observed spatial heterogeneity in the water quality of the WSP, with algal blooms strongly influenced by wind and precipitation. PCA provided insight into the factors/sources responsible for water quality variations within the WSP, mainly related to temperature, pH and nutrients. It

- was revealed using DA that temperature, pH, DO and NO3 were the most significant parameters discriminating between spatial variations within the WSP, in close relation to the algal blooms.

The relationships between algae, climate and water quality parameters provide practical knowledge for the Amherstview WPCP and other passive wastewater treatment systems experiencing similar eutrophication issues in temperate climates. As the constructed wetland was designed to attenuate the increasing pH levels in WSP 1, this study will allow for facility staff to determine the required operational level of the treatment trains in the constructed wetland to provide pH attenuation and prevention of regulatory exceedances. This research can improve management processes for treated effluent discharge into receiving water bodies and reveals important water quality dynamics to guide the operation of passive wastewater treatment systems during periods to algal blooms.

! 66 3.5.! References APHA/AWWA/WEF. (2012). Standard Methods for the Examination of Water and Wastewater.

Standard Methods, 541. http://doi.org/ISBN 9780875532356

Arhonditsis, G. B., Brett, M. T., & Frodge, J. (2003). Environmental control and limnological

impacts of a large recurrent spring bloom in Lake Washington, USA. Environmental

Management. http://doi.org/10.1007/s00267-002-2891-4

Cattaneo, A., Hudon, C., Vis, C., & Gagnon, P. (2013). Hydrological control of filamentous

green algae in a large fluvial lake (Lake Saint-Pierre, St. Lawrence River, Canada). Journal

of Great Lakes Research, 39(3), 409–419. http://doi.org/10.1016/j.jglr.2013.06.005

CH2M Hill. (2011). Technical Assessment to Support Re-Rating the Amherstview WPCP Draft.

Dong, X., & Reddy, G. B. (2010). Nutrient removal and bacterial communities in swine

wastewater lagoon and constructed wetlands. Journal of Environmental Science and Health,

Part A, 45(12), 1526–1535. http://doi.org/10.1080/10934529.2010.506109

Ebrahimi, M., Gerber, E. L., & Rockaway, T. D. (2017). Temporal performance assessment of

wastewater treatment plants by using multivariate statistical analysis. Journal of

Environmental Management, 193, 234–246. http://doi.org/10.1016/j.jenvman.2017.02.027

Graham, J. M., James, M., Lemhi, C. A., Adrian, H. L., & Spencer, D. F. (1995). Physiological

responses to temperature and irradiance in Spirogyra (Zygnematales, charophyceae), 540,

531–540.

Gschlößl, T., Steinmann, C., Schleypen, P., & Melzer, A. (1998). Constructed wetlands for

effluent polishing of lagoons. Water Research, 32(9), 2639–2645.

http://doi.org/10.1016/S0043-1354(98)00047-5

Havens, K., Aumen, N., James, R., & Smith, V. (1996). Rapid Ecological Changes in a Large

Subtropical Lake Undergoing Cultural Eutrophication. Ambio (Vol. 25).

! 67 Havens, K. E., Fukushima, T., Xie, P., Iwakuma, T., James, R. T., Takamura, N., … Yamamoto,

T. (2001). Nutrient dynamics and the eutrophication of shallow lakes Kasumigaura (Japan),

Donghu (PR China), and Okeechobee (USA). Environmental Pollution, 111(2), 263–272.

http://doi.org/10.1016/S0269-7491(00)00074-9

Johnson, R. A., & Wichern, D. W. (2007). Applied Multivariate Statistical Analysis. New York

(Vol. 47). http://doi.org/10.1198/tech.2005.s319

Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educational and

Psychological Measurement, 20(1), 141–151.

Legendre, P., & L., L. (1998). Numerical ecology, 2nd English Edition. Numerical ecology,.

http://doi.org/10.1017/CBO9781107415324.004

Liang, S., Hall, G., & Champagne, P. (2018). Monitoring seasonal variations in treatment

performance of a wastewater stabilization pond with algal blooms and pH fluctuations

Materials & Methods System overview. In 1st WDSA/CCWI Conference Proceedings.

Liu, L., Hall, G., & Champagne, P. (2016). Effects of environmental factors on the disinfection

performance of a wastewater stabilization pond operated in a temperate climate. Water

(Switzerland), 8(1). http://doi.org/10.3390/w8010005

Mahyari, F., Rey, A., Boegman, L., Champagne, P., Mulligan, R., Hall, G., & Silva, A. (2018).

Three-dimensional simulation of hydrodynamics and water quality in a wastewater

stabilization pond Methods Study site. 1st WDSA/CCWI Conference Proceedings, (1), 1–8.

McGarigal, K., Cushman, S., & Stafford, S. G. (2000). Multivariate statistics for wildlife and

ecology research. New York, USA: Springer.

Parinet, B., Lhote, A., & Legube, B. (2004). Principal component analysis: An appropriate tool

for water quality evaluation and management - Application to a tropical lake system.

! 68 Ecological Modelling, 178(3–4), 295–311. http://doi.org/10.1016/j.ecolmodel.2004.03.007

Picota, B., Andrianarisona, T., & Olijnyka, D. P. (2009). Nitrogen removal in stabilisation ponds.

Desalination and Water Treatment, 4, 103–110.

Ragush, C. M., Schmidt, J. J., Krkosek, W. H., Gagnon, G. A., Truelstrup-Hansen, L., &

Jamieson, R. C. (2015). Performance of municipal waste stabilization ponds in the Canadian

Arctic. Ecological Engineering, 83, 413–421. http://doi.org/10.1016/j.ecoleng.2015.07.008

Reinoso, R., Torres, L. A., & Bécares, E. (2008). Efficiency of natural systems for removal of

bacteria and pathogenic parasites from wastewater. Science of the Total Environment,

395(2–3), 80–86. http://doi.org/10.1016/j.scitotenv.2008.02.039

Schueder, M. R., Champagne, P., & Hall, G. (2016). Water quality spatial heterogeneity in

wastewater stabilizations ponds at elevated pH. Queen’s University.

Shrestha, S., Kazama, F., & Nakamura, T. (2008). Use of principal component analysis, factor

analysis and discriminant analysis to evaluate spatial and temporal variations in water

quality of the Mekong River. Journal of Hydroinformatics, 10(1), 43.

http://doi.org/10.2166/hydro.2008.008

Singh, K. P., Malik, A., Mohan, D., & Sinha, S. (2004). Multivariate statistical techniques for the

evaluation of spatial and temporal variations in water quality of Gomti River (India) - A

case study. Water Research, 38(18), 3980–3992.

http://doi.org/10.1016/j.watres.2004.06.011

Singh, K. P., Malik, A., Mohan, D., Sinha, S., & Singh, V. K. (2005). Chemometric data analysis

of pollutants in wastewater - A case study. Analytica Chimica Acta, 532(1), 15–25.

http://doi.org/10.1016/j.aca.2004.10.043

Steinmann, C. R., Weinhart, S., & Melzer, A. (2003). A combined system of lagoon and

! 69 constructed wetland for an effective wastewater treatment. Water Research, 37(9), 2035–

2042. http://doi.org/10.1016/S0043-1354(02)00441-4

Tchobanoglous, G., & Crites, R. (1998). Small & Decentralized Wastewater Management

Systems. Michigan: WCB/McGraw-Hill.

Valiela, I., McClelland, J., Hauxwell, J., Behr, P. J., Hersh, D., & Foreman, K. (1997).

Macroalgal blooms in shallow estuaries: Controls and ecophysiological and ecosystem

consequences. Limnology and Oceanography, 42(5part2), 1105–1118.

http://doi.org/10.4319/lo.1997.42.5_part_2.1105

Wallace, J., Champagne, P., & Hall, G. (2016). Multivariate statistical analysis of water

chemistry conditions in three wastewater stabilization ponds with algae blooms and pH

fluctuations. Water Research, 96, 155–165. http://doi.org/10.1016/j.watres.2016.03.046

Wallace, J., Champagne, P., Hall, G., Yin, Z., & Liu, X. (2015). Determination of algae and

macrophyte species distribution in three wastewater stabilization ponds using metagenomics

analysis. Water (Switzerland), 7(7), 3225–3242. http://doi.org/10.3390/w7073225

Wolfgang, K. H., & Simar, L. (2012). Applied Multivariate Statistical Analysis. Use R!

http://doi.org/10.1007/978-3-642-17229-8

!

!

! 70 Chapter 4. Examining disinfection performance and water quality in a wastewater stabilization pond with algal blooms using multivariate statistical analysis

! ABSTRACT

WSPs can effectively attenuate nutrient loads while also providing environmental conditions suitable for the removal of pathogens through naturally occurring biological, chemical and physical treatment mechanisms. Four bacterial indicators, Escherichia coli (E. coli), Enterococci,

Clostridium perfringens (C. perfringens) and total coliforms (TC), and water quality parameters were selected to assess the disinfection performance of a facultative WSP in eastern Ontario, with excessive algal growth, over 11 months. Spearman’s ρ correlation coefficients and Principal

Components Analysis (PCA) were used to determine significant water quality parameters influencing disinfection performance. The analyses showed strong positive correlations

(ρ>|0.257|, p<0.05) between chlorophyll and temperature, pH, and DO, and strong negative correlations (ρ>|0.257|, p<0.05) between water temperature and E. coli, Enterococci, and C. perfringens. Principal component analysis also revealed that temperature, chlorophyll, pH and

- NO3 -N had accounted for most of the variations, based on retained principal components. This research contributes to the continued improvement of WSPs design and performance.

! 71 4.1.! Introduction

Passive wastewater treatment systems, such as wastewater stabilization ponds (WSPs), are sustainable alternatives to conventional wastewater treatment for remote and rural communities.

(Al-Hashimi & Hussain, 2013; Metcalf & Eddy, 2003). WSPs have the capability to effectively attenuate nutrient loads and remove pathogens present in municipal wastewater through naturally occurring biological, chemical and physical treatment mechanisms (Jiménez, Maya, & Galván,

2007; Maynard et al., 1999; Tchobanoglous & Crites, 1998). In particular, WSP systems have shown to be capable of removing various genera of bacteria, viruses, protozoa and helminthic pathogens, which pose human health risks (Abreu-Acosta & Vera, 2011; Maynard et al., 1999).

Treatment efficiencies of WSPs depend on the complex interactions between biological communities, wastewater quality and the regional climate. Disinfection is dependent on environmental factors, such as sunlight, pH, dissolved oxygen (DO), temperature, and attachment and sedimentation (Awuah et al., 2001; Maynard et al., 1999). Biological communities within

WSPs, such as algae, macrophytes and microorganisms, also contribute to treatment mechanisms.

Indicator organisms a are types of bacteria used to detect and estimate the level of fecal contamination of water. Bacterial indicators, such as Escherichia coli (E. coli), fecal coliforms and total coliforms (TC), are commonly used for pathogen detection in wastewater, evaluating the disinfection performance of WSP systems (Boutilier, Jamieson, Gordon, Lake, & Hart, 2009).

However, these traditionally used fecal indicator organisms may not be capable of representing overall pathogen levels (Burkhardt, Calci, Watkins, Rippey, & Chirtel, 2000; Molleda, Blanco,

Ansola, & de Luis, 2008). As a result, a number of potential indicator organisms have been proposed in addition to those conventionally employed. Enterococci have been suggested as additional bacterial indicators due to their higher survival rates at high pH levels compared to

! 72 fecal coliforms (Awuah et al., 2001). Additionally, Clostridium perfringens (C. perfringens) was also used as its spore phase has been observed to be highly resistant to different environmental conditions, representing more resistant protozoan and helminthic pathogens such as

Cryptosporidium, Giardia and helminth eggs (Abreu-Acosta & Vera, 2011; Dunlop, McMurray,

Hamilton, & Byrne, 2008; Lanao et al., 2010; Mosteo et al., 2013). In this study, two traditional indicators, E. coli and TC, and the two potential indicators, Enterococci and C. perfringens, were used to provide a more comprehensive understanding of disinfection performance in the WSP.

WSPs are complex systems whose performance is affected by many interacting biochemical and physical processes. The use of analytical tools is crucial to developing a better understanding of the interacting factors and are used in modeling and performance prediction in wastewater management. One of the increasingly popular approaches used in environmental engineering is Principal Component Analysis (PCA). PCA is widely used in ecological and biogeochemical studies and provides an objective statistical method for processing large, multi- parameter data sets (Bengraïne & Marhaba, 2003; Perkins & Underwood, 2000; Reisenhofer et al., 1995). This statistical method transforms a dataset into a series of uncorrelated linear combinations of the original variables, termed Principal Components (PCs), which account for most of the variability and reduces the complexity of a multidimensional system. More recently,

PCA has also been applied to the characterization, monitoring, process control and modeling of wastewater treatment systems (Boon, De Windt, Verstraete, & Top, 2002; Dong & Reddy, 2010;

Iordache & Dunea, 2013; Miettinen, Hurse, Connor, Reinikainen, & Minkkinen, 2004; Ouali,

Jupsin, Ghrabi, & Vasel, 2014; Wallace, Champagne, & Hall, 2016).

The main purpose of this study was to use a multivariate statistical approach to determine significant water quality parameters influencing treatment performance, particularly disinfection performance, of a WSP system with excessive algal growth. For this purpose, 15 physical,

! 73 chemical, and biological water quality parameters were sampled from a WSP over a period of 11 months. The study site has had a history of algal blooms and high pH, common for many WSP systems with warmer seasons. Multivariate statistical techniques were used to discern the complex interactions between parameters from field data.

4.2.! Materials 4.2.1.! System overview

A WSP at the Amherstview municipal water pollution control plant (WPCP), located in southeastern Ontario, Canada, was monitored from October 2016 to August 2017. The WPCP treats an average of 3500 m3/day of municipal wastewater and consists of primary treatment, and secondary aeration tanks. The wastewater is then piped into two in-series facultative WSPs and a surface-flow constructed wetland. The three passive treatment systems total a volume of 170,00

0m3 and have operating depths of approximately 1.61 m, 1.42 m, and 1.17 m, respectively, and surface areas of 48,600 m2, 28,200 m2, and 31,200 m2, respectively (Figure 4.1). At the

Amherstview WPCP, the pH limit is 9.5 and E. coli discharge limit is 200 CFU/100mL.

Figure 4.1. Overall schematic of Amherstview WPCP and locations of the six sampling points.

! 74 4.2.2.! Water quality analysis

As shown in Figure 4.1, the water quality parameters were characterized at six surface points within WSP 1 using field probes or according to Standard Methods for the Examination of

Water and Wastewater (APHA, 2005) on a weekly basis. The water quality parameters, measured using Hydrolab DS5 – Multiparameter Data Sonde (Sutron, Sterling, VA, USA), included pH,

DO and temperature. Flourometric chlorophyll and phycocyanin concentrations were also determined using the Hydrolab DS5, as indicators of algae and cyanobacteria, respectively. Water samples were also collected from influent (Inf) to effluent (Eff) and analyzed on a weekly basis for the warm season (May to August, 2017) when water temperatures were above 20°C. Samples were taken at Inf 1 and Eff 1 for the cold season (October 2016 to April 2017) when temperatures were below 15°C. Standard methods and Hach kits (Colorado, USA) were used to determine

+ nitrogen (N) species, including ammonium (NH4 ) (APHA Standard Method 4500-NH3-H),

- nitrate (NO3 ) (APHA Standard Method 4500C) and total nitrogen (TN) (TNT Persulfate

Digestion Method, Hach Method 10071). These methods were also used for phosphorous (P)

3- species, including phosphate/orthophosphate (PO4 ) (APHA Standard Method 4500) and total phosphorous (TP) (U.S. EPA Reactor Digestor Method, Hach Method 8190). Organic matters were reported in terms of total chemical oxygen demand (COD) (U.S. EPA Reactor Digestion

Method, Hach Method 8000). Quality assurance and control were performed using standardization procedures, blank measurements and duplicate samples.

4.2.3.! Detection of indicator organisms

E. coli, Enterococci, C. perfringens and TC were used as pathogen indicators of disinfection performance within the WSP system. To quantify the indicator organisms, wastewater samples were membrane filtered and cultivated on agar plates. The agar plates were

! 75 prepared using Chromocult Coliform-Agar (Merck, Germany), Chromocult Enterococci-Agar

(Meck, Germany) and m-CP Agar base (MilliporeSigma, Canada) (Lanao et al., 2010; Ouali et al., 2014; Reinoso, Blanco, Torres-Villamizar, & Bécares, 2011). Chromocult Coliform-Agar can visually differentiate E. coli (violet to blue colonies) from other coliforms (red colonies). TC was calculated as a sum of both E. coli and other coliforms. Enterococci appeared to be yellow colonies on Chromocult Enterococci-Agar plates while C. perfringens colonies turned pink when exposed to ammonium hydroxide vapor on m-CP Agar plates. Plate counting was performed after a 24-hour incubation period at 37°C for E. coli, TC and Enterococci, and a 48-hour anaerobic incubation at 44 °C for Clostridium perfringens.

4.2.4.! Statistical Analysis

SPSS Statistics 24 and CANOCO 5.0 were used for statistical analysis. Multivariate

Spearman's ρ correlation coefficients between variables were used as general indicators in statistical dependence, determining preliminary strong and weak correlations. Spearman’s ρ correlation coefficients have values between -1 and 1, with values between 0 and 1 indicate positive correlations and values between -1 and 0 indicate negative correlations. Correlation coefficient values greater than the critical value indicate strong correlations. In this study, the critical value was determined to be 0.257 suggesting a strong correlation, based on N-2 d.f.’s, with N=100 and α=0.01 (Ramsey, 1989).

The data set was normalized and outliers reduced to produce a homogenized variance of distribution and avoid misclassification due to different orders of magnitude of numerical values.

Imputation was used to estimate missing values due to inaccessibility of the site during cold weather and subsequent ice formation. PCA was used to summarize the majority of the variability and reduce the number of dimensions in order to determine which variables are to be

! 76 kept for analysis, thereby strengthening reliability of results. The resulting PCs contributing to variance were selected using the Kaiser criterion (Kaiser, 1960). The retained variables were considered to account for most of the variation in the data, while discarded variables were considered redundant (King & Jackson, 1999). The retained PCs were extracted and rotated, using the Varimax normalization, to determine the factor loadings for each water quality parameter (Kaiser, 1960). The loadings in PCs indicate the contribution or weight of corresponding parameter to the PC and its relation between parameters.

4.3.! Results and discussion

4.3.1.! Summary of WSP system treatment performance

In general, higher removal efficiency of both nutrient and indicator organisms was achieved in the warm season, with lower concentrations of parameters observed in the final effluent (Table 4.1). In the warm season, the influent and effluent had similar temperatures

(~21°C); while in the cold season, the effluent was approximately 2°C higher than the influent. pH in both the influent and effluent was generally higher under warmer temperatures, 7.86 and

8.14, respectively, than under colder temperatures, 7.12 and 7.70, respectively. Conversely, surface DO levels were noticeably higher in both the influent and effluent samples in winter and spring months as DO is strongly and primarily affected by temperature. Chlorophyll and phycocyanin, indicators for algae and cyanobacteria, respectively, were found in highest concentrations (34.56 µg/L and 19626.27 cell/mL, respectively) in the influent, where algal blooms were observed to start, during the summer months. However, chlorophyll and phycocyanin levels in the warm season were comparable to those in the cold season in the effluent.

! 77 Table 4.1. Summary of water quality parameters, with mean values and standard deviations, measured in both influent and effluent of the WSP system over the sampling period in both warm and cold seasons.

Warm season Cold season Influent Effluent Overall Influent Effluent Overall change change Temperature (°C) 21.18±1.77 21.35±1.26 -1% 9.65±4.22 7.81±4.51 19% pH (units) 7.86±0.22 8.14±0.74 0.04% 7.12±0.38 7.70±0.68 0.08% DO (mg/L) 7.19±2.86 6.82±2.93 5% 10.18±2.44 12.10±3.43 -19% Chlorophyll (µg/L) 34.56±80.61 1.19±0.81 97% 2.09±3.64 1.42±0.96 32% Phycocyanin 19626.27±49076.28 1209.82±579.04 94% 2317.29±3466.17 1396.69±718.06 40% (cell/mL) 3- PO4 (mg/L) 0.59±0.15 0.50±0.17 16% 0.52±0.25 0.50±0.28 3% TP (mg/L) 2.09±0.36 1.71±0.39 18% 1.54±0.43 1.48±0.43 4% + NH4 (mg/L) 0.67±0.36 0.77±0.42 -15% 0.57±0.74 9.38±19.00 -1557% NO3- (mg/L) 13.26±3.9 7.75±3.41 42% 4.22±5.02 1.31±0.74 69% TN (mg/L) 16.47±5.14 9.66±3.44 41% 17.50±6.21 13.00±2.34 26% COD (mg/L) 21.56±9.78 19.11±9.78 11% 17.30±6.24 21.56±8.28 -25% E. coli 2.17±0.86 1.03±0.52 1.14log 3.55±0.87 2.63±0.77 0.92log (logCFU/100mL) CFU/100mL CFU/100mL C. perfringens 3.10±0.93 2.60±1.23 0.50log 4.11±0.79 3.32±0.64 0.79log (logCFU/100mL) CFU/100mL CFU/100mL Enterococci 2.06±0.86 1.40±0.35 0.66log 2.91±0.63 1.83±0.90 1.08log (logCFU/100mL) CFU/100mL CFU/100mL TC 2.10±0.73 0.92±0.41 1.18log 2.79±0.58 1.73±0.56 1.06log (logCFU/100mL) CFU/100mL CFU/100mL

3- In this WSP system, the nutrient loadings of PO4 , TP, ammonium, and TN entering the system were comparable in both seasons, except for nitrate. The average concentration of influent nitrate in warm season was 13.26 mg/L, while the average nitrate concentration in the influent in

3- + - cold season was 4.22 mg/L. Similarly, the removal rates of PO4 , TP, NH4 , NO3 , TN and COD were higher in the warm season compared to in the cold season. However, the removal efficiency of nitrate in the warm season was 42%, while the removal efficiency was 69% in the cold season.

The average COD concentration was higher in the effluent than in the influent in the cold season, likely due to the decay of algae in the cold season. An increase of 34.6% in effluent COD concentrations was also noted by Wallace et al. (2016), whose work was conducted in the same

WSP during July-October 2012.

! 78 For the indicator organisms, both the influent and effluent concentrations were higher when comparing the cold season to the warm season. The removal rates of both TC and

Enterococci were similar, regardless of the season. Conversely, in the warm season, the removal efficiencies of E. coli and C. perfringens were 1.14 log CFU/100mL and 0.50 log CFU/100mL, respectively, and were 0.92 log CFU/100mL and 0.58 log CFU/100mL, respectively, in the cold season. These observed results may raise concerns and challenges with respect to the operation of the WSP system to meet E. coli discharge limits (200 log CFU/100mL) during the winter season.

4.3.2.! Multivariate statistical analysis

4.3.2.1.! Spearman’s Correlation Coefficient ! Spearman’s ρ correlation coefficient was conducted to examine the types of correlations between water quality parameters, summarized in Table 4.3. Strong positive correlations between chlorophyll and water temperature (ρ=0.257, p<0.01), pH (ρ=0.383, p<0.01) and DO (ρ=0.260, p<0.01) indicated the observed algal activity in summer months. During warmer periods, algal blooms can occur, resulting in increased pH and DO. Wang et al. (2009) noted similar strong positive correlations between these parameters. A strong positive correlation (ρ=0.701, p<0.01) between phycocyanin and chlorophyll was also observed, indicating growth of both algae and cyanobacteria.

Bacterial indicator organisms positively correlated with each other (Table 4.2).

Enterococci strongly and positively correlated with C. perfringens as they are both resistant to environmental stresses (Dunlop et al., 2008). Additionally, both had weaker correlations with both E. coli and TC, suggesting that they can be considered as additional bacterial indicators. E. coli, Enterococci and C. perfringens all strongly negatively correlated with water temperature, with ρ=-0.576 (p<0.01), ρ=-0.329 (p<0.01) and ρ=-0.203 (p<0.05), respectively, suggesting

! 79 higher temperatures are conducive to removal of bacterial indicators. In a laboratory study, Liu et al. (2017) investigated the effect of temperature (4 °C and 20 °C) on the removal and inactivation of the same four indicator organisms, as with this study in municipal wastewater, and noted that the inactivation rates of E. coli, TC and Enterococci were much lower under 4°C than under

20°C. Additionally, in a study by Mezrioui et al. (1995) , the removal rate of E. coli and other fecal coliforms increased when temperature was increased from 8°C to 23°C. This finding demonstrates that temperature is an important environmental factor contributing to disinfection in

WSP systems.

Strong negative correlations were found between pH and E. coli, Enterococci and C. perfringens, suggesting that pH is an important environmental factor in disinfection. Davies-

Colley et al. (1999) found that the inactivation rate of E. coli increased when pH was above 8.5.

Moreover, algae can be considered beneficial for disinfection in WSPs as its presence contributed to high pH. The strongly negative correlation between Enterococci and DO (ρ=-0.254, p<0.05) indicated that high DO levels aids in removal of Enterococci. This result was consistent with a bench scale experiment, where the highest inactivation rate of Enterococci was found at high DO concentrations (20 mg/L) (Liu et al., 2017).

! 80 Table 4.2. Spearman’s ρ correlation coefficients between all parameters.

Chloro- Phyco- pH DO PO 3- TP NH + NO - TN COD E. coli TC Enterococci C. perfringens phyll cyanin 4 4 3 Temperature 0.493** -0.311** 0.257** 0.265** -0.145 0.000 0.184* 0.661** 0.483 0.142 -0.576** -0.005 -0.329** -0.203* pH 0.427** 0.383** 0.332** -0.328** -.270** 0.089 0.525** 0.360 0.018 -0.538** -0.151 -0.550** -0.409** DO 0.260** 0.139 -0.084 -0.092 -0.182 0.065 0.166 -0.193 0.111 -0.081 -0.254* -0.142 Chlorophyll 0.701** 0.087 0.067 -0.232* 0.384** 0.180 -0.030 -0.191 0.079 -0.021 -0.091 Phycocyanin -0.033 0.033 -0.054 0.321** 0.261 0.099 -0.301** 0.043 -0.115 -0.122 3- ** ** ** ** ** PO4 0.736 -0.450 -0.364 0.176 0.096 0.195 0.085 0.389 0.271 TP -0.157 -0.257** 0.286 0.172 0.047 -0.058 0.325** 0.212* + ** NH4 0.421 0.533 0.047 -0.053 0.045 -0.132 0.197 - ** NO3 0.310 -0.003 -0.206 0.192 -0.292 0.075 TN 0.175 0.316 0.029 0.088 0.088 COD -0.131 -0.106 0.003 0.028 E. coli 0.509** 0.661** 0.675** TC 0.277** 0.329** Enterococci 0.542**

Note: |ρ| larger than critical value indicates strong correlation, and is bolded in red. ** indicates correlation is significant at the 0.01 level (2-tailed) * indicates correlation is significant at the 0.05 level (2-tailed)

! 81 4.3.2.2.! Principle Components Analysis

PCA was applied to the entire normalized data set to compare relations and groupings of water quality parameters. The Kaiser criterion was applied to determine the number of PCs to be retained (Kaiser, 1960). Under this criterion, only PCs with eigenvalues greater than 1.0 will be accepted as sources of variance, based on the rationale that selected PCs must have a variance at least as large as that of a single standardized original variable to be acceptable (Table 4.3).

