PERFORMANCE EVALUATION OF PILOT-SCALE CONSTRUCTED WETLANDS FOR THE TREATMENT OF DOMESTIC WASTEWATER IN ADDIS ABABA

PERFORMANCE EVALUATION OF PILOT-SCALE CONSTRUCTED WETLANDS
FOR THE TREATMENT OF DOMESTIC WASTEWATER
IN ADDIS ABABA, ETHIOPIA

Draft
Of
DISSERTATION

By:
Belachew M D

SUPERVISORS:
Prof. Memory Tekere (Supervisor)
Dr. Fasil Asefa (Co-supervisor)

FEBRUARY, 2018
ABSTRACT
PERFORMANCE EVALUATION OF PILOT-SCALE CONSTRUCTED WETLANDS
FOR THE TREATMENT OF DOMESTIC WASTEWATER
IN ADDIS ABABA, ETHIOPIA

Belachew M D
PhD, Department of Environmental Science (UNISA)
Supervisor: Prof. Memory Tekere UNISA
Co-supervisor: Dr. Fasil Asefa (AAU)
February, 2018
Experimental study was carried out to evaluate the performance of pilot scale subsurface flow constructed wetlands for the treatment of domestic wastewater in Addis Ababa, Ethiopia. In 2014, three parallel sets of pilot scale subsurface flow constructed wetland systems which consist of one Horizontal Flow Constructed Wetland (HFCW), one Vertical Flow Constructed Wetland (VFCW) and one hybrid of HFCW and VFCW were built at Kotebe Wastewater Treatment Plant (WWTP) in Addis Ababa. The wetland systems had identical wetland fill media and macrophytes but with different wastewater flow types. All the three wetland cells were filled with gravel and planted with Cyprus papyrus. The total surface area of the wetland systems was 72 m2 (24 m2 for each) and the wetland systems were designed to treat 3.15 m3 (70 PE) of domestic wastewater per day. To evaluate the performance of the pilot scale CW systems, samples were taken from the influent and effluents every 15 days and taken to the laboratory of Addis Ababa Environmental Protection Authority (EPA) and Ethiopian Construction Design and Supervision Works Corporation for analysis. Meanwhile, wastewater parameters which could result change of concentration with variation of environmental factors were measured onsite.
The objective of this study was to evaluate the performance of subsurface flow CW in treating domestic wastewater under the existing climatic conditions of Addis Ababa and also, the effect of different seasons and wastewater flow types for the removal of pollutants from wastewater was determined in the study. Furthermore nutrient uptake by macrophytes and the rate of areal removal rate were also quantified.
Based on the results of the first one year monitoring period, the annual average removal efficiency of the HFCW, VFCW and hybrid of the two were: BOD (89.1%, 92.2% and 93.4%), COD (80.6%, 82.1% and 84.0%), TSS (89.1, 83.8% and 84.7%), NH4+ (58.6%, 66.2% and 65.4%), NO3- (64.0%, 71.5% and 73.5%), TN (49.1%, 54.9% and 58.7%), PO43- (45.4%, 50.3% and 48.4%), TP (58.0%, 51.7% and 54.4%) and FC (97.95%, 96.6% and 96.5%) respectively. The hybrid of HFCW and VFCW wetland systems showed the highest removal efficiency for most pollutants considered in this study because of the exploitation of specific advantages of individual system.
Concerning the nutrient content of the wetland plant /Cyprus papyrus/, the average TN contents of the root (below-ground plant part) and stem (above-ground plant part) were 1.56% and 2.27% for the HFCW, 1.75% and 2.74% for the VFCW and 1.80% and 2.63% for the hybrid systems, respectively. Meanwhile, the average TP contents of the root and stem of the wetland plant were 0.139% and 0.064% for the HFCW, 0.167% and 0.067% for the VFCW and 0.115% and 0.065% for the hybrid systems, respectively. From the result, it was observed that the TN content of the above-ground part of the wetland plant was higher than the TN content of below-ground plant part. Conversely, the TP content of the below-ground plant part was higher than the TP content of above-ground part.
In general, the result showed that properly designed constructed wetland systems could be used as effective wastewater treatment method in Ethiopia.

Key words: Wastewater treatment, subsurface flow constructed wetlands, media/substrate, macrophytes, Cyprus papyrus, influent, effluent, wastewater flow type, nutrient uptake, loading rate, pollutant removal, seasonal performance.

DEDICATION
I dedicate this thesis to my late mother whom I missed in the middle of my study. I wish she came back and stayed a while at this end.

ACKNOWLEDGEMENTS
There were many people standing beside me when I have gone through the long process of my PhD study. I truly realize that their continuous support, encouragement and motivation are the secret behind my effort for making achievable what seemed to be unachievable. I am very indebted to all and I will always be thankful for their contribution. I thank you!
I would like to express my heartfelt gratitude to my supervisor Professor Memory Tekere and co-supervisor Dr. Fasil Assefa for the relentless support, advice, encouragement, guidance, invaluable comments, criticism, and follow up of my study progress. I consider myself as a very lucky person to have you throughout the process of my work. Thank you again!
I am very grateful to Wollo University (WU), Federal Democratic Republic of Ethiopia Ministry of Education (FDRE MoE), and University of South Africa (UNISA) for giving me the opportunity to pursue my PhD study. This work was funded by WU and Bursary research fund. I would like to thank all the staff of the offices for granting and timely facilitating the financial support so that I could continue with the work without interruption. I am also indebted to Addis Ababa Water and Sewerage Authority (AAWSA) for giving me permission to carry out my research in the compound of Kotebe WWTP and Pesticide Action Nexus Ethiopia (PAN-Ethiopia) for providing laboratory equipments that are used to take onsite measurements in the field.
I am thankful to Addis Ababa Environmental Protection Authority (AA EPA) and Ethiopian Construction Design and Supervision Works Corporation for analyzing the samples during the monitoring period and National Meteorological Agency (NMA) for providing the available meteorological data. I gratefully acknowledge the staff of UNISA Akaki branch and library of Economic Commission for Africa (ECA) for giving permission to freely access the library services.
I would like to extend many thanks to my family members for the worship, understanding and enthusiasm throughout the many years of my study. My special father, Dagne Belachew, your heartening words have been helping me to regain energy during those times which were full of challenges. Though I have no words which are strong enough to express my internal feeling, thank you, Abiye!
This would have not been possible without the unreserved support and devotion of my family and colleagues: Linger Dagne, Getie Dagne, Birhanu Minalu, Getahun Kume, Atalo Belay, Zemenu Genet, Wubirst Gebyehu, Tena Gebeyehu, Endashaw Gebeyehu, Henok Gizachew, Dawit Gizachew and Yonas Gizachew. Thank you all! Many thanks go to Tadesse Amera, and Nigussie Habtemariam for being always in the front with solutions whenever there were challenges in the process of my study.
I am fortunate to have a very considerate wife, Yeshiwork Gebeyehu, who has been untiringly helping me with all the tough field work. I am highly grateful to you, Yeshiye. My kids Tsion and Soliana, I don’t exactly know how many times I declined from answering your early day’s request because of the long process of my study. Tsion, I never forget your frequent questions; what are you thinking about? When will you stop your travel to the site? I think this end will give you the summarized answer for all of your questions you have had so far in the future if not now.
Last but not the least; I want to thank the almighty GOD for being always with me. Thank you for giving solutions for each and every up and down.

Table of Contents
Contents Pages
ABSTRACT II
DEDICATION IV
ACKNOWLEDGEMENTS V
TABLE OF CONTENTS VII
LIST OF FIGURES XI
LIST OF TABLES XIV
LIST OF ABBREVIATIONS XV
CHAPTER ONE 1
INTRODUCTION 1
1.1 General Information 1
1.2 Objective of The Study 6
1.2.1 General Objective 6
1.2.2 Specific Objectives 6
1.3 Research Questions 7
1.4 Scope of The Study 7
CHAPTER TWO 9
LITERATURE REVIEW 9
2.1 Basic Information about Wetlands 9
2.2 Constructed Wetlands as an Attractive Technology for WWT; An Overview 10
2.3 Advantages of Constructed Wetlands 12
2.4 Types of Constructed Wetlands 13
2.4.1 Free Water Surface Systems 14
2.4.2 Subsurface Flow (SSF) Systems 17
2.5 Design Factors of Constructed Wetlands 21
2.5.1 Wetland Plants 22
2.5.2 Substrate 28
2.5.3 Retention Time 29
2.5.4 Water Depth 29
2.5.5 Seasons of The Year 30
2.6 Pollutant Removal Mechanisms in Constructed Wetlands 31
2.6.1 Physical Mechanisms 32
2.6.2 Biological Mechanisms 33
2.6.3 Chemical Mechanisms 34
2.7 Removal of Pollutants in Constructed Wetlands 36
2.7.1 Removal of Organic Compounds 36
2.7.2 Suspended Solids 37
2.7.3 Removal of Nutrients 39
2.7.4 Fecal Coliform Removal 54
2.7 Reaction Kinetics 56
2.8 Application of Constructed Wetlands 59
2.8.1 Application of Constructed Wetlands for Domestic Wastewater Treatment 60
2.8.2 Application of Constructed Wetlands for Storm Water Treatment 62
2.8.3 Constructed Wetlands for Industrial Wastewater Treatment 63
2.8.4 Application of Constructed Wetlands for Agro-Industrial Wastewater Treatment 64
2.8.5 Constructed Wetlands for Leachate Treatment 65
2.8.6 Constructed Wetlands for Acid Mine Drainage Treatment 66
2.8.7 Constructed Wetlands for Agricultural Runoff Treatment 66
2.8.8 Constructed Wetlands for Pesticide Treatment 67
2.9 Studies on Performance of Constructed Wetlands in Ethiopia 69
CHAPTER THREE 70
MATERIALS AND METHODS 70
3.1 Description of the Study Area 70
3.2 Field Visit and Site Selection 71
3.3 Construction of The Wetland System 72
3.3.1 Determination of the Size of Wetland Cells 72
3.3.2 Hydraulic Retention Time of The Wetland Systems 73
3.3.3 Layout and Configuration of the Pilot-Scale Constructed Wetlands 73
3.3.4 Site Clearing and Excavation 75
3.3.5 Type of Filter Media Used 75
3.3.6 Plant Species Used in the Study and Planting Procedures 76
3.3.7 Installation of Inlet and Outlet Pipes 76
3.3.8 Lining of the Wetland Beds 77
3.3.9 Sedimentation Tank 78
3.4 Monitoring of the Constructed Wetland Systems 79
3.5 Sampling and Laboratory Analysis 79
3.6 Source of Meteorological Data 80
3.7 Data Analysis 80
3.8 Total Cost of Construction 81
CHAPTER FOUR 81
RESULT AND DISCUSSION 82
4.1 Characteristics of the Domestic Wastewater 82
4.2 Meteorological Data of the Study Area 85
4.3 Results of Performance Monitoring of the Constructed Wetlands 86
4.3.1 Temperature, pH, Electrical Conductivity and Dissolved Oxygen Levels 86
4.3.2 Removal of Biochemical Oxygen Demand (BOD5) 87
4.3.3 Removal of Chemical Oxygen Demand (COD) 93
4.3.4 Removal of Total Suspended Solids (TSS) 97
4.3.5 Removal of Ammonium (NH4+) 102
4.3.6 Removal of Nitrate (NO3-) 107
4.3.7 Removal of Total Nitrogen (TN) 112
4.3.8 Removal of Phosphate (PO43-) 117
4.3.9 Removal of Total Phosphorous (TP) 122
4.3.10 Removal of Fecal Coliforms (FCs) 127
4.4 Kinetic Parameters Determination 131
4.5 Plant Tissue Nutrient (N and P) Content 133
CHAPTER FIVE 137
CONCLUSION AND RECOMMENDATIONS 137
5.1 CONCLUSION 137
5.2 RECOMMENDATIONS 138
REFERENCES: 139
ANNEXES 159

List of Figures
Figures Pages
Figure 2.1: Free water surface flow constructed wetland 16
Figure 2.2: Subsurface flow constructed wetland 18
Figure 2.3: Nitrogen transformation in constructed wetlands 42
Figure 3.1 Sketch map of configuration of the pilot scale constructed wetland systems applied at Kotebe WWTP 75
Figure 4.1: Concentration of BOD5 of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW systems applied at Kotebe WWTP, Addis Ababa, Ethiopia. 94
Figure 4.2: Seasonal BOD5 removal efficiencies of the horizontal vertical and hybrid pilot scale CW systems employed at Kotebe WWPT, Addis Ababa, Ethiopia. 95
Figure 4.3: BOD5 loading rate (g/m2.d) against the BOD5 removal rate (g/m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 96
Figure 4.4: COD Concentration of the influent and effluents of the horizontal, vertical and hybrid pilot scale CWs employed at Kotebe WWTP, Addis Ababa, Ethiopia. 99
Figure 4.5 Seasonal COD removal efficiencies of the horizontal, vertical and hybrid pilot scale CWs employed at Kotebe WWTP, Addis Ababa, Ethiopia. 99
Figure 4.6: COD loading rate (g/m2.d) against the COD removal rate (g/m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 100
Figure 4.7: Concentration of TSS of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW system employed at Kotebe WWTP, Addis Ababa, Ethiopia. 104
Figure 4.8: Seasonal TSS removal efficiencies of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 104
Figure 4.9: TSS loading rate (g/m2.d) against the TSS removal rate (g/m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 105
Figure 4.10: Concentration of NH4+ of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW sytems employed at Kotebe WWTP, Addis ababa, Ethiopia. 109
Figure 4.11: Seasonal NH4+ removal efficiencies of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 109
Figure 4.12: NH4+ loading rate (g/m2.d) against the NH4+ removal rate (g/m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 110
Figure 4.13: Concentration of NO3- of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 114
Figure 4.14: Seasonal NO3- removal efficiencies of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 114
Figure 4.15: NO3- loading rate (g/m2.d) against the NO3- removal rate (g/m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 115
Figure 4.16: Concentration of TN of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 119
Figure 4.17: Seasonal TN removal efficiencies of the horizontal, vertical and hybrid CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 119
Figure 4.18: TN loading rate (g/m2.d) against the TN removal rate (g.m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 120
Figure 4.19: Concentration of PO43- of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 124
Figure 4.20: Seasonal PO43- removal efficiencies of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 124
Figure 4.21: PO43- loading rate (g/m2.d) against the PO43- removal rate (g/m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 125
Figure 4.22: Concentration of TP of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 128
Figure 4.23: Seasonal TP removal efficiencies of the horizontal, vertical and hybrid pilot scale CW systems employed in Addis Ababa, Ethiopia. 129
Figure 4.24: TP loading rate (g/m2.d) against the TP removal rate (g/m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 130
Figure 4.25: Concentration of FC of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 133
Figure 4.26: Seasonal FC removal efficiencies of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 133
Figure 4.27: FC loading rate against FC removal rate of the horizontal, vertical and hybrid flow pilot scale CW systems applied at Kotebe WWTP, Addis Ababa, Ethiopia. 134
Figure 4.28: Percentage of TN in plant tissue of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 139
Figure 4.29: Percentage of TP in plant tissue of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 139

List of Tables
Tables Pages
Table 2.1 the role of plants (macrophytes) used in constructed wetlands 24
Table 2.2: Recommended emergent plant species for constructed wetlands 27
Table 2.3: Summary of pollutant removal mechanisms in constructed wetlands 36
Table 2.4: Kadlec and Knight K-C* model design parameters 59
Table 2.5: Application of constructed wetlands for the treatment of different wastewater 60
Table 3.1: Summary of the characteristics of the pilot scale constructed wetland systems applied at Kotebe WWTP, Addis Ababa, Ethiopia. 79
Table 3.2: Summary of the laboratory methods and instruments used to measure wastewater parameters during the monitoring period. 80
Table 3.3: Cost summary of the construction of the pilot scale constructed wetland systems applied at Kotebe WWTP, Addis Ababa, Ethiopia. 81
Table 4.1: Raw domestic wastewater characteristics before primary treatment, Addis Ababa-Ethiopia. 82
Table 4.2: Domestic wastewater characteristics after primary treatment, Addis Ababa, Ethiopia. 82
Table 4.3: Domestic wastewater characteristics of some other countries 84
Table 4.4: Daily average data of rainfall and ambient air temperature for the sampling dates. 89
Table 4.5: Mean influent values of temperature, pH, electric conductivity and dissolved oxygen of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 90
Table 4.6: Average values of BOD5 concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 94
Table 4.7: Average values of COD concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems applied at Kotebe WWTP, Addis Ababa, Ethiopia. 98
Table 4.8: Average values of TSS concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 103
Table 4.9: Average values of NH4+ concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 108
Table 4.10: Average values of NO3- concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 113
Table 4.11: Average values of TN concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 118
Table 4.12: Average values of PO43- concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 123
Table 4.13: Average values of TP concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia. 128
Table 4.14: Average values of FC concentration (CFU/100ml) and removal efficiencies (%) of the pilot scale CW systems applied at Kotebe WWTP, Addis Ababa, Ethiopia. 132
Table 4.15: Summary of areal removal rate constants, K (m/d) of the measured parameters considered in this study. 135
Table 4.16: Average values of plant tissue nutrient content (%) of the pilot scale CW systems applied at Kotebe WWTP, Addis Ababa, Ethiopia. 138

List of Abbreviations
AAU Addis Ababa University
BOD5 5-day Biological Oxygen Demand
COD Chemical Oxygen damand
CSA Central Statistics Agency
CWs Constructed Wetlands
ECA Economic Commission for Africa
EEC European Economic Community
EPA Environmental Protection Agency
EU European Union
FDRE Federal Democratic Republic of Ethiopia
FWSCW Free Water Surface Constructed Wetlands
GEPD Georgia Environmental Protection Division
GWP Global Water Partnership
HFB Horizontal Flow Bed
HSSF Horizontal Subsurface Flow
ICCIWMEE International Conference on Chemical, Integrated Waste Management and
Environmental Engineering
MoE Ministry of Education
N Nitrogen
Ortho-P Ortho-phosphate
P Phosphorous
PE Population Equivalents
RC Ramsar Convention
SS Suspended Solids
SSFB Subsurface Flow Bed
SSFCW Subsurface Flow Constructed Wetland
TSS Total Suspended Solids
UN United Nations
UNESCO United Nations Education Science and Culture Organization
UNISA University of South Africa
USCS United States Commercial Service
VSSF Vertical Subsurface Flow
TSS Total Suspended Solids
WI Wetland International
WRI World Resource Institute