Interpretation of the PCs is reliant on loadings, which indicate the original variables contributing most towards a certain component (Peres-Neto et al., 2017). The factor loading reveals the type of relation the specific variable has on the respective PC and relation between variables.

Variables with factor loadings exceeding |0.50| are considered to be significant loadings (King and Jackson, 1999).

Table 4.3. Descriptive statistics of selected PCs for all measured water quality parameters.

Cumulative PC Eigenvalue Variance (%) Variance (%) 1 3.40 31.21 31.21 2 1.45 13.27 44.48 3 1.29 11.87 56.34 4 1.08 9.93 66.28 5 1.06 9.14 75.42 6 0.74 6.79 82.22 7 0.63 5.80 88.02 8 0.44 4.00 92.01 9 0.29 2.67 94.68 10 0.19 1.70 96.38 11 0.17 1.58 97.96 12 0.10 0.87 98.84 13 0.07 0.62 99.46 14 0.04 0.41 99.87 15 0.01 0.13 100.00 Bold PCs represent retained PCs, based on the Kaiser criterion.

! 82 In this study, five PCs with eigenvalues exceeding 1 were retained, accounting for

75.42% of the total variance in the data (Table 4.3). Remaining PCs were discarded as they did not comprise of unique sources of variance in the water quality. While all 15 variables were included in the five selected PCs, only certain variables possessed significant loads (>|0.50|) within each PC (Table 4.4). PC 1 accounted for 31.2% of the variance of the data and had high significant loadings of -0.77, -0.71, -0.65, 0.70, 0.83, 0.64, and 0.69 for temperature, pH,

- chlorophyll, NO3 -N, E. coli, C. perfringens, Enterococci and TC, respectively. Alternatively, PC

3- + - 2 represented 13.3% of total variance and had high loadings for PO4 , TP, NH4 and NO3 . PC 3

3- contained high loadings for phycocyanin and PO4 and accounted for 11.9% of the variance in water quality parameters. PC 4, which represented 9.9% of variance, had significant loadings for

DO and TC. Lastly, PC 5 explained 9.1% of total variance in water quality, with high loadings of

0.59, 0.54 and -0.71 for phycocyanin, COD and TC, respectively.

PC 1 represents the primary factors involved in the disinfection of bacterial indicators. E. coli, C. perfringens, Enterococci and TC all had significant positive loadings of 0.70, 0.83, 0.64

- and 0.69, while temperature, chlorophyll, pH and NO3 had significant negative loadings of -

0.77, -0.71, -0.60 and -0.65, respectively. It was noted by Curtis et al. (1992) that in high pH environments, microorganisms have decreased resistance to the effects of solar radiation and induce production certain oxygen forms toxic to bacteria. Furthermore, in situ WSP studies conducted by Pearson et al. (1987) showed that faecal coliform counts were lowest under high pH, temperature, algae and DO conditions. In particular, pH values exceeding 9 increased faecal

- coliform die-off, especially under nutrient-poor conditions, such as low NO3 -N levels (Hirn,

Viljamaa, & Raevuori, 1980). Results from PC 1 are supported by the Spearman ρ correlation coefficient results and numerous studies that reported increased fecal coliforms removal with

! 83 increasing temperature, algae and pH (Curtis et al., 1992; Hirn et al., 1980; Mara et al., 1992;

Picot, Andrianarison, & Gosselin, 2005).

Table 4.4. Rotation PCA loading matrix for retained PCs and loadings for respective water quality parameters.

Principal Component 1 2 3 4 5 Temperature -0.77 -0.17 0.34 0.27 -0.21 pH -0.71 0.44 0.30 -0.21 0.08 DO 0.40 0.51 -0.06 -0.70 0.21 Chlorophyll -0.60 0.06 0.08 0.28 0.59 Phycocyanin -0.44 0.29 0.51 0.08 -0.10 PO₄³⁻ 0.27 -0.51 0.58 -0.39 0.01 TP 0.20 -0.60 0.43 -0.28 0.07 NH₄+⁺ 0.17 -0.57 0.13 0.39 0.03 NO₃- -0.65 0.59 0.02 -0.06 -0.06 TN -0.14 -0.06 0.16 -0.15 0.09 COD 0.07 0.06 0.11 0.08 0.54 E. coli 0.70 0.05 0.11 0.08 -0.01 C. perfringens 0.83 0.35 0.14 0.18 0.24 Enterococci 0.64 0.39 0.42 0.22 0.26 TC 0.69 -0.03 0.38 0.52 -0.71 Total Variance Explained Eigenvalues 3.4 1.4 1.3 1.1 1.1 Variances (%) 31.2 12.3 11.9 9.9 9.1 Cumulative Variance (%) 31.2 44.5 56.3 66.3 75.4 Note: The bold marked loads indicate absolute component loadings higher than |0.5|, considered significant contributors to the variance in water quality.

Variable loadings from PC 2 represented the nitrogen and phosphorus removal processes

- + present in the system. NO3 negatively correlated with NH4 , representing nitrification processes were occurring within the WSP. The high DO concentrations and oxidative conditions observed in the WSP were conducive to nitrification processes. Additionally, nitrifying bacteria were shown to have an optimal pH of approximately 6.8 to 7.8 and optimal temperature range of 15 to

25°C for maximum specific growth rate (Antoniou et al., 1990). These ranges are consistent with observed results as the average temperature of the system during the warm season (21.2°C).

! 84 However, at higher temperatures, the number of heterotrophic bacteria decreased in lagoons due to glucose competition between bacteria and algae, such as C. vulgaris, in a neutral pH environment (Mayo & Noike, 1996).

3- PC 3 loadings for PO4 and phycocyanin further support nutrient removal via biological processes involving algae and also, cyanobacteria. Benthic filamentous cyanobacteria are capable of nitrogen and phosphorus uptake and assimilation in wastewater treatment systems, reducing nutrient loadings. In particular, a study conducted by Chevalier et al. (2000) examined four strains of cyanobacteria and showed that they were capable of growth from 15 and 25°C, but limited at 5°C. This suggests that algae and cyanobacteria are more prevalent in warm climates, common during summer and fall in northern climates, such as this field study site.

In PC 4, DO and TC were shown to have opposite significant loadings of -0.70 and 0.52, respectively, suggesting that DO is an important factor in removal of TCs. It has been reported that both high DO (20mg/L) and low DO (1.0 mg/L) could facilitate their removal and inactivation of TCs while medium DO (8.5 mg/L) levels supported E. coli and TC survival (Liu et al., 2016a). This suggests the presence of algae, leading to high DO concentrations, facilitates the removal of TCs in wastewater (Curtis et al., 1992; Liu et al., 2016a).

PC 5 further supported existing roles of algae in the removal of TC and COD. It has been observed in numerous studies that algal photosynthesis leads to increased pH levels, leadings to die-off of fecal coliforms (Abdel-Raouf et al., 2012; Schumacher, Blume, & Sekoulov, 2003).

Furthermore, while COD removal can be predominantly attributed to bacterial metabolism, algae have also been shown to reduce COD through nutrient uptake and accumulation of suspended solids from wastewater on the surface of algal mats (Lau, Tam, & Wong, 1995).

! 85 4.4.! Conclusion

This study used multivariate statistical methods to evaluate the treatment performance a

WSP system, with the focus on discerning the factors and mechanisms contributing to disinfection processes in passive wastewater treatment systems. 15 water quality parameters were monitored over an 11-month sampling period and statistical tests were performed to determine the main factors affecting the removal and inactivation of indicator organisms and pathogens.

Strong positive correlations between chlorophyll and water temperature, pH and DO demonstrated algal activity in warm season, which resulted in elevated pH and DO. Strong negative correlations between water temperature and E. coli, Enterococci and C. perfringens indicated that high temperature would increase the removal rate of bacterial indicators. High pH can also promote the removal of E. coli, Enterococci and C. perfringens, while high DO can facilitate the inactivation of Enterococci. PCA did not result in considerable data reduction as it required 13 parameters to explain the 75% of the variation in data. However, five PCs obtained indicate that the parameters responsible for disinfection performance are mainly related to algae,

- temperature, pH and NO3 .

This study examined the treatment performance of WSPs, raising concerns with respect to the operation of the system in colder temperatures, as well as identifying important environmental factors involved in disinfection mechanisms. Ultimately, the results of this study provided valuable insight and understanding to the growing body of literature on passive wastewater treatment systems in order to improve the design and operation of WSPs.

! 86 4.5.! References

Abdel-Raouf, N., Al-Homaidan, A. A., & Ibraheem, I. B. M. (2012). Microalgae and wastewater

treatment. Saudi Journal of Biological Sciences, 19(3), 257–275.

http://doi.org/10.1016/j.sjbs.2012.04.005

Abreu-Acosta, N., & Vera, L. (2011). Occurrence and removal of parasites, enteric bacteria and

faecal contamination indicators in wastewater natural reclamation systems in Tenerife-

Canary Islands, Spain. Ecological Engineering, 37(3), 496–503.

http://doi.org/10.1016/j.ecoleng.2010.12.014

Al-Hashimi, M. A. I., & Hussain, H. T. (2013). Stabilization pond for wastewater treatment.

European Scientific Journal May, 9(14), 1857–7881.

Antoniou, P., Hamilton, J., Koopman, B., Jain, R., Holloway, B., Lyberatos, G., & Svoronos, S.

A. (1990). Effect of temperature and on the effective maximum specific growth rate of

nitrifying bacteria. Water Research, 24(1), 97–101. http://doi.org/10.1016/0043-

1354(90)90070-M

Awuah, E., Anohene, F., Asante, K., Lubberding, H., & Gijzen, H. (2001). Environmental

conditions and pathogen removal in macrophyte- and algal-based domestic wastewater

treatment systems. Water Science and Technology, 44(6), 11 LP-18. Retrieved from

http://wst.iwaponline.com/content/44/6/11.abstract

Bengraïne, K., & Marhaba, T. F. (2003). Using principal component analysis to monitor spatial

and temporal changes in water quality. Journal of Hazardous Materials, 100(1–3), 179–

195. http://doi.org/10.1016/S0304-3894(03)00104-3

Boon, N., De Windt, W., Verstraete, W., & Top, E. M. (2002). Evaluation of nested PCR-DGGE

(denaturing gradient gel electrophoresis) with group-specific 16S rRNA primers for the

! 87 analysis of bacterial communities from different wastewater treatment plants. FEMS

Microbiology Ecology, 39(2), 101–112. http://doi.org/10.1016/S0168-6496(01)00198-2

Boutilier, L., Jamieson, R., Gordon, R., Lake, C., & Hart, W. (2009). Adsorption, sedimentation,

and inactivation of E. coli within wastewater treatment wetlands. Water Research, 43(17),

4370–4380. http://doi.org/10.1016/j.watres.2009.06.039

Burkhardt, W., Calci, K. R., Watkins, W. D., Rippey, S. R., & Chirtel, S. J. (2000). Inactivation

of indicator microorganisms in estuarine waters. Water Research, 34(8), 2207–2214.

http://doi.org/10.1016/S0043-1354(99)00399-1

Chevalier, P., Proulx, D., Lessard, P., Vincent, W. F., & de la Noüe, J. (2000). Nitrogen and

phosphorus removal by high latitude mat-forming cyanobacteria for potential use in tertiary

wastewater treatment. Journal of Applied Phycology, 12(2), 105–112.

http://doi.org/10.1023/A:1008168128654

Curtis, T. P., Mara, D. D., & Silva, S. a. (1992). Influence of pH, Oxygen, and Humic substances

on ability of sunlight to dammage fecal-Coliforms in waste stabilisation pond water.

Applied Environmental Microbiology, 58(4), 1335–1343.

Davies-Colley, R. J., Donnison, A. M., Speed, D. J., Ross, C. M., & Nagels, J. W. (1999).

Inactivation of faecal indicator micro-organisms in waste stabilisation ponds: Interactions of

environmental factors with sunlight. Water Research, 33(5), 1220–1230.

http://doi.org/10.1016/S0043-1354(98)00321-2

Dong, X., & Reddy, G. B. (2010). Nutrient removal and bacterial communities in swine

wastewater lagoon and constructed wetlands. Journal of Environmental Science and Health,

Part A, 45(12), 1526–1535. http://doi.org/10.1080/10934529.2010.506109

Dunlop, P. S. M., McMurray, T. A., Hamilton, J. W. J., & Byrne, J. A. (2008). Photocatalytic

! 88 inactivation of Clostridium perfringens spores on TiO2electrodes. Journal of

Photochemistry and Photobiology A: Chemistry, 196(1), 113–119.

http://doi.org/10.1016/j.jphotochem.2007.11.024

Hirn, J., Viljamaa, H., & Raevuori, M. (1980). The effect of physicochemical, phytoplankton and

seasonal factors on faecal indicator bacteria in northern brackish water. Water Research,

14(3), 279–285. http://doi.org/10.1016/0043-1354(80)90098-6

Iordache, Ş., & Dunea, D. (2013). Multivariate analysis applied to physicochemical parameters

of treated wastewater effluents discharged in Teleajen River, 58, 717–723.

Jiménez, B., Maya, C., & Galván, M. (2007). Helminth ova control in wastewater and sludge for

advanced and conventional sanitation. Water Science and Technology, 56(5), 43–51.

http://doi.org/10.2166/wst.2007.555

Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educational and

Psychological Measurement, 20(1), 141–151.

King, J. R., & Jackson, D. A. (1999). Variable selection in large environmental data sets using

principal components analysis. Environmetrics, 10(1), 67–77.

http://doi.org/10.1002/(SICI)1099-095X(199901/02)10:1<67::AID-ENV336>3.0.CO;2-0

Lanao, M., Ormad, M. P., Goñi, P., Miguel, N., Mosteo, R., & Ovelleiro, J. L. (2010).

Inactivation of Clostridium perfringens spores and vegetative cells by photolysis and TiO2

photocatalysis with H2O2. Solar Energy, 84(4), 703–709.

http://doi.org/10.1016/j.solener.2010.01.023

Lau, P. S., Tam, N. F. Y., & Wong, Y. S. (1995). Effect of algal density on nutrient removal

from primary settled wastewater. Environmental Pollution, 89, 59–66.

Liu, L., Hall, G., & Champagne, P. (2016). Effects of environmental factors on the disinfection

! 89 performance of a wastewater stabilization pond operated in a temperate climate. Water

(Switzerland), 8(1). http://doi.org/10.3390/w8010005

Liu, L., Macdougall, A., Hall, G., & Champagne, P. (2017). Disinfection Performance in

Wastewater Stabilization Ponds in Cold Climate Conditions: A Case Study in Nunavut,

Canada. Environments, 4(4), 93. http://doi.org/10.3390/environments4040093

Mara, D., Mills, S., Pearson, H., & Alabaster, G. (1992). Waste Stabilization Ponds - a Viable

Alternative for Small Community Treatment Systems. Journal of the Institution of Water

and Environmental Management, 6(1), 72–78. http://doi.org/10.1111/j.1747-

6593.1992.tb00740.x

Maynard, H. E., Ouki, S. K., & Williams, S. C. (1999). Tertiary lagoons: A review of removal

mechanisms and performance. Water Research, 33(1), 1–13. http://doi.org/10.1016/S0043-

1354(98)00198-5

Mayo, A. W., & Noike, T. (1996). Effects of temperature and pH on the growth of heterotrophic

bacteria in waste stabilization ponds. Water Research, 30(2), 447–455.

http://doi.org/10.1016/0043-1354(95)00150-6

Metcalf & Eddy. (2003). Wastewater Engineering: Treatment and Resource Recovery (Fourth).

McGraw-Hill Science.

Mezrioui, N., Oufdou, K., & Baleux, B. (1995). Dynamics of Non-O1 Vibrio-Cholerae and

Fecal-Coliforms in Experimental Stabilization Ponds in the Arid Region of Marrakesh,

Morocco, and the Effect of Ph, Temperature, and Sunlight on Their Experimental Survival.

Can J Microbiol, 41(6), 489–498. http://doi.org/10.1139/m95-065

Miettinen, T., Hurse, T. J., Connor, M. A., Reinikainen, S.-P., & Minkkinen, P. (2004).

Multivariate monitoring of a biological wastewater treatment process: a case study at

! 90 Melbourne Water’s Western Treatment Plant. Chemometrics and Intelligent Laboratory

Systems. http://doi.org/10.1016/j.chemolab.2003.11.012

Molleda, P., Blanco, I., Ansola, G., & de Luis, E. (2008). Removal of wastewater pathogen

indicators in a constructed wetland in Leon, Spain. Ecological Engineering, 33(3–4), 252–

257. http://doi.org/10.1016/j.ecoleng.2008.05.001

Mosteo, R., Ormad, M. P., Goñi, P., Rodríguez-Chueca, J., García, A., & Clavel, A. (2013).

Identification of pathogen bacteria and protozoa in treated urban wastewaters discharged in

the Ebro River (Spain): water reuse possibilities. Water Science and Technology, 68(3), 575

LP-583. Retrieved from http://wst.iwaponline.com/content/68/3/575.abstract

Ouali, A., Jupsin, H., Ghrabi, A., & Vasel, J. L. (2014). Removal kinetic of

<em>Escherichia coli</em> and enterococci in a laboratory pilot scale

wastewater maturation pond. Water Science and Technology, 69(4), 755 LP-759. Retrieved

from http://wst.iwaponline.com/content/69/4/755.abstract

Pearson, H. W., Mara, D. D., & Bartone, C. R. (1987). Guidelines for the minimum evaluation of

the performance of full-scale waste stabilization pond systems. Water Research, 21(9),

1067–1075. http://doi.org/10.1016/0043-1354(87)90028-5

Peres-Neto, P. R. ., Jackson, D. A., & Somers, K. M. (2017). Giving Meaningful Interpretation to

Ordination Axes$: Assessing Loading Significance in Principal Component Analysis.

Ecology, 84(9), 2347–2363.

Perkins, R. G., & Underwood, G. J. C. (2000). Gradients of chlorophyll a and water chemistry

along an eutrophic reservoir with determination of the limiting nutrient by in situ nutrient

addition. Water Research, 34(3), 713–724. http://doi.org/10.1016/S0043-1354(99)00228-6

Picot, P., Andrianarison, T., & Gosselin, J. P. (2005). Twenty years’ monitoring of Mèze

! 91 stabilisation ponds: part I--Removal of organic matter and nutrient. Water Science and

Technology, 51(12), 23.

Ramsey, P. H. (1989). Order Correlation. Journal of Educational Statistic, 14(3), 245–253.

Reinoso, R., Blanco, S., Torres-Villamizar, L. A., & Bécares, E. (2011). Mechanisms for

Parasites Removal in a Waste Stabilisation Pond. Microbial Ecology, 61(3), 684–692.

http://doi.org/10.1007/s00248-010-9791-6

Reisenhofer, E., Picciotto, A., & Li, D. (1995). A factor analysis approach to the study of the

eutrophication of a shallow, temperate lake (San Daniele, North Eastern Italy). Analytica

Chimica Acta, 306(1), 99–106. http://doi.org/10.1016/0003-2670(94)00584-9

Schumacher, G., Blume, T., & Sekoulov, I. (2003). Bacteria reduction and nutrient removal in

small wastewater treatment plants by an algal biofilm. Water Science and Technology,

47(11), 195 LP-202. Retrieved from http://wst.iwaponline.com/content/47/11/195.abstract

Shrestha, S., & Kazama, F. (2007). Assessment of surface water quality using multivariate

statistical techniques: A case study of the Fuji river basin, Japan. Environmental Modelling

and Software, 22(4), 464–475. http://doi.org/10.1016/j.envsoft.2006.02.001

Tchobanoglous, G., & Crites, R. (1998). Small & Decentralized Wastewater Management

Systems. Michigan: WCB/McGraw-Hill.

Wallace, J., Champagne, P., & Hall, G. (2016). Multivariate statistical analysis of water

chemistry conditions in three wastewater stabilization ponds with algae blooms and pH

fluctuations. Water Research, 96, 155–165. http://doi.org/10.1016/j.watres.2016.03.046

Wallace, J., Champagne, P., & Hall, G. (2016). Time series relationships between chlorophyll-a,

dissolved oxygen, and pH in three facultative wastewater stabilization ponds. Environ. Sci.:

Water Res. Technol., 2(6), 1032–1040. http://doi.org/10.1039/C6EW00202A

! 92 Wang, G. X., Zhang, L. M., Chua, H., Li, X. D., Xia, M. F., & Pu, P. M. (2009). A mosaic

community of macrophytes for the ecological remediation of eutrophic shallow lakes.

Ecological Engineering, 35(4), 582–590. http://doi.org/10.1016/j.ecoleng.2008.06.006

! 93 Chapter 5. Monitoring seasonal variations in treatment performance of a wastewater stabilization pond with pH fluctuations

ABSTRACT

In 2012, Environment Canada updated the Wastewater System Effluent Regulations in an effort to reduce the 150 billion litres of untreated wastewater being discharged into Canadian waters annually. The revised regulations will result in the commissioning of over $20 billion in wastewater infrastructure upgrades for municipalities. Many small, rural and remote municipalities use passive wastewater treatment systems, such as wastewater stabilization ponds

(WSPs), as sustainable alternatives to conventional wastewater treatment due to their ease of operation, minimal energy input and low costs. However, since WSPs are open systems, they are susceptible to variations in external conditions. In particular, they are conducive to algal blooms and high pH events during the summer seasons, with warmer temperatures and higher hydraulic retention times. Temperature, pH, dissolved oxygen, Escherichia coli, nitrate and total phosphorus were collected from a WSP system in eastern Ontario, with excessive algal growth, over a five-year period. The removal efficiencies of various water quality parameters and indicator organisms for each season were used to determine seasonal treatment and disinfection performance of the system. Multivariate statistical tests and time series analyses were used to determine the strength and type of relationships influencing the WSP treatment for different seasons. Nitrate and E. coli removal were shown to be lowest during the winter periods at 95.6% and 27.9%, while total phosphorus remained consistent throughout the monitoring period. E. coli removal was shown to be significantly negatively correlated with pH (ρ =-0.268, p=0.05) and

! 94 DO (ρ=-0.390, p=0.01), using Spearman’s correlation coefficient. Seasonal Kendall tests revealed dissolved oxygen levels and nitrate concentrations both significantly decreased during the fall period. This research will be used directly to inform the monitoring program for the WSP system at the site and contributes to the continued improvement of WSP design and performance.

! 95 5.1.! Introduction

Wastewater management is a prevalent and challenging issue due to the compounding pressures of aging infrastructure, resource scarcity and climate change. In Canada, over 150 billion litres of untreated and raw sewage were discharged directly into rivers and surrounding oceans annually. In response, federal Wastewater Systems Effluent Regulations (WSER) were released in 2012, outlining mandatory minimum effluent quality standards following secondary wastewater treatment (Government of Canada, 2016). In Ontario, the Ministry of Environment and Climate Change (MOECC) has implemented additional criteria for wastewater quality parameters, with additional requirements for pH, total phosphorus (TP) and Escherichia coli (E. coli). These stringent regulations will result in municipalities allocating an estimated $20 billion in wastewater infrastructure upgrades (Canadian Water Network, 2014).

Passive wastewater treatment systems, such as wastewater stabilization ponds (WSPs), are low cost, low energy and sustainable solutions for rural and remote communities requiring additional treatment. WSPs can effectively attenuate nutrient loads while also providing environmental conditions suitable for the removal of pathogens through naturally occurring biological, chemical and physical processes. The effectiveness of WSPs depends on the complex interactions between biological communities, comprised mainly of algae, plants and bacteria. In particular, algae can effectively remove nutrients and aid in disinfection processes (Reinoso et al., 2008). Algae produce dissolved oxygen (DO), which is involved in photo-oxidation processes. Higher DO concentrations have been observed to decrease in E. coli populations

(Curtis et al., 1992). Additionally, algae can form a mutualistic relationship with aerobic bacteria, crucial for nutrient removal from municipal wastewaters. However, longer hydraulic

! 96 retention time common in WSP systems and the eutrophic nature of wastewater often lead to algal blooms and pH fluctuations during warmer periods (Davis et al., 2009).

Consequently, as WSP performance is strongly influenced by both environmental conditions and water quality parameters; they can be highly complex open systems, presenting unique challenges for monitoring and performance prediction. Characterization of wastewater, using multivariate statistical analyses, can provide an understanding of the water quality processes occurring in WSPs and guidance for operational changes or further monitoring.

Traditionally, wastewater monitoring, control and performance assessment have been primarily focused on conventional treatment (Dong & Reddy, 2010; Miettinen et al., 2004).

This study will focus on evaluating the treatment performance and seasonal dynamics of water quality parameters in a passive wastewater treatment system. The Amherstview Water

Pollution Control Plant, a WSP system located in southeastern Ontario, was selected as the study site to better understand these multivariate, time-dependent relationships between parameters over a long-term period. Over the past decade, excessive algal growth and high pH fluctuations have been observed at the Amherstview WSPs, a common operational and treatment issue for

WSPs located in temperate climates (Liu et al., 2016b; Wallace et al., 2016). This research aims to add to the body of literature pertaining to passive wastewater monitoring and treatment performance. A more comprehensive understanding of the long-term, seasonal trends in water quality is required for effective treatment performance. This paper will ultimately contribute to the growing field of study on wastewater treatment and management for small, remote and rural communities.

! 97 5.2.! Materials & Methods

5.2.1.! System overview Amherstview Water Pollution Control Plant (WPCP) consists of a direct treatment process, followed by passive treatment in two in-series facultative WSPs and a surface- flow constructed wetland (Figure 5.1). The WPCP treats an average of 3500 m3/day of municipal wastewater. The three passive treatment systems have operating depths of approximately 1.61 m,

1.42 m, and 1.17 m, and surface areas of 48 600 m2, 28 200 m2, and 31 200 m2, respectively, reaching a combined volume of 170 000 m3 (CH2M Hill, 2011) and an overall hydraulic retention time (HRT) of approximately 27 days. The final effluent is discharged into the

Bayview Bog wetland. WSP 1 contains a horizontal baffle along the length of the pond and has been observed to have historical occurrences of green algal blooms, predominantly from June to

October (Wallace et al., 2016).

Figure 5.1. Aerial view of Amerstview Water Pollution Control Plant. 5.2.2! Field Sampling

Monitoring was carried out from April 2011 to March 2015 to capture longer term, seasonal variations in climate. Weekly samples were collected from the plant secondary effluent

! 98 (SE) and the WSP 1 final effluent (FE), with biweekly sampling during the winter period due to the presence of surface ice cover (Figure 5.2). Water quality parameters including, temperature,

- pH, Escherichia coli (E. coli), dissolved oxygen (DO), total phosphorus (TP) and nitrate (NO3 ), were analyzed from samples collected at the effluent weirs between 10 am and 12 pm. E. coli was used as a common indicator organism for disinfection performance. pH and temperature were recorded using a HQ40d (Hach, Loveland, CO), a portable pH, conductivity, DO, ORP and

ISE multi-parameter meter. DO was monitored using a Model 3100 portable dissolved oxygen analyzer (Insite IG, Slidell, LA, USA). Standard methods and Hach kits (Colorado, USA) were used to determine nitrate (NO3-- N) (APHA Standard Method 4500C) and total phosphorous

(TP) (U.S. EPA Reactor Digestor Method, Hach Method 8190). E. coli enumeration was carried out using the membrane filtration method, adhering to the Standard Methods for the Examination of Water and Wastewater (APHA/AWWA/WEF, 2012).

Figure 5.2. Schematic of Amherstview WPCP and locations of plant secondary effluent (SE) and final effluent (FE) sampling points.

! 99 5.2.3.! Statistical Analysis

Statistical tests were performed using SPSS Statistics 24.0 software. Seasonal averages and removal efficiencies were calculated based on the data collected on the following dates:

Spring from March 20 to June 20, summer from April 21 to September 20, autumn from

September 21 to December 20 and winter spanned from December 21 to April 20. Average removal efficiencies were calculated based on corresponding (SE) and WSP final effluent (FE) values. Student’s t-tests were conducted to determine whether changes within overall and seasonal data sets were statistically significant at the p<0.05 level. Spearman's ρ correlation coefficient test was used as a general indicator of statistical dependence between variables, determining preliminary strong and weak correlations. Spearman’s ρ correlation values between 0 and 1 indicate positive correlations, while values between -1 and 0 indicate negative correlations. Critical values were determined to be 0.257, suggesting a strong correlation, based on N-2 d.f.’s, with N=100 and α=0.05 (Ramsey, 1989). Seasonal Kendall test, a special type of the Mann-Kendall test, is a nonparametric test that discerns significant (p<0.05 level) monotonic linear or non-linear trends.

5.3.! Results & Discussion

- 5.3.1.! Time Series of temperature, pH, DO, E. coli, TP, and NO3

Time series data collected for water quality parameters from April 2011 to March 2015 are shown in Figure 5.3. pH fluctuations occurred regularly, with pH increasing and reaching a maximum during the summer and early fall, coinciding with the occurrence of excessive algal growth. Regular incidences of pH exceedances above 9.5 units were observed over the monitoring period (Ontario Ministry of Environment and Climate Change, 2016). From 2001 to

2015, several spikes in E. coli concentrations were also observed, exceeding the MOECC E. coli

! 100 limit of 200 CFU/100 mL (Ontario Ministry of Environment and Climate Change, 2016). Higher levels of TP were noted to mostly occur during the summer and fall periods, although no

- regulatory exceedances occurred during this period. NO3 concentrations appeared to be cyclical in nature, with maximum and minimum periods contrasting effluent temperature levels.

Figure 5.3. Weekly final effluent concentrations in WSP1 from April 2011 to March 2015.

! 101 5.3.2.! Seasonal variations in treatment efficiency

Overall and seasonal means, corresponding standard deviations and treatment efficiencies are summarized in Table 5.1. DO concentrations were lowest during the summer and highest during the fall. High fluctuating levels of DO in the overall and seasonal averages suggest oxygen supersaturation in the water column, supporting the observed algal blooms as photosynthetic activity generate oxygen as a by-product. Additionally, a high average pH of 9.2 during the summer further corroborates the observed algal blooms as algal consumption of CO2 during photosynthesis leads to an increase in OH- production. A previous study conducted at

WSP 1 revealed similar results, with strong statistical relationships observed between chlorophyll-a, an indicator of green-algae, and DO and pH during the period of high algal growth

- (Wallace et al., 2016). During the summer, NO3 removal rate (95.6%) were the highest, in

- comparison to the other seasons. Both E. coli and NO3 removal efficiency rates were lowest in the winters, at 95.6% and 27.9, respectively.

Table 5.1. Weekly final effluent concentrations in WSP1 from April 2011 to March 2015. Overall and seasonal averages and removal efficiencies for final effluent parameters.

Parameter Overall Spring Summer Fall Winter Temperature FE 11.2±9.1 11.2±9.1 22.0±2.4 8.5±5.7 1.4±2.1 (°C) Change 13.8%* 3.5% -11.8%* 45.5 * 82.9%* FE 9.0±0.9 8.5±1.0 9.2±1.0 8.9±0.6 8.2±0.6 pH (units) Change -30.4%* -36.9%* -36.8%* -30.8%* 35.9%* FE 9.5±3.6 9.0±2.9 4.3±2.7 10.4±2.9 10.5±4.0 DO (mg/L) Change - 87.4%* - 129.6%* - 52.9%* -232.1%* -96.2%* E. coli FE 29.9±55.4 24.7±33.3 53.4±58.5 20.3±26.9 26.9±46.4 (CFU/100mL) Removal 98.6% 98.9% 99.8%* 98.8% 95.6%* FE 0.3±0.2 0.3±0.1 0.4±0.2 0.4±0.2 0.2±0.1 TP (mg/L) Removal 44.2%* 43.6% 38.9% 49.2%* 42.5% FE 6.0±5.5 6.8±6.0 1.0±3.3 4.8±3.2 12.3±2.9 NO - (mg/L) 3 Removal 71.4%* 62.3%* 95.6%* 75.1%* 27.9%* Note: * indicates seasonal differences are significant at the 0.05 level

! 102 5.3.3.! Correlations between water quality parameters Spearman’s ρ correlation coefficient test was conducted to determine the strength and types of relationships between water quality parameters (Table 5.2). Strong positive correlations were observed between temperature and pH (ρ=0.477, p=0.01), while temperature strongly

- negatively correlated with DO (ρ=-0.520, p=0.01) and NO3 (ρ=-0.883, p=0.01). It is well established that DO is dependent both on temperature and hydraulic mixing as DO concentrations are lower during warmer, stagnant periods and higher during cooler, aerated

- waters. NO3 was observed to have a negative correlation with pH (ρ=-0.441, p=0.01) and a positive correlation with DO (ρ=0.564, p=0.01). Nitrification rates have been revealed to be

- highest within a pH range of 7.5 to 9.0, resulting in higher NO3 amounts remaining in wastewater (Glass & Silverstein, 1998). Thus, the high DO concentrations and oxidative conditions observed in the WSP were conducive to nitrification processes. Additionally, nitrifying bacteria were shown to have an optimal pH of approximately 6.8 to 7.8 and optimal temperature range of 15 to 25°C for maximum specific growth rate (Antoniou et al., 1990). This is consistent with observed results as the average temperature of the system in the warm season remained within the optimal temperature range.

Table 5.2. Spearman’s ρ correlation coefficients between all water quality parameters.

- Temperature pH DO TP NO3 pH 0.477** DO -0.520** 0.0127 TP 0.235* -0.188* -0.195* - NO3 -0.883** -0.441** 0.564** -0.265** E. coli 0.237* -0.268* -0.390** 0.455** -0.318** Note: |ρ| larger than critical value (ρ=0.257) indicates strong correlation, with strong positive correlations in green and strong negative correlations in red * indicates correlation is significant at the 0.05 level (2-tailed) ** indicates correlation is significant at the 0.01 level (2-tailed)

! 103 E. coli had a negative correlation to pH (ρ=-0.268, p=0.01) and DO (ρ=-0.390, p=0.01), suggesting that pH and DO are important factors in pathogen removal. pH levels higher than 8.5, as seen in this WSP, have been noted to be effective for the removal fecal coliforms, with pH fluctuations negatively affecting E. coli survival (Awuah et al., 2001; Davies-Colley et al.,

1999). Moreover, fluctuations in pH have been shown to negatively affect the survival of E. coli

(Awuah et al., 2001). Bacterial disinfection via photo-oxidation also requires the presence of oxygen to form reactive oxygen species (ROS). Thus, it has been suggested that higher DO levels increase E. coli removal though light-activated disinfection (Davies-Colley, Bell, &

Donnison, 1994). Furthermore, a synergistic increase in E. coli removal was observed with increased levels of both DO and pH (Davies-Colley et al., 1999).

5.3.4.! Seasonal trends

Seasonal Kendall tests revealed the directional nature of water quality parameters for each season, as shown in Table 5.3. Temperature increased and decreased as expected in a temperate region. pH levels were observed to significantly decrease over time during both fall

- and winter. DO had an increasing trend during the summer months while NO3 appeared to decline during the fall. E. coli concentrations significantly decreased during the summer and increased during the winter.

Table 5.3. Seasonal Kendall test of trends for final effluent parameters in order of spring, summer, fall and winter (respectively).

Temp pH DO E. coli TP NO - 3 Season S S F W S S F W S S F W S S F W S S F W S S F W Trend ↑* - ↓* - - - ↓* ↓* - ↑* - - - ↓* - ↑* ------* ↓* - Note: * indicates correlation is significant at the 0.05 level

! 104 Declining trends in pH for both the fall and winter can be largely attributed to changes in algal activity. Optimum temperature for algal growth can range 20°C to 30°C of different algal species and have severely limited growth rates below 8°C (Butterwick, Heaney, & Talling, 2005;

Singh & Singh, 2015). Average fall and winter temperatures were below optimal conditions, resulting in minimal algal growth. As algal activity decreased, CO2 consumption decreased along with OH- production, subsequently leading to a pH decline. E. coli concentrations decreased over the summer and increased during the winter periods. A monitoring study conducted on the same

WSP further corroborates the observed fluctuations in E. coli concentrations and removal efficiencies (Liang, 2018). Higher wastewater temperatures have been shown to be more conducive to pathogen removal. Removal efficiencies of fecal coliforms, such as E. coli, increased by 1 log (CFU/100mL) with higher wastewater temperatures (15 to 18°C) when

- compared to colder temperatures (2 to 8°C) (Ulrich et al., 2005). Additionally, NO3 concentrations decreased during the fall seasons. Ammonification and denitrification processes are sustained in colder temperatures as ammonifying and denitrifying bacteria, such as

Pseudomonas, Achromobacter and Bacillus, are capable of growth at 4°C. As the average fall temperature range of the WSP was 8.5°C, denitrifying bacteria and nitrate removal activity could likely be sustained with decreasing temperature (Yao, Ni, Ma, & Li, 2013).

5.4.! Conclusions This study evaluated the treatment performance a WSP system, with the focus on the seasonal fluctuations in treatment performance of a WSP experiencing excessive algal growth problems. Algal activity was shown to have significantly impacted pathogen disinfection removal in summer seasons while temperature, pH, and DO were primary treatment mechanisms for pathogen and nutrient removal. Seasonal Kendall tests revealed that nitrate and E. coli

! 105 concentrations are highly variable in WSPs and exhibit seasonal patterns. Thus, algal dynamics and its role in treatment should also be considered. There also potential areas of concerns and challenges with respect to the operation of the WSP system to meet discharge limits during the winter seasons. Thus, WSPs are shown to be sensitive to location and climate, making wastewater treatment using natural systems more challenging. This research contributes to the continued improvement of low cost and effective wastewater treatment options, such as WSPs, which are especially important for resource-limited rural and Indigenous communities.

! 106 5.5.! References Antoniou, P., Hamilton, J., Koopman, B., Jain, R., Holloway, B., Lyberatos, G., & Svoronos, S.

A. (1990). Effect of temperature and ph on the effective maximum specific growth rate of

nitrifying bacteria. Water Research, 24(1), 97–101.

APHA/AWWA/WEF. (2012). Standard Methods for the Examination of Water and Wastewater.

Standard Methods, 541. http://doi.org/ISBN 9780875532356

Awuah, E., Anohene, F., Asante, K., Lubberding, H., & Gijzen, H. (2001). Environmental

conditions and pathogen removal in macrophyte- and algal-based domestic wastewater

treatment systems. Water Science and Technology, 44(6), 11 LP-18. Retrieved from

http://wst.iwaponline.com/content/44/6/11.abstract

Butterwick, C., Heaney, S. I., & Talling, J. F. (2005). Diversity in the influence of temperature

on the growth rates of freshwater algae, and its ecological relevance. Freshwater Biology,

50(2), 291–300. http://doi.org/10.1111/j.1365-2427.2004.01317.x

Canadian Water Network. (2014). 2014 Canadian Municipal Water Priorities Report. Retrieved

from http://www.cwn-rce.ca/initiatives/municipal-consortium/2014-canadian-municipal-

water-priorities-report/

CH2M Hill. (2011). Technical Assessment to Support Re-Rating the Amherstview WPCP Draft.

Curtis, T. P., Mara, D. D., & Silva, S. a. (1992). Influence of pH, Oxygen, and Humic substances

on ability of sunlight to dammage fecal-Coliforms in waste stabilisation pond water.

Applied Environmental Microbiology, 58(4), 1335–1343.

Davies-Colley, R. J., Bell, R. G., & Donnison, A. M. (1994). Sunlight inactivation of enterococci

and fecal coliforms in sewage effluent diluted in seawater. Applied and Environmental

Microbiology, 60(6), 2049–2058.

Davies-Colley, R. J., Donnison, A. M., Speed, D. J., Ross, C. M., & Nagels, J. W. (1999).

! 107 Inactivation of faecal indicator micro-organisms in waste stabilisation ponds: Interactions of

environmental factors with sunlight. Water Research, 33(5), 1220–1230.

Davis, T. W., Berry, D. L., Boyer, G. L., & Gobler, C. J. (2009). The effects of temperature and

nutrients on the growth and dynamics of toxic and non-toxic strains of Microcystis during

cyanobacteria blooms. Harmful Algae, 8(5), 715–725.

http://doi.org/10.1016/j.hal.2009.02.004

Dong, X., & Reddy, G. B. (2010). Nutrient removal and bacterial communities in swine

wastewater lagoon and constructed wetlands. Journal of Environmental Science and Health,

Part A, 45(12), 1526–1535. http://doi.org/10.1080/10934529.2010.506109

Glass, C., & Silverstein, J. (1998). Denitrification kinetics of high nitrate concentration water:

pH effect on inhibition and nitrite accumulation. Water Research, 32(3), 831–839.

http://doi.org/10.1016/S0043-1354(97)00260-1

Government of Canada. (2016). Wastewater Systems Effluent Regulations: SOR/2012-139.

Règlement sur les effluents des systèmes d ’ assainissement des eaux usées: DORS/2012-

139. Government of Canada Justice Laws Website. Retrieved from http://laws-

lois.justice.gc.ca/eng/regulations/SOR-2012-139/

Liu, L., Hall, G., & Champagne, P. (2016). Effects of environmental factors on the disinfection

performance of a wastewater stabilization pond operated in a temperate climate. Water

(Switzerland), 8(1). http://doi.org/10.3390/w8010005

Miettinen, T., Hurse, T. J., Connor, M. A., Reinikainen, S.-P., & Minkkinen, P. (2004).

Multivariate monitoring of a biological wastewater treatment process: a case study at

Melbourne Water’s Western Treatment Plant. Chemometrics and Intelligent Laboratory

Systems. http://doi.org/10.1016/j.chemolab.2003.11.012

! 108 Ontario Ministry of Environment and Climate Change. (2016). Design Guidelines For Sewage

Works. Retrieved January 27, 2018, from https://www.ontario.ca/document/design-

guidelines-sewage-works-0

Ramsey, P. H. (1989). Order Correlation. Journal of Educational Statistic, 14(3), 245–253.

Reinoso, R., Torres, L. A., & Bécares, E. (2008). Efficiency of natural systems for removal of

bacteria and pathogenic parasites from wastewater. Science of the Total Environment,

395(2–3), 80–86. http://doi.org/10.1016/j.scitotenv.2008.02.039

Singh, S. P., & Singh, P. (2015). Effect of temperature and light on the growth of algae species:

A review. Renewable and Sustainable Energy Reviews, 50, 431–444.

http://doi.org/10.1016/j.rser.2015.05.024

Ulrich, H., Klaus, D., Irmgard, F., Annette, H., Juan, L. P., & Regine, S. (2005). Microbiological

investigations for sanitary assessment of wastewater treated in constructed wetlands. Water

Research, 39(20), 4849–4858. http://doi.org/10.1016/j.watres.2004.07.020

Wallace, J., Champagne, P., & Hall, G. (2016). Multivariate statistical analysis of water

chemistry conditions in three wastewater stabilization ponds with algae blooms and pH

fluctuations. Water Research, 96, 155–165. http://doi.org/10.1016/j.watres.2016.03.046

Wallace, J., Champagne, P., & Hall, G. (2016). Time series relationships between chlorophyll-a,

dissolved oxygen, and pH in three facultative wastewater stabilization ponds. Environ. Sci.:

Water Res. Technol., 2(6), 1032–1040. http://doi.org/10.1039/C6EW00202A

Yao, S., Ni, J., Ma, T., & Li, C. (2013). Bioresource Technology Heterotrophic nitrification and

aerobic denitrification at low temperature by a newly isolated bacterium , Acinetobacter sp .

HA2. Bioresource Technology, 139, 80–86. http://doi.org/10.1016/j.biortech.2013.03.189

!

! 109

Chapter 6. Investigating long-term seasonal dynamics in a wastewater stabilization pond with algal blooms using multivariate statistical analysis

ABSTRACT

WSPs can effectively attenuate nutrient loads and remove pathogens through naturally occurring biological, chemical and physical treatment mechanisms. Water quality parameters were collected from a WSP system with extensive algal growth over a five-year period.

Significant seasonal variations in removal efficiencies and trends of water quality parameters were observed. Multivariate statistical analysis, including Spearman’s correlation, Factor

Analysis/Principal Components Analysis and Discriminant Analysis, were used to determine the factors affecting wastewater quality variation. Temperature exhibited a strong positive correlation to pH (ρ=0.477, p<0.01) and negative correlations with DO (ρ=-0.390, p<0.01).

Seasonal Kendall tests revealed increasing trends in E. coli and DO concentrations throughout the winter while pH levels decreased. FA/PCA conducted on the overall, spring, summer fall and winter data sets accounted for 77.28%, 68.44%, 71.50%, 70.79% and 68.19% of the total variance, respectively. DA delineated four parameters, namely temperature, pH, DO and TP, responsible for temporal variations in wastewater quality and correct assigned 74% of data collected. The multivariate statistical methodologies presented offer an insightful approach to monitoring wastewater quality where large datasets are generated. This research will be used directly to inform the monitoring program for the WSP system at the site and contributes to the continued improvement of seasonal WSP operations in temperate climates.

! 110 6.1.! Introduction

Wastewater stabilization ponds (WSPs) effectively facilitate treatment, reducing nutrient loadings and removing pathogenic organisms, via naturally-occurring biogeochemical processes.

WSPs are capable of treating wastewater with a wide range of characteristics, ranging from raw to secondary domestic wastewaters, under a variety of climactic conditions in arctic, temperate and tropical regions (Faleschini & Esteves, 2017; Hayward, Jamieson, Boutilier, Goulden, &

Lam, 2014; Liu, Macdougall, Hall, & Champagne, 2017; Maynard, Ouki, & Williams, 1999;

Ragush et al., 2015). Treatment performance effectiveness may vary vastly depending on location, seasonality, and regional climate. For instance, temperate climates are characterized by distinct, seasonal fluctuations in temperature, types of precipitation and solar radiation. As a result, WSPs often experience periods of warmer temperatures and longer hydraulic retention times (HRT) during the summer months. As a consequence, treatment performance and biological processes, such as algal dynamics, in WSPs can be affected by seasonal variations.

Characterization of water quality parameters using multivariate statistical analysis, can provide an understanding of the biogeochemical processes occurring in WSPs and guidance for operational changes or further monitoring. Multivariate statistical techniques, such as correlation tests, principal component analysis (PCA), factor analysis (FA) and discriminant analysis (DA), can guide the interpretation of large, complex data sets to better understand spatial and temporal changes in ecological systems (Shrestha & Kazama, 2007). In particular, these analytical tools are used in water quality monitoring and facilitate the identification of possible factors that could influence specific water systems, providing valuable information for water resources management. These methods are used in freshwater and groundwater research, exploring and evaluating the impacts of polluting and natural sources on water quality (Arhonditsis et al., 2003;

! 111 Baborowski, Simeonov, & Einax, 2012). In recent years, these methods have also been applied to wastewater monitoring, control and performance assessment, including passive treatment systems. In particular, they have been used to examine nutrient removal, impacts of wastewater effluent into receiving waters, and passive treatment systems experiencing algal blooms (Dong &

Reddy, 2010; Iordache & Dunea, 2013; Wallace et al., 2016).

Amherstview Water Pollution Control Plant (WPCP) employs a passive wastewater treatment system and is located in southeastern Ontario, Canada in a temperate climate.

Excessive algal growth and high pH fluctuations have been observed at the Amherstview WSP, a common operational issue for passive wastewater treatment systems which also been shown to positively affect disinfection performance (Liu, Hall, & Champagne, 2016; Wallace et al., 2016).

A previous study conducted at Amherstview WPCP was conducted examined seasonal variations in government-regulated water quality parameters (Liang et al., 2018). The results revealed significant seasonal variations in treatment performance and required further exploration of a more comprehensive water quality data set to determine which factors were primarily responsible for these observed temporal variations.

This study will investigate the role of seasonality in the long-term treatment performance and fluctuations in a comprehensive water quality data set collected from Amherstview WPCP.

Removal efficiencies, Seasonal Kendall tests, as well as multivariate statistical methods were used to extract meaningful information regarding wastewater quality parameters to discern factors responsible for temporal variations in wastewater quality. A comprehensive understanding of seasonal variations in wastewater quality and treatment performance is required to inform the effective, long-term management of passive wastewater treatment systems in temperate climates.

! 112 6.2.! Materials 6.2.1.! System overview Amherstview WPCP is located in southeastern Ontario, Canada in a temperate climate region, characterized by hot-summer humid continental climate and four distinct seasons: spring, summer, fall and winter. The WPCP consists of a direct activated sludge treatment process, followed by passive treatment in two in-series facultative WSPs and a surface-flow constructed wetland (Figure 6.1). The WPCP treats an average of 3500 m3/day of municipal wastewater. The three passive treatment systems have operating depths of approximately 1.61 m, 1.42 m, and

1.17 m, and surface areas of 48 600 m2, 28 200 m2, and 31 200m2, and a combined hydraulic retention time (HRT) of approximately 27 days (CH2M Hill, 2011). WSP 1 was observed to have experienced repeated occurrences of surface green algal blooms, predominantly from June to October (Liu et al., 2016).

Figure 6.1. Geographic location and schematic of Amherstview WPCP, including location of plant secondary effluent (SE) and final effluent (FE) sampling points (Image adapted from Natural Resource Canada).

! 113 6.2.2.! Field Sampling Sampling was carried out by Loyalist Township staff and the monitoring period lasted from April 2011 to March 2015. Data was categorized into four seasonal groups of Spring

(March 21 to June 21), summer (June 22 to September 22), autumn (September 23 to December

21) and winter (December 22 to March 20). Weekly samples were collected from the plant secondary effluent (SE) and the WSP 1 final effluent weir (FE) between 10am to 12pm.

Monitored water quality parameters included temperature, pH, dissolved oxygen (DO),

- + Escherichia coli (E. coli), nitrate (NO3 ), ammonium (NH4 ), total phosphorus (TP), carbonaceous biological oxygen demand (cBOD), total suspended solids (TSS), and alkalinity.

Descriptive statistics of overall and seasonal water quality parameters are provided in Table 6.2. pH and temperature were recorded using a HACH portable pH meter HQ11d and DO was measured using a HACH portable DO meter HQ30d (Hach, Loveland, CO). HACH kits

3- (Colorado, USA) were used to determine PO4 (HACH method 8048), alkalinity (HACH method 10239) and TP (U.S. EPA Reactor Digestor Method, Hach Method 8190). Standard

Methods for the Examination of Water and Wastewater, SM9222B, SM2540D, SM4110C,

SM4500 and SM5210B, were used for E. coli (a bacterial indicator organism for disinfection

- + performance), TSS, NO3 , NH4 and cBOD, respectively (APHA/AWWA/WEF, 2012).

6.2.3.! Statistical Analysis

6.2.3.1.! Data treatment and calculations All statistical computations were performed using SPSS Statistics 24.0 and CANOCO

5.0. Each season was assigned a numerical value in the data files (spring - 1; summer - 2; autumn

- 3 and winter - 4) corresponding to the season and was considered pairwise with all measured parameters. Average removal efficiencies were calculated based on corresponding secondary

! 114 treatment (SE) and WSP final effluent (FE) values. Student’s t-tests were conducted to determine whether removal efficiencies were significantly different from initial SE sample concentrations.

Kaiser-Meyer-Olkin (KMO) and Bartlett’s test of sphericity were conducted to examine the suitability of the data for PCA/FA. KMO indicates sampling adequacy and the proportion of common variance. Values close to 1 generally indicate that PCA/FA may be useful. In this study, the KMO value was 0.82. Bartlett’s test indicates whether a correlation matrix is an identity matrix, demonstrating that variables are unrelated. In this study, p<0.05, indicating significant relationships existed among variables.

Spearman's ρ correlation coefficient test revealed statistical dependence between variables, determining preliminary positive (between 0 and 1) and negative correlations (between

-1 and 0). Critical values were determined to be 0.257, suggesting a moderately strong correlation, based on N-2 d.f.’s, with N=100 and α=0.05 (Ramsey, 1989). Seasonal Kendall test, a special type of the Mann-Kendall test, is a nonparametric test that analyzes data for significant monotonic linear or non-linear, increasing or decreasing trends. A separate Mann-Kendall trend test on each season was conducted, where trends were determined within the season.

6.2.3.2.! Factor Analysis/Principal Component Analysis

FA and PCA are methods used to reduce the number of variables in order to determine which variables are responsible for majority of the variance in the data. PCA/FA were applied to data standardized through z-scale transformation to avoid misclassifications arising from the different orders of magnitude of both numerical values and variance of the parameters

(Simeonov et al., 2003). In PCA, a set of correlated variables is orthogonally transformed to a set of uncorrelated linear combinations referred to as principal components (PCs) using a covariance matrix. The Kaiser criterion allows for PCs with Eigenvalues exceeding 1.0 to be retained to

! 115 reduce redundancy (Kaiser, 1960; King & Jackson, 1999).The retained PCs were extracted and rotated to determine the factor loadings for each water quality parameter (Kaiser, 1960). The first

PC represents the linear combination of the variables with maximal variance and represents the largest variability of the original data set and subsequent PCs and so on. The parameter loadings of retained PCs indicate the contribution or weight of corresponding parameter to the PC and allows principal parameters to be inferred (Ouyang, 2005).

6.2.3.3.! Discriminant Analysis

DA is used to determine whether a variable known a priori is effective in predicting category membership for a set of observations. DA uses the original data to construct a

Discriminant Factor (DF), a linear combination of independent variables that will discriminate between categories of the dependent variable. The first DF maximally separates the groups, followed by second DF selected on the basis that it is maximally uncorrelated with the first DF.

This procedure continues until the maximum number of DFs produced, determined by the number of predictors and categories in the dependent variable (Alberto et al., 2001; Singh et al.,

2004). The classification table compares the predicted versus actual group membership, and reveals how well the derived function predicts group membership. In forward stepwise DA mode, variables are included step-by-step starting with the highest significance and moves downward, with all variables retained. Alternatively, backward stepwise mode uses the opposite approach and variables are removed beginning with the least significant variables.

6.3.! Results & Discussion 6.3.1.! Time series of wastewater parameters

Descriptive statistics of each water quality parameter are provided in Table 6.1 and time series plots of wastewater quality parameters can be seen in Figure 6.2. While the average pH of

! 116 the system was 8.9, season fluctuations in pH occurred. High average pH of 9.2 and 9.0 in the spring and summer as well as high summer and fall alkalinity levels of 124.5 mg/L CaCO3 and

127.4 mg/L CaCO3, further supported the observed algal blooms as the carbonate–bicarbonate equilibrium was impacted by algal photosynthetic activity (Figure 6.1). As the carbonate– bicarbonate equilibrium was impacted by algal photosynthetic activity, increased production of

- - OH and HCO3 ions were observed based on high alkalinity levels, which in turn affected the pH of the system (Ragush & Jamieson, 2016). During the winters, mean alkalinity levels reduced to

99.6 mg/L CaCO3, corresponding with the minimal algae observed on the WSP surface.