CHAPTER ONE
INTRODUCTION
General Information
Wastewater production inevitably exists following the consumption of water by human being for different purposes. Currently, the volume of wastewater production is considerably large and of course it continues to increase and get higher from time to time as a result of growing of population size, unintended rapid urbanization and industrialization and improvement of people’s living standards (Wallace, 2004; Mekala, et al, 2008; Corcoran et al, 2010). As indicated by UNESCO (2012), the estimation of domestic wastewater generation in urban areas lies in the range between 150 and 250 m3 per day, which can have the potential to create a serious concern these days.
In spite of this fact, the existing situation with regard to sanitation coverage enlightens serious problems. Of the total wastewater volume produced globally, more than 80% of the wastewater does not receive any level of treatment and it is being directly discharged into water bodies or leaches into the surface of the Earth. The practice of wastewater treatment in developing countries is very minimal, and therefore the fraction of wastewater which is being discharged without treatment is estimated to be 90% or more. The situation in Africa is perhaps worse in this regard (Corcoran et al, 2010; Biswas, 2010; UNESCO 2012).
At present, there may be infrastructures or collection systems in urban areas which can be used to collect wastewater even though water bodies such as lakes, rivers and oceans are being highly polluted due to the practice of discharging untreated or partially treated wastewater. Thus, a grave water crisis may be resulted in most developing countries from this phenomenon if the trend is not reversed in the future (Biswas, 2010).
Domestic wastewater is mainly a mixture of human feces, urine and gray water (Charles and Ian, 2009), and the major constituents of typical wastewater that are considered as pollutants of concern include: oxygen-demanding materials, pathogens, nutrients, suspended solids, sediments, grease/oil, heavy metals and other hazardous materials (Manahan, 2000; Kayombo et al, 2003). The throwing away of contaminants with untreated domestic wastewater from urban areas into water bodies is causing public health problem and environmental pollution (John, 2009; Biswas, 2010; Kurniadie, 2011; Saravanan, Peter and Janos, 2011). The question of proper wastewater treatment using effective treatment methods is widely promoted in recent times because of river quality deterioration, especially those rivers found nearby water sources intended to be used for water supply (Katayon, 2008; Biswas et al. 2008; UN-Habitat, 2008; Katayon, 2008; John, 2009; GWP, 2009).
Wastewater treatment is usually carried out primarily to remove contaminants that may have negative impacts on public health and on the environment (Thomas and William, 2001). Nevertheless, water reclamation is another opportunity standing for giving solution by exploiting new sources of water, particularly in areas where there is shortage or lack of fresh water (Scheierling et al, 2011; Lavrnic and Mancini, 2016).
Reclaimed water can be used for non-potable purposes such as cooling, irrigation, toilet and urinal flushing, ground water recharge, fire protection and construction. Water reclamation provides economical, environmental, and health-related benefits among others. In view of that, the EU water framework directive disseminated in 2000 highly influences the intention of wastewater reclamation and reuse in the member countries of the European Union (WHO/UNEP, 1997; Shutes, 2001; GWP, 2009; Redder et al, 2010; Murray and Ray, 2010; Mustafa, 2013).
Since wastewater production is characterized by less variability with different seasons, it is a dependable source for water and nutrients. Wastewater reuse for irrigation is prevalent worldwide and it accounts 10% of the total irrigated surface. It is used by households practicing agriculture in and around the urban areas and employed to improve agricultural productivity and food security. The livelihoods of low income households in urban areas depend on wastewater reuse. Countries like Jordan, Chile and Israel are well known for their successful achievement in wastewater reclamation and the need for continuous improvement over many years is learned from the experiences of these countries in order to achieve safe wastewater irrigation. There are 3,000 water reclamation sites identified worldwide even though the figures can be quickly obsolete (Bixio, 2005; Jimenez, 2006; Mara et al, 2007; Murray and Ray, 2010; Scheierling et al, 2011; Wu et al, 2015;).
Then, to resolve the public health and environmental problems associated with overthrowing of discharging of untreated wastewater and to reclaim and use the enormous amount of wastewater, appropriate treatment technologies must be evaluated and applied. The developed countries in Europe, North America, and Australia have applied different treatment technologies to treat various types of wastewater at the desired level before being discharged to the environment. For instance, United States alone treat nearly 150 billion liters of wastewater per day by using more than 15,000 wastewater treatment plants. Treating of wastewater is obligatory to comply with the stringent discharge standards set by each country (Mekala et al. 2008; Charles and Ian, 2009).
Many of the developing nations of Africa and Asia have not been able to treat their wastewater to the desired levels and this is attributed mainly to lack of adequate funds, high treatment costs of the conventional treatment systems, lack of skilled manpower and fast escalating of wastewater volumes (Nhapi and Gijzen, 2005; Schertenleib, 2005; Mekala et al, 2008).
On the other hand governments in developing countries have more pressing needs such as dealing with war and conflicts, health care and food security. Hence, wastewater treatment is usually low in the list of priorities (Massoud, Tarhini, and Nasr, 2008). Consequently, they are discharging untreated domestic wastewater into the nearest water bodies or leaching into the environment despite of the negative impact that can create to the environment. For instance, in China, the amount of wastewater discharged to the environment was 65 billion tons by the year 2011. The problem is expected to get worse unless appropriate measures are taken to control and treat effluents (John, 2003; UN-Habitat, 2008, USCS, 2013).
In Africa, South Africa has shown remarkable development by constructing infrastructures for wastewater treatment that bring the country at the front line in the continent although informal settlement around urban areas poses significant challenges. The country has about 900 wastewater treatment plants with the capacity of treating 5,000,000 to 7,000,000 m3 of wastewater each day (UN-HABITAT, 2008). However; a large number of people are without basic sanitation facilities in countries found in sub-Saharan Africa. The gap between many African countries and the developed countries such as North America, Europe, Australia and Japan with regard to wastewater management is implausible. The later allocate and utilize huge fund to protect water bodies and the environment from pollution while morbidity and mortality are the common features for millions in the aforementioned poorer countries because of absence of basic sanitation (Abdel-Halim and Rosenwinkel, 2005; UN-HABITAT, 2008).
The problem in Ethiopia has no special features comparing to other developing countries. Except a very limited effort made by Addis Ababa Water and Sewerage Authority (AAWSA) and some other institutions to treat domestic wastewater, almost all towns throughout the country have not yet utilized any of the natural or modern mechanical technologies that are being used to treat domestic wastewater. Wastewater stabilization ponds have been employed in Addis Ababa since the end of 1970s. However, the proportion of domestic wastewater treated by the plant is not greater than 5% of the total volume of domestic wastewater generated in the city (Getahun and Adinew, 1999). Besides, three more treatment plants using stabilization ponds were constructed during a few years back. Alebel et al, (2009) revealed that in Ethiopia, both domestic and industrial wastewater is discharged mostly untreated into the nearby rivers, which are used as sources for multiple purposes including domestic uses and irrigation. Smallholder farmers in and around the cities have used the untreated wastewater for crop production.
Even if, conventional wastewater treatment methods demand high cost and considerable energy, they were popular until recent times and used to treat domestic wastewater for long period of time. However, the demand of high cost and energy by modern mechanical treatment methods becomes a ringing bell to seek other alternatives that can fill the gaps. Then constructed wetlands, the biological treatment technologies, are standing as effective treatment methods and are being applied for the treatment of various types of wastewater since their recognition in the 1960s. Constructed wetlands are the simple, low cost, long-lived and an eco technology wastewater treatment system that can be widely applied in developing countries. The application of constructed wetlands is particularly attractive in regions with water shortage since effluent from the treatment plant can be used for irrigation (Belmont et al, 2004; Chen et al, 2006; Trivedy and Siddharth, 2010; Azni et al, 2010; Kurniadie, 2011; Zapater-Pereyra et al, 2013; Prashant et al, 2013; Amaral et al, 2013; Dimuro et al, 2014; Wu et al, 2015; Deeptha, Sudarsan and Baskar, 2014; Vergeles et al, 2015).
Constructed wetlands (CWs) system is one of the natural treatment systems that rely on natural processes utilizing systems composed of plants, support medium and microorganisms, which are cost-efficient and present good operational steadiness. Due to their ease of construction and operation, they are a viable alternative for developing countries with tropical climates (Sarmento, Borges and Matos, 2012). In view of this, developing countries are realizing that the application of natural systems instead of energy demanding conventional technology for wastewater treatment has a number of advantages (Shutes, 2001). Properly designed and operated constructed wetlands could also be used for secondary and tertiary wastewater treatment (Korkusuz, Beklioglu and Demirer, 2004). Study results indicated that phytoremediation is not only a feasible environmental remediation alternative but also presents numerous related advantages (Erakhrumen, 2007).
Both natural and constructed systems offer potentially cheaper and low-energy treatment alternatives to treat wastewaters. However; constructed wetlands are able to attract more attention in recent works because of the regulatory requirements to the discharged effluent and the treatment processes can proceed under more controlled manner (Haberl et al, 2003; Robert, 2004; Rousseau, 2008; Charles and Ian, 2009; Vymazal, 2009; Trivedy and Siddharth, 2010; Van, 2010; Siti, 2011). Moreover, CWs can be successfully used for secondary and tertiary treatment if the design and operation condition is done in the right way (Mustafa, 2013). CWs can be defined as:
‘a natural, low-cost, eco-technological biological wastewater treatment technology designed to mimic processes found in natural wetland ecosystems, which is now standing as the potential alternative or supplementary systems for the treatment of wastewater’ (UN-HABITAT, 2008: p.3).
CWs were recognized as an appropriate alternative treatment technology and they were initially used for treating municipal or domestic wastewater. But, at present, their potential of treating wastewater from various sources leads to the rapidly expansion and applicability of the technology in almost every country (Robert, 2004; Vymazal, 2009; Vymazal, 2014). Now, wetlands have been used for the treatment of virtually all types of wastewaters including municipal, industrial, agricultural, acid mining drainage, animal, and leachate as well as storm water. Pollutants found in wastewater can be converted in-to harmless byproducts or essential nutrients for biological productivity by applying CWs (Haberl, Perfler and Mayer, 1995; Thomas and William, 2001; Kayombo et al. 2003; Liu et al. 2008; Robert and Scott, 2009; Azni et al. 2010; Vymazal, 2010).
The extent of its applicability ranges from small scale level to serve single household or institution to large-scale centralized municipal systems. To effectively use the system, there are different configurations, scales, and designs and the efforts to optimize the performance is still in progress in every corner of the world. From the previous trend, the free water surface system is extensive in North America while the subsurface flow system is predominantly used in Europe. As different types of constructed wetlands are employed for primary, secondary or tertiary treatment level, the right design or configuration should be chosen and applied based on the treatment objective. The right design and configuration and proper operation help to overcome the influences that may occur by differences of environmental factors (Kayombo et al. 2003; UN-HABITAT, 2008; Kadlec and Scott, 2009; Azni et al. 2010; Van, 2010).
Following the recognition of purification capability of wetlands in the 1950s, they are used extensively in developed countries. The adoption of wetlands technology for wastewater treatment in developing countries has been carried out at very gradual rate. In fact, Jan and Lenk (2008) revealed that ‘the planned use of wetlands for meeting wastewater treatment and water quality objectives has been seriously studied and implemented in a controlled manner merely during the past few decades’. The fundamental processes that are carried out in constructed wetlands have not yet adequately known even after years of experiences, practical application and a number of studies. Because, compared to other treatment technologies, CWs depends on the interaction of various components in it (Stottmeister et al, 2009). Undeniably, however, CWs have gained increasing attention all over the world these days (Liu et al, 2008; Russo, 2008; UN-HABITAT, 2008; Zhang et al, 2015).
Meanwhile, a number of studies related to the performance of pilot-scale or full-scale constructed wetlands have been carried out in many African countries such as South Africa (Schulz et al, 2003), Egypt (Hussein and Ahmed, 2012; Abou-Elela and Hellal, 2012), Tunisia (Abidi et al, 2009), and Nigeria (Erakhrumen and Agbontanor, 2007). Similar studies have been conducted in the Eastern African countries; e.g. in Uganda (Kyambadde et al. 2004), Kenya (Kelvin and Tole, 2010; Odinga, Otieno and Adeyemo, 20011) and Tanzania (Mashauri, 2000; Mairi et al, 2012). Most of the studies were conducted in Tanzania, Egypt and Kenya.
In Ethiopia, the full-scale application of this technology for wastewater treatment is restricted only to a few institutions and some studies related to the performance of CWs have been conducted in recent times (Birhanu, 2007; Asaye, 2009; Tadesse, 2010; Kenatu 2011).

1.2 Objective of the Study
1.2.1 General Objective
The general objective of the study was to evaluate the performance, and basic design and operation criteria of pilot-scale constructed wetlands for the treatment of domestic wastewater. In addition to this, the research aimed to quantify the effect of different seasons of the year on the treatment performance of constructed wetlands based on the major wastewater parameters that are taken into consideration in this study. It was also aimed to determine the performance difference of horizontal, vertical and hybrid subsurface flow constructed wetlands used similarly in the prevailing climatic conditions in Addis Ababa, Ethiopia.
1.2.2 Specific objectives
The specific objectives of the study were:
To evaluate the performance of pilot scale constructed wetlands in treating domestic wastewater;
To compare the performance of horizontal flow, vertical flow and hybrid of the horizontal and vertical flow constructed wetlands;
To evaluate and compare the effect of different seasons of the year on the performance of the pilot scale constructed wetlands;
To determine the areal removal rate constants of wastewater pollutants and compare those figures with the literature values;
To determine the amount of nitrogen and phosphorous in the root (below-ground plant part) and the stem (above-ground plant part) of the wetland plants/Cyprus papyrus after the operation period of one year; and
To forward appropriate recommendations for further studies based on the results of the study.

1.3 Research Questions
In consideration of attaining the stated objectives of the study, the following are the research questions on which the study focused:
How efficient could constructed wetlands be in treating domestinc wastewater?
Could the different wastewater flow types have effects on constructed wetlands applied for the treatment of domestic wastewater under similar environmental conditions?
Would the performance of constructued wetland systems have different values at various seasons of the year?
Would the values of areal removal rate constants of wastewater pollutants be different from literature values?
Could the different wetland plant parts have different nitrogen and phosphorous content?

1.4 Scope of the Study
The existing situations with regard to wastewater production, the major causes for increasing the volume of wastewater and the possible negative impacts that can be created by discharging untreated wastewater to the nearby environment were discussed in the introduction part. The aim of this chapter was to highlight the major challenges our environment is facing as a result of improper management of wastewater; and the natural treatment methods such as constructed wetlands which are standing as credible alternatives to address the challenges in the case of developing countries. The disparity of the magnitude of the burden on the developed and developing countries was also shown.
In the literature review part, the most important points about constructed wetlands are discussed. The basic concepts concerning the types, functions, values, removal mechanisms and their effectiveness in removing wastewater pollutants of previous studies are presented in detail.
In the materials and methods part, basic information about the study area is presented and the approaches followed to design the pilot-scale constructed wetlands are illustrated in brief and the materials used for the construction are listed out. The wastewater parameters which were identified to evaluate the performance of the wetland system are presented in this chapter. Moreover, the methods of laboratory analysis were indicated for all parameters.
In the results and discussion part, the results of the laboratory analysis conducted during the monitoring period, between November 2015 to November 2016, to evaluate the performance of the pilot scale constructed wetlands of Addis Ababa are presented and discussed in the chapter. The data obtained from the laboratory analysis of the determination of nitrogen and phosphorous removal by wetland plants uptake are indicated and discussed.
Finally, the conclusion is made based on the results of the study and the recommendations forwarded for further studies are described in the last chapter. The knowledge and experiences learnt at the time of the implementation and monitoring of operation of the three pilot scale constructed wetlands of at Kotebe WWTP in Addis Ababa are also included in this chapter.