High fluctuating levels of DO in the overall and seasonal averages suggest oxygen supersaturation of the water column, supporting the observed algal blooms as photosynthetic activity generates oxygen as a by-product (Figure 6.2c). A previous study conducted at WSP 1 revealed similar results, with strong statistical relationships observed between chlorophyll-a, pH and, DO during the period of algal bloom (Wallace, Champagne, & Hall, 2016). It was also determined that both nitrogen and phosphorus were available in excess of the concentrations, providing conditions suitable for high algal activity in the WSP (Wallace et al., 2016).

While average E. coli concentrations remained fairly constant, several spikes occurred during the winter months, reaching or exceeding the MOECC discharge limit of 200

CFU/100mL during the period from 2001 to 2015. These results may raise concerns and challenges with respect to WSP operation during the winter season.

! 117 Table 6.1. Descriptive statistics of overall and seasonal wastewater quality parameters monitored at WSP 1.

- + Temperature pH E. coli DO TSS TP NO3 NH4 Alkalinity cBOD (°C) (units) (log CFU /100mL) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L CaCO3) (mg/L) SE FE SE FE SE FE SE FE SE FE SE FE SE FE SE FE SE FE SE FE

Min 6.00 0.10 6.68 7.20 3.2 0 0.50 0 0.20 0 0.04 0.05 0.01 0.01 0 0.01 52.60 38.90 0.50 0.50 Max 22.40 29.70 7.91 10.68 3.4 3.42 8.60 19.30 25.80 32.50 2.60 1.16 29.30 19.70 0.74 1.12 195.80 175.00 15.00 10.00 Mean 14.05 11.67 7.32 8.87 3.3 0.26 4.72 8.95 6.62 5.21 0.56 0.31 19.77 6.26 0.06 0.21 102.27 119.24 3.55 2.09 Overall S.D. 4.91 9.07 0.27 0.84 2.3 2.41 1.94 3.91 4.84 5.81 0.36 0.19 5.45 5.66 0.09 0.22 29.51 28.81 3.23 1.96 Min 6 0.9 6.69 8.61 3.3 0 0.2 1 0 0.6 0.11 0.06 6.2 0.6 0.00 0.00 71.12 55.05 0.50 0.50

Max 17.80 23.20 7.68 10.68 3.3 0.99 11.09 19.30 20.60 22.70 2.60 0.52 27.40 12.10 0.74 0.37 173.10 146.70 15.07 9.10 Mean 3.3 Spring 12.27 12.70 7.23 9.55 0.87 5.56 11.45 7.17 5.92 0.50 0.14 17.50 4.67 0.08 0.09 125.08 95.15 3.93 2.45 S.D. 3.43 7.04 0.26 0.55 0.00 1.34 2.35 4.13 5.49 5.16 0.51 0.10 5.54 2.78 0.16 0.11 28.46 31.52 5.21 2.27 Min 3.3

17.1 15.3 6.7 7.99 0 1.71 0 0.2 0 0.17 0.14 0.05 0.05 0.00 .00 53 50 0.50 0.54 Max 22.4 29.7 7.84 11.07 3.3 2.6 5.5 9.0 15.7 23.5 1.96 0.95 29.3 19.7 0.245 1.05 195.8 175 11.12 10.12 Mean 19.73 22.51 7.28 9.62 3.3 1.7 3.42 5.38 4.66 4.19 0.63 0.39 21.81 1.19 0.05 0.18 98.88 127.37 3.67 2.85 Summer S.D. 1.40 2.65 0.31 0.61 0.00 1.91 1.02 2.09 3.76 6.21 0.35 0.23 5.63 3.76 0.05 0.25 36.87 27.37 3.15 2.68 Min 17.10 1.80 6.70 7.88 3.3 0 1.71 1.00 0.20 0.00 0.17 0.13 0.05 0.51 0.00 0.00 53 62.12 7.97 4.02

Max 22.40 20.08 7.84 10.07 3.3 2.6 5.50 13.40 15.70 32.50 1.96 0.85 29.30 12.70 0.25 0.76 195.80 151.20 0.50 0.50 Fall Mean 19.77 22.08 7.28 9.01 3.3 1.7 3.46 9.80 4.61 4.26 0.62 0.34 21.85 5.07 0.05 0.16 98.81 124.46 7.32 1.14 S.D. 1.39 5.43 0.31 0.81 0.00 1.91 1.01 3.08 3.81 6.77 0.36 0.15 5.71 3.76 0.06 0.20 37.40 24.07 3.86 1.24 Min 6.00 6.10 0.10 6.87 7.20 0 0.00 0.70 1.40 0.30 0.30 0.07 0.06 8.80 7.80 0.00 0.00 75.21 2.01 0.50

Max 11.2 10.80 3.20 8.98 9.54 2.6 160.00 17.00 17.40 11.80 13.30 0.42 0.42 19.30 19.30 1.12 1.05 169.70 7.10 5.00 Mean Winter 8.34 8.23 1.2 8.01 8.19 1.9 26.86 10.74 10.54 3.32 4.15 0.25 0.23 12.99 12.29 0.25 0.19 99.59 4.29 1.60 S.D. 1.12 1.09 0.89 0.49 0.55 2.46 46.48 3.31 4.01 2.55 3.40 0.10 0.11 2.91 2.88 0.24 0.23 23.65 1.60 1.27

! 118

Seasonal fluctuations in TP were observed, with lowest TP concentrations in the spring

(seasonal average of 0.14 mg/L), reaching a variable maximum during the summer and

+ decreased during the winter. NH4 and TSS concentrations were fairly consistent with several

high spikes. Average cBOD levels remained fairly constant, with the most visible fluctuations

during the summer, in corroboration with high levels of organic matter from increased algae and

plant biomass (Figure 6.2h). Alkalinity levels were found to be in higher concentrations (average

of 127.4 mg/L CaCO3) during summer when biological nitrification rates were highest.

a)# b)#

c)# d)#

! 119 e)# f)#

g)# h)#

i)# j)#

Figure 6.2. Time series plots depicting seasonal variations of a) temperature, b) pH, c) DO, d) E. - + coli; e) NO3 ; f) NH4 ; g) TP; g) cBOD; h) TSS; i) alkalinity in the final effluent water quality. 6.3.2.! Treatment performance of WSP Overall and seasonal removal efficiencies of MOECC-monitored wastewater effluent

parameters for WSP 1 can be seen in Table 6.3. E. coli removal was highest in the spring and

summer and lowest during the winter, coinciding with pH levels withinh# the system. High pH levels and frequent fluctuations have been noted to negatively affect E. coli survival (Awuah,

Anohene, Asante, Lubberding, & Gijzen, 2001; Liu et al., 2016). cBOD removal rates were

! 120 highest in the fall, with a 71.1% decrease from secondary effluent to the final effluent of WSP 1, remaining consistent with algae decomposition. TSS removal was lowest during the summer as algal growth and decay can contribute much of the TSS loading of a WSP effluent, reaching up

- to 50 to 60% of the TSS concentrations in high rate aerobic WSPs (Shammas et al., 2008). NO3 removal rates were highest during summer and fall, at 95.6% and 75.1% respectively, while TP removal remained fairly constant.

Table 6.1. Removal efficiencies for MOECC effluent discharge parameters.

Overall Spring Summer Fall Winter N=216 (N=52) (N=62) (N=55) (N=47) Log Removal E. coli 1.5±2.4 1.42±1.32 1.6±1.9 1.6±1.9 1.3±2.1 (log CFU/100mL) Removal 2.0±2.1 1.1±2.2 1.7±2.9 3.4±1.8 1.2±1.3 cBOD % 39.4% 29.1% 35.0% 71.1% 55.0% Removal 2.6±6.9 3.2±3.8 1.0±7.0 5.1±8.2 4.7±6.6 TSS % 34.6% 49.5% 21.2% 58.2% 53.2% Removal 0.3±0.3 0.2±0.2 0.2±0.4 0.3±0.3 0.2±0.2 TP % 44.2% 43.6% 38.9% 49.2% 42.5% Removal 13.7±9.0 11.6±10.1 20.9±8.2 14.5±4.4 4.9±5.0 NO - 3 % 71.4% 62.3% 95.6% 75.1% 27.9% Removal 1.2±0.7 0.1±0.2 3.6±2.3 4.7±07. 0.2±0.2 NH + 4 % -241% -79% -200.9% -397.8% -246.1% Note: Underlined values indicates removal is significant at the 0.05 level (2-tailed). Significance of overall removal values were calculated based on plant secondary effluent & WSP 1 final effluent. Significance of seasonal removal values were calculated based on overall and seasonal averages. 6.3.3.! Analysis of intra-seasonal trends Seasonal Kendall tests revealed the directional nature of water quality parameters for each season (Table 6.4). Any upward and downward trends significant at the p<0.05 level is discerned, providing insight into the intraseasonal variations in treatment performance. No significant monotonic trends were observed for TP, alkalinity and cBOD. Declining trends in pH for both the fall and winter can be largely attributed to decreased in algal activity in colder temperatures as algae have severely limited growth rates in lower temperatures below 8°C

(Butterwick, Heaney, & Talling, 2005; Singh & Singh, 2015). E. coli concentrations decreased

! 121 over the summer and increased during the winter periods. A study conducted on the same WSP further corroborates the observed fluctuations in E. coli concentrations (Liu et al., 2016).

Table 6.2. Seasonal Kendall test of trends for final effluent parameters in order of spring, summer, fall and winter (respectively).

- + Temperature pH E. coli DO TSS TP NO NH4 Alkalinity cBOD 3 Season S S F W S S F W S S F W S S F W S S F W S S F W S S F W S S F W S S F W S S F W Trend ↑ - ↓ - - - ↓ ↓ - ↓ - ↑ - ↓ - ↑ ↓ - ↓ ------↓ - - ↓ ↓ ------DO had an increasing trend during the summer months while NO3 appeared to decline during the fall. DO levels decreased towards the end of the summer as photosynthetic activity decreased. However, during the winter period, the water column becomes more saturated as the colder winter temperatures allow for more atmospheric O2 to remain in the water column.

- NO3 concentrations decreased throughout the fall as ammonification and denitrification processes can be sustained in colder temperatures. As the average fall temperature range of the

WSP was 8.5°C, denitrifying bacteria are still capable of nitrate removal activity as they are

+ capable of growth at temperatures as low as 4°C (Yao et al., 2013). NH4 concentrations are declined towards the end of the summer and fall seasons, likely attributed to algal uptake and volatilization of unionized NH3. At elevated pH levels, algae metabolizes CO2, shifting the

+ NH4 -NH3 equilibrium towards increased NH3 concentration, resulting in increased NH3 availability for algal uptake and potential for volatilization (Rockne & Brezonik, 2006).

6.3.4.! Correlations between water quality parameters

Preliminary temporal variations of wastewater quality parameters were evaluated using a correlation matrix. Spearman ρ correlations were used to study the correlation structure between variables, accounting for non-normal distribution, and relationships between water quality parameters (Alberto et al., 2001). Strong positive correlations were observed between temperature and pH (ρ=0.477, p=0.01), while temperature strongly negatively correlated with

! 122 - DO (ρ=-0.520, p=0.01) and NO3 (ρ=-0.883, p=0.01). It has been well established that DO is dependent both on temperature and hydraulic mixing, and as expected, water columns were less

DO saturated during warmer, stagnant periods and higher during cooler, aerated waters. Two previous studies conducted at the WSP revealed similar strong correlations between temperature, pH and DO (Liang et al., 2018; Schueder et al., 2016).

6.3.5.! Factor Analysis/Principal Component Analysis

FA/PCA were performed on the overall and seasonal data sets to determine the structure of the relationships between water quality parameters (Table 6.5.). An eigenvalue provides a measure of the significance of the PC and PCs with eigenvalues exceeding 1.0 are retained, according to the Kaiser criterion (Kaiser, 1960). The variable loadings from the retained PCs can be used to determine the relative importance of a variable within the PC. In the PCA conducted on the entire normalized data set, four PCs were retained, explaining 77.3% of the total variance.

Three PCs were retained for spring and fall, with a cumulative variance of 68.4% and 70.8%, respectively. Four PCs were retained in the summer and fall, accounting for 71.5% and 68.2% of the variance, respectively.

In the PCA for the overall data set, PC1, accounting for 27.5% of the total variance, had high positive loadings for temperature and pH, suggesting that these parameters accounted for most of the variation within the WSP. In the spring, PC1, accounting for 31.7% of variance in data, showed temperature and pH to have strong negative loadings while E. coli had strong positive loadings, representing primary disinfection mechanisms. Other studies have outlined the role of temperature and pH in the removal of pathogens in naturalized systems (Huang et al.,

2017; MacDougall et al., 2017; Maynard et al., 1999). During the summer, PC1, which accounted for 29.8% of the variance, exhibited strong positive loadings for TSS, TP and cBOD,

! 123 representing the effects of high algal activity on water quality within the WSP. High TSS and cBOD levels have been shown to be a result of increased algal biomass (Ragush et al., 2015;

Wang et al., 2010). PC2 had opposite loadings for E. coli and DO, as DO is involved photo- oxidation processes contributing to pathogen removal (Liu, Au, Anderson, Lam, & Wu, 2007).

In the fall PCA, E. coli, TSS and DO were shown to have positive loadings in PC1, showing similar results as the summer analysis. A study conducted in 2016 revealed similar results as reductions in TSS levels positively correlated with E. coli removal (Huang et al., 2017). In the winter, PC1, explaining 30.2% of the variance in data, had strong opposite loadings for pH and

- NO3 . Denitrification rates have been noted to be highest within a pH range of 7.5 to 9.0,

- resulting in higher NO3 amounts remaining in wastewater (Glass & Silverstein, 1998). This would suggest that the high DO concentrations and oxidative conditions observed in the WSP were conducive to nitrification processes.

! 124 Table 6.3. Factor loading matrix and explained variance of water quality parameters for overall period and four seasons.

Overall Spring Summer Fall Winter PC 1 2 3 4 1 2 3 1 2 3 4 1 2 3 1 2 3 4 Temperature 0.92 -0.01 -0.22 -0.05 -0.87 0.36 -0.15 0.45 0.01 -0.12 -0.61 0.91 -0.13 0.14 -0.18 -0.01 -0.71 0.36 pH 0.61 -0.65 -0.11 -0.03 -0.91 -0.07 -0.08 -0.65 0.62 0.17 -0.19 -0.88 0.06 -0.33 -0.83 -0.17 0.09 -0.01 E. coli -0.01 0.59 -0.11 -0.56 0.63 0.16 -0.52 0.06 -0.76 0.13 -0.03 0.68 0.34 0.20 0.29 -0.59 0.47 -0.04 DO -0.56 -0.46 0.47 0.06 0.27 -0.82 0.18 0.12 0.83 -0.31 0.02 0.81 -0.26 -0.18 -0.42 0.27 0.05 -0.62 TSS 0.20 0.24 0.84 -0.09 -0.01 -0.67 0.11 0.85 0.23 0.21 0.06 0.58 0.69 -0.01 -0.66 -0.08 0.30 0.21 TP 0.39 0.67 0.16 0.30 0.36 0.68 -0.24 0.80 0.25 0.17 0.13 -0.64 0.22 0.27 0.60 -0.27 0.44 -0.11 - NO3 -0.89 0.15 0.05 0.11 0.78 -0.07 0.19 -0.35 0.37 -0.25 0.47 0.22 -0.55 0.02 0.87 -0.01 0.02 0.04 + NH4 -0.19 0.69 -0.21 -0.32 0.29 0.72 0.38 0.21 -0.39 -0.28 0.61 -0.36 0.30 0.69 0.75 -0.14 -0.26 0.23 Alkalinity 0.04 0.46 -0.21 0.78 0.73 0.12 -0.23 -0.10 0.27 0.76 0.33 0.95 -0.14 0.01 0.26 0.77 0.31 0.05 cBOD 0.44 0.18 0.72 0.01 0.24 -0.64 -0.05 0.84 0.10 0.30 0.03 0.49 0.71 -0.28 -0.40 -0.17 0.42 0.60 Eigenvalue 27.47 2.22 1.63 1.14 3.48 2.79 1.25 3.28 2.15 1.29 1.14 4.91 1.73 1.04 3.32 1.68 1.41 1.09 Variance % 27.47 22.2 16.26 11.37 31.67 25.39 11.38 29.82 19.59 11.73 10.37 44.65 15.71 10.48 30.18 15.24 12.84 9.93 CV% 27.47 49.65 65.91 77.28 31.67 57.05 68.44 29.82 49.40 61.14 71.50 44.65 60.36 70.79 30.18 45.43 58.26 68.19

! 125

6.3.6.! Discriminant Analysis DA was carried out to determine the most significant parameters associated with temporal variation. Seasons were used as temporal grouping variables, while the water quality parameters were considered independent variables (Shrestha et al., 2007). DA was performed on each raw data matrix using standard, forward stepwise and backward stepwise modes in constructing discriminant functions to evaluate seasonal variations in water quality. The results obtained from each mode of DA are displayed in Table 6.6. Three DFs were found for each mode, explaining 100% of the total variance within seasons. With the standard and forward stepwise mode, the first DF explained 75.3% of the total variance between seasons; the second

DF explained 15.0%, while the third DF explained 9.6% of the total variance between the seasons. With the backward stepwise approach, DF 1 accounted for 86.3% of variance while the

DF 2 and DF 3 explained 11.2% and 2.4%, respectively, of total variance.

Table 6.4. Eigenvalues for the discriminant factors (DFs).

Standard Mode Forward Stepwise Mode Backward Stepwise Mode Variance Variance CV Variance CV DF Eigenvalue CV (%) Eigenvalue Eigenvalue (%) (%) (%) (%) (%) 1 4.10 75.3 75.3 4.10 75.3 75.3 3.80 86.3 86.3 2 0.82 15.0 90.4 0.82 15.0 90.4 0.50 11.2 97.6 3 0.53 9.6 100.0 0.53 9.6 100.0 0.11 2.4 100

Classification function coefficients were obtained from standard, forward stepwise and backward stepwise DA modes (Table 6.7.). The classification matrix is used to assess the performance of each DA mode, revealing the percentage of correctly predicted cases versus actual group membership. The standard and forward stepwise mode assigned 78.6% of the cases correctly using all of the 11 parameters while the backward stepwise DA mode yielded 74.0% of cases correctly using only four discriminant parameters, temperature, pH, DO, and TP (Table

6.8). Thus, the backward stepwise mode was the most effective method of discriminating

! 126 variables as only four variables were used to account for the correct assignations, with only a

4.6% difference of correctly assigned cases compared to the standard and forward stepwise modes, which used all 11 parameters. The DA results revealed that temperature, pH, DO and TP were the most significant parameters discriminating between the four seasons, responsible for most of variations in temporal water quality over the monitoring period.

Similar to the Spearman’s ρ correlation coefficient test, temperature, pH and DO were strongly correlated to each other and were also shown through DA to be the primary factors affecting system dynamics. The average temperature and pH levels were observed to be highest in the summer, coinciding with the occurrence of excessive algal growth, and lowest during the winter season (Figure 6.2). Conversely, DO concentrations were lowest in the summer, with an average of 5.4 mg/L. A clear inverse relationship between temperature and DO was observed, attributed to the seasonality effect, as warmer water becomes more easily saturated with oxygen and holds lower concentrations of DO. Additionally, TP concentrations reached a maximum during the summer and declined in the fall and winter, coinciding with the observed algal blooms

(Figure 6.2). Phosphorus uptake increases during growth, while during periods of algal decay, phosphorus release and complexation/precipitation processes occur. Thus, during periods of high photosynthetic activity, rapid algal growth and decay could have contributed to higher and fluctuating TP levels throughout the summer (Perkins & Underwood, 2000).

! 127 Table 6.5. Classification functions for discriminant analysis of spatial variations of water quality.

Parameters Standard Mode Forward Stepwise Mode Backward Stepwise Mode Winter Spring Summer Fall Winter Spring Summer Fall Winter Spring Summer Fall

Temperature -1.50 -0.81 -1.37 -1.84 -1.50 -0.81 -1.37 -1.83 -1.49 -0.83 -1.36 -1.81 pH 40.28 39.51 41.07 40.26 40.28 39.51 41.07 40.26 27.23 26.80 28.29 26.85 E. coli 0.14 0.17 0.16 0.15 0.14 0.17 0.16 0.15 DO 0.37 -0.38 -0.26 -0.13 0.37 -0.38 -0.27 -0.13 -0.34 -1.00 -0.90 -0.77 TSS -0.04 -0.04 -0.02 -0.21 -0.042 -0.04 -0.02 -0.21 TP 29.93 47.53 43.68 37.34 29.93 47.53 43.68 37.34 47.29 63.542 58.88 52.55 - NO3 1.66 1.44 1.43 1.75 1.66 1.44 1.43 1.75 + NH4 36.50 28.76 31.53 33.39 36.50 28.76 31.53 33.39 Alkalinity 0.16 0.22 0.19 0.17 0.16 0.22 0.19 0.17 cBOD 1.29 0.54 0.49 1.30 1.29 0.54 0.50 1.30 Constant -202.47 -208.37 -208.21 -198.68 -202.47 -208.37 -208.21 -198.68 -120.66 -125.52 -128.39 -112.05

Table 6.6. Classification matrix for Discriminant Analysis (DA) of temporal variation in WSP 1.

Standard Mode Forward Stepwise Mode Backward Stepwise Mode Season Correct Correct Correct Spring Summer Fall Winter Spring Summer Fall Winter Spring Summer Fall Winter Spring 64 0 16 20 64 64 0 16 20 64 52.8 2.8 30.6 13.9 52.8 Summer 0 89.7 10.3 0 89.7 0 89.7 10.3 0 89.7 0 92.5 7.5 0 92.5 Fall 3.3 16.7 70 10 70 3.3 16.7 70 10 70 12.2 14.6 61 12.2 61 Winter 10.7 0 0 89.3 89.3 10.7 0 0 89.3 89.3 21.1 0 2.6 76.3 76.3 Total 78.6 78.6 74.0 Note: Values displayed as percentage (%) of seasonal samples.

! 128 6.4.! Conclusions In this study, multivariate statistical techniques were used to evaluate the seasonal fluctuations in treatment performance of a WSP experiencing excessive algal blooms. While algal blooms can be an operational issue, they were shown to have a significant positive impact on wastewater dynamics, particularly regarding disinfection performance and removal of nutrient loadings. Temperature, pH and DO were identified to be important factors associated with the

- - removal of E. coli. Seasonal Kendall tests revealed that NO3 E. coli, TSS, TP and NO3 concentrations exhibited increasing and decreasing trends within each season. Although the factor analysis/principal component analysis did not result in a significant data reduction, it helped to identify the factors/sources responsible for temporal variations in water quality, mainly related to temperature, pH and nutrients. Four parameters, temperature, pH, and DO and TP, correctly assigned 74% of the data set and were shown to be the primary factors responsible for seasonal variations in water quality, using DA.

Broader application of these multivariate statistical methods could be particularly useful in the assessment of treatment performance and understanding water quality for effective wastewater management. In particular, wastewater systems in temperate climates experience strong seasonal variations in treatment performance that should be carefully monitored in order to remain within regulatory limits. The identification of water quality parameters affecting the spatial and temporal dynamics in this WSP can aid in the operation of the system and, ultimately, the effective treatment of the wastewater. This research contributes to the continued improvement of low cost, effective wastewater treatment technologies, such as WSPs, which are especially important for resource-limited small, remote and rural communities.

! 129 ! ! ! !

6.5.! References

Alberto, W. D., Marı́a del Pilar, D., Marı́a Valeria, A., Fabiana, P. S., Cecilia, H. A., & Marı́a de

los Ángeles, B. (2001). Pattern Recognition Techniques for the Evaluation of Spatial and

Temporal Variations in Water Quality. A Case Study:: Suquı́a River Basin (Córdoba–

Argentina). Water Research, 35(12), 2881–2894. http://doi.org/10.1016/S0043-

1354(00)00592-3

APHA/AWWA/WEF. (2012). Standard Methods for the Examination of Water and Wastewater.

Standard Methods, 541. http://doi.org/ISBN 9780875532356

Arhonditsis, G. B., Brett, M. T., & Frodge, J. (2003). Environmental control and limnological

impacts of a large recurrent spring bloom in Lake Washington, USA. Environmental

Management. http://doi.org/10.1007/s00267-002-2891-4

Awuah, E., Anohene, F., Asante, K., Lubberding, H., & Gijzen, H. (2001). Environmental

conditions and pathogen removal in macrophyte- and algal-based domestic wastewater

treatment systems. Water Science and Technology, 44(6), 11 LP-18. Retrieved from

http://wst.iwaponline.com/content/44/6/11.abstract

Baborowski, M., Simeonov, V., & Einax, J. W. (2012). Assessment of Water Quality in the Elbe

River at Water Conditions Based on Cluster Analysis, Principle Components

Analysis, and Source Apportionment. Clean - Soil, Air, Water, 40(4), 373–380.

http://doi.org/10.1002/clen.201100085

Butterwick, C., Heaney, S. I., & Talling, J. F. (2005). Diversity in the influence of temperature on

the growth rates of freshwater algae, and its ecological relevance. Freshwater Biology,

! 130 ! ! ! ! 50(2), 291–300. http://doi.org/10.1111/j.1365-2427.2004.01317.x

CH2M Hill. (2011). Technical Assessment to Support Re-Rating the Amherstview WPCP Draft.

Davis-Colley, R. J., Donnison, A. M., & Speed, D. J. (2000). Towards a mechanistic

understanding of pond disinfection. Water Science and Technology, 42(10–11), 149 LP-158.

Retrieved from http://wst.iwaponline.com/content/42/10-11/149.abstract

Dong, X., & Reddy, G. B. (2010). Nutrient removal and bacterial communities in swine

wastewater lagoon and constructed wetlands. Journal of Environmental Science and Health,

Part A, 45(12), 1526–1535. http://doi.org/10.1080/10934529.2010.506109

Faleschini, M., & Esteves, J. L. (2017). Full-scale maturation ponds working below a latitude of

43°S in a semiarid area: seasonal performance and removal mechanisms. Water and

Environment Journal, 31(1), 47–56. http://doi.org/10.1111/wej.12219

Glass, C., & Silverstein, J. (1998). Denitrification kinetics of high nitrate concentration water: pH

effect on inhibition and nitrite accumulation. Water Research, 32(3), 831–839.

http://doi.org/10.1016/S0043-1354(97)00260-1

Hayward, J., Jamieson, R., Boutilier, L., Goulden, T., & Lam, B. (2014). Treatment performance

assessment and hydrological characterization of an arctic tundra wetland receiving primary

treated municipal wastewater. Ecological Engineering, 73, 786–797.

http://doi.org/10.1016/j.ecoleng.2014.09.107

Huang, Y., Truelstrup Hansen, L., Ragush, C. M., & Jamieson, R. C. (2017). Disinfection and

removal of human pathogenic bacteria in arctic waste stabilization ponds. Environmental

Science and Pollution Research, 1–13. http://doi.org/10.1007/s11356-017-8816-9

Iordache, Ş., & Dunea, D. (2013). Multivariate analysis applied to physicochemical parameters of

! 131 ! ! ! ! treated wastewater effluents discharged in Teleajen River, 58, 717–723.

Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educational and

Psychological Measurement, 20(1), 141–151.

King, J. R., & Jackson, D. A. (1999). Variable selection in large environmental data sets using

principal components analysis. Environmetrics, 10(1), 67–77.

http://doi.org/10.1002/(SICI)1099-095X(199901/02)10:1<67::AID-ENV336>3.0.CO;2-0

Liang, S., Hall, G., & Champagne, P. (2018). Monitoring seasonal variations in treatment

performance of a wastewater stabilization pond with algal blooms and pH fluctuations

Materials & Methods System overview. In 1st WDSA/CCWI Conference Proceedings.

Liu, L., Hall, G., & Champagne, P. (2016). Effects of environmental factors on the disinfection

performance of a wastewater stabilization pond operated in a temperate climate. Water

(Switzerland), 8(1). http://doi.org/10.3390/w8010005

Liu, L., Macdougall, A., Hall, G., & Champagne, P. (2017). Disinfection Performance in

Wastewater Stabilization Ponds in Cold Climate Conditions: A Case Study in Nunavut,

Canada. Environments, 4(4), 93. http://doi.org/10.3390/environments4040093

Liu, W., Au, D. W. T., Anderson, D. M., Lam, P. K. S., & Wu, R. S. S. (2007). Effects of

nutrients, salinity, pH and light:dark cycle on the production of reactive oxygen species in

the alga Chattonella marina. Journal of Experimental Marine Biology and Ecology, 346(1–

2), 76–86. http://doi.org/10.1016/j.jembe.2007.03.007

MacDougall, A., Champagne, P., & Hall, G. (2017). The effects of climate and humic substances

on disinfection performance in Arctic wastewater stabilization ponds. Queen’s University.

Retrieved from

! 132 ! ! ! ! https://search.proquest.com/docview/1886374899?accountid=6180%5Cnhttp://dw2zn6fm9z

.search.serialssolution.com?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-

8&rfr_id=info:sid/Dissertations+%26+Theses+%40+Queen%27s+University+at+Kingston

%2C+Ontario&rft_val_f

Maynard, H. E., Ouki, S. K., & Williams, S. C. (1999). Tertiary lagoons: A review of removal

mechanisms and performance. Water Research, 33(1), 1–13. http://doi.org/10.1016/S0043-

1354(98)00198-5

Ouyang, Y. (2005). Evaluation of river water quality monitoring stations by principal component

analysis. Water Research, 39(12), 2621–2635. http://doi.org/10.1016/j.watres.2005.04.024

Perkins, R. G., & Underwood, G. J. C. (2000). Gradients of chlorophyll a and water chemistry

along an eutrophic reservoir with determination of the limiting nutrient by in situ nutrient

addition. Water Research, 34(3), 713–724. http://doi.org/10.1016/S0043-1354(99)00228-6

Ragush, C. M., & Jamieson, R. C. (2016). Oxygen Dynamics and Carbon Removal in Municipal

Waste Stabilization Ponds in Arctic Regions by. Dalhousie University.

Ragush, C. M., Schmidt, J. J., Krkosek, W. H., Gagnon, G. A., Truelstrup-Hansen, L., &

Jamieson, R. C. (2015). Performance of municipal waste stabilization ponds in the Canadian

Arctic. Ecological Engineering, 83, 413–421. http://doi.org/10.1016/j.ecoleng.2015.07.008

Ramsey, P. H. (1989). Order Correlation. Journal of Educational Statistic, 14(3), 245–253.

Rockne, K. J., & Brezonik, P. L. (2006). Nutrient Removal in a Cold-Region Wastewater

Stabilization Pond: Importance of Ammonia Volatilization. Journal of Environmental

Engineering, 132(April), 451–459. http://doi.org/10.1061/(ASCE)0733-

9372(2006)132:4(451)

! 133 ! ! ! ! Schueder, M. R., Champagne, P., & Hall, G. (2016). Water quality spatial heterogeneity in

wastewater stabilizations ponds at elevated pH. Queen’s University.

Shammas, N., Wang, L. K., & Wu, Z. (2008). Waste Stabilization Ponds and Lagoons. In

Biological Treatment Processes - Volume 8. New Delhi, India: Humana Press.

Shrestha, S., & Kazama, F. (2007). Assessment of surface water quality using multivariate

statistical techniques: A case study of the Fuji river basin, Japan. Environmental Modelling

and Software, 22(4), 464–475. http://doi.org/10.1016/j.envsoft.2006.02.001

Simeonov, V., Stratis, J. A., Samara, C., Zachariadis, G., Voutsa, D., Anthemidis, A., …

Kouimtzis, T. (2003). Assessment of the surface water quality in Northern Greece. Water

Research, 37(17), 4119–4124. http://doi.org/10.1016/S0043-1354(03)00398-1

Singh, K. P., Malik, A., Mohan, D., & Sinha, S. (2004). Multivariate statistical techniques for the

evaluation of spatial and temporal variations in water quality of Gomti River (India) - A case

study. Water Research, 38(18), 3980–3992. http://doi.org/10.1016/j.watres.2004.06.011

Singh, S. P., & Singh, P. (2015). Effect of temperature and light on the growth of algae species:

A review. Renewable and Sustainable Energy Reviews, 50, 431–444.

http://doi.org/10.1016/j.rser.2015.05.024

Wallace, J., Champagne, P., & Hall, G. (2016). Multivariate statistical analysis of water

chemistry conditions in three wastewater stabilization ponds with algae blooms and pH

fluctuations. Water Research, 96, 155–165. http://doi.org/10.1016/j.watres.2016.03.046

Wallace, J., Champagne, P., & Hall, G. (2016). Time series relationships between chlorophyll-a,

dissolved oxygen, and pH in three facultative wastewater stabilization ponds. Environ. Sci.:

Water Res. Technol., 2(6), 1032–1040. http://doi.org/10.1039/C6EW00202A

! 134 ! ! ! ! Wang, L., Min, M., Li, Y., Chen, P., Chen, Y., Liu, Y., & Ruan, R. (2010). Cultivation of green

algae Chlorella sp. in different wastewaters from municipal wastewater treatment plant.

Applied Biochemistry and Biotechnology, 162(4), 1174–1186.

http://doi.org/10.1007/s12010-009-8866-7

Yao, S., Ni, J., Ma, T., & Li, C. (2013). Bioresource Technology Heterotrophic nitrification and

aerobic denitrification at low temperature by a newly isolated bacterium , Acinetobacter sp .

HA2. Bioresource Technology, 139, 80–86. http://doi.org/10.1016/j.biortech.2013.03.189

! !

! 135 ! ! ! ! Chapter 7. Summarizing Conclusions!

7.1. Thesis Summary

Adequate treatment of wastewater is fundamental to human health, water accessibility, improving aquatic environments and conservation of natural resources. As the human population continues to grow and natural resources continue to dwindle in the face of climate change, water pollution is becoming a more pervasive and detrimental dilemma. In particular, municipal wastewater is widely produced and often discharged following inadequate treatment in large volumes. As such, wastewater is an important target for treatment, monitoring and mitigation.

Passive wastewater treatment systems, such as wastewater stabilization ponds (WSPs), are sustainable alternatives to conventional wastewater treatment for small and rural communities.

WSPs have been shown to effectively remove nutrients, pollutants and pathogens present in municipal wastewater through naturally occurring biological, chemical and physical treatment mechanisms. However, passive wastewater treatment systems are open systems involving complex interactions that are subject to climatic variability, bringing challenges in their operation and treatment performance.

A case study approach was undertaken, using an existing WSP to investigate water quality, treatment performance and factors contributing to system dynamics. Amherstview Water

Pollution Control Plant (WPCP) located in eastern Ontario, Canada, has a WSP system that experiences seasonal periods of excessive algae growth and high pH levels. These issues have raised concerns and challenges with regard to meeting regulatory discharge limits as stipulated in the operating permit of the treatment facility. ! 136 ! ! ! ! In Chapter 3, an exploratory approach to determining the spatial and temporal variations in algal blooms was employed. Water quality parameters were collected on a weekly basis and meteorological factors were collected hourly from May to August 2017, during the predominant algal development period. Multivariate statistical methods, namely PCA and DA were used to determine the main factors affecting algal blooms. Wind direction, speed and precipitation were

- shown to affect spatial distribution of surface algal cover while temperature, pH, DO and NO3 , used for 79.7% of correct assignations, were key factors responsible for temporal algal blooms dynamics.

In Chapter 4, an 11-month study was conducted to examine the disinfection performance of the WSP. Four bacterial indicators, Escherichia coli (E. coli), Enterococci, Clostridium perfringens (C. perfringens) and total coliforms (TC), were selected to assess the disinfection performance of the pond. Strong positive correlations between chlorophyll and temperature, pH, and DO, and strong negative correlations (|ρ|>0.257, p<0.05) between temperature and both traditional and potential indicator organisms, and a strong negative correlation (|ρ|>0.257, p<0.05) between pH and bacterial indicators more revealed. Moreover, the strong positive correlations between chlorophyll and water temperature (ρ=0.277, p<0.05), pH (ρ=0.383, p<0.05) and DO (ρ=0.260, p<0.05) were consistent with the observed algal activity during the summer

- months. PCA revealed that temperature, chlorophyll, pH and NO3 -N had the most impact on variations in water quality during monitoring period.

- In Chapter 5, pH, DO, Escherichia coli, NO3 -N and TP were monitored over a five-year period. The removal efficiencies of various water quality parameters and indicator organisms for each season were used to determine seasonal treatment and the disinfection performance of the

! 137 ! ! ! ! - system. NO3 -N and E. coli removal were shown to be lowest during the winter periods at 95.6% and 27.9%, while TP remained consistent throughout the monitoring period. E. coli removal was shown to be significantly negatively correlated with pH (ρ =-0.268, p<0.05) and DO (ρ=-0.390, p<0.05), corroborating with results in Chapter 3. Seasonal Kendall tests revealed that DO levels

- and NO3 -N concentrations both significantly decreased during the fall period.

In Chapter 5, Spearman’s correlation ρ, Seasonal Kendall test, PCA and DA were conducted using a comprehensive water quality data set collected from the same WSP system over a five-year period. The removal efficiency and Seasonal Kendall trends of water quality parameters for each season were used to determine seasonal fluctuations in the system.

Temperature exhibited a strong positive correlation to pH (ρ=0.477, p<0.05) and negative correlations with DO (ρ=-0.390, p<0.05). Seasonal Kendall tests revealed increasing trends in E. coli and DO concentrations throughout the winter while pH levels decreased. DA delineated four parameters, namely temperature, pH, DO and TP, were used to represent 74% correct assignations, indicating that these parameters accounted for majority of the variations observed in seasonal water quality.

WSPs are an effective technology for nutrient removal and disinfection, with algae playing a large role in treatment. However, these systems can be sensitive to seasonal variations in treatment depending on location and climate, making wastewater treatment using natural systems more challenging. This research contributes to the continued improvement of low cost and effective wastewater treatment options, such as WSPs, which are especially important for resource-limited rural and Indigenous communities. These studies also illustrate the effectiveness of multivariate statistical techniques for analysis and interpretation of large data sets. Broader

! 138 ! ! ! ! application of these statistical methods could be particularly useful in the assessment of treatment performance and understanding temporal variations in water quality for effective wastewater management.

7.2.! Future Work

There are a number of recommendations and areas of further investigation that can be synthesized from this research. The research conducted in Chapters 5 and 6 included the full- scale monitoring of a WSP. From the results in Chapter 5, temperature, pH, DO and TP were determined to be the primary factors affecting seasonal variations in water quality. A bench-scale experiment, with a factorial design, could focus on varying temperature, pH, DO and TP, while maintaining other water quality parameters. Considering each value individually could provide insight into the quantitative effect of each parameter on treatment performance.

A multi-year simulation could be conducted, incorporating the two existing WSPs and the newly constructed wetland to provide a better understanding of the water quality dynamics and treatment performances of the entire, passive wastewater treatment system. This research could provide more detailed conclusions regarding algal blooms and pH fluctuations prior to final discharge into the natural environment.

Additionally, a full characterization of biological communities within the WSP could enhance knowledge regarding treatment performance, extending beyond nutrients and bacteria, with respect to emerging contaminants. A complete overview of the microbiome, using metagenomic analysis, in the WSP could provide valuable insight into the specific interactions involving algae and their effects on water quality dynamics. As emerging contaminants, such as endocrine disrupting compounds, are a potential human health and ecological issue that has not ! 139 ! ! ! ! been fully studied, a community profile of the structure and function of different biological communities could reveal important relationships between algae, bacteria and emerging contaminants. This would contribute to the body of knowledge related to the removal of emerging contaminants using passive wastewater treatment systems, potentially providing a cost- effective method of reducing risks to both human health and aquatic ecosystems.

7.2. Contributions to Science and Engineering

Passive wastewater treatment approaches, such as WSPs, are cost-effective options for the municipal wastewater resource management in small, rural and remote communities. However, treatment performance and system dynamics are susceptible to climatic and ecological variability. As such, there are continued advancements towards the improvement of passive wastewater treatment and further developing insights into improving their performance effectiveness and efficiency. The concerns associated with algal blooms support the need for new investigations that examine the treatment mechanisms and dynamics of water quality parameters in WSPs.

The research conducted as part of this thesis reveals a number of significant contributions to the field of science and engineering, providing valuable insights and understanding to the growing body of literature. First, an enhanced understanding of pathogen removal mechanisms and factors associated with the presence of algae in WSPs was obtained. The presence of algae likely plays an important role in providing sufficient pH levels and DO concentrations to facilitate the removal and inactivation of bacterial indicator organisms. Additionally, pollutant removal was often lower during the winter compared to the summer, coinciding with algal activity and associated high pH levels. Additionally, pH, DO, TP and temperature were shown to ! 140 ! ! ! ! be the primary factors affecting water quality. Therefore, algae can be considered beneficial to disinfection and pollutant removal processes in WSP systems.

Additionally, strong seasonal variations in treatment performance and water quality were observed in the system studied, located in a temperate region with four distinct seasons. These conclusions highlight a need for more updated wastewater effluent regulations in temperate climates. In Canada, the current federal WSER contain single effluent discharge limits for E. coli,

+ TP, NH4 , BOD and TSS for all wastewater treatment systems. These regulatory limits do not take into account the effect of seasonal variations and climate on passive wastewater treatment systems. Therefore, there may be a need for different effluent limits, based on seasons and location, to ensure all municipalities are able to meet requirements while remaining aware of finance and resource limits in certain regions.

The application of multivariate statistics and time series techniques contributes to the growing body of literature using statistical analyses to better understand and improve wastewater treatment. Multivariate statistics, reveal patterns difficult to discern in large data sets and identify key parameters affecting system dynamics. The statistical methodologies presented offered an insightful approach to the monitoring of wastewater quality using full datasets of water quality parameters, thus providing for a whole systems approach in evaluating the naturalized systems.

Time series analyses provided insight into time-dependent statistical relationships and identified dynamic time lags between parameter trends. While these sets of techniques are frequently used in ecological studies and in evaluations of engineered systems, this research included novel applications of these statistical methods in passive wastewater treatment systems.

! 141 ! ! ! ! The knowledge gained through this study can be applied to improve the operation and design of WSPs in temperate climates, to identify correlations between wastewater quality factors and, more importantly, to provide practical guidance regarding nutrient and pathogen removal from wastewater in a cost-effective and sustainable manner. This research will be used directly to inform the monitoring program for the WSP system at the site. Ultimately, this research contributes to the continued improvement of WSP operations in temperate regions, ensuring that they remain resilient and adaptable in the face of a changing climate.

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !

! 142 ! ! ! ! !

APPENDIX A

A.1. Pre-Screening Data Calculations Prior to multivariate analysis, data for each parameter was analyzed for normality in order to determine whether data transformation should be applied. Data that fits a normal distribution can be analyzed using parametric methods, including multivariate methodologies, thus providing more statistical power for interpretation (Reisfield & Mayeno, 2013). Normality was assessed by evaluating the Shapiro-Wilk test for a normal distribution fit as this method can be applicable to small sample sizes (n<50) as well as large sample sizes (n>2000).

! 143 ! ! ! ! Table A.1. Shapiro-Wilk test of normality for water quality data found in Chapter 3.

Statistic df Sig. Temperature 0.795 8 0.026 pH 0.796 8 0.026 DO 0.881 8 0.019 Chlorophyll 0.475 8 0 Phycocyanin 0.473 8 0 3- PO4 0.827 8 0.055 TP 0.89 8 0.033 + NH4 0.485 8 0 - NO3 0.575 8 0 TN 0.909 8 0.049 COD 0.74 8 0.006 E. coli 0.912 8 0.369 TC 0.786 8 0.02 Enter 0.932 8 0.532 C. perfringens 0.99 8 0.995

! 144 ! ! ! ! Table A.2. Shapiro-Wilk test of normality for water quality data found in Chapter 4.

Statistic df Sig. Temperature 0.897 32 0.005 pH 0.955 32 0.002 DO 0.954 32 0.019 3- PO4 0.973 32 0.059 TP 0.905 32 0.008 + NH4 0.616 32 0 - NO3 0.978 32 0.037 COD 0.912 32 0.013

! 145 ! ! ! ! Table A.3. Shapiro-Wilk test of normality for water quality data found in Chapter 6.

Statistic df Sig. Temperature 0.795 8 0.026 pH 0.796 8 0.026 E. coli 0.881 8 0.019 DO 0.475 8 0 TSS 0.473 8 0 TP 0.89 8 0.033 - NO3 0.575 8 0 - NO2 0.909 8 0.049 COD 0.74 8 0.006 Alkalinity 0.912 8 0.369 + NH4 0.786 8 0.02 cBOD 0.932 8 0.532

! 146 ! ! ! !

A.2. Criteria for Spearman’s ρ correlation coefficient Pearson’s correlation coefficient test evaluates the linear relationship between two continuous variables while Spearman correlation coefficient is based on the ranked values rather than the raw data. The data set used is classified as non-parametric as a specific probability distribution could not be applied to it. As a result, Spearman rank correlation coefficient was determined to be more appropriate for the data sets. To determine the validity and significance of correlations, Spearman ρ values were sorted by p-value, with a small p-value indicating that the correlation is significant. The resulting Spearman ρ values are then compared against with the critical Spearman ρ value for a two- tailed test at the 99% confidence level. Those values with a small p-value and above the critical ρ value were considered valid and significant.

! 147 ! ! ! ! A.3. Data Screening for Principal Component Analysis The pairwise estimation method was used in PCA analysis of the imputed dataset, based on the criteria of the data having more than 10 columns (variables). The robust method was deemed inappropriate for this data, as it works on the assumption of the presence of outliers in the dataset, which is not feasible for field sampling data. Additionally, PCA analysis itself assumes the absence of outliers in the data, which is appropriate for measured field data (Suhr, 2005). The weight, frequency, and by qualifiers in the multivariate and PCA platforms do not apply and were not used, as all sample entries (dates) are equally weighted and frequent. PCA results were interpreted based on the correlations between parameters, a unitless measurement, required for this dataset which consists of various units for parameters.

! 148 ! ! ! ! A.4. Variables Entered/Removed in Discriminant Analysis F to Remove values are useful for describing what happens if a variable is removed from the current model (given that the other variables remain). A variable is entered into the model if its F value is greater than the Entry value and is removed if the F value is less than the Removal value. Entry must be greater than Removal, and both values must be positive. To enter more variables into the model, lower the Entry value (IBM, 2011).

! 149 ! ! ! ! Table A.4. Variables entered/removed at each step in the backward stepwise mode for water quality data set from Chapter 3.

Step Entered Wilks' Lambda df1 df2 df3 Exact F df1 df2 Sig. 1 pH 0.812 1 1 58 13.389 1 58 0.001 2 NO3 0.633 2 1 58 16.492 2 57 0 3 DO 0.573 3 1 58 13.914 3 56 0 4 WT 0.529 4 1 58 12.244 4 55 0 At each step, the variable that minimizes the overall Wilks' Lambda is entered. a Maximum number of steps is 20. b Minimum partial F to enter is 3.84. c Maximum partial F to remove is 2.71. d F level, tolerance, or VIN insufficient for further computation.

! 150 ! ! ! ! Table A.5. Variables entered/removed at each step in the backward stepwise mode for water quality data set from Chapter 6.

Step Entered Wilks' Lambda df1 df2 df3 Exact F df1 df2 Sig. 1 WT 0.812 109.146 3 108 0 109.146 3 108 2 TP 0.633 48.104 6 214 0 48.104 6 214 3 DO 0.573 35.724 9 258.127 0 35.724 9 258.127 4 pH 0.529 27.569 12 278.095 0 27.569 12 278.095 At each step, the variable that minimizes the overall Wilks' Lambda is entered. a Maximum number of steps is 22. b Minimum partial F to enter is 3.84. c Maximum partial F to remove is 2.71. d F level, tolerance, or VIN insufficient for further computation.

! 151 ! ! ! ! A.5. Numerical Values for Seasonal Kendall Tests Table A.6. Seasonal Kendall tests performed on effluent water quality parameters in Chapter 6.

Spring Summer Fall Winter S : 43 -1 -29 7 Temperature p: 0 0.46 0.0004 0.33806 Z : 0.73532 S : -2080 -4869 9397 272 pH Z : 1.379 4.4607 8.0033 0.49352 P: 0.16788 8.17E-06 1.21E-15 0.62165 S : 369 -3720 751 1712 DO Z : 0.66304 6.4941 1.4592 3.9303 p: 0.50731 8.35E-11 0.14451 8.48E-05 S : -1482 245 -3344 529 TSS Z : 2.1856 0.37764 5.4337 0.8332 p: 0.028845 0.7057 5.52E-08 0.40473 S : -657 -466 -980 -1076 TP Z : 3.6351 2.518 5.4226 6.5362 p: 0.00027789 0.011802 5.87E-08 6.31E-11 S : 163 -942 -909 -679 - NO3 Z : 0.82159 5.3351 5.0257 4.1193 p: 0.41131 9.55E-08 5.02E-07 3.80E-05 S : -2 -292 -127 -170 Alkalinity Z : 0.0046702 1.3057 0.76566 1.2989 p: 0.99627 0.19167 0.44388 0.19399 S : -485 160 -729 -618 E. coli Z : 2.7546 0.88053 4.0402 3.4955 p: 0.0058762 0.37857 5.34E-05 0.00047324 S : 16 23 -35 -9 cBOD Z : 0.47261 0.61986 0.69063 0.15431 p : 0.63649 0.53535 0.4898 0.87736 S : -62 -100 -224 -90 + NH4 Z : 1.4262 2.3147 3.7944 1.5133 p: 0.15381 0.020631 0.000148 0.13019

! 152 ! ! ! ! A.6. Meteorological Data Temperature, precipitation and wind data was measured at the Kingston Airport, which is approximately 5 km east of the site. It was provided by Environment Canada online and was quality checked, as the Kingston Airport station is subject to review by the National Climate

Archives division of Environment Canada.

!

!

!

!

!

!

!

!

!

!

!

!

!

!

! 153 ! ! ! ! A.7. Works Cited IBM. (2011). IBM SPSS Statistics 20 Core System User ’ s Guide. Spss, 418. Retrieved from

ftp://public.dhe.ibm.com/software/analytics/spss/documentation/statistics/20.0/en/client/Ma

nuals/IBM_SPSS_Statistics_Core_System_Users_Guide.pdf

Reisfield, B., & Mayeno, A. N. (2013). Computational Toxicology: Volume II. Computational

Toxicology (Vol. 930). New York, USA: Springer. http://doi.org/10.1007/978-1-62703-059-

5_21

Suhr, D. (2005). Principal component analysis vs. exploratory factor analysis. SUGI 30

Proceedings, 203–230. http://doi.org/10.1177/096228029200100105

! ! !

! 154 ! ! ! ! APPENDIX B

B.1. Water Quality in the Americas – Canada

Pascale Champagne1*, Pratik Kumar2 Geoffrey Hall1, Agnieszka Cuprys2,Shuang Liang1, Mehrdad Taheran2, David Blair1, Mitra Naghdi2 Megan McKellar1 and Satinder Kaur Brar2*

1Department of Civil Engineering, Queen’s University, 58 University Ave, Kingston, Ontario, Canada K7L 3N9 2INRS-ETE, Université du Québec, 490, Rue de la Couronne, Québec, Canada G1K 9A9

* Corresponding Authors

*1Phone: (613) 533-3053; Fax: (613) 533-2128; Email: [email protected]

*2Phone: 1 418 654 3116; Fax: 1 418 654 2600; E-mail: [email protected]

! 155 ! ! ! ! 1: Introduction Canada is known to have significant freshwater reserves with 9% of the world's renewable freshwater. However, 85% of the Canadian population in more urban centers within 100 km of the southern border with the United States and the majority of its freshwater flows northward, hence freshwater is not always readily available in the more highly populated areas leading to higher freshwater use-to-availability ratios in drainage basins located in the southern part of the country (Environment and Climate Change Canada, 2017a). In general, the supply of safe potable water is universal in most Canadian municipalities (Environment Canada, 2011a), with the exception of drinking water on First Nations land, which has been deemed to be substandard for extended periods of time (Levasseur and Marcoux, 2015). From a wastewater treatment perspective, most Canadian municipalities (~75%) are serviced by centralized sewer and treatment systems, with the remaining 25% of the population, generally located in small, remote or rural communities served by individual or decentralized septic disposal systems (Environment and Climate Change Canada, 2017d)

Because of the perceived abundance of freshwater, and the generally good quality of various water sources (lakes, rivers, and groundwater), water usage in Canada tends to be relatively high compared to Europe. In addition, water tariffs have traditionally been low to the point that less than 50% of infrastructure maintenance and operating costs can be covered through tariffs and the rest is financed through taxes (Environment and Climate Change Canada, 2016e). This scenario can lead to cases of increasingly deteriorating infrastructure, which can have detrimental impacts on human health and receiving environments unless federal, provincial and regional governing bodies and agencies take action and remain committed to ensuring source and receiving water protection and maintain a mandate to provide access to and deliver safe potable water (Environment and Climate Change Canada, 2016c).

! 156 ! ! ! ! 2: Authorities and governance of water 2.1 Legal framework In Canada, the governance of water resources is divided across federal and provincial governments while the Canadian Constitution largely controls governance jurisdiction. The Constitution prescribes responsibility of internal water resources management to each province, covered under section 109 of the Constitution Act of 1867 (Government of Canada, 1982). Ontario, Quebec, Nova Scotia and New Brunswick were originally covered under this Act, and the remaining provinces of British Columbia, Alberta, Saskatchewan, Manitoba, Prince Edward Island and the Yukon Territory were included at later dates. Governance of water resources in the Northwest Territories, Nunavut and First Nations land within provinces fall under the Federal Government. Transboundary water resource jurisdiction also falls to the Federal Government (Dunn et al. 2014).

Beyond decentralization, fragmentation also exists within individual provinces, across departments and agencies tasked with monitoring and protecting water resources (Hill et al., 2008). Due to the fragmented nature of water governance in Canada, there are no enforceable national water quality regulations (Bakker and Cook, 2011). However, there are national water quality guidelines in place that are intended to act as benchmarks for all jurisdictions across the country. The Canadian Council of Ministers of the Environment (CCME), an intergovernmental organization with representation from the federal, provincial and territorial environmental ministries, created the Canadian Water Quality Guidelines for the Protection of Aquatic which set out maximum concentrations of different water quality parameters in natural water bodies (CCME, 1999). Nevertheless, these guidelines are non-binding, national benchmarks for drinking water (Government of Canada, 2017a; Health Canada, 2010a).

Within the federal and provincial mandates, regulations are in place that may directly or indirectly affect water quality. For example, federal jurisdiction is exercised over navigation, which through the Canada Shipping Act, controls the discharge of ballast water and pollution from ships (Government of Canada, 2015a). Federal jurisdiction also covers fisheries through the

! 157 ! ! ! ! Fisheries Act, centered on pollution prevention for the protection of fish stocks (Fisheries and Oceans Canada, 2010; Government of Canada, 2016b). The Canadian Environmental Protection Act also provides protections for pollution entering a water body (Government of Canada, 2014). In some instances where transboundary waters are concerned, joint provincial and federal cooperation is in place in order to ensure that water quality is maintained in those bodies of water. The Canada-Ontario Agreement on the Great Lakes is an example of this cooperation and ensures that Great Lakes water quality and the aquatic environment remains a priority and is used to promote water quality (Environment and Climate Change Canada, 2017a).

At the provincial level, a variety of water quality regulations are in place to specifically target drinking water and wastewater treatment that could affect water quality in the environment. For instance, in Ontario, the Clean Water Act lays out specific regulations, which are designed to protect the drinking water source for communities across the province (Government of Ontario, 2017a). As in Ontario, and due to the responsibility for water resource management at the provincial level, a diversity of regulations may be found across the country, which contribute to water quality in natural waters (Hill et al. 2013). These may include nutrient management regulations, natural resource exploitation legislation, and development guidance protocols.

2.2 Participation of stakeholders Water governance in Canada is managed on a number of levels, incorporating both Non- Government Organizations (NGOs) and university institutions. In some cases, NGOs participate in the protection and monitoring of water quality. For instance, in Ontario, Conservation Authorities are tasked with oversight for the protection of water quality through the regulation of development in many parts of the province (Government of Ontario, 2017b). Source Water Protection Committees, managed by the Conservation Authorities, are mandated through the Ontario Clean Water Act (Government of Ontario, 2017a). In this instance, multi-stakeholder committees identify and protect drinking water sources by ensuring that sources of water pollution are identified, and mitigation plans are in place.

! 158 ! ! ! ! In Canada, academic collaborations on water quality occur across many levels, including between universities, government agencies, industry, and non-profit organizations. Funding mechanisms are also in place to facilitate multi-stakeholder water quality research at the federal and provincial levels.

2.3!Monitoring water quality Monitoring of water quality, as with water resources management in general, falls to the federal and provincial governments (Dunn et al., 2014). Due to this decentralization, databases on water quality monitoring are also fragmented. In many cases, water quality monitoring networks are in place that can provide time-series data on a number of water quality parameters for various municipal, regional and provincial jurisdictions, NGOs and conservation authorities. At the federal level, there are sites across water bodies within federal jurisdiction, which are monitored for chemical parameters through the Freshwater Quality Monitoring and Surveillance (FWQMS) program (Environment and Climate Change Canada, 2017b). In addition, the Canadian Aquatic Biomonitoring Network (CABIN) is a federal program established to monitor the health of aquatic ecosystems through biological indicators (Environment Canada, 2012). Water quality monitoring is also conducted at the provincial level in a variety of forms. For instance, the Ontario Ministry of the Environment and Climate Change (MOECC) operates the Provincial Water Quality Monitoring Network (PWQMN) and the Provincial Groundwater Monitoring Network (PGMN), which monitor nutrient and chemical parameters for surface water and ground water, respectively. Other provinces across the country have similar water quality monitoring programs.

3: Main issues that impact water quality Although Canada leads the world in terms of clean water sources and thousands of freshwater lakes, it still faces challenges pertaining to the water pollution issues. Significant industrial development such as mining, agricultural practices, and municipal non-regulatory practices over the years are the primary reasons for this cause. Based on the National Pollutants Release Inventory (NPRI) reports, water sources, such as Saint John River, Great Lakes, and St. Lawrence River are under “very high” threats of pollution. Major pollution sources ! 159 ! ! ! ! contaminating these water sources include the pulp and paper industry, oil industry, forestry, agriculture, technical industrial wastes, wood-related industry and runoff from local airports (World Water Federation of Canada, 2016). Lake Erie, which is part of Great Lakes, was in 1960s termed as the dead lake because of severe phosphorus levels caused by nutrient-rich municipal wastewater discharges. The Great Lakes basin is not only important for Canada but the United States of America as well. Major significant stressors include tributary dams which bring nutrients and impede flow of water, zebra and quagga mussels which have colonized all five of Great Lakes, and agricultural and municipal runoffs which carry high nutrient loads (Walker, 2013a). These must be regulated or controlled in order to sustain water quality. Besides the advisories and continuous regulatory measures, drinking water sources face many challenges. Different activities that contribute to water pollution are mainly due to anthropogenic practices.

3.1!Eutrophication Eutrophication is a phenomenon in which the overgrowth of aquatic plants and algae, called bloom, occurs due to excessive nutrient conditions, leading to the depletion of oxygen level and hence causing the death of aquatic species. It is one of the major causes of depleting water quality in Canada. The principal sources of nutrients affecting water quality in Canada include municipal wastewater runoffs, agricultural waste flow and field runoff, and other untreated waste effluents (Dorcy, 2006; Conference Board of Canada, 2013). Lake Winnipeg, the world's 10th largest freshwater lake basin, covers one million square kilometers with an area extending over four provinces in Canada viz. Alberta, Saskatchewan, Manitoba, and part of Ontario along with four American states, is affected by eutrophication every summer. Since 1900, the abundance of such blooms has increased by 300-500%. In summer-autumn climatic conditions, these blooms can release cyanotoxins due to overgrowth of toxigenic strains of cyanobacteria which over a period of time might lead to acute to chronic diseases in humans and aquatic species. Since 1990, the concentration of cyanobacteria has increased by over 1000% in Lake Winnipeg making it one of the most ecologically compromised lakes in the world (Xu et al., 2015). Similarly, several studies have discussed the sustained problem related to eutrophication in the St. Lawrence River. They

! 160 ! ! ! ! observed excess filamentous algae growth, recommending further work be undertaken to assess its severity and the reasons for these occurrences (Munawar and Heath, 2008).

Lake Erie, one of the most heavily used drinking water sources in Canada serving more than half a million residents experiences regular cyanobacteria blooms. The release of cyanotoxins from these blooms may be of concern for compliance with the WHO recommendation of < 1 µg/L for the drinking water systems. Drinking water treatment facilities, depending on the processes used, may not be adequate for the removal of cyanotoxins and hence may pose a threat to the end- users.

Various recommendation and protection acts need to be taken by the provincial governments. Manibota government in 2014 has brought the toughest wetland protection in Canada. A five- year plan, worth 300 million dollars, is designed to save Lake Winnipeg by reducing the amount of cyanobacteria growth that poisons the lake water. These surface water strategies are aimed at dealing particularly with the drainage issues and wetland management. However, these nutrient management strategies will require challenging goal-setting approaches marked by proper monitoring and assessments.

3.2!Natural contaminants Natural contaminants include minerals, dead animals, animal feces, turbidity, and contaminants evolving from natural disasters, such as volcanic eruption and forest fire. Canada has five active volcanic areas all of which are in British Columbia and Yukon Territory. Potential eruptions could unearth underground minerals and chemicals, such as benzo[a]pyrene, which would flow into the nearby water sources and deplete the water quality. Dead creatures such as birds and their feces may also affect groundwater quality.

Also, human burial affects groundwater quality as bio-oxidation of proteins and fats can be problematic after 10 years (Bouwer, 1978). However, a survey conducted in Ontario regarding burial practices noted that cemeteries might not be a significant contributory source of

! 161 ! ! ! ! contaminants, such as formaldehyde to groundwater (Chan et al., 1992). Nevertheless, these natural contaminant sources cannot be ignored as a possible source of water quality depletion in different regions and provinces of Canada.

3.3!Agrochemicals A large amount of pesticide and fertilizer usage for agricultural purposes has also created some negative impacts, including the degradation of water quality and formation of algal blooms in the aquatic environment. Pesticide contaminants may enter waterbodies by several routes including overland flow from agricultural fields, leaching to groundwater resources and application near ditches and . The diversity of this contamination in Canada can be illustrated by looking at the province of British Columbia, where about $22 million/year is spent on 120 distinct types of pesticides (Chen et al., 2016). Since 2003, there have been several different pesticides found in the St. Lawrence River near Quebec City (Quebec, Canada) and at the mouths of some tributaries of Lake Saint-Pierre (He et al., 2015).

Pesticides are also used at the household level, primarily by homeowners that have lawns or gardens; 32% of households across Canada, with Saskatchewan being reported to be the highest (52%), and New Brunswick to be the lowest (19%) among all the provinces (Statistics Canada, 2013a; 2013b). As a result of the increased impact of pesticides on water quality in Canada, as of April 2015, the use of some of the chemical pesticides have been banned in seven provinces including New Brunswick and Quebec (Government of Manitoba, 2014).

Canada, being one of the largest producers and exporters of agricultural products around the globe, is also a large user of fertilizers. The adverse impact of pesticides on water quality is mainly associated with pesticide formulation, contaminants in the active ingredients, the use of additives, and by-products formed during spraying. Canada, where one-tenth of jobs are related to the forest sector, involves a high use of insecticides applied by aircraft over large land masses (Lopez-Gresa et al., 2002). Atrazine, a pesticide widely found in Canadian drinking water supplies, has been reported to lead to sexual deformities and other reproductive problems in frogs

! 162 ! ! ! ! at the ppb (parts per billion) level (Boyd, 2006). Currently, Canadian drinking water regulations are less stringent than European Union regulations with respect to allowable concentrations of pesticide residues in water. In some cases, this discrepancy is 50-1000 times higher in Canada than the European Union, which impacts the overall drinking water quality source.

3.4!Heavy metals The presence of heavy metals in surface and groundwater sources has been a fundamental problem in Canada for decades. For example, since 1966, the south end of Buttle Lake, a region of historic mining activity, has been linked to the presence of copper and zinc in many oligotrophic lakes in Vancouver Island, Canada (Roch et al., 1985). The use of metals in the industrial, domestic, agricultural and medical application has led to a wide range of impacts on water bodies across the country. Tests performed for 49 different makeup items showed evidence of the presence of heavy metals in their composition (such as arsenic, cadmium, lead, mercury, nickel), which eventually reached the downstream water sources (Zhao et al., 2011). Many Canadian provinces (e.g., British Columbia, Alberta, Manitoba, Newfoundland, New Brunswick) have noted the presence of arsenic in groundwater, exceeding the regulatory limit of 10 µg/L (Ae-Ngibise et al., 2015). Even some Canadian bottled water has been found to contain metal ions, such as copper (88-147 µg/L) and nickel (16-35 µg/L), likely due to their presence in the feed water before the distillation process (Dabeka et al., 2002).

3.5!Deforestation Canada is among the world leaders in sustainable forest management where only 0.02% of its forest area is under deforestation (Psomas and Kessissoglou, 2013). However, some 8% of world virgin forests were degraded between 2000 and 2013, almost three times the size of Germany in terms of area (Huffington Post Canada, 2014). Both water quality and quantity are affected by deforestation. Precipitation runoff has been noted to affect surface water quality due to effects of disruption of the forest canopy, as their removal may increase the intensity of storm runoff. The protection of water quality is associated with the forest management practices to ensure water quality. In this respect, the management of forests in northwestern Quebec has been considered

! 163 ! ! ! ! essential from a long-term perspective as runoff can carry many unwanted chemicals and other foreign contaminants into surface drinking water sources (Sánchez-Monedero et al., 2001). British Columbia comprises almost 58% of the forest cover in Canada, and a high rate of deforestation activities was a concern. However, with recent sustainable efforts and programs, the activity has been decreased.

Deforestation is controlled under Canada`s forest laws, which manage and control environmental, social and economic needs and ensure sustainable forest management practices in Canada. Provinces and territories have jurisdiction over forests and they are tasked with developing and enforcing laws, regulations and policies regarding the use and protection of the Canadian forests. Proper and sustainable practices of forest harvesting have shown to significantly benefit the water quality and protect streams and lakes in forest lands.

3.6!Salinization Salinization is one of the critical factors influencing the depletion of the water quality in Canada. During the winter, the usage of salts on roads generates a high salinity load in receiving water environments. Salinization has been reported to an extent where salinity levels equivalent to the oceans have been observed in some streams during the winter period, and highway de-icing salts have been found to be the major source of salinization. A study conducted by the University of Toronto observed that a scenic lagoon on the shore of Lake Ontario was being polluted due to salinization, where approximately 7600 tons of salt was applied on community roads each year. Half of which was reported to be due to groundwater seeps, which in turn flowed to the tributaries, and the half directly to the lake as (Sharma et al., 2009). According to reports by Environment Canada, the excessive usage of salts, eventually entering the water system, has the potential to be quite harmful to humans and other aquatic habitats, where salty water can affect human kidneys and kill freshwater fish.

! 164 ! ! ! ! 3.7!Wastewater discharges Unregulated wastewater discharge from municipalities and other sources, such as agricultural runoff and industrial discharge, is a major concern for Canadian water quality. According to CBC news reports, more than 205 billion liters of untreated and raw sewage were noted to flow directly into Canadian rivers and surrounding oceans (Thompson, 2016). The percentage of untreated wastewater varies in Canada, with the least noted for Saskatchewan (approx. zero) and highest for Nova Scotia, accounting for over 20%. Increasing demands for wastewater treatment plant infrastructure upgrades (around 850-1000 wastewater treatment plants need upgrading) to meet federal wastewater regulation is becoming a challenge for municipalities as well as for industry. This is particularly true in small, remote and rural communities. This eventually leads to unregulated discharge into freshwater sources, which pollutes the (James and Leffler, 2013). Insufficient wastewater or no treatment of wastewater can lead to serious deterioration of surface water. High levels of organic matter can cause a decline in dissolved oxygen levels upon degradation, creating anoxic conditions and water which is unfit for drinking. Inadequate tertiary treatment, marked by high nitrogen and phosphorous levels may cause eutrophication as discussed earlier and reduce dissolved oxygen concentrations.

3.8!Leaching Leaching in groundwater and water bodies is an important phenomenon affecting Canadian water quality. Although the number of farms has been reduced in Canada over the past 40 years, those remaining have grown much larger, requiring higher amounts of mineral fertilizers. This has increased the leaching process due to the surface runoff carrying the agrochemicals and nutrients and is affecting the quality of both surface and ground waters.

The leaching phenomenon is especially prominent in the higher rainfall areas of Canada, where it reaches beyond the rooting depth of vegetation (Lefebvre et al., 2000). Leaching of nitrogen seeps into groundwater, which eventually meets the lakes, rivers or oceans. The problem is attributed to nutrients like phosphorus and nitrogen, which sorb onto lake sediments and their

! 165 ! ! ! ! slow release may trigger the eutrophication. The sorption of nutrients onto soil sediments can be a problem for many years, and hence requires proper control and management measures. In addition, mining practices, generally in the form of mine drainage, have played an important part in the contribution of metals to water resources. For instance, in British Columbia, near Brenda Mines, exposed rocks located near the closed mine site produced mine drainage during rainfall and snow melts. Especially the open-pit excavation method of mining involves a large amount of waste rocks which, if handled improperly, promotes leaching to the ground water. In Canada, for every ton of copper extracted from ore, 99 tonnes of waste rocks are produced. Certain mining practices which impact water quality are as follows: a)! : This phenomenon occurs due to the formation of sulphuric acid when sulphides present in rock come in contact with water or air. The problem escalates mainly due to the open-pit mining process where sulfides are exposed to oxygen in the air and water. When the acidification occurs, a naturally occurring bacteria Thiobacillus ferroxidans may thrive to build a more oxidative environment which further promotes leaching into the water sources. Acid mine drainage severely degrades the water quality and affects the aquatic ecosystem. Tulsequah mine in British Columbia was closed in 1957, yet it is still leaching acid water towards the Alaska region in the U.S. b)! Processing chemicals and heavy metals: Various reagents and chemicals are used to process out the target mineral from the ore. Use of cyanides or sulphuric acid and their improper handling in the industries impacts the water quality and increases the toxicity level. Heavy metals such as arsenic, cobalt, cadmium, and silver are leached and washed out by the carrying surface water where leaching is favored in the acidic environment. Thus, mining industries need to adhere to the regulations and proper management practices as laid by the Environment Canada to ensure the safety of drinking water sources requiring minimal treatment process.

3.9!Emerging contaminants Emerging contaminants are generally comprised of chemicals used in plasticizers, flame retardants, pharmaceuticals, personal care products, and are not commonly monitored. These

! 166 ! ! ! ! chemicals may pose an environmental threat to aquatic ecosystems even at very low concentrations.

For example, perfluorooctanoic acid (PFOA) has a maximum acceptable concentration of 0.2 µg/L in drinking water. Water quality analyses were performed on both treated and untreated samples at a few locations in Calgary and Quebec drinking water treatment plants. PFOA was detected in almost 75% of the test samples with a median value of 2.5 ng/L (Turel et al., 1996). In Ontario, five tap water samples were reported with PFOA at a concentration of 2.1 ng/L (Mak et al., 2009), while low concentrations (0.2 ng/L) were reported for Calgary and Vancouver (MetroVancouver, 2016). Although these concentrations are well below the harmful exposure level, it should be noted that PFOA has a tendency to bioaccumulate in living cells and tissues. Overuse of drugs is also threatening water quality where compounds, such as antibiotics, hormones, steroids, acetaminophen, and anti-epileptic compounds eventually find their way into drinking water systems. The Great Canadian lakes were found to contain all these compounds at a concentration high enough to be categorized under environmental concern (Crowe, 2014). A Canadian study reported a wide range of pharmaceutical compounds in river water in southwestern Ontario (Chan et al., 2014). Chemicals, such as the diabetic drug metformin, the acid reflux drug ranitidine, and the diuretic hydrochlorothiazide, which had never before been found in Canadian water bodies are now increasingly being detected.

Drinking water treatment plants do not have processes specifically designed to remove emerging contaminants and existing conventional treatment units are ill-equipped to treat them efficiently. Hence, continuous monitoring is necessary to assess and control the rate of water quality depletion. Improper disposal of pharmaceutical and personal care products (PPCPs) into is also a major concern and requires education of public to deal with this issue. As a result, high concentrations of PPCPs are found in wastewater (Dore, 2015). A study on the Canadian Great Lakes revealed that out of 42 common chemical contaminants, removal in activated sludge systems was less than 50% (Arvai et al., 2014). In addition to the optimization and improvement of biological processes, wastewater treatment plants need to be

! 167 ! ! ! ! upgraded from lagoon/primary treatment to at least secondary treatment facilities. In the long run, the aim is to follow water treatment regulations and improve quality, thereby, lessening the overall impact on the environment. It would be beneficial to protect the quality of water sources downstream of wastewater treatment facilities (Dore, 2015).

4: Social and economic aspects 4.1 Health In Canada, the most common threats to water sources and aquatic ecosystems are waterborne pathogens (National Water Research Institute, 2001). Disease outbreaks are often related to insufficient water treatment, which was the main cause of dysentery, cholera and typhoid fever epidemics in Canada during the 19th century (Rutty and Sullivan, 2009). However, advances in sanitation and water treatment ameliorated public health and decreased mortality rates such that, today, these diseases have been practically eradicated in Canada.

Between 1974-2001, Shigella and Salmonella were a cause of 9 of 16 confirmed, proposed or suspected drinking water outbreaks in Canada (Schuster et al., 2005). During the same time, the Campylobacter caused 24 outbreaks following Giardia with 51 outbreaks. However, the most serious waterborne outbreak in recent history took place in May 2000 in Walkerton, Ontario where the drinking water system was contaminated with pathogenic strains of Escherichia coli O157:H7 and Campylobacter jejuni. It led to the death of seven people, while almost half of the Walkerton citizens (over 2300 people) became infected (O’Connor, 2002). After this incident, more people started to purchase bottled water, which is still believed to be safer. Around 20% of Ontario residents claimed that tap water could pose moderate or serious threat to their health (Dupont et al., 2010). Studies have shown that as a result of this concern, Ontarians consume more tap water substitutes than residents in the rest of Canada. An increase in the use of tap filters or boiled water to avoid microbial contamination also have been observed. Health Canada recognizes that safe drinking water is a core public issue. Hence, it collaborates with the provincial and territorial governments to develop the guidelines that would minimize the risk of poor water quality. The most important parameter is the microbiological quality and the

! 168 ! ! ! ! presence of Escherichia coli, enteric viruses, and protozoa. The presence of various chemicals and radiological compounds such as chlorine residuals, lead and other heavy metals and trace organic compounds are also under survey (Health Canada, 2013).

The quality of drinking water affects humans in a direct way. The drinking water pollution by atrazine (50-649 ng/L), widely used in agriculture, was positively correlated with stomach cancer occurrence in Ontario, between 1987-1991 (Leeuwen et al., 1999). However, the quality of wastewater treatment may pose the adverse effects on health. The release of the contaminants to surface water may lead to negative effects on fish and shellfish population which may result in restrictions of their harvesting and consumption. For instance, the presence of perfluorooctane sulfonate (PFOS), known persistent pollutant, leads to bioaccumulation and considered to be a dangerous compound to humans and other living organisms (Health Canada, 2007). In general, PFOS concentrations in water (2011-2016) and fish tissue (2011-2014) have been below Federal Environmental Quality Guidelines. However, in some cases, concentrations were above wildlife diet guidelines, which posed a potential health risk (Environment and Climate Change Canada, 2016b). The discharge of heavy metals results in a chronic and acute effect on aquatic species as well as humans. The mercury poisoning is relatively rare in Canada. However, exposure to increased levels of mercury may result in health problems ranging from rashes to birth defects, or even death in extreme cases.

4.2! Indigenous people and water There are three groups of Canadian aboriginal people. The First Nations and Inuits are the Indigenous people of Canada, who lived there before the European’s arrival. Métis are the descendants of First Nations and European settlers. The beliefs of aboriginal communities are strongly connected with water as they consider it as the lifeblood of the land. The high-quality water is essential not only for ecological and health aspects but also to perform traditional activities. They use it for transportation, drinking, purification, cleansing or for cultivating the plants and animals as the source of food and medicine. However, there are many water-related challenges that Indigenous people must face. The most recent incident has taken place in Ontario

! 169 ! ! ! ! with Grassy Narrows First Nation community. Between 1963-1970, up to 10 metric tonnes of mercury were discharged into the Wabigoon-English River system, in Ontario by Reed Ltd (ECCC, 1980). The Grassy Narrows First Nation community is located 150 km downstream from the company’s site, and the Wabigoon River is their main fish source. The decontamination of the polluted area was performed due to the elevated concentration of mercury in the environment and toxic effect on the aquatic life. In the 1970s, the health survey was conducted from First Nation’s volunteers, and mercury was detected in their blood (up to 600 ppb). According to a 2017 report, the former Reed Ltd site is still discharging mercury into Wabigoon River, and the mercury’s concentration downstream the site exceeds the Ministry’s Severe Effects Level. The concentration range of mercury from collected fish samples was above “do not eat” benchmark for the sensitive population. Thus, the Grassy Narrows First Nation community is still directly affected by the mercury pollution.

As mentioned before, the beliefs of Indigenous people of Canada are strongly water-dependent. The aboriginal women play a crucial role due to their sacred relationship with water. Thus, they are strongly affected by water quality due to their water-related life-styles, including pregnancy, child delivery and bathing rituals, which are the part of childbearing. Moreover, they are known as ‘water caretakers’. Their role in the aboriginal community is to care for and protect water by talking circles, water walks or traditional activities. However, the European colonization resulted in discrimination and disenfranchised of women which was followed by disconnection with the land and their water-governance. Additionally, men usually occupy the leadership positions, where the women are not allowed to enter. Based on the Indian Act, the aboriginal women are excluded from decision-making procedures (Department of Justice Canada, 1985). This frequently resulted in the creation of high risk in drinking water systems in Indigenous communities. Recently, the aboriginal women across Canada have started to draw attention to water issues that affect not only Indigenous people but the whole country. They believe that their traditional role as the water caretakers should be restored so that they can share their history and knowledge with other Indigenous and non-Indigenous communities.

! 170 ! ! ! ! The Government of Canada and the aboriginal communities are collaborating to improve the various aspects of their life. Indigenous and Northern Affairs Canada, with the support of Health Canada, is responsible for water treatment on First Nations reserves, excluding British Columbia (Health Canada, 2009; 2010b; 2011; 2012; 2017). Between 2006-2014, around $3 billion was invested to assist in the upgrading, building or designing of water and wastewater treatment systems (Government of Canada, 2015b). Progress in water quality improvement was observed between November 2015 and January 2017 by a notable decrease in long-term drinking water advisories in First Nation communities (Black and McBean 2016; 2017b; Government of Canada, 2017b). Eighteen long-term (10 years) drinking water advisories have been lifted in this period and it is anticipated that all drinking water advisories will be removed by 2021. However, drinking water systems are still frequently considered as high risk. It is a result of many factors, like marginalization through placing Indigenous communities on reserves. In addition, population densities in Indigenous communities tend to be relatively low translating to a scarce community fund to support the construction, maintenance, and operation of adequate water treatment plants. Moreover, despite the attempts of the Government of Canada, the average income of Indigenous people tends to be below the Canadian average. Between 2004-2014, 2/3 of Canadian Indigenous people lived under at least one water advisory (Sawchuck, 2011; Black and McBean, 2016; 2017). 98% of on-reserve households have minimal water services, where only 54% of houses have access to fresh water, delivered by piped systems, as shown in Figure 1.

2%

36%

54%

8%

piped water system sewage hauled by truck septic and other individual wastewater systems

Figure 1: First Nation on-reserve household minimum water service, based on data from Sawchuk, 2011. ! 171 ! ! ! ! 4.3!Public education To improve the quality of Canadian waters, many agencies, councils, and institutes have been formed, with the common goal of public education, particularly targeted at Canadian youth, regarding the importance of water conservation and protection. One of the best examples is the Children's Water Education Council (CWEC), a charitable organization that organizes water- based programs and supports multiple water-related festivals across the province of Ontario (Children’s Water Education Council, 2017), where students are educated about the importance of conservation and protection of water resources. The activities of similar groups have contributed to Canadian water quality improvements resulting from increased public awareness.

4.4!Rural and urban areas The increase in urbanization and population growth strongly affects the quality of Canadian waters. The run-off from cities is introduced into rivers and lakes, which results in increased nutrient levels, sediments, animal wastes, pathogens, recalcitrant organics, petroleum products and other emerging contaminants. They are factors responsible for water quality deterioration. The greatest concerns are the Great Lakes basins, which are the most populated areas and include 9 of Canada’s 20 largest cities, including Toronto, Windsor, Oshawa, Hamilton, and Mississauga (Government of Canada, 2015b; Environment and Climate Change Canada, 2017c). On the other hand, large cities have access to resources and technologies that can effectively treat source water, enabling them to provide a high quality of drinking water. Some Canadian treatment plants even exceed water quality standards and regulations stipulated by governmental guidelines. On the other hand, water quality in rural areas tends to be much lower, and often do not meet these guidelines (Peterson and Torchia, 2008). This has been attributed to the close proximity between water supplies and livestock and . Small (rural) water systems usually do not have financial, technical, institutional capacity to create centralized water treatment facilities. Hence, each Canadian province has managed this critical issue differently. For instance, British Columbia has exemptions for smaller systems (Cook et al., 2013). There are also Rural Water Quality Programs, which offer technical and financial support for rural landowners to improve their water quality in several provinces (Peterson and Torchia, 2008).