?
CHAPTER TWO
LITERATURE REVIEW
2.1 Basic Information about Wetlands
Natural wetlands commonly represent the areas which are found between land and water bodies. They are intermediary areas linking the two surfaces and have been recognized as natural resource throughout the history of mankind. The submissive nature of wetlands arising from different factors makes almost impossible to give one general definition. So that, they can be defined in several ways depending on those factors which include: personal perspective, existing water and plant condition, landscape position/geographic setting, and wetland diversity and function (Thomas and William, 2001; Scholz, 2006; Kadlec and Wallace, 2009). Although there are a variety of ways to define the wetland system, the most widely agreed definition was formulated by the International Union for the Conservation of Nature and Natural Resources (IUCN) in the Ramsar Convention, in 1980. According to the convention, wetlands are defined as:
‘areas of marsh, fen, peatland or water, whether natural or artificial, permanent or temporary, with water that is static or flowing, fresh, brackish or salt, including areas of marine water the depth of which at low tide does not exceed six meters.’
Article 1.1
Wetlands are well known for giving magnificent services as biological filters to protect water resources of both surface and groundwater. Natural wetlands have acted as ecological buffer all the time to protect the environment. But conducting researches and the advancement of using wetland treatment technology for treating various wastewaters is relatively a new incident started in the early 1950s, in Germany. In the United States, researches on wetlands began in the late 1960s and extended in scope during the 1970s. Following this, treating wastewater using wetlands was emerged as an appropriate alternative technology globally (Thomas and William, 2001).
Wetlands have distinctive characteristics which make them different from major ecosystems plainly known (Kadlec and Wallace, 2009). As nearly all forms of biological productivity highly depends on the amount of water available, water is the most important factor affecting the wetland environment and the associated forms of life (RC, 2013). Ramsar convention (2013) pointed out that:
‘Wetlands are among the world’s most productive environments. They are cradles of biological diversity, providing the water and primary productivity upon which countless species of plants and animals depend for survival. They support high concentrations of birds, mammals, reptiles, amphibians, fish and invertebrate species. Wetlands are also known to be important storehouses of plant genetic material.’
(Ramsar convention, 2013: p.8).
In view of this, wetlands are recognized for giving protection for the environment especially water bodies by removing pollutants in discharged wastewater ranging from rainfall runoff to strong wastewater such as community sewage over many years (Robert, 2004). They have been employed as convenient wastewater discharge sites for as long as sewage has been collected, at least 100 years in some locations (Kadlec and Wallace, 2009), so that they have been receiving increasing attention as effective alternatives for wastewater treatment (Charles and Ian, 2009). In this regard, bacterial metabolism and physical sedimentation are the two major processes acknowledged for the performance of wetlands in treating wastewater (Kadlec and Wallace, 2009).
Furthermore, immense and highly indispensable services for human well-being and poverty reduction are offered by wetland ecosystems (Kent, 2001; WRI, 2005; Vymazal, 2010). Some of the most important ecosystem services include: aquatic and wildlife habitat (Forman and Godron, 1986; Kent, 1994; Vymazal, 2010; Si et al, 2014), educational and scientific venue (Drake and Vicario, 1994), flood flow alteration (Vymazal, 2010), groundwater recharge, elemental transformation, particle retention and sources of raw materials, recreation, and soil stabilization (Kent, 2001; WRI, 2005).
2.2 Constructed Wetlands as an Attractive Technology for WWT; An Overview
Constructed wetlands (CWs) are artificial or engineered wetlands designed and constructed to reproduce and improve the processes of wastewater treatment which take place in natural wetlands (Mara, 2003). They are basins having low depth, filled more often with sand or gravel as filter media and planted with aquatic plants. The wastewater to be treated is conveyed through the inlet pipes into the basins and flows through the substrate (media) or over the surface, and finally the effluent is discharged out of the system through the outlet pipes that are designed to keep the depth of the wastewater in the basin. Azni et al, (2010) pointed out that CWs can be created from existing marshlands or built at any land with limited alternative uses. The most important sections of a constructed wetland include; Basin, Substrate, Vegetation, liner, inlet/outlet pipes (UN-HABITAT, 2008). Constructed wetlands have been successfully used as treatment systems for domestic wastewater effluent, from single-residence (Gikas and Tsihrintzis, 2010) wetlands to large municipal wastewater treatment facilities (Kent, 2001).
Seidel (1953) as cited in Kadlec and Wallace (2009) was the person to begin experimenting aquatic plants to improve water quality, and she is acknowledged for the development of constructed wetlands in Europe. CWs can function without any electromechanical devices such as aerators to supply oxygen and therefore they are usually called natural treatment systems. Wetland plants are the main sources of oxygen for the oxidation of organic pollutants by the heterotrophic bacteria (Trivedy and Sidharth, 2010).
Rousseau (2008) pointed out that if constructed wetlands are designed and maintained carefully, they can yield an effluent which meets reuse requirements and concurrently provide some opportunities to recycle nutrients and to accommodate wildlife. Advances in the innovative design and operation of CWs have greatly increased contaminant removal efficiencies, thereby improving the sustainable applications of these treatment systems. For treatment wetlands, optimization and innovation in design, operation, and maintenance, could help to make this treatment technology much more attractive (Wu et al, 2015).
Based on energy synthesis, CWs are found to be less energy-intensive with relative low Ecological Waste Removal Efficiency (EWRE), and environmental friendly and cheaper than Cyclic Activated Sludge Systems (CASS) either in construction or during operation and maintenance, implying that CWs still gain some advantages over the conventional treatment plants, especially in the developing countries (Zhou et al, 2007; Vymazal, 2010). Amaral et al (2013) revealed that CWs have gained attention of the present-day for alternative wastewater treatment method since they are simple to operate and maintain and require low resource consumptions.
Although, developing countries which are in need of such effective treatment methods do not give adequate attention (Kivaisi, 2001), the application of CWs as wastewater treatment technology has been rapidly grown in many parts of the world starting from 1985. This is because, for one thing, the construction of the technology can be made by using human labor and locally available materials, providing low cost and low maintenance but high technology alternative for developing countries (Hench et al, 2003; Konnerup, Koottatep and Brix, 2009). Secondly, wetland systems are able to achieve high treatment performance as a result of complex hydrological and biological processes which take place in removing contaminants even though they are mechanically simple (Merlin, Pajean and Lisollo, 2002; Kadlec and Wallace, 2009).
In addition to their potential as effective treatment technology, CWs can also provide other ecosystem services such as conservation of biodiversity. Therefore, in view of sustainability, they are considered as a good sanitation solution with great potential of protecting natural resources and the environment. Solar energy is the only energy source on which the system depends for pollutant removal processes and hence this ensures its sustainability (Merlin and Lisollo, 2010). However, institutional limitations, relatively large land requirement and less public awareness are expected to be the major ongoing challenges that possibly avert the wider application of CWs (Liu et al, 2008; Wendong et al, 2014; Langergraber, 2014).
The performance of CWs in tropical areas, where most of the developing countries are located (Kivaisi, 2001) and extreme cold temperature is absent, is satisfactory and it has more or less steady performance throughout the year. There is rapid ecological succession with tropical biodiversity mix in those areas where the temperature is high, so that, treatment of domestic wastewater can be achieved to an acceptable standard by natural processes. Greater treatment efficiencies as a result of the complex natural processes can be achieved by CWs than widely used waste stabilization ponds (Kelvin and Tole, 2011).
Zurita, Anda and Belmont (2009) described that the type of flow in CW system is the most important factor that affects the rate of removing wastewater contaminants. Higher removal rate of almost all pollutants except NO3-, TN and TSS can be achieved by using Vertical Flow Constructed Wetlands (VFCWs). Because there is better aeration/oxygen supply/ that enables to enhance wastewater nitrification in VFCWs.
2.3 Advantages of Constructed Wetlands
Nowadays, CWs are able to replace conventional wastewater treatment methods, and treat not only domestic wastewater but also other wastewaters from various sources. They are designed and constructed to take the advantage that virtually all processes occur naturally in a more controlled environment (Haberl et al, 2003; Anna et al, 2008; Azni, 2010). Hence, CWs have a number of advantages including: they have a long life time with minimum maintenance requirement, they can treat broad spectrum of contaminants in wastewater simultaneously to an acceptable level, and they can play a major role in increasing biodiversity and then to provide ecosystem services in themselves and the surrounding environment (Habler et al, 2003; Siracusa and Rosa, 2006; Azni et al, 2010).
According to (Kayombo et al, 2003; Russo, 2008; Massoud, Tarhini and Nasr, 2009), the advantages of CWs are summarized as follows;
can often be less expensive to build than other treatment options,
can be built and operated simply,
utilize natural processes,
their operation and maintenance expenses (energy and supplies) are low,
are able to tolerate fluctuations in flow and pollutant concentration,
are able to treat wastewaters with very different constituents and concentration,
are characterized by a high process stability (buffering capacity),
are characterized by low excess sludge production,
Facilitate water reuse and recycling, and
Provide other indirect benefits such as green space, wildlife habitats and recreational and educational areas.
2.4 Types of Constructed Wetlands
Constructed wetlands are designed and constructed in various ways according to the theoretical basis of earlier studies to take the advantage of many complex processes that occur naturally by microbial community, vegetation, substrate and wastewater (USEPA, 2000; UN-HABITAT, 2008; Vymazal, 2008). The present wetlands are designed to employ specific characteristics for improved treatment capacity (Kadlec and Wallace, 2009). Haberl (1999) as cited in UN-habitat (2008) described that CWs have diverse design configurations and the basis for their classification comprise of:
Life form of the dominating macrophytes (free-floating, emergent, submerged),
Flow pattern in the wetland systems (FWS flow; SS flow: horizontal and vertical),
Type of configurations of wetland cells (hybrid systems, one-stage, multi-stage systems),
Type of wastewater to be treated,
Treatment level of wastewater (primary, secondary or tertiary),
Type of pretreatment,
Influent and effluent structures,
Type of substrate (gravel, soil, sand, etc.), and
Type of loading (continuous or intermittent loading).
But the type of water flow is the most important factor among others that are used as the basis for the classification of constructed wetlands (Vymazal, 2008). Although constructed wetlands have a lot of characteristics in common, based on the water flow type, they are in practice classified into two general types: Free Water Surface (FWS) wetlands (also called surface flow (SF) wetlands) and Subsurface Flow (SSF) wetlands (also known as Vegetated Submerged Bed (VSB) systems) (Kayombo et al. 2003; Russo, 2008; Kadlec and Wallace, 2009).
The water flows above the ground and it is exposed to the atmosphere in free water systems. However, in subsurface flow systems, the water usually flows through a bed made with a porous media such as sand, gravel or aggregates and hence, the water is not exposed to the atmosphere (USEPA, 2000; Kayombo et al. 2003; Russo, 2008; Kadlec and Wallace, 2009). Concerning their application distribution, Cole (1998) as cited in Thomas and William (2001) described that subsurface flow wetlands are the common systems in Europe for treating domestic wastewater while the free water systems types are widely used in North America.
The selection of either of the system during the design phase should focus on the required quality of the effluent coming out of the wetland system. For instance, in areas where there is enough space and when the removal of merely SS and BOD is needed, SFSs with one-unit SSF type can be adequate. But multistage or combined systems having both horizontal flow (HF) types and vertical flow (VF) types should be used in areas where there is more stringent discharge limits. On the other hand, phosphorous adsorption capacity and hydraulic conductivity are the two major criteria used in the selection of the medium used in SSF constructed wetlands. As multistage systems contain more treatment beds, they are more expensive than one-unit system. But they are still notably cheaper than the conventional treatment technologies (Brix, 1995).
2.4.1 Free Water Surface systems
Free water surface (FWS) constructed wetlands are areas of open water systems and there is no any barrier between the water surface and the surrounding atmosphere. The appearance and function of these systems closely resemble natural wetlands in that: they have open-water which is in contact with the atmosphere, emergent vegetation, varying water depths, and other typical wetland features. They are designed in such a way that the processes taken place in natural wetlands can be used effectively in the treatment of wastewater (USEPA, 2000; Charles and Ian, 2009).
The common features of FWS include; a basin with impermeable bottom, inlet structures, open-water areas with vegetation, and outlet structures. The size, shape, and complexity of the system design usually attributed to the characteristics of the site than the perceived criteria prior to its construction (USEPA, 2000; Thomas and William, 2001; Thullen, Sartoris and Walson, 2002; Charles and Ian, 2009; Azni et al, 2010). Free water surface wetlands are distinguished by a relatively shallow layer (usually ranges between 0.3 to 0.4m) of surface water flowing over the impermeable soil with low flow velocity. Macrophytes play a major role in regulating the flow of water in a long basin which helps to maintain the plug-flow condition in the system (Beharrell, 2004; Polprasert, 2004).
These types of wetlands are chosen in many parts of the world since they are cheaper than other treatment methods including SSF constructed wetlands, and beyond this the values of wetland habitat and reuse opportunities are highly associated with FWS systems (USEPA, 1993; Beharrell, 2004; Mohammadpour et al, 2014; Han et al, 2014). A wide-range of wildlife including insects, amphibians, birds, reptiles and mammals are greatly attracted by FWS wetlands (Kadlec and Knight, 1996). FWS systems are extensively used in North America typically for treating large flow municipal wastewater. The predominant wetland type is FWS and they are applied at larger sizes even if wetlands of smaller sizes are used in some localities (Polprasert, 2004; Beharrell, 2004).
Every part of the world including the coldest northern hemisphere can be suitable site to employ FWS wetlands for wastewater treatment. However, ice formation can be occurred in extreme cold conditions and the ice can cover the water surface, which results in decreasing of the transfer of oxygen from the atmosphere to the wetland system. This situation has a negative impact on some removal processes. For example the rate of conversion of nitrogen decreases in this kind of situation. Conversely, TSS removal efficiency increases under ice comparing to summer season. In general, the removal efficiency is higher in warm seasons, so that collecting and storing wastewater at cold seasons and treating during the warm seasons is the best approach (Kadlec and Wallace, 2009).
Flocculation and sedimentation are the two most important processes in FWS wetlands which are known for their key role in removing pollutants in wastewater while the wastewater flows through stands of wetland vegetation. Sometimes, the physical pollutant removal processes in some FWS systems can be complemented by aerobic bio-oxidation process (US EPA, 2000). So, the retention time of the wastewater to be treated in a FSW constructed wetland is the most important factor in evaluating the effectiveness of the purification capacity of the system (Su et al, 2009).
In FWS system, there is high probability of human exposure to disease causing micro-organisms present in wastewater. As a result, the system is not commonly considered as a good alternative for secondary treatment (USEPA, 2000; Kadlec and Wallace, 2009). But effluents coming out of other treatment technologies such as lagoons, trickling filters, and activated sludge can be further polished by FWS systems (Kadlec and Wallace, 2009).
As reported by Siti et al (2011) the SFCW system are not reliable treatment method for treating wastewater with high ammonium concentration particularly in situations where there is short retention time. Therefore, to increase oxygen concentration within the system, and then to enhance the capacity of ammonia removal, intermittent or batch flow type to alternating basin should be encouraged. Thullen, Sartoris and Walson, (2002) noted that the presence of macrophytes in FWS systems improves the performance in treating concentrated ammonia in secondary-treated effluent.

Figure 2.1: Free water surface flow constructed wetland (Russo, 2008; pp. 329)
The FWS system is further divided into three sub-groups based on the various vegetation types dominantly grown in the wetland system (Brix, 1993; USEPA and USDA-NRCS, 1995; Vymazal et al, 1998b; USEPA, 2000; Russo, 2008).
A floating macrophyte system – these systems make use both of floating species that are rooted in the substrate (e.g. Nymphae spp Nuphar spp. (waterlilies), Potamog etonnatans (pondweed), Hydrocotyle vulgaris (pennyworth)) and species which are free floating on water surface (e.g. Eichhornia crassipes (water hyacinth), Lemna spp., Spirodella spp. (duckweed));
A submerged macrophyte system – the plants used in these systems have their photosynthetic tissue entirely submerged with the flowers being exposed to the atmosphere. Two types of submerged aquatics are usually recognized: the elodeid type (e.g. Elodea spp., Myriophyllum aquaticum (parrot feather), Ceratophyllum spp.) and the isoetid (rosette) type (e.g. Isoetes, Littorella, Lobelia);
A rooted emergent macrophyte system – these systems use plants which are the dominating form of life in the natural wetlands. Plants grow at well above the water level, producing aerial stems and an extensive root and rhizome system. These comprise species like the Phragmites australis (common reed), Thypha spp. (cattails), Scirpus spp. (bulrushes), Iris spp. (blue and yellow flags) Juncus spp. (rush), Saggitaria latifolia (duck potato), Phalaris arundinocea (reed canary grass), Carex spp. (Sedges), Zizania aquatica (wild rice), Eleocharis spp. (Spikerushes) and Glyceria spp. (mannagrasses).
2.4.2 Subsurface Flow (SSF) systems
The other wetlands type is subsurface flow (SSF) system, which is designed to create subsurface water flow and to keep the wastewater to be treated below the surface of the bed. These systems are supportive to avoid bad odor and other nuisance condition that may cause disease incidence as a result of pathogenic microorganisms found in wastewater. The system commonly uses gravel, aggregates, sand or soil as a porous media, on which the macrophytes are rooted and grown (USEPA, 1993; Kayombo et al, 2003). SSF systems are usually planted with emergent wetland vegetation (Cooper et al, 1996; Vymazal et al, 1998a; Russo, 2008).
In SSF wetland, the bed is filled with a porous media and the depth of the media is about 0.3 to 0.6m deep and the bottom is covered with geosynthetic impermeable layer to prevent underground infiltration or seepage. The bed should have 1% slope at the bottom to avoid water flows over the bed. In the meantime, perforated pipes are buried at the inlet zone to keep maximum flow through the treatment zone and the effluent is then collected by the outlet pipes buried at the base of the media, which is about 0.3-0.6 m below the surface of the bed. Then, the wastewater to be treated flows from the inlet the outlet direction under the surface of the wetland bed (Kadlec and Wallace, 2009; Nelson et al, 2009; Azni et al, 2010).
The application of SSF systems is extensive in Europe, Australia, South Africa, and nearly every country of the world and they are the most common treatment plants at this time. One of the peculiar features of SSF systems, unlike the free water surface flow system, is that insect vectors do not get any opportunity to breed in the system as there is no contact between the water column and the surrounding atmosphere. So, the likelihood of incidence of public health problem which is associated with the application of SSF constructed wetlands is extremely low. As a result, they can offer better option for primary wastewater treatment (Kayombo et al, 2003; Robert, 2004; Kadlec and Wallace, 2009).
Compared to the free water surface system, the performance of subsurface flow constructed wetlands for nutrient removal is higher. The removal efficiency of constructed wetlands is dependent on aerobic and anaerobic condition within the wetland cells and the differences of water flow in the system (Li et al, 2007). Yang et al (2014) pointed out that even slightly polluted drinking water source could be effectively treated by applying SSF systems and consequently the quality of drinking water source could be improved which in turn reduces the burden on drinking water treatment.
Basically, SSF constructed wetlands are further grouped into two categories based on the direction of water flow. These are vertical up or down flow (VF) and horizontal flow (HF) types (Cooper et al, 1996; Vymazal et al, 1998a). In addition, employing of the hybrid/combined system of the two wetland types to effectively exploit the advantages of each type is becoming common practice in many areas now a day (Merlin, Pajean and Lisollo, 2002).

Figure 2.2: Subsurface flow constructed wetland (Russo, 2008; p. 329)

2.4.2.1 Horizontal Subsurface Flow Constructed Wetlands (HSSF CWs)
In horizontal subsurface flow system, the influent from the inlet zone flows horizontally thorough the porous media below the surface of the bed until it gets to the outlet zone. At the time of its flow, the wastewater undergoes in different processes that are taken place in aerobic, anoxic, and anaerobic zones of the bed. The aerobic zones are those sites around roots and rhizomes where oxygen is released into the media of the constructed wetland. The effluent or the treated wastewater is then collected in the outlet zone before leaving the system (Cooper et al, 1996; Vymazal et al, 1998; Russo, 2008).
Toscano et al (2015) described that the performance of HSSF constructed wetland in reducing major physical, chemical, and microbiological concentration of contaminants in municipal wastewater is very high. He also emphasized the active role of vegetations in the removal processes of pollutants in wastewater treatment using wetland system. HSSF constructed wetlands showed high and steady removal performance over many years of operation for organic pollutants like BOD5, COD, TSS and oil and grease with satisfactory effluent quality for being discharged into the environment. But their performance is poor in terms of contaminants in wastewater such as phosphorous, ammonium nitrogen and organic matter due to oxygen deficiency (Haberl, Perfler and Mayer, 1995; Vymazal, 2005; Naz et al, 2009; Cakir, Gidirislioglu and Cebi, 2015; Costa et al, 2015; Albalawneh et al, 2016). Despite this, (Costa et al, 2015) pointed out that good P removal efficiency (70%) and fair N removal efficiency (40%) can be achieved by using HSSF constructed wetlands.
2.4.2.2 Vertical Subsurface Flow Constructed Wetlands (VSSF CWs)
In vertical SSF CWs, whether it is ascending or descending, vertical direction of flow through the media is established by using various designs of wastewater feeding or collection mechanisms. This can be achieved by applying the wastewater to be treated into the cell intermittently or by burying inlet pipes into the bed at certain depths. This kind of wetland system is known as “infiltration wetlands” since wastewater infiltration occurs through the medium (Cooper et al, 1996; Vymazal et al, 1998).
The performance of VF bed is significantly better than HF for the removal of BOD5, COD, Kjeldahl-nitrogen and ammonia-nitrogen. This is mainly occurred as the unsaturated flow condition in VF bed presents more oxygen for the oxidation-reduction potential to take place in VSSF wetland (Pandey et al, 2013). Kurniadie (2011) mentioned that the effective removal of organic matter, nutrients and pathogenic bacteria can be achieved by the proper application of VSSF CWs planted with macrophytes and the reduction of concentration of COD, NO3-N, PO4-P, and total coliforms in the final effluent is very low.
In a wetland system, carbon degradation is carried out chiefly by bacteria while fungi have minor role. In VSSF CWs, as more than 80% of the growth or multiplication of microbes is taken place inside the first 10 cm of the filter media, the depth of the filter media should not be less than 10 cm to maintain steady performance and filtration process (Tietz et al, 2008). Additionally, the design with two-stage VSSF CW can enhance the performance in treating wastewater (Xie et al, 2011; Langergraber, Pressl and Habler, 2014).
In VF constructed wetlands, DO levels increase initially and then decrease vertically from top to bottom. There is a positive correlation between the levels of DO and the biofilm mass, showing the presence of other sources of oxygen supply in addition to the oxygen in the influent, particularly in the upper part of the wetland bed. This incident supports the assumption that the major oxygen source for VFCWs is atmospheric reoxygenation, and of course the contribution of atmospheric reoxygenation in the process of domestic wastewater treatment is more than 99.9% of the total oxygen supply to the VFCWs. The upper part of VFCWs usually encompasses 0 – 10cm below wastewater distribution system is supplied by just about 50% of atmospheric reoxygenation (Ye et al, 2012).
2.4.2.3 Hybrid Constructed Wetlands
A number of treatment processes which take place in CWs can be more effective in removing pollutants in wastewater when different wetland types combine. This is possible as a result of the occurrence of supplementary abiotic/biotic pollutant removal pathways which is attributed to different physico-chemical conditions present at different wetland configurations. For instance, anaerobic removal pathways are predominant in HSSF CWs, while VSSF CWs are more appropriate for pollutants that can be easily biodegraded under aerobic conditions. Similarly, FWS wetlands can take the advantages of the effect of photo-oxidation and other processes for the removal of emerging organic contaminants (Avila et al, 2014). Moreover, Masi and Martinuzzi (2007) revealed that it is possible to reduce the total surface area when hybrid configuration is employed and consequently the water loss via evapotranspiration is decreased. In general different environmental conditions such as aerobic, anoxic, and anaerobic conditions can increase the performance of CWs (Haberl et al, 2003).
Hybrid constructed wetland system is being applied by combining various types of constructed wetlands in order to achieve higher treatment efficiency especially for nitrogen removal. Hybrid systems combining VF and HF beds are the most common ones and proved to be more efficient for practical application (Cooper et al, 1996; Vymazal et al, 1998; Merlin, Pajean and Lissolo, 2002; Vymazal, 2005).
High loads of organic matter, nitrogen, suspended solids, pathogens and chemicals can be removed using the hybrid subsurface CW system and the efficiency is high and steady during both cold and warm seasons. But N transformation and concentration is affected by total carbon concentration available in the system (Rousseaou, Vanrolleghem and Pauw, 2004; Fabio and Martinuzzi, 2007; Abidi et al, 2009; Vymazal and Kropfelova, 2015; Zhang et al, 2016). Abidi et al (2009) reported that HF bed seems to be a promising design for denitrification while VF bed a potential design for the process of nitrification. But in general, VF-HF hybrid system has shown great potential for the accomplishment of nitrification to the level that is required although its capability for the removal of nitrate nitrogen is not good. In the meantime, significant differences of pollutant removal processes are demonstrated where the VF and HF beds are alternated, so that the desired configuration should be chosen to attain highest removal efficiency (Gaboutloeloe et al, 2009).
2.5 Design Factors of Constructed Wetlands
Several complex processes are taken place in constructed wetlands treating wastewater. Therefore, the removal of pollutants in wastewater is resulted from the supportive and mutually dependent actions of the following components. These are: substrate, vegetation, and microorganisms. So, the selection of those components based on adequate knowledge and skill helps to achieve high system performance. In line with this, various design factors of CWs are being considered in order to optimize the capability of the wetland system. On the other hand, the key design factors include: wetland plants, substrates, retention time, and water depths (Dordio and Carvalho, 2013; Upadhyay, Bankoti and Rai, 2016).
Alley et al (2013) pointed out that giving adequate attention to seasonal factors for instance temperature and evapotranspiration during designing helps to optimize the wetland performance in removing targeted pollutants for the seasons. Similarly, Valsero, Cardona and Becares (2012) revealed that pollutant removal efficiency of constructed wetlands is marked by seasonal difference. On the other hand, the effect of temperature is not clearly known and the available data on it is sometimes contradictory, and also contradictory formulas have been formulated by wetland designers to determine the hydraulics and size of CWs in areas which have different climatic conditions (Siti et al, 2011). Generally, CW configurations applied under similar environmental and hydraulic load greatly vary (Valsero, Cardona and Becares, 2012).

2.5.1 Wetland Plants
Wetland plants and its litter are among essential components of CWs system in improving the performance and giving attractive aesthetic value. They make most of the major visible structure of CWs. Wetland plants or macrophytes can grow well in wetland system and show significant removal efficiency in treating different wastewaters. It is realized that they are used in virtually all wetland types for increasing the performance and getting better effluent quality which can meet the discharge requirements. The reports of many studies on planted and unplanted wetland system has concluded that the performance of wetlands is high in the presence of macrophytes (USEPA, 1993; Thomas and William, 2001; UN-HABITAT, 2008; Kadlec and Wallace, 2009; Upadhyay, Bankoti and Rai, 2010).
Wetlands are typically dominated by macrophytes. Wetland plants can acclimatize water saturated environment and they can also tolerate anaerobic environment caused by the excess water content. Compared with terrestrial plants, they show a worldwide similarity. This similarity overrules climatic conditions and is imposed by a free water supply common characteristics and oddly harsh chemical environment that must be tolerated by plants. Macrophytes develop different functional mechanisms to survive the unfavorable environmental conditions (Russo, 2008). Besides this, the basic nutritional requirements of macrophytes and other terrestrial plants are similar (Robert, 2004). In general, pollutants and nutrients in wastewater are taken up by aquatic plants, as a removal pathway in treatment processes that are taken place in constructed wetlands (Sarmento, Borges and Matos, 2012; Bialowiec, Albuquerque and Randerson, 2014).
Zhang, Gersberg and Keat (2009) reported that the removal efficiency of planted CWs is higher than unplanted CWs for certain pollutants such as TN and NH4-N although the role of plants for the removal of BOD5, COD and TP is limited. Moreover, the performance of planted wetlands in removing nitrogen is usually found to be efficient and steady in all months of the year (Lee and Scholz, 2007; Fonkou et al, 2011; Abou-Elela and Hellal, 2012; Mesquita, Albuquerque and Nogueira, 2012).
Wetland plants have vital roles in providing attachment site for microorganisms, sufficient surface area for pollutant adsorption, and diffusion of atmospheric oxygen to the rhizosphere, adequate hydraulic residence time, and trapping and settlement of suspended wastewater constituents as a result of resistance to hydraulic flow. The stem and leaves in the water column on the other hand, help for improved sedimentation and used by microorganisms as a substrate for their multiplication. All these things can have an effect on the plant-microorganisms-wastewater interactions and then treatment performance of the wetland system. Therefore, the proper structural development and the general growth rate of wetland plants supposed to be applied in constructed wetland system are of highly important (Joseph et al, 2004; Chazarenc, Merling and Gonthier, 2004; GEPD, 2010; Dong et al, 2016).
Plants in aquatic environment can also have great transpiration potential. Evapotranspiration causes low treatment efficiency in CWs since it increases the concentration of dissolved compounds as water volume decreases, and creates the accumulation of pollutants in soil (Bialowiec, Albuquerque and Randerson, 2014).
Based on total solids (TS) analyses, the existence of wetland plants increases accumulation/production of solids and intensifies clogging, generating greater headloss and possible surface flow at the inlet of the planted wetlands, when compared to the unplanted systems (De Paoli and Sperling, 2013).
The role of macrophytes in the course of action of wastewater purification should not be undermined while CWs are employed as method of treatment (Dong et al, 2016). In general, the effects of wetland vegetations can be summarized as follows (Kadlec and Wallance, 2009; GEPD, 2010):
The plant growth cycle seasonally stores and releases nutrients, thus providing a “flywheel” effect for a nutrient removal time series.
The creation of new, stable residuals accretes in the wetland. These residuals contain chemicals as part of their structure or in absorbed form, and hence accretion represents a burial process for nitrogen.
Submersed litter and stems provide surfaces on which microbes reside. These include nitrifiers and denitrifiers, and other microbes that contribute to chemical processing.
The presence of vegetation influences the supply of oxygen to the water. Emergent vegetation blocks the wind, and shades out algae, presumably lowering re-aeration. Floating vegetation may provide a barrier to atmospheric oxygen transfer. Submerged vegetation may provide photosynthetic oxygen supply directly in the water. To some limited extent, plant oxygen flux supplies protective oxidation in the immediate vicinity of plant roots.
The carbon content of plant litter supplies the energy need for heterotrophic denitrifiers.

Table 2.1 the role of plants (macrophytes) used in constructed wetlands
Part of wetland plant Role
Aerial plant tissues light attenuation: reduced growth of phytoplankton
influence on microclimate: insulation during winter
reduced wind velocity; reduced risk of resuspension of solids
aesthetic appearance
nutrient storage
Plant tissue in water filtering effect: filter out large debris
reduced current velocity: increased rate of sedimentation: reduced risk of resuspension
surface area for attached microorganisms
excretion of photosynthetic oxygen: increased aerobic degradation
nutrient uptake
Roots and rhizomes stabilizing the sediment surface: less soil erosion
prevents the medium for clogging in vertical flow systems
release of oxygen increase organic degradation and nitrification
nutrient uptake
secretions of antibiotics for detoxification of root zone: pathogen removal
Source: Russo, 2008; p.216
There are reports from several studies on the applicability of special plant species and their capability to improve the removal efficiency when compared to others. As far as known, the type of plants used plays a minor role for domestic wastewater treatment. But the selection of the right plant species can have significant role when unique organic compounds and/or heavy metals are found in the wastewater. In view of this, both the plant productivity and the pollutant removal efficiency are relevant in finding an appropriate plant for a given application (Haberl et al, 2003).
Haberl et al (2003) revealed that treatment efficiency can be improved if a combination of different environmental conditions (e.g. aerobic, anoxic, and anaerobic conditions) is provided and/or different plant species are used. However, there is very few known about the effects of the combination of plants regarding treatment efficiency.
In the selection of appropriate wetland plants for CW system, the most important and widely used criteria include (Russo, 2008; Merchand, et al, 2010):
Ecological acceptability, that is, no significant weed or disease risks or danger to the
Ecological or genetic integrity of surrounding natural ecosystems;
Tolerance of local climatic conditions, pests and diseases;
Tolerance of pollutants and hypertrophic water-logged conditions;
Ready propagation, and rapid establishment, spread and growth; and
High pollutant removal capacity, either through direct assimilation or storage, or
Indirectly by enhancement of microbial transformations.
Additionally, the following points should also be taken in to consideration while wetland plants are selected (Azni et al, 2010):
The species available or suitable for the proposed wetland site,
The substrate on which the plants will prefer to grow (e.g., sand, mud, clay, peat),
Aerobic vs. anaerobic conditions and when and where this is likely to occur within the wetland,
The depth of water in which the plants normally grow, e.g., shallow or deep water,
The frequency and depth of inundation, and
Periods of drying and the ability of the plants to withstand drying.
Wetland plants are typically classified into three broad types based on their growth form. These are: floating, submerged, and emergent plants (Robert, 2004; Russo, 2008; Kadlec and Wallace, 2009; Azni et al, 2010):
Floating. These plants are not attached to the wetland media; rather they freely float at the water surface. Hyacinth, pennywort, and duckweed are the best examples of floating species. Free floating plants are able to use oxygen and carbon dioxide directly from the surrounding atmosphere and mineral nutrients from the water.
The roots of floating plants are directed downwards into the water column while other part of the plants where photosynthesis is carried out is found at or right above the water surface. The process of photosynthesis is accomplished by taking up nutrients from the water, through the root system and using atmospheric oxygen and carbon dioxide. The root system in the water is an ideal place for adsorption or filtration and at the same time for bacterial growth. The development of plant roots, functions as treatment medium, can be affected by those factors such as quality of wastewater, temperature and frequency of harvesting.
The entrance of sunlight into the water column of the wetland system is appreciably limited in the presence of floating plants and the exchange of gases between the atmosphere and water is highly hampered as well. Thus anoxic or anaerobic conditions as a result of organic loading rate and the chosen floating plant species and coverage density can be created while the wastewater becomes algae free.
Submerged. The submerged plant species can be either attached to the substrate or free floating although the stems and leaves of the plants are submerged permanently. Those plants whose flowers may be emergent are grouped under submerged plants.
These plants can be submerged in the water column or rooted in the bottom sediments. The photosynthetic parts of the plants are usually found in the water. They are theoretically considered as an attractive alternative to polish effluents. But the practical importance can be compromised since there is a possibility of the plants to be harmed by anaerobic condition and to be covered by excessive algal growth.
Submerged plants have a key role in removing organic nitrogen. The removal of ammonia in the presence of submerged plants is associated with photosynthetic processes. Unlike floating species, submerged species use carbon dioxide found in the water in photosynthesis, the processes that cause the raising of pH and then removing of gaseous ammonia through diffusion into the atmosphere. Ammonia, the gaseous form of nitrogen is usually known for its toxic effect for fish.
Emergent. These types of plants are commonly attached to the substrate of the wetland system. Their stems and leaves of the plants extend above the surface or float on the surface. Plants which can be grouped under emergent are overwhelmed either occasionally or permanently.
The plant root zone (rhizosphere) is the only site for the plants and the wastewater to be treated to get in contact since the flow of the wastewater is through the gravel or aggregates. The symbiotic relationships established between wetland plants with bacteria and fungi, excretion of root exudates and transfer of oxygen affect the surrounding environment of the root zone (Scott, 2004). An important role is played by fine root than the role played by the entire root system in wastewater treatment and seasons and plant growth can affect removal efficiency (Yang et al, 2007).

Table 2.2: Recommended emergent plant species for constructed wetlands (Source: Handbook of environmental engineering, Volume 11, p.335)
Recommended species Maximum water deptha Notes

Arrow arum (Peltandra 12 in. Full sun to partial shade. High wildlife value. Foliage and
virginica) rootstocks are not eaten by geese or muskrats. Slow grower.
pH: 5.0–6.5
Arrowhead/duck potato 12 in. Aggressive colonizer. Mallards and muskrats can rapidly
(Saggitaria latifolia) consume tubers. Loses much water through transpiration
Common three-square 6 in. Fast colonizer. Can tolerate periods of dryness. High metal
bulrush (Scirpus pungens) removal. High waterfowl and songbird value
Softstem bulrush (Scirpus 12 in. Aggressive colonizer. Full sun. High pollutant removal. Provides
validus) food and cover for many species of birds. pH: 6.5–8.5
Blue flag iris (Iris 3–6 in. Attractive flowers. Can tolerate partial shade but requires full sun
versicolor) to flower. Prefers acidic soil. Tolerant of high nutrient levels
Broad-leaved cattailb 12–18 in. Aggressive. Tubers eaten by muskrat and beaver. High pollutant
(Typha latifolia) treatment, pH: 3.0–8.5
Narrow-leaved cattailb 12 in. Aggressive. Tubers eaten by muskrat and beaver. Tolerates
(Typha angustifolio) brackish water. pH: 3.7–8.5
Reed canary grass 6 in. Grows on exposed areas and in shallow water. Good ground
(Phalaris arundinocea) cover for berms
Lizard’s tail 6 in Rapid grower. Shade tolerant. Low wildlife value except for
(Saururus cernuus) wood ducks
Pickerelweed (Pontedaria 12 in. Full sun to partial shade. Moderate wildlife value. Nectar for
cordata) butterflies. pH: 6.0–8.0
Common reedb 3 in. Highly invasive; considered a pest species in many states. Poor
(Phragmites australis) wildlife value. pH: 3.7–8.0
Soft rush (Juncus effuses) 3 in. Tolerates wet or dry conditions. Food for birds. Often grows in
tussocks or hummocks
Spikerush (Eleocharis 3 in. Tolerates partial shade
Palustris)
Sedges (Carex spp.) 3 in. Many wetland and several upland species. High wildlife value
for waterfowl and songbirds
Spatterdock (Nuphar 5 ft (2 ft min.) Tolerant of fluctuating water levels. Moderate food value for
luteum) wildlife, high cover value. Tolerates acidic water (to pH 5.0).
Sweet flag (Acorus 3 in. Produces distinctive flowers. Not a rapid colonizer. Tolerates
calamus) acidic conditions. Tolerant of dry periods and partial shade.
Low wildlife value
Wild rice (Zizania 12 in. Requires full sun. High wildlife value (seeds, plant parts, and
aquatica) rootstocks are food for birds). Eaten by muskrats. Annual,
non-persistent. Does not reproduce vegetatively
aThese depths can be tolerated, but plant growth and survival may decline under permanent inundation at
these depths.
bNot recommended for storm water wetlands because they are highly invasive, but can be used in treatment wetlands if approved by regulatory agencies.
2.5.2 Substrate
The difference in pollutant retention capacity of various materials is attributed to the different characteristics they have (Dordio and Carvalho, 2013). Hence, one of the most important things in the application of CW technology is the scientific and practical selection of materials used as a substrate (Lu et al, 2016). Then, the most important approach in determining the usefulness and applicability of certain substrate in CWs is finding of the middle ground between adsorption capacity and hydraulic conductivity. There are also other factors which should be given the required attention. These are: local availability, cost, saturation time and recyclability of saturated filter media (Dordio and Carvalho, 2013).
Once the material is selected, the central single property that must be carefully evaluated when applying as media for constructed wetland is the texture, particularly distribution of the grain size. Danish EPA (1999), as cited in Arias, Bubba and Brix (2001), recommends the particle size distribution in terms of D10 and D60, which are the typical in particle size distribution. Accordingly, the effective grain size d10 should be in range of 0.3±2.0 mm, d60 between 0.5 and 8 mm, whereas the uniformity coefficient d60/d10 should be less than four in order to decrease the occurrence of clogging by ensuring sufficient hydraulic conductivity. In other words, the use of coarse media can maintain operation of CWs for long period of time by avoiding clogging (Meyer et al, 2013).
Throughout the period of setup of CW operation, porosity of the selected median, the growth of plant roots, and formation of biofilm show continuous progress until the operation reaches steady state phase. Accumulation of suspended matter, expansion of roots of plants and attaching of biofilm on the surfaces of the substrate are known to cause lessening in porosity for CWs bed and the media porosity value reaches steady phase after certain operation period. The growth of plant roots and bacterial biofilm attached to different plant parts and the substrate at the setup period increase contaminants accumulation, biodegradation, and finally treatment efficiencies (Zidan et al, 2015).
Shutes (2001) pointed out that wastewaters which contain pollutants such as heavy metals and hydrocarbons in it can cause accumulation of these pollutants in the substrate which in due course requires transport and disposal to a sanitary landfill site. However, substrates which are applied in CWs which are constructed for the purpose of treating domestic or agricultural wastewater, relatively free of toxic substances, can be used without substitution for a number of years.
Most of the time, the common media which are used in CWs include different soils, sand, gravels, and crushed stones, either alone or in combination (Kayombo et al, 2003). But other extra properties of the media shall be considered if there is a need to enhance the removal of pollutants (for example P and N) (Arias, Bubbo and Brix, 2001). For example, Calcium based materials such as calcite and marble (Brix, Arias and Bubba, 2001) are known for their superior capacity in removing phosphorous while zeolite shows higher ammonium nitrogen removal efficiency (Zou et al, 2012).
2.5.3 Retention Time
The wastewater to be treated must stay long enough in the wetland system before the completion of the treatment process, and this length of time is usually known as hydraulic retention time. It is one of the key factors to estimate the performance of a CW system. The physical, chemical, and biological processes which are carried out to remove pollutants in wastewater are greatly influenced by the retention time of the water: i.e. if there is longer retention time, the removal rate of all pollutants, except total coliforms and total suspended solids, will be faster. The removal of COD is highly sensitive to retention time. However, too long retention time may be associated with negative effects (Kayombo et al, 2003; Katayon et al, 2008; UN-Habitat, 2008; Su et al, 2009; Masi, Caffaz and Ghrabi, 2013; Cakir, Gidirislioglu and Cebi, 2015).
2.5.4 Water Depth
Water depth is another essential design criterion taken in to account in optimizing the pollutant removal efficiency of CWs system (Garcia et al, 2005). The movement or flow of water in CWs is slow since they have saturated media or shallow water depths. Then, the low water depth and the slow water flow create suitable environment for sediments to settle down while the water flows through the wetland system. Moreover, sufficient contact time among the water, substrate, and wetland surfaces can be obtained in these situations. So, a large variety of substances in wastewater can be decomposed by the action of microbial community as a considerable mass of organic and inorganic materials are available (Azni et al, 2010).
Subsurface flow CWs which have a water depth of 0.27m are more efficient for the removal of BOD5, COD and ammonia, than SSF CWs with a water depth of 0.5m. In general, as the water depth increases from 0.27m to 0.5m, the pollutant removal efficiency decreases. Hydraulic loading rate and/or areal organic loading rate are other factors in regulating treatment efficiency of SSF. Subsurface flow with a medium size of 3.5mm produced effluents of better quality than SSF with a medium size of 10 mm; but the differences were smaller in comparison to the effect of water depth and HLR (Garcia et al, 2004; Garcia et al, 2005). Water depth is assumed to have an effect on biochemical reactions which are key processes for organic matter degradation. The removal of organic matter is achieved more importantly by biochemical reactions like methanogenesis and sulphate reduction in CWs with a water depth of 0.5m than in those wetlands with a water depth of 0.27m (Garcia et al, 2005).
2.5.5 Seasons of the Year
The available literatures regarding the impacts of seasonal variations on the performance of CWs in treating different wastewaters are not consistent (Siti et al, 2011). Vymazal (2014) reported that there is no significant difference between the average outflow concentrations of all monitored parameters in summer and winter periods. For instance, CWs in mountainous regions in the Czech Republic showed very good treatment effect with overall treatment efficiencies between 88% and 94% for BOD5, 67% and 85% for COD and 74% and 96% for TSS. The removal of these parameters was stable during the year and during the time of operation.
However, a number of studies are in favor of the thought that seasonal removal of constituents from wastewater using CWs can change with location and targeted constituent, so an initial pilot-scale study could be beneficial prior to construction of a full-scale system to estimate the removal rate coefficients of targeted constituents (Alley et al, 2013). Seasonal variations seem to affect the performance of CWs technology even if the performance is not consistent for all wastewater parameters. Lower effluent concentrations were observed during the warm period, especially for TN and NO3-N, whereas the performances improved as wastewater temperature rises. The removal efficiency of NH4-N can be affected by seasonal differences to a lesser degree in the VF than the HF. The observation depict that the performance of CWs system during summer season is higher than winter season (Zhiwen et al, 2006; Prachaska, Zouboulis and Eskirdge, 2007; Roussou et al, 2008; Mustafa et al, 2009; Mietto et al, 2015; Ramprasad and Philip, 2016).
Wu et al (2014) described that the operation of CWs at cold climate is a big challenge. A number of adaptation mechanisms are commenced through change of design, natural or artificial thermal insulation and upgraded operation approach, for example artificial aeration. Wetlands with green-house structures can be considered in highly frigid climatic conditions despite of the high investment cost.
In the meantime, the influence of climate on the removal of BOD and TSS by physical mechanisms such as sedimentation and flocculation is less. The absence of plant cover in colder seasons could permit the occurrence of atmospheric re-aeration and solar insolation, but the wetland hydraulics can be changed and solar insolation, atmospheric re-aeration and biological activities can be limited as a result of ice cover. However, there is no influence of the insulating layer created by ice cover on physical processes including filtration, sedimentation, and flocculation (USEPA, 2000).
As in other biological processes, growth rates in aquatic plant systems depend on temperature and the vegetated system show a much better performance during the warmer months of the year (Karathanasis, Potter and Coyne, 2003; Wang and Li, 2014). The mean TN and TP removals were high in summer (23%) and fall (45%), respectively. Lee et al (2013) pointed out that the dependence of removal efficiency on temperature is significant due to plant uptake, which plays a significant role in nutrient removal. Greater bacterial activity is shown during the warmer season than the colder one (Chon and Chon, 2015). So, warmer climate improves performances, especially for nitrification (Masi, Caffaz and Ghrabi, 2013; Molle et al, 2015).
Rai et al (2015) reported that accumulation of trace element in summer season was high in comparison to winter. A substantial change from winter to summer was observed for Zn (68.40–83.48%), As (63.18 – 82.23%) and Cr (64.5 – 81.63%) while other trace elements showed little difference.
2.6 Pollutant Removal Mechanisms in Constructed Wetlands
A number of complex processes engaging all physical, chemical, and biological mechanisms are undertaken in constructed wetland system to transform and separate various pollutants found in wastewaters. The different removal mechanisms can happen simultaneously or sequentially as the wastewater to be treated gets into the treatment system. Even though the removal processes of the wetlands system are identified, the measurement of these processes to obtain accurate quantitative value is becoming an existing challenge in most cases. The internal interactions, the external input parameters, and characteristics of the wetland are basic factors among others on which the major removal mechanisms and their reaction sequence depends on (USEPA, 2000; Qasaimeh, Alsharie and Masoud, 2015; Gokalp et al, 2016).
In general, there are two key mechanisms in almost all wastewater treatment systems: namely, pollutant transformations and liquid/solid separations. Gravity separation, absorption, filtration, adsorption, ion exchange, stripping, and leaching are commonly involved in separations, where as transformations are resulted from chemical reactions, including re-dox reactions, precipitation, flocculation, acid-base reactions, or biochemical reactions taking place under anaerobic, anoxic, or aerobic conditions. Therefore, in CW system, both mechanisms can play a key role in the removal of pollutants while the wastewater gets into the wetland and stays for certain period of time (USEPA, 2000).
Adequate knowledge of the fundamental physical, chemical and biological processes governing the performance of wetlands raises the acceptance and application of CWs in a large extent. Similarly, in order to understand the structure and functions of the wetland system, a working knowledge of biogeochemical cycling, the movement and transformation of nutrients, metals and organic compounds among the living and non-living components of the ecosystem is required. This level of understanding and practical knowledge is helpful for evaluating the performance of CWs to remove pollutants (Russo, 2008).
2.6.1 Physical Mechanisms
Filtration and sedimentation are the two most significant processes representing the physical mechanisms. The physical removal of wastewater pollutants related to particulate matter is carried out by filtration and sedimentation and these two processes are considered as highly efficient. The presence of plant biomass in all wetlands, and also the substrate in the case of SSF CWs are the chief factors promoting the physical contaminant removal processes in wetland technology (Kayombo et al, 2003; Russo, 2008).
Slow flow velocity really helps to improve sedimentation for the removal of suspended solids in CWs system. The flow of wastewater is hindered in the wetlands because of the resistance occurred by wetland plants, and this event is responsible to advance sedimentation of SSs. Furthermore, floating plants can also have a primary role in the removal of suspended solids by limiting re-suspension of particulate matter already settled at the treatment bed. On the other hand, similar to the processes taking place in filtration, the media applied to the wetland system is another removal pathway of suspended solids (Russo, 2008).
Another route of physical removal mechanism in wetlands is volatilization, the process of diffusing of dissolved compounds in wastewater into the atmosphere. Several organic compounds and certain simpler inorganics formed as a result of mineralization like ammonia are volatile compounds, which are lost to the atmosphere from CWs. Because of its potential to cause air pollution, volatilization is not a preferred removal process (Russo, 2008).
2.6.2 Biological Mechanisms
There are six biological reactions indentified to be highly significant in the performance of CWs technology. These are: photosynthesis, respiration, fermentation, nitrification, denitrification and microbial phosphorous removal (Mitchell, 1996). Carbon and oxygen are added to the wetland system as a result of photosynthesis process carried out by algae and wetland plants nitrification process is enhanced by carbon and oxygen. All living forms in the system oxidize organic carbon (respiration) to produce energy, CO2 and water. Bacteria, algae, fungi, and protozoa are the typical living forms of wetland systems and they require favorable conditions for optimal performance. Fermentation is another process which refers to the microbial decomposition of organic carbon without the presence of oxygen. Nitrification/denitrification processes are also accomplished by microorganisms to remove nitrogen. Moreover, dissolved nutrients and other contaminants in wastewater are taken up by wetland plants to produce additional biomass (Kayombo et al, 2003; Azni et al, 2010).
Wetlands can rapidly remove readily degradable organic C compounds typically found in municipal wastewater. Unlike this, various recalcitrant organic compounds such as lignin-based compounds and products of petroleum can also be removed by the help of microbial action in wetlands, even if the rate of removal for these compounds are substantially lower compared to readily biodegradable compounds (Azni et al, 2010).
Enzymatic activity is one of the biological processes in CW which plays a major role in releasing nutrients from organic substances and higher nutrient loading into wetlands reduced the nutrients removal efficiency of wetlands. Also, enzyme activity is continuously contributing to the release of inorganic nutrients which may reduce the wetlands efficiency (Baddam et al, 2016).
It is obvious that a quantifiable amount of pollutant uptake and storage occur through microorganisms. But the decomposition of organic compounds throughout the conversion from complex to simple molecules is achieved by the metabolic processes undertaken by microbes. This is the most significant biological mechanism by which a wide range of pollutant organic compounds can be removed. In fact, the removal rate and efficiency of microbial degradation of organic carbon vary depending on the type of compounds (Russo, 2008). Nitrogen gas, in the form of N2, is released from nitrate and ammonium as a result of microbial metabolism in wetlands, and lost to the atmosphere. This depicts that microbial metabolism can also remove inorganic nitrogen (Vymazal, 2007).
Additionally, plant uptake for the removal of inorganic pollutants, is probably the most widely recognized biological process. Of the contaminants of wastewater, some contaminants like ammonium, nitrate, and phosphate are essential nutrients for plants while others are toxic metals or compounds. Most wetland plants can uptake both types of contaminants and even the accumulation of considerable amounts of toxic chemicals in the plants tissue can occur over time (Russo, 2008; Kadlec and Wallace, 2009; Qasaimeh, Alsharie and Masoud, 2015).
2.6.3 Chemical Mechanisms
The removal of pollutants can also be achieved by a number of chemical processes, and the chief chemical processes which are involved in the removal of contaminants are sorption and precipitation. Reduction-oxidation reactions, complexation, and hydrolysis are also other chemical processes which result in pollutant conversions/transformations which are prerequisite for contaminant removal through precipitation or adsorption. Certain unique groups of pollutants can be removed because of chemical processes such as photolysis and ionic exchange with mineral components of the substrates in CWs (Russo, 2008).
Sorption, one of the chief chemical processes can be described as:
‘Sorption refers simultaneously to both adsorption and absorption phenomena, and the term is used whenever the extent to which each phenomenon is responsible for the compound’s removal is not clear or well defined. These chemical processes occur at the surfaces of plants roots and substrate, resulting in a short-term retention or long-term immobilization of the contaminants.’ Raymundo (2008)
The sorption removal process which occurs in wetlands is influenced by the following three factors (Russo, 2008):
Substrate characteristics (texture, content in organic matter and ion exchange properties),
Wastewater characteristics (dissolved organic matter content, pH and electrolyte composition), and
Pollutant characteristics; when pollutants have acid-base properties, the pH in the liquid compartment can be the cause to influence the degree of sorption process to mineral surfaces through ion exchange properties.
Sometimes, the binding of contaminants to mineral surfaces is facilitated by the occurrence of complexation, although the complexation phenomenon itself also helps to increase the solubility of pollutants in another time (Russo, 2008).
Precipitation, which depends on re-dox condition and pH, is also another chemical process by which long-term pollutant removal is resulted. The formation of insoluble oxides and hydroxides by hydrolysis and the conversion of soluble ionic compounds into neutral insoluble forms in protonation state are the two important conditions for precipitation to take place. In some cases, sorption process can assist co-precipitation event to happen (Russo, 2008). Metals which are found in the water column can be precipitated as insoluble compounds (EPA, 1995).
The direction of most processes and reactions such as biological transformation, partitioning of ionized and unionized forms of acids and bases, cation exchange, and solubility of gases in wetland system is highly influenced by the pH of water and soils (Kayombo et al, 2003) .