! 172 ! ! ! ! 4.5!Investment in water quality programs Water quality in Canada has remained generally stable over the last 10 years. This can be attributed to the growing number of projects related to water issues, which have focused on the prevention of pathogenic organism intrusion, decreasing the concentration of pollutants, preserving aquatic wildlife, as well as reducing algal blooms and phosphorous inputs in the aquatic ecosystems. Table 1 presents examples of different projects, funds, and organizations that aim to improve the water quality in Canada.

Table 1. Examples of investment programs in Canadian waters. Data based on the Federal Sustainable Development Strategy Report (Government of Canada, 2015b) Aquatic Organization/ Stewardships/ Additional data compartment Fund Canada-Ontario Agreement on Great - protection and preservation of Great Lakes Lakes Water Quality and Ecosystem - $1.5 million (2014) for projects referring Great Lakes Health (2014) to Canadian areas of concern The Great Lakes Sustainability Fund St. Lawrence Action Plan 2011-2026 by - Over 50 projects on the conservation of St. Lawrence River Government of Canada with biodiversity, improvement of water quality, Government of Quebec the sustainable use of the river - prevention of 4040 kg/year of phosphorous the Lake Simcoe/ South-eastern Lake Simcoe and South- from entering the Lake Simcoe Georgian Bay Clean-up Fund eastern Georgian Bay - 124 kg/year of phosphorus prevented from Stewardship’s projects entering the South-eastern Georgian Bay - prevention of 14 800 kg/year of phosphorous from entering the Lake Stewardship’s projects with the aid of Winnipeg and its rivers Lake Winnipeg the Lake Winnipeg Basin Stewardship - main goal: to reduce phosphorous level by Fund 50 % to level before 1990 (0,1 mg/L to 0,05 mg/L) by 2017

4.6!Industry, mining and agriculture The constant development of agricultural and industrial areas, as well as mining, quarrying, and oil sands sectors has significantly affected the quality of Canadian waters. All of these sectors use water and then discharge it, which is the primary source of contaminants including heavy metals, toxic recalcitrant compounds, and phosphorus and nitrogen. However, the trend in direct release of contaminants into water significantly decreased between 2006 and 2015 due to governmental regulations, connected with broader worldwide environmental concerns (Environment and Climate Change Canada, 2015a). The latest report from the National Pollutant Release Inventory

! 173 ! ! ! ! of Canada (2015) revealed that the manufacturing sector released about 9,222 tons of pollutants (7% of total reported substances) directly into the water. The mining and quarrying sector discharged 3,921 tones (3%), while the oil and gas extraction sector released 2,442 tones (2%) (Environment and Climate Change Canada, 2015b). The number of reporting facilities increased between 2013-2015, which probably contributed to the increase in the total overall direct release of toxic compounds (e.g., acrylonitrile, nonylphenol and its ethoxylates). However, in some cases, such as isoprene or toluene diisocyanate, a decreasing trend of total direct release was observed. To estimate the quality of Canadian waters, the Compound index of Water Quality has been created. It includes four indicators: nitrogen, phosphorous, coliforms and pesticides levels and these are connected to the general overview of the environmental impact of the agriculture sector. Between 1981 and 2006 the index was at a ‘desired’ level (the highest) (Agriculture and Agri-Food Canada, 2016). However, data from 2011 revealed that index moved to a ‘good’ level (one point lower). The decline in all four indicators was attributed to wider fertilizer and manure usage, a greater amount of livestock production and pesticide treatments.

5: Implementation of the Sustainable Development Goals 6 (SDG-6) 5.1 Universal and equitable access to safe and affordable drinking water Water supply and sanitation in Canada is perceived as universally accessible and of high quality. Water appears to be an abundant and endless resource, with over 7% of the world’s renewable freshwater and over 243,000 km of coastline. This is also attributed to reliable water supplies, advanced water treatment technologies, adequate water operator knowledge and training, and access to extensive land management decision-making (CCME, 2006). In 2009, a large majority (88.9%) of the population was served by water distribution systems, with 10.5% on private and 0.6% receiving trucked water. Of the population served using piped water systems, 94.0% were also provided with water treatment, adhering to the Guidelines for Canadian Drinking

Water Quality, established by the Federal-Provincial-Territorial Committee on Drinking Water. These guidelines, pertaining to health effects, aesthetics, and operational procedures, are based on state of the art scientific research (Government of Canada, 2016a). However, Indigenous communities are largely excluded from equitable access to drinking water

! 174 ! ! ! ! and continue to face long-standing accessibility and sanitation issues. Indigenous peoples, comprised of First Nations, Métis and Inuit, constitute 4.3% of the total Canadian population (Environment Canada, 2011). In 2011, a study led by the Department of Aboriginal Affairs and Northern Development inspected water and wastewater systems located in Indigenous communities to assess overall human health risk. Of the 807 systems, 39% were categorized as high overall risk and 34% were categorized as medium overall risk, demonstrating the systemic and long-lasting accessibility issues and drinking water advisories associated with water in remote Indigenous communities in Canada (Black and McBean, 2016; 2017b; Department of Aboriginal Affairs and Northern Development, 2011). While new long-term drinking water advisories may continue to occur, INAC has developed an Action Plan to eliminate all long-term drinking water advisories affecting public water systems financially supported by INAC by 2021 (Department of Aboriginal Affairs and Northern Development, 2011). The federal government is actively working with provinces and territories, as well as Public Health Agency of Canada to provide ongoing funding, education, monitoring and service support on reserves to ensure that water treatment systems are improved and that the water quality levels specified in the Guidelines for Canadian Drinking Water Quality can be achieved.

Aside from remote First Nations communities, urban municipalities nationwide are also experiencing emerging water accessibilities issues. These problems are primarily associated with aging water infrastructure attributed to historical underinvestment in infrastructure maintenance and increasing system operation costs. As customer fees increase in an attempt to finance system maintenance and encourage water conservation, there is a growing concern that these adjustments may further marginalize lower socioeconomic groups. Various strategies to provide security for low-income communities include customer assistance funds. Other pricing reforms, such as seasonal surcharges and peak-load pricing, have been suggested as potential solutions (Canadian Municipal Water Consortium, 2015)

! 175 ! ! ! ! 5.2 Access to adequate and equitable sanitation and hygiene Nearly 75% of Canadians are serviced by municipal sewer systems while the remaining 25% of the Canadian population is served by septic disposal systems. While access to adequate sanitation and hygiene may seem to be universal, equitable access and water quality problems are still prevalent in some small, remote and rural communities in Canada. Between 1974 and 1996, over 200 outbreaks of infectious diseases associated with drinking water were reported with over 8000 confirmed cases linked to these outbreaks (Todd and Chatman, 1996). One of the most significant regulatory and impactful cases of a drinking water contamination occurred in 2000, in Walkerton, Ontario, where seven people died, and 2500 became ill due to E. coli O157:H7 contamination (Environment Canada, 2001).

While the majority of the Canadian population may not encounter health issues surrounding water pollution, vulnerable populations including infants, elderly, pregnant women and people with pre-existing health problems are at a higher risk of waterborne diseases and outbreaks. First Nations people are also uniquely susceptibility to adverse health effects from water pollution due to their strong dependence on natural resources such as water, for traditional, cultural and subsistence uses (Mascarenhas, 2007). Waterborne infections and disease outbreaks are considered to be 26 times greater in First Nations communities (Patrick, 2011). The emergence and spread of waterborne pathogens are likely to be a growing problem in Canada and worldwide with increasing human population densities and water pollution, as a result of the greater demands that will be placed on aging Canadian wastewater treatment infrastructure (Côté, 2005).

5.3 Improve water quality by reducing pollution Over 150 billion liters of untreated and undertreated wastewater is released into Canadian waterways annually making municipal wastewater discharges one of the largest sources of pollution in Canadian waters. Wastewater effluents contribute to approximately 75% of water pollution followed by forest-related products (12%), minerals and metals (4%), oil and gas (<1%), and electricity generation (<1%) (Natural Resources Canada, 2009). Municipal wastewater is composed of sanitary sewage and stormwater and can contain grit, debris,

! 176 ! ! ! ! suspended solids, disease-causing pathogens, organic wastes, nutrients, and about 200 identified chemicals. Although toxic effluents are heavily regulated in Canada, the release of nutrients, such as phosphorus and nitrogen, into the environment is common. High nutrient concentrations often lead to eutrophication, a serious water quality issue for the Prairie provinces, southern Ontario, and Quebec (Environment and Climate Change Canada, 2017d). In 2012, the Federal Government, in collaboration with the provinces, territories, engaged municipalities, Indigenous communities, conservation authorities and other non-governmental organizations, established the country’s first national standards for wastewater treatment, the WSER, to reduce water pollution and risks to aquatic and human health (Government of Canada, 2017c).

5.4 Increase water efficiency In 2013, approximately 38,300 million liters of water were withdrawn from Canada's rivers, lakes, groundwater and oceans (Environment and Climate Change Canada, 2017a). The thermal power generation industry withdrew the largest quantity of water, mainly used for cooling and steam production for electricity-generating turbines, followed by manufacturing, municipalities, commercial/industrial agriculture, mining and oil and gas sectors. The residential sector continues to account for the bulk of municipal water use (57.4% in 2009) (Natural Resources Canada, 2009), while the commercial/institutional sector follows with 18.7% of total water use. Leaks and system flushing/maintenance account for 13.3% (Environment Canada, 2011b).

The 2016-2019 Federal Sustainable Development Strategy outlines plans to monitor average percentage decreases in the volume of water leakage and infiltration, and a number of systems have improved the quality of wastewater effluent or stormwater discharge as a result of funded investments and incentives. Furthermore, the Clean Water and Wastewater Fund received $2 billion, to provide communities with more reliable water and wastewater systems and ensure both drinking water and wastewater effluents meet legislated standards before the end of 2016–2017, under Phase 1 of Infrastructure Canada’s two-phased infrastructure pertaining to green infrastructure development (Natural Resources Canada, 2005).

! 177 ! ! ! ! 5.5 Integrated water resources management at all levels Canada has been engaged in integrated water resources management (IWM) for a number of years. The Canada Water Act enables co-operative agreements for consultation and collaboration between the federal, provincial, and territorial governments in matters relating to water resources; an example would be the IJC a bi-national organization established by the United States and Canadian governments under the Boundary Waters Treaty of 1909. The Great Lakes Water Quality Agreement (GLWQA) of 1978 was later added to encompass additional responsibilities. The updated GLWQA includes chemical contamination management, ecosystem protection, nutrient loading monitoring, groundwater surveying and mitigation of invasive species. The IJC strives to prevent and resolve disputes surrounding water quality and resources in boundary regions. Furthermore, for specific water issues or watersheds, Canadian provinces and U.S. states work together in various bi-national initiatives and forums. For instance, in 1987, the Canada - United States GLWQA identified 43 AOCs across the Great Lakes. Through the GLWQA, Canada and the United States consults with State and provincial governments; Tribal, First Nations and Métis governments; Municipal governments, watershed management agencies, and other local public agencies, in order to develop programs that conserve and restore water in the region (Davies and Mazumder, 2003; Environment and Climate Change Canada, 2016b; 2016c; 2016d; 2016e; 2016f; Shrubsole et al., 2017).

Additionally, the Canadian Council of Ministers of the Environment and the Canadian Council of Resource Ministers provides a platform for intergovernmental discussion on regional and national environmental issues including water management. The federal and provincial/territorial governments have a national hydrometric agreement on water quantity use and monitoring. Regional water management is also achieved through organizations such as the Prairie Provinces Water Board, which ensure that interprovincial surface waters and groundwaters are equitably shared by Canada's Prairie provinces (Environment and Climate Change Canada, 2015e). In Ontario, Conservation Authorities oversee integrated water resource management efforts with municipalities in highly populated watersheds across the province. Numerous non-governmental

! 178 ! ! ! ! watershed stewardship bodies are also active in many areas, such as the Lake Ontario Waterkeepers and Earthroots (Environment Canada, 2011b).

5.6 Protect and restore water-related ecosystems Since 2010, various biological and ecological effects have been observed, such increased native species mortality, species range expansions and contractions, as well as changes in fish size, assemblages, and community structure. (Environment Canada, 2011b). Although ecosystem- based management measures have been implemented for most species, recovery has been limited in most cases. Some marine mammals that were exploited (e.g., the beluga and narwhal in the Arctic; sea otters; and killer, humpback and gray whales in the Pacific) are showing signs of recovery. However, the same cannot be said for many fish stocks on both the Atlantic and Pacific coasts that have been subjected to extensive commercial overexploitation. To overcome or stabilize these declines, new fisheries closures were announced in September 2016 under the 2016-2019 Canadian FSDS (Environment and Climate Change Canada, 2016d).

The status and trends of Canadian marine ecozones are also changing. The FSDS Target 4.5 has called for 10% of total coastal and marine territory to be conserved in MPAs and other effective area-based conservation measures by 2020. MPAs were established by the United Nations Convention on Biological Diversity in 2004 through the Program of Work on Protected Areas in habitats significant to fish and marine mammals (United Nations Development Program, 2009). However, most MPAs do not encompass commercial fishing, and one even allows for oil and gas development within its boundaries. From 2016 to 2017, several additional MPAs were designated in Scott Islands in British Columbia, Hecate Strait, and Queen Charlotte regions of British Columbia (Environment and Climate Change Canada, 2015d).

Eutrophication in lakes and other water bodies continue to be an issue for Canada. While there has been an overall decrease in phosphorus levels over the past 40 years, recent levels vary greatly from year to year, and phosphorus levels generally remain relatively high in the offshore waters of Lake Erie. The status of the St. Lawrence River continues to be monitored using

! 179 ! ! ! ! indicators outlined in the Overview of the State of the St. Lawrence and five Canadian Great Lakes Areas of Concern (Environment and Climate Change Canada, 2015d).

5.7 International cooperation in developing areas The CIDA is the primary organization that oversees and supports sustainable development activities focused on the alleviation of poverty and equitable distribution of resources. CIDA works with both private and public organizations in Canada and in developing countries along with international organizations and agencies. CIDA collaborates with Statistics Canada, the United Nations, the Organisation for Economic Co-operation and Development and other international agencies (Weston and Pierre-Antoine, 2003). For instance, in 2015, Canada contributed $14.25 million to a World Bank initiative, through the ASPIRE program, to promote private-public-partnerships in Indonesia. Within these partnerships, a number have focused on water infrastructure development and accessibility issues (Canadian International Development Platform, 2016).

However, in recent years, Canada’s aid strategy has changed with a decline in contributions towards clean water and sanitation. According to the OECD, in 2014, Canada only contributed 1.3% of its total aid towards water supplies and sanitation. By comparison, other OECD members, on average, contributed 4.0% of their total aid to water-related initiatives (OECD, 2017). However, with the introduction of the United Nations sustainable development goals, Canada has pledged more funding and initiatives. For instance, from 2015 to 2020, Canada will provide $17 million for the Food Security Innovation and Mobilization initiative in Peru, Bolivia and Burkina Faso that promotes groundwater extraction and hydroponic greenhouse technologies in order to improve water accessibility and food security (Government of Canada, 2017d). A greater focus on water-related assistance may be required as global shortages in water accessibility and sanitation services continue to increase.

! 180 ! ! ! ! 5.8 Support and strengthen the participation of local communities There are a number of eNGOs dedicated to improving water quality and accessibility through various monitoring programs, community outreach efforts and collaborations with government organizations in Canada. In fact, public participation is now mandated in some jurisdictions, such as in Ontario and Manitoba. In other jurisdictions, such as Alberta and Quebec, programs are voluntary in nature (Morin and Cantin, 2009; Bakker and Cook, 2011).

Since the 1980s, eNGOs have taken on an increasing role in environmental and water governance in Canada. These eNGOs are involved in consultation processes, advocacy, and monitoring programs while operating as watchdogs with considerable legal capacity (Morin and Cantin, 2009). This demonstrates an increasing emphasis on integrated management of environmental issues and cross-sectoral engagement in policy-making and implementation.

Currently, there is a trend towards connecting community networks with ecosystem services for watershed governance. For instance, a number public groups affiliated with local grass-root organizations were noted as major stakeholders actively involved developing and revising the conservation plan for the Oak Ridges Moraine, an ecologically important geological landform in the Mixedwood Plains of south-central Ontario. Several of these organizations developed community monitoring programs, focusing on water quality and reporting of native and invasive species. These groups are an integral part of the dissemination of scientifically relevant data to government agencies and providing information that can affect policy changes (Pollock and Whitelaw, 2005).

!

! 181 ! ! ! ! 6: Successful experiences in improving water quality 6.1: Policy changes 6.1.1. Source Water Protection Plans In May 2000, the groundwater supplying the municipal well for Walkerton, Ontario became contaminated with E. coli O157:H7 resulting from the multiple failed protection measures. Seven people died, and thousands became ill after drinking the contaminated water, serving as the catalyst for major reconstruction in drinking water protection policy (Conservation Ontario, 2013). Following a public inquiry, recommendations for a multi!barrier approach to providing safe drinking water were collected. In 2006, the Ontario government passed the Clean Water Act, focusing on protecting sources and supplies of water that provide drinking water to communities (Conservation Ontario, 2013). The Act also focused on increasing the awareness and protection of private drinking water wells that rely on regional groundwater sources. Additionally, funding was provided to strengthen drinking water source protection initiatives and existing legislation. Under the Clean Water Act, 19 source protection areas or regions were established in Ontario to carry out drinking water risk assessments and source protection plans (Ministry of Environment and Climate Change, 2017e).

Each region has been required to conduct an assessment of drinking water sources describing physical features and water resources, delineating vulnerable areas, identifying activities and potential threats, and outlining recommendations (Conservation Ontario, 2013). From this initial assessment, the regional Conservation Authority will develop a source protection plan. These source water protection initiatives aim to minimize contamination and overuse of drinking water sources (Water Quality and Health Council, 2012). Multi-stakeholder source protection committees, with representation from municipalities, economic sectors, and the general public, have also been created and are responsible for the protection of drinking water resources.

6.1.2. Implementation of Secondary Treatment In July 2012, Environment Canada updated its wastewater regulations with the aim of reducing the 150 billion litres of untreated sewage and wastewater released into waterways on an annual basis. The WSER (SOR/2012-139), under the Fisheries Act (R.S.C., 1985, c. F-14), specify ! 182 ! ! ! ! minimum treatment standards, monitoring protocols and testing methods for systems over 100 cfm/day excluding the Northwest Territories, Nunavut, and north of 54th parallel in Quebec and Newfoundland (Government of Canada, 2017c). Prior to 2012, primary treatment was the most common method of municipal wastewater treatment throughout Canada, and the implementation of mandatory secondary treatment was a significant step forward for many regions. Secondary treatment processes, as outlined in the legislation, allows for a 95% removal of pollutant mass, and reduces oxygen-consuming and solid materials, nutrients and bacteria using physical and biological methods (Government of Canada, 2017c). The regulations identified between 850- 1000 treatment facilities and systems to be at high, medium, and low risk, with requirements to meet these new standards by 2020, 2030, and 2040 respectively (Neegan Burnside Ltd., 2011). The addition of a biologically aerated facility in each system was found to reduce potential eutrophication problems associated with high nutrient levels. Biological aeration and substituent sludge removal were also integrated with recovery programs to produce biogas for winter heating and fertilizer for agricultural applications (Neegan Burnside Ltd., 2011).

6.1.3. Improving Water Quality in First Nations Communities The concerns regarding water quality in First Nations Communities primarily surrounds unsafe drinking water and the lack of funding and resources for new projects. In 2008, the First Nations Water and Wastewater Action Plan (FNWWAP) was introduced with the goal of providing funding for the creation of new water treatment facilities and operational training for First Nations communities (Indigenous and Northern Affairs Canada, 2012). With the support of Aboriginal Affairs and Northern Development Canada (AANDC) and Health Canada, the action plan facilitated the investment support into water projects over 4 years. Additionally, the Safe Drinking Water for First Nations Act was created with the goal of providing safe, clean and reliable drinking water through the development of treatment systems and source water protection on reserves (Neegan Burnside Ltd., 2011).

! 183 ! ! ! ! 6.2 Case Studies 6.2.1. Innovative Technologies: Saskatoon, Saskatchewan With an increasing population, Saskatoon (Saskatchewan) has prioritized drinking water quality, with goals that extend well beyond standards to prepare for future development. The Saskatoon Potable Water Supply System services a large area surrounding the city, delivering water to businesses and industrial users, as well as selling water to SaskWater, responsible for the delivery of water to surrounding communities (SaskWater, 2017). Saskatoon, therefore, requires a high functioning system capable of treating and storing large quantities of water. To ensure that the high demand for water would be met, a project plan was devised in 2011 to increase storage capacity, add a high lift pump station and install a UV disinfection system to the supply system (City of Saskatoon, 2014). The project was successfully completed in 2015 at a cost of $77 million, which was funded by the City of Saskatoon, Government of Canada and Province of Saskatchewan (City of Saskatoon, 2016).

While Saskatoon’s water treatment already complied with Canadian regulations, the municipality aimed to be adaptive to development and emergencies. The update comprised of a 7-step process, beginning with the screening and oxidation of organic compounds, followed by coagulation and water softening, chlorine and fluoride disinfection, filtration and additional chlorine disinfection. Water is then pumped into the distribution system or into storage . The new UV disinfection stage was placed between water and final chlorination. The UV disinfection system consisted of a series of medium pressure UV lamps designed to inactivate viruses and bacteria not properly removed through chlorination (City of Saskatoon, 2016). Storage capacity was increased from 42.1 million litres to 71.8 million litres (City of Saskatoon, 2016). The new high lift pump stations increased the efficiency of each storage location. The updated Saskatoon facilities exceed the Canadian Drinking Water Guidelines (City of Saskatoon, 2015).

6.2.2. Water Crisis: Constance Lake First Nation, Ontario One First Nation community with historically poor drinking water quality is Constance Lake First Nation (CLFN) (Indigenous and Northern Affairs Canada, 2012). The first water treatment

! 184 ! ! ! ! system installed adequately disinfected and distributed water from Constance Lake to the community but had discolouration issues. However, when Constance Lake suffered from algal bloom events, water quality was impacted, and the community was placed on a series of water boiling advisories in 2005, 2007, 2008, and 2010. A state of emergency was finally declared in 2014 (Constance Lake First Nation, 2016a). The drinking water source was changed from a surface water lake to a groundwater aquifer. While this mitigated the algal bloom issue, high levels of metals (such as iron and manganese) and high phosphorus concentrations gave rise to potential health issues. These ongoing issues were further exacerbated by a lack of proper training and education for system operators. The upgraded final WTP added a green sand filter that successfully reduced heavy metal concentrations, chemical requirements and maintenance costs. Aside from the technological innovation and upgrades to the system, a new community- based Water Management Action Plan was developed, and a training program was implemented for the employees of the WTP. The CLFN WTP was fully commissioned and functional as of March 2016, lifting the water boil ban that had been in place since 2014 (Constance Lake First Nation, 2016b).

6.2.3. Source Water Contamination: Howe Sound, BC The Britannia Mine, operational from 1904 to 1974, extracted copper, lead, zinc, gold, silver, and cadmium in the Howe Sound region, north of Vancouver. The Britannia Mine was once the largest copper mine in the British Empire. Prior to comprehensive legislation regarding procedures for mine closures and disposals (Ednie, 2006), approximately 5 billion liters of acidic, metal-laden water from underground workings, , and waste dumps entered. This was equivalent to approximately 250,000 kg of dissolved metals entering the water body annually, making it the one of the largest point source pollution of metals in North America. A sterilized zone extended over 1 km of the shoreline adjacent to the discharge river due to acid mine drainage and the dissolved heavy metals (Rhatiga, 2016). The heavy precipitation and steep coastal topography resulted in unsuccessful containment efforts (McCandleless, 2016). In 2003, the BC provincial government secured a public-private partnership contract with EPCOR for the construction and operation of a water treatment plant (Britannia Mine Museum,

! 185 ! ! ! ! 2017, 2017). The treatment facility, which opened in 2005 and cost $20 million, was designed to treat 4.2 billion liters annually and remove 226,000 kg of heavy metal contaminants. The acidic waste is first treated using high-density lime sludge in bioreactor tanks. The slurry rapidly increases the pH, precipitating out metals such as copper, zinc, iron, aluminum, manganese, and cadmium, ensuring that no regulatory exceedances occur. Future improvements to the treatment system include increased diversion of groundwater around ore-bodies and mine shafts, increased removal of manganese, and recovery of metals from sludge (EPCOR, 2017).

! 186 ! ! ! ! References Agriculture and Agri-Food Canada. 2016. Environmental Sustainability of Canadian Agriculture: Agri-Environmental Indicator Report Series—Report #4. Retrieved on June 30 2017 from http://www.agr.gc.ca/eng/science-and-innovation/agricultural-practices/ environmental- sustainability-of-canadian-agriculture-agri-environmental-indicator-report-series-report- 4/?id=1467307820931

Ae-Ngibise, K.A., P.L. Kinney, K.P Asante, D. Jack, E.A. Boamah, R. Whyatt, M. Mujtaba, A. Manu, S. Owusu-Agyei and B.J. Wylie. 2015. Pesticide Exposures in a Malarious and Predominantly Farming Area in Central Ghana. African Journal of Environmental Science and Technology. 9(8): 655-661.

Arvai, A., G. Kecka, S. Jasim, H. Melcer and M.T. Laitta. 2014. Protecting Our Great Lakes: Assessing the Effectiveness of Wastewater Treatments for the Removal of Chemicals of Emerging Concern. Water Quality Research Journal of Canada. 49(1): 23-31. Assembly of First Nations. 2007. Expert panel report on safe drinking water for First Nations. Retrieved from http://64.26.129.156/cmslib/general/water-panel-report.pdf Attawapiskat First Nation. 2015. Attawapiskat First Nation FYNE Membrane Pilot System Study. Showcasing Water Innovation- Government of Ontario. Retrieved June 22, 2017 from http://www.attawapiskat.org/wp-content/uploads/20150614CaseStudyAFN-FYNE- Membrane-System-Final_ENGLISHf.pdf Bakker, K. and C. Cook. 2011. Water Governance in Canada: Innovation and Fragmentation. International Journal of Water Resources Development. 27(2): 275-289.!! Black, K. and E. McBean. 2016. Increased Indigenous Participation in Environmental Decision- Making: A Policy Analysis for the Improvement of Indigenous Health. International Indigenous Policy Journal. 7(4): Article 5. Black, K. and E. McBean. 2017a. Indigenous Water, Indigenous Voice–A National Water Strategy for Canada’s Indigenous Communities. Canadian Water Resources Journal. 42(3): 248-257.