Table 2.3: Summary of selected pollutant removal mechanisms in constructed wetlands
Pollutant Removal mechanism
Biochemical Oxygen
Demand (BOD) Oxidation
Absorption
Filtration
Sedimentation
Microbial decomposition
Suspended solids (SS) Filtration
Sedimentation
Nitrogen (N) Adsorption
Assimilation
Absorption
Ammonification–nitrification–denitrification
Heavy metals Adsorption
Cation exchange
Bioaccumulation
Pathogenic bacteria and viruses Adsorption
Predation
Sedimentation
Sterilization by UV
Other pollutants Precipitation
Evaporation
Evapotranspiration
Source: Handbook of Environmental Engineering, Volume I, p.329
2.7 Removal of Pollutants in Constructed Wetlands
2.7.1 Removal of Organic Compounds
Although physical processes (sedimentation and filtration) are capable of removing organic matter near the inlet of SSF, the major removal process of organic matter in CW system is microbial degradation. Domestic wastewater usually contains readily biodegradable dissolved organic compounds and when the dissolved organic matter gets into the biofilms attached on the root system, submerged plant stems and litter, and the surrounding fill media, the biodegradation process takes place. Macrophytes supply oxygen to the root system and give support medium for the occurrence of microbial degradation (Kayombo et al, 2003; Robert, 2004; Scott, 2004; Garcia et al, 2005; Azni et al, 2010). But the role of plant uptake for the removal of BOD5 and COD is much lesser (Zhang, Gersberg and Keat, 2009).
The carbon content of organic matter reaches roughly 45–50% and a wide range of microbes in wetlands utilize the carbon as a source of energy. The process of breaking down of organic carbon to CO2 by microorganisms to obtain energy for growth requires dissolved oxygen (Azni et al, 2010).
In fact, both aerobic and anaerobic decomposition processes can take place in the removal of organic matter. However, the nature of biochemical reactions depends on the conditions created as a result of the rate of oxygen transfer in the wetland and the load of internal and external organic load. If there is adequate oxygen supply, aerobic decomposition is so rapid that the accumulation of organic matter in the wetland is small. But if the rate of oxygen transfer cannot fulfill the oxygen requirements, the removal process is anaerobic decomposition which results the accumulation of organic matter in the wetland (Scott, 2004; Garcia et al, 2005).
Generally, there are reports which describe that the performance of CWs for the removal of BOD5 and COD from municipal/domestic wastewater is significantly high (Vymazal and Kropfelova, 2009; Azni et al, 2010; Vymazal, 2014). Even, the HFCWs, in which the rate of transfer of oxygen in the system is expected to be relatively low, can show high performance in removing BOD5 and COD and result quality effluent that fulfills the requirement for discharge in terms of the two parameters (Cakir, Gidirislioglu and Cebi, 2015; Albalawneh et al, 2016).
On the other hand, the degradation of organic matter by microorganisms is influenced by climatic conditions and consequently the rate of degradation in tropical areas is higher than temperate or cold areas (Anish, Ajay and Bijay, 2012). Song et al (2006) reported that removal efficiencies for BOD5 and COD vary from season to season (10% less efficient in winter and autumn compared to summer and spring). In the contrary, Paing et al (2015) revealed that temperature has no effect on the removal of BOD5, COD, and many other parameters.
2.7.2 Suspended Solids
Suspended solids are the typical contaminants in wastewater and are originated from either internal or external sources. The external sources are usually the influent and atmospheric inputs whereas planktons and animal and plant detritus are created within the wetland. Wastewater commonly constitutes 99.9% water and the rest 0.1% is solids. Hence, suspended solids are essential parameter in water quality monitoring and therefore applied to measure the quality of influent and effluent, and also to evaluate the performance of many processes (Thomas and William, 2001; Kayombo et al, 2003; Frank, 2003; Kadlec and Wallace, 2009).
Contaminants such as organic compounds, nutrients, and heavy metals are constituents of suspended solids. These contaminants can exist in particulate form in wastewater or they can be bound to other particulate matter either physically or chemically. Therefore, the removal of contaminants from wastewater and water source through sedimentation of suspended solids can be effective in conditions where the mass of the contaminant load binds with particulate matters (Thomas and William, 2001).
Wetlands are able to remove suspended solids from wastewater efficiently (Thomas and William, 2001; Avsar et al, 2007; Kadlec and Wallace, 2009). The system has normally extended HRTs and low flow velocity, which create desirable conditions for easily removal of settleable solids by gravitational settlement. Alternatively, processes including biodegradation, adsorption on submerged parts of the plant and wetlands media, filtration, and flocculation/precipitation are involved in the removal of colloidal or non-settling solids. In the removal of suspended solids, the nature and size of contaminant solids and the type fill media are the major factors on which the practical action of each removal mechanism depends (Kayombo et al, 2003; Robert, 2004; Vymazal, 2008; Azni et al, 2010).
Lana et al (2013) pointed out that the decreasing volume of wastewater to feed the wetland by increasing the batch frequency possibly raises the HRT. Then when the HRT becomes longer, the contact between bacteria in the wetland system and wastewater becomes better and the retention of suspended solids will be improved. This condition enables the system to present better pollutant removal efficiency.
In general, the settling rate of particles and the wetland length highly affects the efficiency in removing SSs. The process of sedimentation is thought to be irreversible. Nevertheless, suspended solids may be released from the sediment as re-suspension as a result of high flow velocity, wind driven turbulence, animals and humans disturbance, and gas lift occurred by oxygen, methane and carbon dioxides. Particle re-suspension does not occur in HSSF and VF CWs although excessive biological growth, creating head loss through the wetland system, leads to overland flow in HSSF system and complete failure of VF wetland (Thomas and William, 2001; Kadlec and Wallace, 2009).

2.7.3 Removal of Nutrients
The other pollutants of concern for water bodies are nutrients; for the most part, nitrogen and phosphorous. Usually, either phosphorous or nitrogen is the limiting factor in an aquatic environment. Therefore, discharging of untreated wastewater, containing excess nutrients, into water bodies results the onset of eutrophication in water bodies. Moreover, it can cause other adverse effects, for example, ammonia toxicity to aquatic life and public health problem because of the presence of excessive nitrate in drinking water (Russo, 2008). Nicholas (2002) depicts that taking out N or P from wastewater is referred to as nutrient removal.
Different forms of nutrients undergo through different routes of transformation among different wetland sections including substrate, plants, water, and litter. Then, wetlands are considered as nutrient sink following this transformation pathway for nutrient cycling. CWs are considered to be less efficient in removing nutrients (Rousseau, Vanrolleghem and Pauw, 2004; Vymazal, 2005; Konnerup, Koottatep and Brix, 2009; Abou-Elela et al, 2014).
However, they still have a good stand to be reliable alternative treatment methods to alleviate the problems associated with indiscriminate discharge of nutrients. The mechanisms for the removal of nutrients include: direct plant uptake, uptake by algae and bacteria, chemical precipitation, soil absorption, nitrification/denitrification, and reduction by fish and insect uptake (Li et al, 2008; Russo, 2008; Kadlec and Wallace, 2009; Qasaimeh, Alsharie and Masoud, 2015).
Nutrient uptake by plants shows a higher fraction of N removal in FWS, and a higher fraction of P removal in SSF CWs. The contribution of plants in the removal of nutrients is considered to be high (Konnerup, Koottatep and Brix, 2009; Lee et al, 2013; Bialowiec, Albuquerque and Randerson, 2014; Yucong et al, 2016; Wang et al, 2016). In order to increase the nutrient removal rate by plant uptake, it is suggested that the treatment regions, in the CW need to be covered by plants (Lee et al, 2013). Even though the capacity of plants to take up nutrients varies from species to species, it is more dependent on individual plant biomass irrespective of plant type, i.e., on the size of individual plants or plant density (Adhikari et al, 2011; Dzakpasu et al, 2015).

2.7.3.1 Nitrogen removal
Nitrogen is among the most important constituents in wastewater since it causes eutrophication, toxicity to aquatic life and undesirable consequence on the level of oxygen in the water bodies (Kadlec and Wallace, 2009). Proteinaceous matter and urea are the two principal components in which nitrogen can be available in domestic wastewater. If the condition in the pretreatment is kept anaerobic, ammonia, ammonium (NH4+) is produced as a result of the breaking down of protein and urea and the remaining organic N will be converted to ammonium by the process of ammonification (Scott, 2004).
The presence of N in the environment has many forms even though the transformations among those forms may occur quickly. Substantial quantity of both organic and inorganic N forms can exist in wastewater. In wetlands that are designed to treat municipal or domestic wastewater, the most significant N forms are ammonia (NH4+), nitrite (NO2?), nitrate (NO3?), nitrous oxide (N2O), and dissolved elemental N or dinitrogen gas (N2). In most cases, dissolved forms of N are commonly present in wetlands even if little particulate N can exist in settled wetland surface waters. Nitrite and, particularly, nitrate nitrogen are often found in waters where there is adequate oxygen while ammonium is the most common in anaerobic wetland soils (Thomas and William, 2001; Tanner, 2004; Kadlec and Wallace, 2009).
The total or dissolved forms of N which are determined using common analytical methods (APHA, 1998) include:
Nitrate
Nitrite
Ammonia
Total Kjeldahl nitrogen (TKN) = (organic +ammonia nitrogen)
So, the following formulas can be derived from the above basic measurements (Kadlec and Wallace, 2009):
Oxidized nitrogen = nitrate + nitrite Eqn. (2.1)
Inorganic nitrogen = oxidized nitrogen +ammonia Eqn. (2.2)
Organic nitrogen = TKN – ammonia Eqn. (2.3)
Total nitrogen = TKN + oxidized nitrogen Eqn. (2.4)

The major transformation processes functioning in constructed wetlands are:
ammonification (organic N ? NH4+), nitrification (NH4+ ? NO2- ? NO3-) denitrification (NO3 – ? N2O ? N2), biological fixation (N2 ? organic N), nitrate ammonification (NO3- ? NH4+), anaerobic ammonia oxidation (ANAMMOX, NH4+ ? N2) and volatilization (NH4+ ? NH3) (Vymazal, 2008).
Nitrogen compounds also enhance plant growth, which in turn stimulates the biogeochemical cycles of the wetland. The wetland N cycle is very complex, and control of even the most basic chemical transformations of this element is a challenge in ecological engineering (Kadlec and Wallace, 2009; Despland et al, 2014). VF CWs remove ammonia-N successfully but very limited denitrification takes place in these systems. On the other hand, HFCWs provide good conditions for denitrification but the ability of these systems to nitrify ammonia is very limited. Therefore, various types of CWs may be combined (hybrid systems) with each other in order to exploit the specific advantages of the individual systems (Vymazal, 2007).
The level of N removal in wetlands system is relatively high, despite of the fact that the natural background level in the effluent is frequently greater than 1 mg/L due mainly to breaking down and release of the native organic matter (Thomas and William, 2001). There are many N removal mechanisms involved in CWs (figure 3). These include: volatilization, ammonification, nitrification/denitrification, uptake by plants and adsorption with wetland. The major removal mechanism in most of the CWs is microbial nitrification/denitrification. Ammonia is oxidized to nitrate by nitrifying bacteria in aerobic zones. Nitrates are converted to dinitrogen gas by denitrifying bacteria in anoxic and anaerobic zones (UN-Habitat, 2008; Wu et al, 2013).

Anaerobic zone Volatilization matrix absorption Biomass uptake Aerobic zone

Biomass uptake

Figure 2.3: Nitrogen transformation in constructed wetlands (Cooper et al, 1996)
2.7.3.1.1 Volatilization
Diffusion of dissolved compounds into the atmosphere is one of the possible mechanisms of contaminant removal in wetlands and the process is referred to as volatilization. Quite a lot of organic compounds are readily lost to the atmosphere from wetlands and other water surfaces since they are volatile. Volatilization of unionized ammonia (NH3) can result in considerable N removal if the pH of the water is high (greater than about 8.5). But if the water pH is low or neutral, ammonia N occurs virtually totally in the ionized form (ammonium, NH4+) which is not volatile (Thomas and William, 2001; Tanner, 2004; Kadlec and Wallace, 2009). In wetlands, high pH is created during the day time as a result of photosynthesis by algae and submerged macrophytes (Russo, 2008; Vymazal, 2008).
Vymazal (2008) explained that volatilization of ammonia is a physicochemical process in which ammonium N is known to be in equilibrium between gaseous and hydroxyl forms as indicated by the following equation (Vymazal, 2008):
NH3 (aq) + H2O ? NH4+ + OH- eqn. (2.5)
The removal of NH3 through volatilization from flooded soils and sediments are insignificant if the pH value is below 7.5 and very often losses are not serious if the pH is below 8.0. At a pH value of 8.0 approximately 95% of the ammonia N is in the form of NH4+. At pH of 9.3 the ratio between ammonia and ammonium ions is 1:1 and the losses via volatilization are considerably high (Vymazal, 2008).
In general, the rate of volatilization is controlled by the following factors: the NH4+ concentration in water, wind velocity, solar radiation, pH values, temperature, solar radiation, the nature and density of vegetation and the capacity of the system to change the pH value in diurnal cycles (absence of CO2 increases volatilization) (Russo, 2008).
2.7.3.1.2 Ammonification
Ammonification (mineralization) is a process whereby N-containing organic compounds for example proteins, amino sugars, and nucleic acids are biologically degraded to ammonium (NH4+) (WI, 2003). It is the primary step in the mineralization of organic matter and can take place under aerobic or anaerobic conditions. The group of heterotrophic microorganisms is normally deemed to be involved in ammonification process (Scott, 2004; Kadlec and Wallace, 2009). Therefore, in order to achieve higher ammonia utilization in the wetland, more favorable environmental conditions for ammonia oxidizing bacteria should be established in the HSSF wetland (Truu et al, 2005).
The common ammonification reactions are shown below (Kadlec and Wallace, 2009):
Urea breakdown
NH2CONH2 + H2O ? 2NH3 + CO2 Eqn. (2.6)
Amino acid breakdown
RCH(NH )COOH + H2O ? NH3 + CO2 Eqn. (2.7)
Animal and plant tissues and excreted urea are the main sources of nitrogenous organic compounds. Domestic wastewater contains almost all N in the form of organic N or ammonia (Scott, 2004; Kadlec and Wallace, 2009). Eventually all organic N is degraded into ammonia (NH3), during pretreatment or soil-based treatment processes (Scott, 2004). Ammonium (NH4+) is primarily resulted from mineralization of N within flooded wetland soils. The rhizome and the root systems of macrophytes can absorb the soil-bound ammonium and another process which is carried out by anaerobic microorganisms can again reconvert this ammonium to organic matter (Scholz, 2006).
Under aqueous conditions, ammonium (NH4+) is formed by the rapid hydrolysis of ammonia (NH3), as shown in the following equation (Scott, 2004).
NH3 + H2O ? NH4+ + OH- Eqn. (2.8)
To all intents and purposes, the conversion of virtually all N to ammonium (NH4+) form can be deemed before the occurrence of further treatment (Scott, 2004). Then the ammonium mineralized from N-containing organic compounds does not stay for long period of time in the soil. Rather, it will be converted quickly to other forms of N in the soil/plant system via different processes (Vymazal, 2008).
The rate of mineralization in treatment wetlands system depends on various factors including microbial biomass, C/N ratio of the residue, temperature, available nutrients, pH value, extracellular enzyme, soil conditions such as texture and structure, soil redox conditions (Reddy and Patrick, 1984; Reddy and D’Angelo 1997). The pH range 6.5 and 8.5 is the optimum pH value for the process of mineralization. Unlike a number of microbiological processes, ammonification requires a temperature range of 40 to 60oC even though it is not likely to acquire these temperatures in the field (Reddy and Patrick, 1984) (Vymazal, 2008) (Russo, 2008).
Kinetically, the rate of ammonification is very faster than nitrification reaction (Kadlec, and Knight 1996), and it takes place at all degrees of soil aeration, although the rate varies depending on the level of aeration. It goes on at a much slower rate in flooded soil system than in drained-soil system (Reddy, 1982; Reddy and Patrick, 1983). Meanwhile, mineralization occurs at fastest rate in the oxygenated section, and the rate declines as mineralization changes from aerobic to facultative anaerobic and obligate anaerobic microflora (Reddy and Patrick 1984). Basically, in flooded soils, the aerobic zone has depth less than 1 cm, and hence the role of aerobic mineralization to the total mineralization could be very low, compared to facultative anaerobic and obligate anaerobic mineralization (Vymazal, 2008).
2.7.3.1.3 Nitrification
The process of converting ammonium nitrogen (NH4+), by chemoautotrophic bacteria to nitrate (NO3-) with nitrite (NO2-) as an intermediate product in the reaction, is known as nitrification (Vymazal et al., 1998; Vymazal, 2007; Russo, 2008). In the first step (Eqn. 2.8), ammonium nitrogen is oxidized to Nitrite by Nitrosomonas and in the second step (Eqn. 2.9); the oxidation of nitrite to nitrate by Nitrobacter is taken place (WI, 2003). During the oxidation process, about 7.14 g of alkalinity as CaCO3 are consumed, and 4.3g of 02 are required to convert 1g of ammonium nitrogen to nitrate. The process of nitrification depends on temperature and pH (EPA, 1993; USEPA, 2000; Scholz, 2006).
Nitrosomonas
2NH4+ + 3O2 ? 2NO2- + 2H2O + 4H+ + energy Eqn. (2.8)
Nitrobacter
2O2- + CO2+ O2 ? 2NO3- + energy Eqn. (2.9)