! 187 ! ! ! ! Black, K. and E. McBean. 2017b. Analysis of Challenges and Opportunities to Meaningful Indigenous Engagement in Sustainable Water and Wastewater Management. Water Policy. wp2017078 Britannia Mine Museum. 2014. Britannia Mine Museum 2014 Press Kit. Retrieved June 15 2017 from http://www.britanniaminemuseum.ca/assets/bmm-2014-press-kit-new.pdf.

Britannia Mine Museum. 2017. Britannia's Environment. Retrieved June 22 2017 from http://www.britanniaminemuseum.ca/assets/Britannia-environment-low-res.pdf. Bouwer, H. 1978. Groundwater Hydrology. McGraw-Hill, Inc. p. 423-424.

Boyd, D.R. 2006. The Water We Drink: An International Comparison of Drinking Water Standards and Guidelines Retrieved May 4 2017 from http://www.davidsuzuki.org/ publications/downloads/2006/DSF-HEHC-water-web.pdf

Canadian International Development Platform. 2016. Canada International Development Platform: Data Report 2016. Retrieved June 22 2017 from http://cidpnsi.ca/data-report- 2016/ Canadian Municipal Water Consortium. 2015. 2015 Canadian Municipal Water Priorities Report. Retrieved June 22, 2017 from http://www.cwn-rce.ca/assets/resources/pdf/2015- Municipal-Priorities-Report/2015-Canadian-Municipal-Water-Consortium-Report-web.pdf CCME. 1999. Canadian Water Quality Guidelines for the Protection of Aquatic Life. Retrieved June 22 2017 from http://ceqg-rcqe.ccme.ca/download/en/312 CCME. 2006. An analysis of Canadian and Other Water Conservation Practices and Initiatives. Retrieved June 15, 2017 from http://www.ccme.ca/files/Resources/water/ water_conservation/kinkead_fnl_rpt_2005_04_2.1_web.pdf

Chan, G.S., M. Scafe, and S. Emami. 1992 Cemeteries and Groundwater: An Examination of the Potential Contamination of Groundwater by Preservatives Containing Formaldehyde. Queen’s Printer for Ontario

! 188 ! ! ! ! Chan, L., O. Receveur, M. Batal, W. David, H. Schwartz and C. Tikhonov. 2014. First Nations Food, Nutrition and Environment Study (FNFNES): Results from Ontario (2011/2012). University of Ottawa. Universite de Montreal. Assembly of First Nations.

Chen, G., X. Liu, D. Tartakevosk and M. Li. 2016. Risk Assessment of Three Fluoroquinolone Antibiotics in the System. Ecotoxicology and Environmental Safety 133: 18-24.

Children's Water Education Council. 2017. Who’s Watching the Water for Our Children. Retrieved on October 31 2017 from: http://www.cwec.ca/

City of Saskatoon. 2014. Saskatoon Water 2014 Annual Water Quality Report. Saskatoon, Saskatchewan. City of Saskatoon. 2015. Saskatoon Water 2015 Annual Report. Saskatoon, Saskatchewan. City of Saskatoon. 2016. Water Treatment Plant - Water Reservoir Storage Expansion Project. Retrieved June 21 2017 from https://www.saskatoon.ca/services-residents/power-water/ water-wastewater/drinking-water/water-treatment-plant Conference Board of Canada. 2013. Water Quality Index. Retrieved March 13 2017 from http://www.conferenceboard.ca/hcp/details/environment/water-quality-index.aspx

Conservation Ontario. 2013. Walkerton and Source Water Protection: Ten Years Later. Retrieved September 24 2017 from http://conservationontario.ca/library?view=document& id=36:walkerton-and-source-water-protection-ten-years-later&catid=59:source-water- protection Constance Lake First Nation. 2016a. A Springboard to the Future-- Restoring Water Security to Constance Lake First Nation. Retrieved June 21 2017 from http://www.clfn.on.ca/ ?pgid=20 Constance Lake First Nation. 2016b. Public Works. Retrieved June 21 2017 from http://www.clfn.on.ca/ ?pgid=20 Cook, C., N. Prystajecky, I. Ngueng Feze, Y. Joly, G. Dunn, E. Kerby, V. Ozdemir and J. Isaac- Renton, 2013. A Comparison of the Regulatory Frameworks Governing Microbial Testing

! 189 ! ! ! ! of Drinking Water in Three Canadian Provinces. Canadian Water Resources Journal / Revue Canadienne des Ressources Hydriques, 38(3):185-195. Crowe, K. 2014. CBCnews - Drinking Water Contaminated by Excreted Drugs a Growing Concern Retrieved May 17 2017 from http://www.cbc.ca/news/health/drinking-water- contaminated-by-excreted-drugs-a-growing-concern-1.2772289

Dabeka, R.W., H.B Conacher, J.F. Lawrence, W.H. Newsome, A. McKenzie, H.P. Wagner, R.K.H. Chadha and K. Pepper. 2002., Survey of Bottled Drinking Waters Sold in Canada for Chlorate, Bromide, Bromate, Lead, Cadmium and Other Trace Elements. Food Additives and Contaminants. 19(8): 721-732.

Davies, J. and Mazumder, A. 2003. Health and Environmental Policy Issues in Canada: The Role of Watershed Management in Sustaining Clean Drinking Water Quality at Surface Sources. Journal of Environmental Management (68): 273-286. Department of Justice Canada. 1985. Indian Act (R.S., 1985, c.1-5), Ontario. Department of Andian Affairs and Northern Development. 2011. National Assessment of Water and Wastewater Systems in First Nations Communities. Orangeville, Ontario. Dorcey, A.H.J. 2006. Water Pollution, Historica Canada. Retrieved March 12 2017 from http://www.thecanadianencyclopedia.ca/en/article/water-pollution/

Dore, M.H. 2015. Water Policy in Canada: Problems and Possible Solutions. Springer International Publishing. Dupont, D., W.L. Adamowicz, and A. Krupnick. 2010. Differences in Water Consumption Choices in Canada: The Role of Socio-Demographics, Experiences and Perceptions of Health Risks. Journal of Water Health 8(4): 671-686. Dunn, G., K. Bakker and L. Harris. 2014. Drinking Water Quality Guidelines Across Canadian Provinces and Territories: Jurisdictional Bariation in the Context of Decentralized Water Governance. International Journal of Environmental Research and Public Health 11: 4634-4651. Ednie, H. 2006. Britannia Mine Reclamation: Showing the World How It’s Done. Canadian Institute of Mining, Metallurgy and Petroleum, Westmount, QC.

! 190 ! ! ! ! Environment Canada. 2001. Threats to Sources of Drinking Water and Aquatic Ecosystem Health in Canada. National Water Research Institute, Burlington, Ontario. NWRI Scientific Assessment Report Series No. 1. 72 p. Environment Canada. 2011a. 2011 Municipal Water Use Report. Retrieved June 15, 2017 from https://www.ec.gc.ca/Publications/B77CE4D0-80D4-4FEB-AFFA-0201BE6FB37B/2011- Municipal-Water-Use-Report-2009-Stats_Eng.pdf Environment Canada. 2011b. Integrated Watershed Management. Retrieved July 17 2017 from https://www.ec.gc.ca/eau-water/default.asp?lang=En&n=13D23813-1 Environment Canada. 2012. Canadian Aquatic Biomonitoring Network Field Manual: Wadeable Streams. Retrieved June 15 2017 from http://publications.gc.ca/collections/ collection_2012/ec/En84-87-2012-eng.pdf Environment and Climate Change Canada. 1990. Environmental Protection Act, section 18, Incident Report No. 6112-AMQL2R. Environment and Climate Change Canada. 2015a. Summary of Data Reported for 2015. National Pollutant Release Inventory (NPRI) Fact Sheet. Retrieved on October 31 2017 from https://www.ec.gc.ca/inrp-npri/D55C89B3-C1CB-48F6-A48D-D53DE8D70498/NPRI %20Factsheet%20-%20Downloadable%20version%20for%20Dec%206%20publication% 20-%20Summary%20of%20Data%20Reported%20for%202015.pdf Environment and Climate Change Canada. 2015b Summary Report: Reviewed Facility-Reported Pollution Data. Retrieved on October 31 2017 from https://www.canada.ca/en/ environment-climate-change/services/national-pollutant-release-inventory/publications/ 2015-summary-report-pollution-data.html Environment and Climate Change Canada. 2015c. Canadian Environmental Sustainability Indicators: Freshwater Quality in Canadian Rivers. Retrieved July 27 2017 from https://www.ec.gc.ca/indicateurs-indicators/default.asp?lang=En&n=68DE8F72-1 Environment and Climate Change Canada. 2015d. 2015 Progress Report of the Federal Sustainable Development Strategy. Retrieved July 27 2017 from https://www.canada.ca/ content/dam/eccc/documents/pdf/federal-sustainable-development-strategy/2015-federal- sustainable-development-strategy/EC_2015-Progress-Report_EN.pdf

! 191 ! ! ! ! Environment and Climate Change Canada. 2015e. Water and Climate Change: How Might These Changes Affect Canada? Retrieved July 27 2017 from https://www.ec.gc.ca/indicateurs- indicators/default.asp?lang=en&n=578AC3D2-1 Environment and Climate Change Canada. 2016a. Polybrominated Diphenyl Ethers (PBDEs) in Fish and Sediment. Retrieved March 30 2017 from https://www.ec.gc.ca/indicateurs- indicators/default.asp? lang=en&n=0970C75C-1 Environment and Climate Change Canada. 2016b. Perfluorooctane Sulfonate (PFOS) in Fish and Water. Retrieved March 30 2017 from https://www.ec.gc.ca/indicateurs-indicators/ default.asp?lang=en&n=9AA98870-1 Environment and Climate Change Canada. 2016c. Achieving a Sustainable Future: A Federal Sustainable Development Strategy for Canada 2016-2019. Retrieved July 27 2017 from http://www.fsds-sfdd.ca/downloads/FSDS_2016-2019_Final.pdf Environment and Climate Change Canada. 2016d. Canadian Environmental Sustainability Indicators: Water Quantity in Canadian Rivers. Retrieved July 27 2017 from www.ec.gc.ca/indicateurs-indicators/default.asp?lang=En&n=D55F8316-1. Environment and Climate Change Canada. 2016e. Canadian Environmental Sustainability Indicators: Water Withdrawal and Consumption by Sector. Retrieved July 28 2017 from www.ec.gc.ca/indicateurs-indicators/default.asp?lang=en&n=5736C951-1. Environment and Climate Change Canada. 2016f. Canadian Environmental Sustainability Indicators: Nutrients in the St. Lawrence River. Retrieved July 28 2017 from www.ec.gc.ca/indicateurs-indicators/default.asp?lang=en&n=F7F5426B-1. Environment and Climate Change Canada. 2017a. Fresh Water Quality Monitoring and Surveillance. Retrieved June 28 2017 from https://www.ec.gc.ca/eaudouce-freshwater/ Environment and Climate Change Canada. 2017b. Canada-Ontario Agreement on Great Lakes Water Quality and Ecosystem Health. Retrieved June 28, 2017 from https://www.ec.gc.ca/ grandslacs-greatlakes/default.asp?lang=En&n=B903EE0D-1 Environment and Climate Change Canada. 2017c. Canadian Environmental Sustainability Indicators: Phosphorus Levels in the Offshore Waters of the Great Lakes. Retrieved June 28 2017 from www.ec.gc.ca/indicateurs-indicators/default.asp?lang=en&n=A5EDAE56-1.

! 192 ! ! ! ! Environment and Climate Change Canada. 2017d. Wastewater. Retrieved June 29 2017 from https://www.ec.gc.ca/eu-ww/default.asp?lang=En&n=5F25AD10-1 Environment and Climate Change. 2017e. Source Protection. Retrieved September 24 2017 from https://www.ontario.ca/page/source-protection EPCOR. 2017. Britannia Mine, BC. EPCOR. Retrieved June 15 2017 https://www.epcor.com/ about/who-we-are/where-we-operate/Pages/britannia-mine.aspx. Fisheries and Oceans Canada. 2010. Canadian Marine Ecosystem Status and Trends Reports. Retrieved June 28 2017 from http://waves-vagues.dfo-mpo.gc.ca/Library/341126.pdf Government of Canada. 1982. Constitution Acts, 1867 to 1982. Retrieved June 28 2017 from http://laws-lois.justice.gc.ca/eng/Const/index.html Government of Canada. 2014. Canadian Environmental Protection Act, 1999. Retrieved June 16 2017 from http://laws-lois.justice.gc.ca/PDF/C-15.31.pdf Government of Canada. 2015a. Canada Shipping Act, 2001. Retrieved June 16, 2017 from http://laws-lois.justice.gc.ca/PDF/C-10.15.pdf Government of Canada. 2015b. Progress Report of the Federal Sustainable Development Strategy: Maintaining Water Quality and Availability. Retrieved on October 31 2017 from https://www.canada.ca/en/services/environment/conservation/sustainability/federal- sustainable-development-strategy/2015-progress-report/water-quality-availability.html Government of Canada. (2016a). Achieving a Sustainable Future: A Federal Sustainable Development Strategy for Canada. Retrieved June 26 2017 from http://fsds-sfdd.ca/ index.html#/en/intro/executive-summary Government of Canada. 2016b. Fisheries Act. Retrieved June 15 2017 from http://laws- lois.justice.gc.ca/PDF/F-14.pdf Government of Canada. 2017a. Guidelines for Canadian Drinking Water Quality Summary Table. Retrieved June 16 2017 from https://www.canada.ca/content/dam/hc- sc/migration/hc-sc/ewh-semt/alt_formats/pdf/pubs/water-eau/sum_guide- res_recom/sum_guide-res_recom-eng.pdf Government of Canada. 2017b. Progress Towards Eliminating Long-Term Drinking Water Advisories in First Nation Communities. Retrieved March 30 2017 from

! 193 ! ! ! ! https://www.aadnc-aandc.gc.ca/eng/1488477520261/1488477554976. Aboriginal Affairs and Northern Development. Government of Canada. 2017c. Wastewater Systems Effluent Regulations: Reporting Overview. Retrieved November 1 2017 from https://www.canada.ca/en/environment-climate- change/services/wastewater/system-effluent-regulations-reporting/overview.html Government of Canada. 2017d. Water in Developing Countries. Retrieved June 26, 2017 from http://international.gc.ca/world-monde/issues_development-enjeux_developpement/ environmental_protection-protection_environnement/water-eau.aspx?lang=eng Government of Manitoba. 2014. Manitoba News release. Retrieved May 21 2017 from http://news.gov.mb.ca/news/index.html?item=30526 Government of Ontario. 2017a. Clean Water Act, 2006. Retrieved June 26 2017 from https://www.ontario.ca/laws/statute/06c22 Government of Ontario. 2017b. Conservation Authorities Act. Retrieved June 26 2017 from https://www.ontario.ca/laws/statute/90c27?search=conservation+authority He, K., A.D. Soares, H. Adejumo, M. McDiarmid, K. Squibb and L. Blaney. 2015. Detection of a Wide Variety of Human and Veterinary Fluoroquinolone Antibiotics in Municipal Wastewater and Wastewater-Impacted Surface Water. Journal of Pharmaceutical and Biomedical Analysis. 106: 136-143. Health Canada. 2007. Perfluorooctane Sulfonate (PFOS) and Health. Retrieved March 30 2017 from http://www.hc-sc.gc.ca/ewh-semt/pubs/contaminants/perflurooctane_sulfonate- eng.php Health Canada. 2009. Drinking Water Advisories in First Nations Communities in Canada: A National Overview 1995-2007. Ottawa, Ontario. Health Canada. 2010a. Guidelines for Canadian Drinking Water Quality. Ottawa, ON: the Federal- Provincial-Territorial Committee on Drinking Water. Health Canada. 2010b. Integration Envelope – Funded Projects. First Nations, Inuit and Aboriginal Health. Ottawa, Ontario. Health Canada. 2011. First Nations, Inuit and Aboriginal Health: Drinking Water and Wastewater. Ottawa, Ontario.

! 194 ! ! ! ! Health Canada. 2012. First Nations, Inuit and Aboriginal Health: Ensuring Safe Drinking Water in First Nations Communities in Canada. Ottawa, Ontario. Health Canada. 2013. Guidance on Providing Safe Drinking Water in Areas of Federal Jurisdiction. Ottawa, Ontario. Health Canada. 2017. Questions and Answers: Drinking Water and Wastewater in First Nations communities south of 60. Retrieved March 16 2017 from: http://www.hc-sc.gc.ca/fniah- spnia/promotion/public-publique/water-eau-eng.php. Hill, C., Furlong, K. Bakker, K. and A. Cohen. 2008. Harmonization vs Subsidiarity in Water Governance: A Review of Water Governance and Legislation in the Canadian Provinces and Territories. Canadian Water Resources Journal 33(4): 315-332. Huffington Post Canada. 2014. Canada Largest Contributor to Deforestation Worldwide: Study. Retrieved May 31 2017 from http://www.huffingtonpost.ca/2014/09/05/canada- deforestation-worst-in-world_n_5773142.html

Indigenous and Northern Affairs Canada. 2012. First Nations Water and Wastewater Action Plan. Government of Canada Retrieved June 6 2017 from http://www.aadnc-aandc.gc.ca/ eng/1313426171775/1313426357946 James, H. and B. Leffler. 2013. Global News - Canada’s Wastewater Regulations. Retrieved June 13 2017 from http://globalnews.ca/news/566131/canadas-wastewater-regulations/ Lefebvre, A., W. Eilers, and B. Chunn. 2000. Environmental Sustainability of Canadian Agriculture: Agri-Environmental Indicator Report Series–Report# 2. Levasseur J. and J. Marcoux. 2015. Bad water: ‘Third World' conditions on First Nations in Canada. CBC News. Retrieved October 30 2017 from http://www.cbc.ca/news/canada/ manitoba/bad-water-third-world-conditions-on-first-nations-in-canada-1.3269500 Lopez-Gresa, M.P., R. Ortiz, L. Perello, J. Latorre, M. Liu-Gonzalez, S. Garcia-Granda, M. Priede and E. Canton. 2002. Interactions of Metal Ions with Two Quinolone Antimicrobial Agents (Cinoxacin and Ciprofloxacin). Spectroscopic and X-ray Structural Characterization. Antibacterial studies. Journal of Inorganic Biochemistry. 92(1): 65-74.

! 195 ! ! ! ! Mak, Y.L., S. Taniyasu, L.W.Y. Yeung, G. Lu, L. Jin, Y. Yang, P.K.S. Lam, K. Kannan and N. Yamashita. 2009. Perfluorinated Compounds in Tap Water from China and Several Other Countries. Environmental Science and Technology. 43(13): 4824-4829. Maria, C.S. 2015. The weather Network, Bloodroot and 7 Other Dangerous Plants: How to Identify Each. Retrieved March 24 2017 from https://www.theweathernetwork.com/news/ articles/toxic-plants-found-in-canada/52042

Mascarenhas, M. 2007. Where the Water Divides: First Nations, Tainted Water and Environmental Justice in Canada. Local Environment, 12(6), 565-577. McCandless, R.G. 2016. Ending Pollution at the Britannia Copper Mine. BC Studies 188: 9-33, 163. MetroVancouver. 2016. Drinking Water Facilities. Retrieved June 13 2017 from http://www.metrovancouver.org/services/water/quality-facilities/facilities-processes/ drinking-water-treatment-facilities/Pages/default.aspx Morin, A. and B. Cantin. 2009. Strengthening Integrated Water Resource Management in Canada. Ottawa, ON: Policy Research Initiative. Munawar, M. and R. Heath. 2008, Checking the Pulse of Lake Erie. Michigan State University Press. Aquatic Ecosystem Health and Management Society. 640 pp.

National Water Research Institute. 2001. Threats to Sources of Drinking Water and Aquatic Ecosystem Health in Canada. Burlington, Ontario. 72 pp. Natural Resources Canada. 2005. Freshwater: The role of contribution of Natural Resources Canada. Retrieved June 13 2017 from http://publications.gc.ca/Collection/M4-16- 2005E.pdf Natural Resources Canada. 2009. Overview of water use by the natural resources sectors. Retrieved June 13 2017 from https://www.nrcan.gc.ca/sites/www.nrcan.gc.ca/files/www/ pdf/com/resoress/publications/wateau/wateau-eng.pdf Neegan Burnside Ltd. 2011. National Assessment of First Nations Water and Wastewater Systems. Department of Indian and Northern Affairs Canada, Ottawa, ON. O'Connor, D.R. 2002. Report of the Walkerton Inquiry: The Events of May 2000 and Related Issues. The Walkerton Inquiry. Toronto. Queen’s Printer for Ontario. 510 pp. ! 196 ! ! ! ! OECD. 2017. Aid to the Water and Sanitation sector. Organisation for Economic Cooperation and Development Stats. Retrieved July 13 2017 from http://www.oecd.org/dac/stats/water- relatedaiddataataglance.htm Patrick, R. 2011. Uneven Access to Safe Drinking Water for First Nations in Canada: Connecting Health and Place Through Source Water Protection. Health and Place, 17: 386-389. Peterson, H. and M. Torchia. 2008. Safe Drinking Water for Rural Canadians. CMAJ 179(1):55. Pollock, R.M. and G.S. Whitelaw. 2005. Community-Based Monitoring in Support of Local Sustainability. Local Environment 10(3): 211-228. Psomas, G. and D.P. Kessissoglou. 2013. Quinolones and Non-Steroidal Anti-Inflammatory Drugs Interacting with Copper(II), Nickel(II), Cobalt(II) and Zinc(II): Structural Features, Biological Evaluation and Perspectives. Dalton Transactions 42(18): 6252-6276. Rhatigan, J. 2016. Afterlife of a Mine: The Tangled Legacies of the Britannia Mine. The University of British Columbia, Vancouver. Roch, M., R.N. Nordin, A. Austin, C.J.P. McKean, J. Deniseger, R.D. Kathman and J.A. McCarter, M.J.R. Clark. 1985. The Effects of Heavy Metal Contamination on the Aquatic Biota of Buttle Lake and the Campbell River Drainage (Canada). Archives of Environmental Contamination and Toxicology. 14(3): 347-362. Rutty, C. and S.C. Sullivan. 2009. This is Public Health: A Canadian History. Canadian Public Health Association.

Sánchez-Monedero, M.A., A. Roig, C. Paredes, M.P. Bernal. 2001. Nitrogen Transformation During Organic Waste Composting by the Rutgers System and its Effects on pH, EC and Maturity of the Composting Mixtures. Bioresource Technology. 78(3): 301-308.

SaskWater. 2017. Potable Water. (SaskWater) Retrieved June 2017 from http://www.saskwater.com/services/our_services/potable_water.php Sawchuk, J. 2011. Social Conditions of Indigenous Peoples. Retrieved June 19 2017 from http://www.thecanadianencyclopedia.ca/en/article/native-people-social-conditions/. Schuster, C.J., Ellis, A. G., Robertson, W. J., Charron, D. F., Aramini, J. J., Marshall, B. J., Medeiros, D. T. 2005. Infectious disease outbreaks related to drinking water in Canada, 1974-2001. Can J Public Health, 96(4): p. 254-8. ! 197 ! ! ! ! Sharma, P.C., A. Jain and S. Jain. 2009. Fluoroquinolone Antibacterials: A Review on Chemistry, Microbiology and Therapeutic Prospects. Acta Poloniae Pharmaceutica. 66(6): 587-604. Shrubsole, D., D. Walters, B. Veale and B. Mitchell. 2017. Integrated Water Resources Management in Canada: The Experience of Watershed Agencies. International Journal of Water Resources Development 33(3): 349-359. Statistics Canada. 2013a. Canadians and Nature: Fertilizers and Pesticides. Retrieved March 17 2017 from http://www.statcan.gc.ca/pub/16-508-x/16-508-x2015007-eng.htm. Statistics Canada. 2013b. Persons in Low Income After Tax (In percent, - 2007 to 2011) Retrieved on June 27 2017 from http://www.statcan.gc.ca/tables-tableaux/sum- som/l01/cst01/ famil19a-eng.htm?sdi=low%20income Statistics Canada. 2013b. Human Activity and the Environment: Freshwater in Canada. Retrieved July 13 2017 from http://www.statcan.gc.ca/pub/16-201-x/16-201-x2017000-eng.pdf Statistics Canada. 2017. Women in Canada: A Gender-Based Statistical Report (89-503-X). Retrieved October 31 2017 from http://www5.statcan.gc.ca/olc-cel/olc.action?objId=89- 503-X&objType=2&lang=en&limit=0 Thompson, E. 2016. CBCnews - Billions of Litres of Raw Sewage, Untreated Wastewater Pouring into Canadian Waterways. Retrieved from May 16 2017 from http://www.cbc.ca/news/ politics/sewage-pollution-wastewater-cities-1.3889072 Todd, E.C.D. and P. Chapman. 1996. Foodborne and Waterborne disease in Canada: Annual Summaries. Laval: Polyscience Publications Inc. Turel, I., N. Bukovec and E. Farkas. 1996. Complex Formation Between Some Metals and a Quinolone Family Member (Ciprofloxacin). Polyhedron. 15(2): 269-275. Turgeon, S., M.J Rodriguez, M. Theriault, P. Levallois. 2004. Perception of Drinking Water in the Quebec City Region (Canada): The Influence of Water Quality and Consumer Location in the Distribution System. Journal of Environmental Management, 70(4): 363-373. United Nations Development Programme (UNDP). 2009. Human development report 2009. Retrieved July 13, 2017 from http://hdr.undp.org/sites/default/files/reports/269/hdr_2009_en_complete.pdf

! 198 ! ! ! ! Van Leeuwen, J. A., Waltner-Toews, D., Abernathy, A., Smit, B., Shoukri, M. 1999. Associations between stomach cancer incidence and drinking water contamination with atrazine and nitrate in Ontario (Canada) agroecosystems, 1987-1991. International Journal of Epidemiology, 28 (5), 836–840.

Water Quality and Health Council. 2012. Tools and Approaches for Source Water Protection in Canada. Governance for Source Water Protection in Canada, Report No. 1.

Weston, A. and D. Pierre-Antoine. 2003. Poverty and Policy Coherence: A Case Study of Canada’s Relations with Developing Countries. The North-South Institute. Retrieved July 13 2017 from http://www.nsi-ins.ca/wp-content/uploads/2012/10/2003-Poverty-and-Policy- Coherence-Final-Report.pdf Walker, N. 2013a. Canadian Geographic - Pollution in the Great Lakes. Retrieved May 12 2017 from https://www.canadiangeographic.ca/article/pollution-great-lakes Walker, N. 2013b. Canadian Geographic - Pollution in the Great Lakes. Retrieved June 19 2017 from https://www.canadiangeographic.ca/article/pollution-great-lakes World Water Federation of Canada. 2016. WWF-Canada Blog:Freshwater - Water Pollution in Canada and What You Need to Know. Retrieved April 2 2017 from http://blog.wwf.ca/blog/ 2016/03/25/what-to-know-about-water-pollution-in-canada Xu, J., Y. Xu, H. Wang, C. Guo, H. Qiu, Y. He, Y. Zhang, X. Li and W. Meng. 2015. Occurrence of Antibiotics and Antibiotic Resistance Genes in a Sewage Treatment Plant and its Effluent-Receiving River. Chemosphere 119: 1379-2385.

Zhao, Y., J. Geng, X. Wang, X. Gu and X. Gao. 2011. Adsorption of Tetracycline onto Goethite in the Presence of Metal Cations and Humic Substances. Journal of Colloid and Interface Science. 361(1): 247-51.

!

! 199 ! ! ! !