Robert (2004) pointed out that nitrification and denitrification processes are the most important mechanisms for nitrogen removal. Nitrification may occur by suspended bacteria or within any aerobic biofilms in aerobic regions of the soil and surface water. Nitrate remains in the water or pore of water of the sediments as it is not immobilized by soil minerals. It may be absorbed by plants or microbes in assimilatory nitrate reduction or may undergo dissimilatory nitrogenous oxide reduction, denitrification (USEPA, 2000; Thomas and William, 2001). A little ammonium nitrogen which exists in NH3 form will be liberated to the atmosphere through volatilization at elevated pH of 10. Generally, the most suitable sites for nitrification to take place are the oxidized layer and the submerged portions of plants (Azni et al, 2010).
Nitrification occurs in virtually all types of CWs; but the availability of oxygen affects the degree of the process. In the majority of CW types nitrification is a limiting process for the removal of nitrogen since NH3-N is the prevailing nitrogen types in various wastewaters, and in general, DO concentrations greater than 1.5mg/L are necessary for nitrification to take place (Ye and Li, 2009; Thomas and William, 2001).
Tuncsiper (2009) revealed that there is higher N removal efficiency in summer as average temperature rises to 23oC. It is also described that HSSF wetland system shows higher NO3- removal while NH4+ removal is lower. Apparently, the system provides suitable environmental conditions for denitrification but limited conditions for nitrification. In contrast, the VSSF wetland system has a higher NH4+ removal efficiency as VSSF system has better aeration.
In a FWS system, nitrification occurs at the aerobic zone in the water column. Then, the nitrate disperses into the sediments where anaerobic conditions required for denitrification exist. In this system, although the concentration of oxygen for nitrification is limited, both nitrification and denitrification processes can occur to remove nitrogen (Robert, 2004). Zhang et al (2012) described that the presence of wetland plants considerably improves both oxidation of ammonia and removal of TP in both batch and continuous types of operation as compared to that for unplanted beds.
The removal rate of N with HF CW system alone is low due to the deficiency of nitrification. On the other hand, single-stage VF CW cannot attain high N removal since environmental conditions that favor denitrification process lack. High removal of N can be realized in hybrid CWs system where the combination of HF and VF beds is applied (Canga et al, 2011).
2.7.3.1.4 Denitrification
The dissimilatory biological reduction of nitrate nitrogen to nitrogen gas under anaerobic or anoxic conditions is termed as denitrification (Eqn. 2.10). In this process, organic carbon is used as electron donor while nitrate act as alternate electron acceptor (EPA, 1993) (Thomas and William, 2001; USEPA, 2000; Kayombo, 2003). Organic compounds are oxidized by chemoheterotrophic denitrifiers for energy and carbon source. Some of these denitrifying bacteria are: Pseudomonas, Micrococcus, Bacillus and Achromobacter (Brady and Weil, 2002) (Russo, 2008). Denitrification is carried out mainly in the sediments of the wetland and in the periphyton films where the availability of carbon is high and the concentration of DO is low. For denitrification to take place the minimum carbon to nitrate-nitrogen ratio would be approximately 1 g C/g NO3-N (USEPA, 2000). The process of denitrification is demonstrated in the following equation:
Denitrifying bacteria
6(CH2O) + 4NO3- ? 6CO2 + 2N2 + 6H2O Eqn. (2.10)
Where CH2O represents biodegradable organic matter (Scott, 2004)
This reaction occurs under anaerobic or anoxic conditions and it is irreversible by its nature. In the reaction nitrogen acts as electron acceptor instead of oxygen. Meanwhile, a number of evidences from pure culture studies have made it clear that denitrification can occur in the presence of oxygen. Therefore, nitrate reduction may start in water logged sediments before the oxygen is depleted (Kuenen and Robertson, 1987; Laanbroek, 1990; Vymazal, 2008). Denitrification does not take place in the presence of oxygen theoretically. But the reaction has been shown in suspended and attached growth treatment systems which constitute fairly low DO concentration,, but not above 0.3–1.5 mg/L (U.S. EPA, 1993; Kadlec and Wallace, 2009)
As denitrification progresses, the N2 is then lost to the atmosphere for permanent removal, and is not stored in the wetland. The removal of ammonium in wetlands can occur as a result of the sequential processes of nitrification and denitrification. Ammonium is transformed to nitrate in aerobic regions of the soil and surface water. Then, the newly formed nitrate may undergo denitrification when it diffuses into the deeper or anaerobic regions of the soil. The coupled processes of nitrification and denitrification are universally important in the cycling and bioavailability of nitrogen in wetland and upland soils (Thomas and William, 2001).
In most CW types, denitrification plays the major role in the removal of nitrogen although the nitrate concentration in wastewater is usually low (Thomas and William, 2001). Environmental factors known to influence denitrification rates include the absence of O2, redox potential, substrate moisture, temperature, pH value, presence of denitrifiers, substrate type, organic matter, nitrate concentration and the presence of overlying water (Vymazal, 2007; Russo, 2008). However, Robert (2004) explained that pH, temperature, organic carbon, nitrate levels, and the ecological interactions and exposure times of the denitrifying bacteria within the system are the key factors. In general, although the reaction dependent on a number of factors, denitrification is the permanent removal of Nitrogen from the system (WI, 2003).
2.7.3.1.5 Plant Uptake
Plants take up nutrients to maintain normal metabolism processes and show an average N:P ratio of about 7:1 under natural conditions. But luxury uptake of N and P by plants can be resulted in situations where there is high concentration of these nutrients (Robert, 2004; Kadlec and Wallace, 2009). Nutrients uptake is usually the function of the roots in wetland soils of nutrients is normally the function of the root systems in the wetland soils and sometimes adventitious roots which are found in the water column. Nutrients may reach up to the leaves and stems of wetland plants (Kadlec and Wallace, 2009).
In addition to microbial removal mechanisms, plant uptake and storage of nutrients in the sediment could be the chief N conversion and removal routes in CWs in treating wastewater (Wu et al, 2013). In planted CWs system, there is higher N and COD removal (Wang et al, 2016), and plant uptake is one of the major means to remove nitrate produced by the process of nitrification (Robert, 2004).
Nitrogen in the mineralized state is taken up by wetland macrophytes and used to build plant biomass. As plants die out, some of the accumulated N in plant tissues can leach into the wetland system, and therefore, there is no net N removal through plant uptake (WI, 2003). But the removal of N by plant uptake can be efficient if plant biomass is regularly harvested and removed from the wetland system (Robert, 2004; Li et al, 2008; Azni et al, 2010). Adhikari et al (2011) pointed out that harvesting of aboveground plant parts is sufficient for N removal since there is usually higher N concentration in those parts. Lee et al (2012) reported that the average phosphorus contents in aboveground tissues of plants obtained ranges between 1.2± 0.7 to 2.4± 1.0 mg/g.
In the process of assimilation, the plants reduce inorganic N to organic N compounds, plant structure. There is significantly high rate of N uptake by wetland plants from water and sediments during the growing season. Increased immobilization of nutrients by microbes and uptake by algae and epiphytes also lead to retention of inorganic N. The net annual uptake of N by macrophytes approximately ranges between 0.5 to 3.3 g N/m2/yr (USEPA, 2000). However, N uptake by wetland plants reduces while its concentration and load in wastewater rises up. This shows that plants capability for N uptake is limited and it can be considered as efficient method under situations where the load of N is minimal (Avsar et al, 2007). Zheng et al (2016) reported that plants nutrients uptake accounted for a higher proportion of the N removal in FWS, and higher proportion of P removal in SSF wetland system.
Generally, macrophytes enhanced N removal and processing while reducing GHG fluxes compared to unplanted CWs (Landry, Maranger and Brisson, 2009). Plants utilize nitrate and ammonium, and decomposition processes release N back to the water. There are two direct effects of vegetation on N processing and removal in treatment wetlands (Kadlec and Wallace, 2009):
The plant growth cycle seasonally stores and releases N, thus providing a “flywheel” effect for a N removal time series.
The creation of new, stable residuals accretes in the wetland. These residuals contain N as part of their structure, and hence accretion represents a burial process for N.
2.7.3.1.6 Matrix Adsorption
Unlike the oxidized forms of N, ammonium nitrogen (NH4+-N) can bind to inorganic and organic solid substrates, because of the positive charge it possesses. Ammonium ion is adsorbed onto active cation exchange sites of the wetland bed matrix. In FWS, the ionized ammonia in water can be removed through exchange with inorganic sediments and plant detritus, or with the wetland media in the case of SSF systems. At a given ammonia concentration in the water column, a fixed amount of ammonia is adsorbed to and saturates the available attachment sites. But when the ammonia concentration in the water column is reduced due to factors such as nitrification, some ammonia will be desorbed to regain the equilibrium with the new concentration. If the ammonia concentration in the water column is increased, the adsorbed ammonia also will increase (Kadlec and Wallace, 2009).
But when the chemistry of the water is changed, the adsorbed ammonia is leached back into the water system since it is bound loosely to the substrate. Furthermore, the sorbed ammonium can be converted to the oxidized form, nitrate if there are conditions such as periodic draining, in which the substrate of the bed is exposed to oxygen. Hence, ammonium adsorption is a reversible removal process (WI, 2003; Kadlec and Wallace, 2009).
2.7.3.2 Phosphorus Removal
Phosphorus (P) is one of the essential macronutrients required by plants for growth, and is a limiting factor for the growth of vegetation (Kadlec and Wallace, 2009). Hence, addition of P to the environment often contributes to the occurrence of eutrophication of lakes and coastal waters (Thomas and William, 2001). A measure of relative ecosystem requirements is the proportion among the nutrient elements in the biomass, which is often represented as a molar proportion of C: N: P =106:16:1, or 41:7:1 on a mass basis (the Redfield ratio). Wastewaters do not have this ratio except by rare chance, and most often, there is excess P in domestic wastewater. The introduction of trace amounts of this element into receiving waters can have profound effects on the structure of the aquatic ecosystem (Kadlec and Wallace, 2009). Removal of P is required where the wastewater effluent is discharged into a lake or into a watercourse which later discharges into a lake (Brix, 2004).
The most reactive forms are the dissolved phosphates, which change hydration in response to pH. The most common species are mono- and dibasic phosphates, which dominate at all typical wetland pH values (4 ; pH 5-15 days) is required to allow the system to operate more in a steady state conditions for treatment of sewage to acceptable levels. In general, treatment wetlands show considerable potential for removing fecal bacteria from domestic wastewater (Fountoulakis et al, 2009; Vallejos, Caballero and Champagne, 2015; Sleytr et al, 2007).
Macrophytes-based systems turned out to be a good alternative for wastewater treatment concerning bacterial removal and water quality. In contrast, those systems without plants show lower efficiencies than their corresponding planted wetlands. It is also found that mean removal efficiencies and surface removal rates turn out to be significantly high in wetlands, and some increases in removal efficiencies are associated with warm season (Mercedes et al, 2008; Foladori, Bruni and Tamburini, 2015; Wu et al, 2016). A wetland tends toward a better performance during the maturity period reached by the system, noticeable through the presence of well-developed macrophytes (Zurita and Carreon-Alvarez, 2015).
According to Sharma and Brighu (2016), the major removal mechanism of microbes occurs due to the release of antimicrobial extract, especially from the rhizomatic part of the plant Canna indica (Abdullah et al, 2012, Gaur et al, 2014). Moreover, increased surface area facilitated by increased and fibrous roots may help to result into higher filtration and adsorption mechanism of microbial removal (Sharma and Brighu, 2016).
The removal of FCs by using treatment wetlands is increased when the following conditions are met (Kadlec and Wallace, 2009).
Longer nominal hydraulic residence time (t) or lower hydraulic loading
Finer bed materials (sand), but only to the extent that the fine bed media does not impair hydraulic performance
Warmer water temperatures
Shallower bed depths
Tuncsiper, Ayaz and Akca (2012) revealed that the HRT and the loading rates are two of the most important factors in removing coliforms although the rate can be affected by a number of other conditions and environmental factors.
In tropical and subtropical climates, it is possible to remove harmful pathogenic organisms and to produce disinfected reclaimed wastewater without using expensive disinfectants, in poor areas where the reclamation of raw wastewater in agriculture is endangering human health (Zurita and Carreon-Alvarez, 2015). Hence, application of wetland system is especially suitable for small communities in developing countries, where the potential health benefits from pathogen removal are considerable (Shutes, 2001).
2.7 Reaction Kinetics
Envisaging the performance of treatment wetlands is based on the theory that the systems act as plug-flow reactors or attached-growth biological reactors, through which the wastewater flows in lock step. Plug flow evidently provides a more suitable description of the pattern of water flow in CWs. The model is first order in the forward direction while the reverse direction is zero order. Therefore, the removal performance equation can be described by employing first order plug flow kinetics (Kadlec and Knight, 1996; Thomas and William, 2001; Kadlec and Wallace, 2009; Azni et al, 2010).
Similarly, the removal of BOD5 in SSF CW system can be explained with first-order plug flow kinetics and the soluble BOD5 is removed as a result of microbial growth attached to the plant roots, stems, leaf litters and substrates. The removal rate of a particular contaminant is directly proportional to the remaining concentration at any point within the wetland bed and it is known as first-order kinetics (Kadlec and Knight, 1996; Kayombo, 2003).
The two idealized mixing theories that can be applied in first order kinetics are:
Completely mixed reactor – the concentration is the same as the effluent concentration at any point in the reactor;
Plug flow – the reactant concentration decreases along the length of the flow path through the reactor.
A number of individual processes such as mass transport, sedimentation, volatilization, and sorption which are taken place in CWs are therefore mainly first-order. So, it is rational to deduce that the processes can behave in a similar manner in combination, at least over some range of pollutant concentration. The local removal rate equation (Kadlec and Wallace, 2009) is:
J = k (C-C*) Eqn (2.12)
Where:
J = removal per unit area, g/m2.d
K= rate coefficient, m/d
C = concentration, g/m3
C* = background constituent concentration, (g/m3)
This rate equation is the most prevalent in treatment wetland literature. Then, combining the basic equation for a plug-flow model with the water mass balance, an exponential relation between inlet and outlet concentrations can be described by integration of the previous equation (Kadlec and Knight, 1996):
K=Q/A ln ((Ci-C*))/((Ce-C*)) Eqn. (2.13)
Where:
As = surface area of a wetland (m2)
Q = input discharge to a wetland (m3/day) = 1.05m3
K= hydraulic loading rate (m/day) = 34 m/yr
Ci = inlet concentration (mg/L) = 200mg/L
Ce = outlet concentration (mg/L) =25mg/L, and
C* = background concentration (mg/L)
To overcome variability as a result of short term variation of inflow, areal removal rate constant is derived using time-averaged data. Despite this, Alley et al (2013) described that location and targeted constituent of wastewater can affect seasonal removal in a CW system and therefore seasonal factors such as temperature can be included in design features to maintain or enhance removal of targeted constituents.
The effluent concentration of targeted constituents is lower during the warm period as the performance is higher during this period compared to the winter period (Mietto et al, 2015; Prachaska, Zouboulis and Eskridge, 2007). Wu et al (2014) pointed out that the operation of CWs at cold climate is a challenge, and consequently various adaptations are initiated through specific design and enhanced operation strategy. The variability among different systems is thus a fact of life for treatment wetlands which are closely related to their climate and surrounding environment (Kadlec and Wallace, 2009). In general, the effect of temperature on areal removal rate constant (k) can be modeled as a modified Arrhenis equation as:
kT =k20 ?(T-20) Eqn. (2.14)
Where:
kT = rate constant at temperature T, d-1
k20 = rate constant at 20 oC, d-1
T = water temperature, oC
? = modified Arrhenius temperature factor, dimensionless
Where, kT is the rate constant at temperature T = T°C and k20 is the rate constant at 20°C. Values of the temperature correction factor (?) have been estimated for data sets with adequate operational temperature data (Kadlec and Wallace, 2009).

Table 2.4: Kadlec and Knight K-C* model design parameters
Parameters KA, 20 ? C* (mg/L)
BOD
TSS
Organic-N
TN
TP
FC
34
1000
17
22
12
75 1.00
1.00
1.05
1.05
1.00
1.00 3.5+0.053 Ci
5.1+0.16 Ci
1.5
1.5
0.02
300 cfu/100mL

Source: Treatment wetlands, by Kadlec and Knight, 1996, p. 217.
2.8 Application of Constructed Wetlands
Constructed wetlands with different designs have long been used primarily for the treatment of municipal or domestic wastewaters. However, because of the unique advantages of lower costs and additional benefits, the application of CWs is getting more attention and they have evolved into a dependable wastewater treatment system for the removal of a wide range of pollutants from a number of wastewater types during the last couple of decades of development. So, they are presently used for a wide variety of pollution, including agricultural and industrial wastewaters, various runoff waters, and landfill leachate (Vymazal, 2009; Haiming et al, 2015; Vymazal, 2014; Vymazal and Kropfelova, 2009). Accordingly, the current literature has been reviewed and the review on the applicability of constructed wetlands for various types of wastewater is presented in the following sections.

Table 2.5: Application of constructed wetlands for the treatment of different wastewater
Type of wastewater Level of application Location References
Pesticide polluted wastewater Operational South Africa Schulz and Peall, 2001
Domestic wastewater Experimental Turkey Korkusuz, Beklioglu and Demirer, 2004
Industrial wastewater Experimental Taiwan Chen et al, 2006
Domestic wastewater Experimental USA Prochaska, Zouboulis, and Eskridge, 2007
Farmyard runoff Operational Ireland Mustafa et al, 2009
Industrial wastewater Operational USA Knox et al, 2010
Animal farm wastewater Experimental Ireland Babatunde et al, 2011
Sugar factory wastewater Experimental Kenya Odinga, Otieno and Adeyemo, 2011
Swine wastewater Operational Brazil Sarmento, Borges and Matos, 2012
Tannery wastewater Experimental Bangladesh Saeed et al, 2012
Textile wastewater Experimental India Sivakumar et al, 2013
Winery wastewater Operational Spain Varga, Ruiz and Soto, 2013
Pulp and paper mill wastewater Experimental India Choudhary, Kumar and Sharma, 2013
Domestic wastewater Operational Tanzania Mahenge, 2014
Acid Mine Drainage Experimental South Africa Seadira et al, 2014
Storm water Operational Australia Mangangka et al, 2015
Agricultural and
Urban runoff Operational USA Pietro and Ivanoff, 2015
Diesel polluted wastewater Experimental Malaysia Al-Baldawi et al, 2015
Landfill leachate Experimental Colombia Madera et al, 2015
Food processing wastewater Operational France Paing et al, 2015
Dairy farm wastewater Operational USA Tuncsiper, Drizo and Twohig, 2015
Municipal wastewater Operational Central Jordan Albalawneh et al, 2016
Landfill leachate
and domestic wastewater Experimental Iran Mojiri et al, 2016
Leachate Experimental Iran Bakhshoodeh et al, 2016

2.8.1 Application of Constructed Wetlands for Domestic Wastewater Treatment
Like many other treatment technologies, treatment wetlands have been employed primarily for the treatment of wastewater from human dwellings and activities. The majority of applications of CW as a treatment method were associated with municipal and domestic wastewater, and the technology is still growing rapidly in many areas (Kadlec and Wallace, 2009; Russo, 2008). Many such systems are currently in use around the world, designed to treat domestic wastewater (Trivedy, 2010).
The number of CWs in use has very much increased in the most recent years. The use of CWs in the United States, New Zealand and Australia is gaining rapid interest. The systems are mainly used in towns and cities of those countries for tertiary treatment to remove nutrients of low concentration. However CWs are usually used in European countries as a method of secondary treatment for domestic of village populations (Russo, 2008).
Likewise, the enormous potential for large-scale treatment and high demand for clean water in the tropical and subtropical areas have become the driving forces for many developing countries in those locations to use the technology in order to solve pollution problems. Unfortunately, the available literatures regarding the application of CWs in those locations are comparatively low. It becomes visible that in some countries, basic researches are being conducted, while the technology has reached pilot and full-scale levels for various applications in other countries (Zhang et al, 2015; Kivaisi, 2001).
In recent times, researches related to the performance of CWs in treating domestic wastewater have been carried out in developing countries such as Tanzania (Mashauri, Mulungu and Abdulhussein, 2000, Mahenge, 2014), Uganda (Kyambadde et al, 2004; Kyambadde, Kansiime and Dalhammar, 2005), Malaysia ( Katayon et al, 2008), Thailand (Konnerup, Koottatep and Brix, 2009), Kenya (Kelvin and Tole, 2011), Cameroon (Fonkou Theophile et al, 2011), Egypt (Abou-Elela and Hellal, 2012; Abou-Elala et al, 2014; Abdelhakeem, Aboulroos and Kamel, 2016), Taiwan (Hsueh et al, 2014) Morocco (Laffat, Ouazzani and Mandi, 2015), and Pakistan (Sehar et al,2015).
To see some of the studies, Mashauri, Mulungu and Abdulhussein (2000) carried out a research on the performance of a horizontal flow constructed wetland applied for the treatment of wastewater effluent from waste stabilization ponds at Dar es Salaam University, Tanzania. The efficiency of the reed bed treatment system at low filtration rate (0.27 m/h) was 80% for TSS, 66% for COD, 90% for TC and 91% for FC. Similarly, in Uganda, Kyambadde et al (2005) assessed the feasibility of nutrient removal from wastewater using horizontal flow CWs. The wetland system was substrate-free and planted with two tropical plants: Cyprus papyrus and Miscanthidium violaceum. Accordingly, Results showed that the removal efficiencies for BOD, NH4–N and TP fractions in papyrus-based CWs were 86.5%, 68.6%, and 69.7% respectively, while for Miscanthidium-based system the values were 53.2%, 45.5%, and 46.7% respectively. Therefore, wetlands planted with Cyprus Papyrus showed higher removal efficiency than Miscanthidium based system.
In Malaysia, three SSF CWs (two planted with Lepironia articulate and one unplanted cell), operated at four different retention time, were tested to evaluate the performance in treating a mild domestic wastewater. The CWs were able to remove about 56–77% of COD, 50–88% of TSS, 20–88% of TP, 27–96% of NH4+ and 99% of total coliforms. The removal rates of COD, TP, and NH4+ were affected by different hydraulic retention times, but the rates for TSS and TC were not affected by the retention time (Katayon et al, 2008).
Konnerup, Koottatep and Brix, (2009) conducted a study on eight horizontal SSF constructed wetlands planted with Canna and Helichonia in Thailand. The system was designed to treat domestic wastewater at four different hydraulic loading rates: 55mm d-1, 110mm d-1, 220 mm d-1, 440 mm d-1. The result showed that the rates of mass removal for TSS, COD, TN and TP varied between 88-96%, 42-83%, 4-37%, and 6-35% respectively, depending on the loading rates. Although the removal of TN in beds planted with Canna was higher than beds planted with Helichonia, the removals of both TN and TP were low in the pilot-scale wetlands. In another study done in Kenya by Kelvin and Tole, (2011), the performance of the tropical constructed wetland was evaluated and comparison with the conventional treatment methods was made. In this study, the tropical CWs achieved the removal efficiency of 96.2% for BOD, 97.6% for COD, 21.4% for TP, 41% for ammonia and 99.99% for FC. Hence, the tropical CW system was found to perform much better than the commonly used waste stabilization ponds.
The performance of Cyprus Papyrus in HSSF and HSF constructed wetlands for the treatment of domestic wastewater was evaluated with the study carried out in Cameroon. As of the result of the study, the reductions of several physico-chemical parameters and FC in vegetated systems were not significantly different as compared with the non vegetated wetland system (Fonkou et al, 2011). In Egypt, Abou-Elela and Hellal, (2012) tested a pilot scale vertical flow constructed wetland for treating primary treated municipal wastewater. The wetland unit was employed with the surface area of 457.56m2 and influent flow rate of 20m3/day. The average removal efficiencies for BOD, COD, and TSS were 90%, 88% and 92% respectively.
2.8.2 Application of Constructed Wetlands for Storm water Treatment
Urbanization creates large impervious areas that increase the quantity and peak rate of runoff. Rainfall then washes deposited materials directly into surface waters, causing stream pollution (Robert, 2001). The composition of storm water varies greatly, depending on the surrounding land use. For example, urban runoff may contain soil particles, dissolved nutrients, heavy metals, oil, and grease. Residential and agricultural runoff may also contain organic matter and pesticides (USEPA, 2000; Thomas and William, 2001). Pollutant concentrations and loads generally range from low levels from undeveloped and park lands, to low density residential and commercial, to higher density residential and commercial, and finally to high density commercial and industrial land uses (Kadlec and Wallace, 2009).
The application of CWs for treatment of combined sewer overflow (CSO) is considered a sustainable approach with relevant potential and it can be interesting in an urban context (Amaral et al, 2013). The use of wetland retention basins for treatment of storm water runoff has become relatively common (Thomas and William, 2001), and FWS wetlands are the nearly exclusive choice for the treatment of urban, agricultural, and industrial storm waters, because of their ability to deal with pulse flows and changing water levels. In contrast to other applications, there is basically no pretreatment for urban stormwaters, if the forebay settling basin is considered part of the wetland (Kadlec and Wallace, 2009).
Gan (2004) stated that wetlands are an effective storm water treatment measure for the removal of fine SSs and associated contaminants, as well as soluble contaminants. These systems utilise a combination of physical, chemical and biological processes in removing storm-water pollutants. They are used as “end-of-pipe” or at “source control measures” (DLWC, 1998). Mangangka et al (2015) highlighted the importance of ensuring that the inflow into the wetland has low turbulence in order to achieve consistent treatment performance for both, small and large rainfall events.
2.8.3 Constructed Wetlands for Industrial Wastewater Treatment
Industrial wastewater is here a loosely defined category, including wastewaters that are not from domestic, municipal, animal, or food product processing (Kadlec and Wallace, 2009). CWs have been increasingly applied in treating various industrial wastewaters with specific characteristics (Al-Baldawi et al, 2013; Vymazal, 2014; Wu et al, 2015). A reliable and stable effluent can be obtained using the CW system for treating industrial wastewater treatment (Chen et al, 2006).
Vymazal (2014) revealed that all types of CWs have been used with most systems being either free water surface CWs with emergent vegetation or horizontal subsurface flow CWs. The use of vertical flow CWs for treatment of industrial wastewaters has been less frequent so far. However, vertical flow CWs have been successfully used for treatment of olive mills wastewaters and also in hybrid CWs. The treatment technology of CWs has evolved in a reliable technology which is nowadays successfully used for many types of industrial effluents.
The application of CWs system for the treatment of various types of industrial wastewaters is listed below.
Acid mine drainage (Johnson and Hallberg, 2005; Seadira et al, 2014)
Control of landfill lecheat (Madera-Parra et al, 2015; Mojiri et al, 2016)
Electroplating wastewater (Sudarsan et al, 2015)
Industrial park wastewater (Chen et al, 2006)
Tannery wastewater (Leta et al, 2004)
Textile wastewater (Sivakumar et al, 2013)
Refinery effluent (Robert and Kadlec, 1996)
Pulp and paper wastewater (Pokhrel and Viraraghavan, 2004)
Winery wastewater (Rozema, Rozema and Zheng, 2016)
2.8.4 Application of Constructed Wetlands for Agro-Industrial Wastewater Treatment
Agro-industrial wastewaters can be very strong in terms of pollutant concentrations and hence can contribute significantly to the overall pollution load imposed on the environment. Examples of agro-industrial wastewaters include those arising from industrial-scale animal husbandry, slaughterhouses, fisheries, and seed oil processing (Jern, 2006).
It has, however, been noted that wastewaters with COD: BOD5 ratios of 3 or lower can usually be successfully treated with biological processes. COD: BOD5 ratios of 3 or lower are encountered in many of the agricultural and agro-industrial wastewaters. But agro-industrial wastewaters need not always have such low COD: BOD5 ratios. For instance tobacco processing wastewater can have a COD: BOD5 ratio of about 6:1. This is a strong wastewater which can be difficult to treat to meet COD discharge limits because residual organics following biological treatment are resistant to further biological treatment (Jern, 2006).
Wastewater from the dairy industries that does not receive some form of biological pre-treatment contain almost twice the content of soluble organics as a percentage of the total load compared with municipal wastewater (Milne and Gray, 2013). Hu et al, (2011) described in his study that significant variety of livestock wastewater in biodegradability during long-time storage was observed and the laboratory results showed that fresh livestock wastewater was readily biodegradable, while it turned to be non-biodegradable after long-time storage.
Despite of these, CWs efficiently reduced BOD5 and TSS in dairy effluent. The BOD5 and TSS reduction efficiencies were significantly greater during the best growing seasons of plants and the best seasons of microbiological activity. CWs have the potential as a recommendable practice for the treatment of BOD5 and TSS contained in dairy farm effluents under cold climate conditions. The BOD5 and the TSS treatments by CWs were enhanced by connecting two CWs in-series (Tuncsiper, Drizo and Twohig, 2015). Cortes-Esquivel et al, (2012) also indicated that CWs can be a very useful tool for the removal of heavy metals like Zn and Cu in swine wastewater. Similarly, wetlands may be used effectively for treatment of animal and aquaculture wastes (Thomas and William, 2001).
2.8.5 Constructed Wetlands for Leachate Treatment
Treatment and disposal of liquid leachates is one of the most difficult problems associated with the use of sanitary landfills for disposal of solid waste. Leachates are produced when rainfall and percolated groundwater combine with inorganic and organic degraded waste. The highly variable nature of solid waste, differences in age and decomposition, and the diversity of chemical and biological reactions that take place in landfills result in a wide range of chemical quality of leachates (Mulamoottil et al., 1998) (Kadlec and Wallace, 2009).
Madera-Parra et al (2015) revealed that CWs effectively remove COD, BOD5 and nutrients (TKN, NH4+–N, PO43–P) from pretreated landfill leachate and except for NH4+–N achieved concentrations below the provisional standard. Additionally, they attain reduction levels similar to those obtained with highly mechanized systems. Hence, the development of these types of CWs at a full scale is an attractive technology for landfill leachate treatment in countries with low resources and high necessities to protect the environment and public health.
It is reported that the levels of particular contaminants in municipal landfill leachate exceed the allowable discharge restrictions for colour, COD, ammonia, Ni, and Cd. Pollutants from landfill leachate wastewater can be removed by using a CW system. CWs show removal efficiency of 90.3%, 86.7%, 99.1%, 86.0%, and 87.1% for colour, COD, ammonia, Ni, and Cd, respectively. Removal efficiencies decrease as leachate ratio in the leachate and wastewater mixture increased (Mojiri et al, 2016; Bakhshoodeh et al, 2016).
2.8.6 Constructed Wetlands for Acid Mine Drainage Treatment
The major producer of acid mine drainage is the mining industry. Waters draining active and, in particular, abandoned mines and mine wastes are often net acidic. Such waters typically pose an additional risk to the environment by the fact that they often contain elevated concentrations of metals (iron, aluminium and manganese, and possibly other heavy metals and metalloids of which arsenic is generally of greatest concern) (Johnson and Hallberg, 2005).
Treatment wetlands typically operate at neutral pH for influents that are not strong acids or bases. This is true for both FWS and SSF CWs, but low influent pH levels are the norm in case of acid mine drainage. Some HSSF wetlands are designed with reactive media in the bed material such as zeolite that can produce high pH effluents. Although CWs often operate at pH 6.5, acid mine drainage wetlands function with incoming pH less than 5, which is commonly regarded as a lower limit for aquatic resource protection (Kadlec and Wallace, 2009).
Microbiological processes that generate net alkalinity are mostly reductive processes and include denitrification, methanogenesis, sulfate reduction, and iron and manganese reduction. Ammonification is also an alkali-generating process. Due to the relative scarcity of the necessary materials (e.g., nitrate), some of these processes tend to be of minor importance in acid mine drainage-impacted environments. However, in as much as both ferric iron and sulfate tend to be highly abundant in acid mine drainage, alkali genesis resulting from the reduction of these two species has a potentially major significance in acid mine drainage-impacted waters. Photosynthetic microorganisms, by consuming a weak base (bicarbonate) and producing a strong base (hydroxyl ions), also generate net alkalinity (Johnson and Hallberg, 2005)
2.8.7 Constructed Wetlands for Agricultural Runoff Treatment
The use of CWs to improve the quality of wastewater or the water from mining exploitations and agriculture is a technology under way of development (Muresan, 2012).Concurrent high removal rate of COD, ammonia, and phosphorous can be obtained in a two-stage CW system, demonstrating its potential use for cost-effective reduction of pollution load of agricultural wastewaters (Babatunde, 2011).
The efficiency of a FWSF CW in treating agricultural discharges was investigated during storm and non-storm events. Overall, the results indicated that the design of the CW system could feasibly function for the retention of typical non-point source pollutants like suspended solids, excess nutrients and organic matters. The hydraulic fluctuations and increase in pollutant concentrations during storm events made the system more efficient in addition to the moderate temperature (greater than 150C) during the storm seasons. Although the overall mass may not have been removed, the levels of pollutants were reduced to appreciable levels. More importantly, CWs contributed to the improvement of stream water quality thereby reducing the potential impact of pollutants downstream (Maniquiz et al, 2012).
The long-term assessment demonstrates that the example integrated CWs system can be considered an effective and sustainable wastewater treatment option for agricultural runoff rich in nutrients (Mustafa et al, 2009). CWs should be considered an option for current agricultural wastewater applications. Many have adopted this view, and, as a result, there are a large amount of full scale, functional CWs found throughout North America, and the world, that are being used to treat various types of wastewater. There is no one CW design that is the most effective for agricultural wastewater, but, rather, each design has strengths and weaknesses so hybrid designs may prove to be the most practical (Rozema et al, 2016).
Scholz et al (2010) described that integrated constructed wetlands (ICW) are capable of treating farmyard dirty water and that they provided a sustainable management option to effectively reduce nutrient and contaminant loss from farmyards to watercourses. Significant concentration reductions in suspended organic matter, nutrients and faecal bacteria between ICW influents and effluents were observed. Surface discharges from the ICW sites had seasonal patterns. None of the farm ICW had surface discharges during most summer months.
To treat low C/N ratio wastewaters, such as nitrate-rich agricultural runoff and polluted groundwater, the carbon source only from the root exudates of macrophytes is not sufficient to maintain a high performance of nitrate removal. Denitrification can be enhanced by the external supply of electron donors via direct organic carbon addition using organic filtration media and/or step feeding operation. However, the potential secondary pollution should be considered. The promotion of autotrophic denitrification, especially via the pathway of microbial anammox, could be a potential promising strategy (Wu et al, 2014).
2.8.8 Constructed Wetlands for Pesticide Treatment
CWs have become the best management practice for pesticide mitigation from non-point source agricultural runoff and drainage in many countries. So far, CWs with free water surface have been primarily used, while the SSF CWs have recently been used as well. As both aerobic and anaerobic processes are involved in pesticide removal hybrid CWs may offer efficient solution. Current survey indicated that removal of pesticides is generally effective, but the efficiency varies widely among pesticides and also among systems for a particular pesticide (Vymazal and Brezinova, 2015).
There are many processes which are responsible for pesticide mitigation such as hydrolysis, photolysis, sedimentation, adsorption, microbial degradation or plant uptake, however, the extent of these processes depends on local conditions, and it is difficult to single out the most important ones (Vymazal and Brezinova, 2015).
There is strong evidence to suggest that the presence of vegetation enhances pesticide retention. The results of the survey revealed that the highest pesticide removal was achieved for pesticides of the organochlorine, strobilurin/strobin, organosphosphate and pyrethroid groups while the lowest removals were observed for pesticides of the triazinone, aryloxyalkanoic acid and urea groups (Vymazal and Brezinova, 2015).
Organophosphorus pesticides were found in the outlet suspended-particle samples, highlighting the retention capability of the wetland. A toxicological evaluation employing a Chironomus bioassay in situ at the wetland inlet and outlet revealed an 89% reduction in toxicity below the wetland during runoff (Schulz and Peall, 2001).
Macrophyte vegetated wetlands have the potential to contribute to aqueous-phase pesticide risk-mitigation….It can be concluded that the conservation and management of vegetation in small drainage channels may be an effective tool to avoid agricultural pesticide contamination of larger receiving water bodies (Schulz et al, 2003).
A long water residence time improves the effectiveness of CWs, favouring relatively fast processes, such as sedimentation and pesticide sorption, and subsequently removing pesticides from the water phase. Thus, these processes result in pesticide storage in the solid phases and can reduce the effectiveness of CWs by saturating the sorption sites. Pesticides can be degraded from the solid phase in CWs during periods without water flow, which reduces their accumulation and risks decreasing their effectiveness over time. In addition, the effectiveness of CWs may be improved by increasing the water residence time (Romain, Sylvie and David, 2015).
2.9 Studies on Performance of Constructed Wetlands in Ethiopia
Even though the potential of the application of CWs for treating various wastewater types is tremendous, they are not commonly used in Ethiopia. Only few institutions have employed operational CWs for treating domestic wastewater, and obviously some other industries such as tanneries, breweries, and coffee mills are starting to implement the system in recent times (Birhanu, 2007; Asaye, 2009; Kenatu, 2011; Connie, 2012).
In the mean time, many studies to evaluate the performance of pilot-scale and in few cases full-scale CWs for different wastewaters have been carried out. Birhanu (2007) conducted a study to evaluate the removal performance of horizontal SSF CWs system constructed at JWBO, for treating domestic wastewater. Based on the results of his study, the average removal efficiency of the treatment system were 99.3% for BOD5, 89% for COD, 85% for TSS, 28.1% for NH4 +-N, 64% for NO3–N, 61.5% for TN, 28% for PO4+, 22.7% for TP, 77.3% for SO4+, 99% for S2-, 94.5% for TC and 93.1% for FC. The result also mentioned that wetland beds planted with Cyprus papyrus showed higher removal efficiency for NO3–N, NH4+-N, TN, PO43-, and TSS than the other wetland cells, while beds planted with Phoenix canariensis showed higher removal efficiency for TP, S2-, BOD5, COD, TC and FC.
Asaye (2009) conducted a study to evaluate selected plant species (Cyprus papyrus, Typha domingensis, Cyperus alopcuroides, Schenoplectus corymbosus, Sesbania sesban, Aeschynomene elaphroxylon) for the treatment of tannery wastewater. Accordingly, he reported that the wetland cell planted with Cyperus Papyrus showed high removal efficiencies for NO3-(73.2%) and NH4+ (26.2%). High removal efficiencies for total Cr (98.4%), COD (68.7%) and S2- (59.2%) resulted in the cell planted with Schenoplectus corymbosus, cell cell planted with Sesbania sesban showed removal efficiencies for SO42- (96.3%), BOD5 (84.7%) and TN (58.3%).
In another experimental study, Kenatu (2011) carried out a study to evaluate the performance of SSF constructed wetlands planted with Canna indica, Phragmite karka, and colocacia gigantean for the treatment of wastewater from Breweries. The result showed that the average removal efficiencies of planted CWs for BOD5, COD, TN, NH4+-N, NO3-N, TP, PH43-, SO42-, S2-, TSS, TDS, and EC were: 74.4%, 78.9%, 77.4%, 62.8%, 55.2%, 68 %, 81.4%, 36.1%, 97.3%, 61.7%, 54.1%, and 27.4% respectively. She also reported that the values of some of the effluent parameters complied with the provisional emission standard limits.
Similarly, the performance of HSSF CWs planted with Phragmites australis in removing heavy metals from landfill leachate was evaluated. The experimental study revealed that the removal efficiencies were 99.33% for Fe, 93.67% for Mn, 89.24% for Pb, 96.14% for Cu and 98.33% for Zn. The result also showed that heavy metal uptake by the root system is higher than the uptake by the stem and leave (Mesele, 2013).
Likewise, in another study, Yezbie (2014) evaluated nutrient uptake efficiency and growth of Cyperus payrus and Phragmites karaka. The results of the study showed that Cyperus papyrus showed higher rate of biomass accumulation as evidenced by increase in shoot and root weights (83.93 gm) compared to Phragmites karka. Cyperus papyrus showed higher accumulation of TP in the root system and TN in leaves than Phragmites karaka. The mean removal efficiencies of the cell planted with C. papyrus were 56.37% for NO3–N, and 84.04% for PO43-, while removal efficiencies were by the cell planted with P. karaka were 58.37% for NO3—N and 65.18% for PO43-.
In this study, the performance of the pilot scale constructed wetland systems comprised of the horizontal flow, vertical flow and the hybrid of the horizontal and vertical flow employed in Kotebe WWTP plant for the treatement of domestic wastewater was evaluated.

CHAPTER THREE
MATERIALS AND METHODS
3.1 Description of the Study Area
The study was conducted in Addis Ababa, the capital city of Ethiopia. Addis Ababa covers a total area of about 540 km2. According to the 2007 census, the population of Addis Ababa is 2.738 million, of which 51.6% are females while 48.4% are males (FDREPCC, 2008). The city lies between 2,200 and 2,500 meters above sea level and the lowest and the highest annual average temperatures are 9.89 and 24.64 °C, respectively (Dillon, 2005; AAWSA, 2002).
Although there are different seasonal classifications along with different geographical areas of the country, Ethiopia has four seasons; namely winter (December, January and February), spring (March, April and May), summer (June, July and August) and autumn (September, October and November). However, the seasons have opposite features compared to temperate regions: winter is the dry season and summer is the heavy rainy season in Ethiopia. Similarly, spring is the autumn season characterized by high temperature with occasional showers and autumn is the spring season in Ethiopia.
According to Daniel et al (2010), Addis Ababa generates an estimated annual volume of 45Mm3 domestic wastewater. However, the existing sewerage system is not adequate and access to piped sewerage system is limited to about 10% (World Bank, 2015). Most of the wastewater from housing units connected to the city’s sewer system is conveyed to Kality wastewater treatment plant that has not improved since 1993 (AAWSA, 2002). About 4,500 m3/day wastewater is transported and treated by Kality wastewater treatment plant (design capacity of 7,600m³/d). In addition to this, about 1,200m3/day transported by truck and treated by the drying beds.
Kotebe treatment plant is the other centralized treatment plant which was initially designed to receive and treat sludge from vacuum trucks that empty dry pit latrines and septic tanks, with annual capacity of approximately 150,000 m3 and it consisted of 20 drying beds and 10 lagoons (NEDECO, 2002). But, the treatment plant has been under continuous expansion and connected to the sewerage system to receive and treat wastewater using stabilization ponds, starting from some years back. Kotebe wastewater treatment plant is located in the north-eastern outskirts of Addis Ababa. It has a capacity of 2,000 m3 per day and serves primarily condominiums which are equivalent to about 5000 households (World Bank, 2015).
Therefore, the pilot-scale constructed wetland systems were implemented at Kotebe WWTP and the specific area was situated at a longitude of 38o51’9.67” E, a latitude of 8o58′ 14.73” N and an altitude of 2,266m.
3.2 Field visit and site selection
Following the perrmission obtained from Addis Ababa Water and Sewerage Authority (AAWSA) to undertake the study at the compound of Kotebe wastewater treatment plant, repeated field visits were conducted at the site to gather detailed information before the application of pilot-scale constructed wetland systems. During the visit, relevant information about the general situations of the area was collected and evaluated against the technical feasibility of the implementation of pilot-scale constructed wetland systems.
Finally, in consultation with the head of the treatment plant, the free area which was located at about 165m down to the sedimentation tank of the main sewerage system was selected for the implemenation. The selected plot was found to be convenient to use gravity flow instead of using a pump to transport the wastewater from the main sedimentation tank of Kotebe WWTP to the wetland systems without any difficulty.
3.3 Construction of the wetland system
The pilot-scale CW systems were implemented at the compound of the existing wastewater treatment plant. The wetland system had three parallel SSF treatment wetland cells with different wastewater flow pattern (one VF, one HF, and one hybrid of VF and HF). Each wetland contained inlet piping, outlet piping, plastic liner to protect ground water pollution and water loss through infiltration, the same filter (gravel) media and emergent vegetation (Cyprus papyrus).
3.3.1 Determination of the size of wetland cells
The wetland system in this research was applied at pilot scale level with a capacity of 70 PE. According to Daniel et al. (2010), a person in Addis Ababa can produce approximately 0.045 m3 wastewater. Based on this figure, the amount of wastewater produced per each day was assumed to be approximately 3.15 m3, and therefore the flow rate of 3.15m3/d was used for the wetland systems used in this study.
The size of the pilot scale constructed wetland was determined based on published first order plug flow kinetic model. In this application Kadlec and Kright method was used considering the BOD removal as described by the “Constructed Wetlands Manual” UN-habitat (2008).
As= Q/K ( ln??Ci-ln?Ce ?) (Kadlec and Kright, 1996a)
Where: As = surface area of a wetland (m2)
Q = input discharge to a wetland (m3/day) = 3.15m3
K = Areal removal rate (m/year) = 34m/year
Ci = inlet concentration (mg/L) = 200mg/L
Ce = outlet concentration (mg/L) =25mg/L
The total area required to treat the wastewater flow rate of 3.15m3/d was calculated as 72m2.
3.3.2 Hydraulic retention time of the wetland systems
The time period that the domestic wastewater resides in the wetland systems was estimated using using Darcy’s formula (US EPA, 1993) as follows:
HRT = nLWd; …………………….. Darcy’s law
Qav
Where: n = effective porosity of the media, % as a decimal (0.23 – .38% for gravel media); 32%
L = Length of the wetland cell (m); 6m
W = Width of the wetland bed (m); 4m
d = average depth of liquid in the wetland (m); 0.55m
Qav = Average of the inflow and outflow; ((Qin + Qout))/2, (m3/d);
= ((1.05+0.85))/2 (m3/d) = 0.95m3/d
= (0.32*6m*4m*0.55m)/(0.95 m3/d)
= 4.224/0.95 days
= 4.5 days

3.3.3 Layout and configuration of the pilot-scale constructed wetlands
In order to achieve the goals of wastewater treatment, the area chosen can be arranged in many different ways. Number of independent flow paths, number of cells and aspect ratios are the basic decisions during wetland cell arrangement (Kadlec and Wallace, 2009).
Regarding the configuration of the wetlands system, the pilot-scale CWs system applied at Kotebe wastewater treatment plant had three subsurface constructed wetland (SSCW) cells. All the three CW cells are parallel and the hybrid CW system had two cells (HFCW and VFCW) connected in series. The only difference in designing the three SSFCW is the wastewater flow pattern. The types of wastewater flows into the wetland system were continuous horizontal flow /horizontal CW/, vertical flow /vertical CW/ and hybrid of horizontal and vertical flow /hybrid CW/. The area of each wetland cell was 24m2 and it has a dimension of 4m width and 6m length. In line with this, the area of the horizontal flow bed in the case of the hybrid system (having horizontal and vertical flow paths arranged in series) was chosen as 8m2 (2m x 4m ), which is half of the area of the vertical flow wetland, 16m2 (4m x 4m).

Wastewater pipe from the sewerage system

Floating valve

Sedimentation tank
Gate valves

Gravel

Hybrid bed Vertical bed horizontal bed
outlet (hybrid) outlet (vertical flow) outlet (horizontal)
Figure 3.1 Sketch map of configuration of the pilot scale constructed wetland systems applied at Kotebe WWTP
3.3.4 Site clearing and excavation
Following the selection of the appropriate site, the area was cleared up using human labour to prepare the site for the construction. The loosen soil was cut and transferred to the nearby idle area. During excavation work, the site was made to have 1% slope, to allow easier water circulation within the bed during the operation phase of the system.

3.3.5 Type of filter media used
There are different types of filter media that can be used in a constructed wetland system designed to treat wastewater. Some of the criteria which are commonly used to choose a filter media include: cost-effectiveness; size (smaller in size), high capacity of phosphorous adsorption, large surface area, high porosity and locally availability.
However, taking the wide application of gravel as a filter media for wetland systems into account (Konnerup et al, 2009; Babatunde et al, 2011; Caselles-Osorio et al, 2011; Sarmento et al, 2012), the filter media used for the pilot scale CW system applied at Kotebe WWTP in Addis Ababa was gravel with different diameter. The size of the filter media was in the range of 10 and 30 mm, the recommended size in most cases (USEPA, 2000). The effective depth of the filter media applied for the pilot-scale constructed wetlands of the study was 0.6m.
3.3.6 Plant species used in the study and planting procedures
Cyprus papyrus commonly called papyrus is among the typical characteristic species of wetland ecosystems of Ethiopia. Since it has been successfully used in wastewater treatment, all the three constructed wetlands were planted with C. papyrus. The rhizomes with shoot were collected from the natural stock/natural wetlands and transported to the study area using vehicles. All the three constructed wetland cells were planted with the rhizomes with shoots of Cyprus papyrus at 0.5m interval between each planted rhizomes: a density of 9/nine/ rhizomes/m2. The plantation was carried out in June, the starting time of the rainy season/summer. Immediately after the completion of the plantation, the wetland beds were filled with tap water and watering of the wetland with rain water and tap water was went on until the wetland plants adapted the environment and grown well.
3.3.7 Installation of inlet and outlet pipes
To transport the wastewater from the main sedimentation tank of the treatment plant to the wetland system, High-density Polyethylene (HDPE) pipe with 2 inch diameter and 10m length was installed to receive wastewater. The pipe was tightly fitted at the wastewater inlet end with a 5L jerrican, having more or less uniform holes around its surface for the entrance of wastewater, but to prevent larger debris or solids from getting into the HDPE pipe. Then the pipe with large diameter was reduced to ¾ inch pipe and the wastewater was transported using gravity to the wetland system, constructed downwards at a distance of about 165m from the main sewerage system. The use of force of gravity to transport wastewater helped to avoid the expenses and the problem that might result due to the use of pumps. It was also advantageous to avoid CO2 emission to the atmosphere.
To regulate the wastewater flow of the wetland system, a floating valve was fitted at the edge of the HDPE pipe which fed the sedimentation tank of the wetland system. When the tank becomes full, to the height of 0.8m, the wastewater in the sedimentation tank picks the floating valve up and the pipe becomes closed. Whereas when the level of the wastewater in the sedimentation tank is lowered, also the floating valve is turned down to open the HDPE pipe that feed the sedimentation tank.
The influent from the sedimentation tank entered into the wetland system by the PVC pipe installed at 10cm above the bottom of the sedimentation tank. It was divided into three branches using T, pipe fitting, to fed the three wetlands. The pipe line at the inlet of each of the wetland was fitted with a gatevalve to regulate the flow of wastewater. As a result, the amount of wastewater that entered into each wetland cell was adjusted using a stop watch and a graduated glass container; i.e. the flow rate was adjusted to be about 0.73 liter/min for each wetland. It means there was a mark on each gatevalve where the exact wastewater flow rate, as mentioned above, was obtained. All the pipes, valves, and the wetland system in general were monitored twice per day to maintain proper wastewater flow rate and functioning of the system.
Polyvinyl Chloride (PVC) pipes of 50mm diameter were used for the flow of wastewater into the CW system and to collect the effluent. The PVC pipes were drilled at every 25cm distance and each hole had 7mm diameter. This was made to ensure an equal wastewater flow distribution at the inlet of the horizontal CW and all over the surfaces of the vertical CW. Therefore, the perforated pipes were installed at the inlet region in case of the horizontal CW while they were installed to run from the inlet to the outlet direction above the surface of the vertical CW. Here, welded metallic rods were used to hold the pipes at about 20cm above the top surface in case of the vertical wetlands. However, after some time, the metallic rods were stolen by metal scavengers and they were replaced by blocks.
In horizontal flow wetlands, the wastewater was fed at the inlet and flow slowly through the porous fill media under the surface of the bed in a more or less horizontal path until it reaches the outlet zone. But the wastewater was fed from the top and then gradually percolates down through the bed and was collected by a drainage network at the base in case of vertical flow wetlands. Generally, PVC/perforated PVC pipes with 50 mm diameter were used to feed the treatment wetlands or to collect the effluent.
3.3.8 Lining of the wetland beds
Sealing of the wetland bed is essential to prevent ground water contamination by wastewater infiltration from the wetland bed. On the contrary, ground water can be infiltrated into the wetland system which in turn can influence the characteristics of the effluent. Therefore to cope up such shortcomings during the operation period, various sealing materials are being used as the conditions are suitable.
In this study, considering its convinience to apply, easily availability, and competent price, synthetic geomembrane was selected to be applied as a sealing material. It is available in the local market with various sizes and it was cut off and sewed using sewing machins to fit the dimensions required for sealing the wetland beds. After the synthetic geomembrane was laid down, the bottom of the pilot-scale constructed wetlands was made to have a slope of 1% to allow easier water circulation within each bed.
3.3.9 Sedimentation tank
A sedimentation tank with dimensions of 3.2m (length), 2.2m (width) and 0.90m (height) was constructed 0.5m above the surface of the wetlands so that the primarly treated wastewater can enter into the wetland cells by the help of gravitional force. It had a total volume of 6.34m3 and an effective volume of about 4.9m3. The sedimentation tank had been used primarily for storing wastewater and settling of solids. The internal wall and the bottom surface were plastered to prevent any water leakage from the tank.
A PVC pipe of 50mm diameter which fed the wetlands was connected to the tank at 10cm above the bottom surface. The first 10cm height from the bottom surface was left for the accumulation of wastewater sludge/ settled solids and the sedimentation tank was provided with an outlet/PVC pipe at the bottom to remove the accumulated sludge. Then, the sludge was regularly removed every two months using the outlet pipe and stored in a pit prepared at 25m from the tank.

Table 3.1: Summary of the characteristics of the pilot scale constructed wetland systems applied at Kotebe WWTP, Addis Ababa, Ethiopia.

Design factors Type of constructed wetlands
Horizontal
Flow CW Vertical
Flow CW Hybrid type
Vertical Flow CW Horiontal Flow CW
Wastewater flow type HF VF VF HF
hydraulic loading (m3/day) 1.05 1.05 1.05 1.05
Width (m) 4 4 4 4
Length (m) 6 6 4 2
Area (m2) 24 24 16 8
Wetlands fill media Gravel Gravel Gravel Gravel
Depth of fill media (m) 0.6 0.6 0.6 0.6
Hydraulic loading rate (m/d) 0.044 0.044 0.066 0.131
Wetland plants C. papyrus C. papyrus C. papyrus C. papyrus
3.4 Monitoring of the constructed wetland system
The wastewater flow rate and the proper functioning of the wetland system were monitored regularly in the morning and in the afternoon times. The most common activities during the monitoring period were checking of the piping system and valves whether they are blocked with solids or not, removing of weds which grew on the wetland beds and protecting the wetlands system from animals.
3.5 Sampling and laboratory analysis
For water quality monitoring, triplicate grab samples were collected from the influent and the three outlets /effluent/ of the wetland systems every two weeks. The water quality parameters such as Temperature, pH, conductivity, DO, BOD5, COD, TSS, NO3–N, NH4+-N, TN, orthophosphate and TP were analyzed at Addis Ababa Environmental Protection Authority (AA EPA) laboratory based on the standard methods for the examination of water and wastewater (APHA, 1999). Temperature, pH, conductivity, and DO were measured onsite to avoid value changes due to variation of environmental factors. Likewise, the analyses of Fecal Coliforms (FC) and the N and P content of wetland plants were done at the laboratory of Water and Energy Design and Supervision Works Sector, Ethiopian construction design and supervision works corporation.
Table 3.2: Summary of the laboratory methods and instruments used to measure wastewater parameters during the monitoring period.
Parameters Laboratory methods/instruments
Temperature, To Portable pH/EC/TDS/oC Meters, HI 9811-5
pH Electrometeric method
Portable pH/EC/TDS/oC Meters, HI 9811-5
Electrical Conductivity, EC Portable pH/EC/TDS/oC Meters, HI 9811-5
Dissolved Oxygen, DO Membrane Electrode DO meter, ExStik* DO600
Biochemical Oxygen Demand, BOD Pressure sensor
Chemical Oxygen Demand, COD Spectrophotometer, DR 3900, Hack method
Total Suspended Solids, TSS Spectrophotometer, DR 3900, Hack method
Nitrate Nitrogen, NO3–N Spectrophotometer, DR 3900, Hack method
Ammonium Nitrogen, NH4+ – N Spectrophotometer, DR 3900, Hack method
Total Nitrogen, TN Spectrophotometer, DR 3900, Hack method
Orthophosphate, PO43- – P Spectrophotometer, DR 3900, Hack method
Total Phosphorous, TP Spectrophotometer, DR 3900, Hack method
Fecal Coliforms, FC Membrane filter technique
Wetland plant P content DTPA extraction, KH2PO4 extraction, Olsen, Kjeldahl digestion Walklay black, Ammonium acetate and instrumental
* GARMIN GPSmap 62 – is used to take GPS readings.
3.6 Source of meteorological data
Average daily rainfall and daily ambient air temperature of the study area during the performance monitoring periond were obtained from the National Meteorological Agency (NMA).

3.7 Data analysis
The data obtained from the study were analyzed by using MS Office Excel 2007 and IBM SPSS Statistics version 21 software package. Mean, standard error, removal efficiencies, linear correlation, and analysis of variance (ANOVA) tests were done using these packages.
3.8 Total cost of construction
The total construction cost for the implementation of the pilot-scale constructed wetland system is indicated in table 3.3. Most of the costs used in conducting this project was covered by the fund obtained from Wollo University (WU) and Bursary Research Fund (BRF).
Table 3.3: Cost summary of the construction of the pilot scale constructed wetland systems applied at Kotebe WWTP, Addis Ababa, Ethiopia.
Description Quantity Unit cost ($) Total cost ($)
Excavation, m3 60 5.25 315.00
Stone, m3 11 10.90 119.90
Blocks, each 400 0.35 140.00
Sand, m3 12 15.20 182.40
Cement, bag 40 18.90 435.00
Geo-synthetic membrane, m2 120 0.90 108.00
Pipes – HDPE, 2 inch diameter, m
HDPE, ¾ inch diameter, m
PVC, 50mm diameter
with 6m length, each 15
200

14 3.05
0.35

4.35 45.75
70.00

60.90
Gravel 4mm diameter, m2 8 16.50 132.00
Gravel 3mm diameter, m3 14 16.50 231.00
Gravel 2mm diameter, m3 8 15.00 120.00
Gravel 1mm, m3 14 11.30 158.20
Manpower: – Mason, contractor
3 laborer, days
12
3.50 260.00
126.00
Total 2,504.15

CHAPTER FOUR
Result and Discussion
4.1 Characteristics of the Domestic Wastewater
The characteristics of the domestic wastewater treated using the pilot-scale subsurface constructed wetlands employed at Kotebe WWTP, the Eastern part of Addis Ababa, was determined. Triplicate samples from the raw domestic wastewater before getting into the sedimentation tank for primary treatment and samples from the wastewater after primary treatment were analyzed and the results are presented in table 4.1 and 4.2 respectively.
Table 4.1: Raw domestic wastewater characteristics before primary treatment, Addis Ababa-Ethiopia.

Descriptive
(mg/L) Wastewater Parameters

BOD
COD
TSS
NO3–N
NH4+-N
TN
PO4-3
TP FC
(CFU/100 ml)
Minimum 166 269 183 0.7 24.6 44 1.50 3.25 85000
Maximum 205 552 449 4.3 72 80 6.44 9.10 220000
Average 187 413 311 2.7 53 67 4.06 6.40 156667
St. dev 19.8 141.6 133.3 1.83 24.83 20.4 2.48 2.95 67885

Table 4.2: Domestic wastewater characteristics after primary treatment, Addis Ababa, Ethiopia.

Descriptive
(mg/L) Wastewater Parameters

BOD
COD
TSS
NO3–N
NH4+-N
TN
PO4-3
TP FC
(CFU/100 ml)
Minimum 132 278 114 1.1 32.9 40 2.30 3.85 91000
Maximum 195 461 243 7.5 68 89.6 6.88 9.30 133000
Average 170 369 178 4.93 50 68 4.44 6.87 112333
St. dev 33.65 91.50 64.5 3.38 17.57 25.4 2.30 2.77 21008

As shown in table 4.1 and 4.2, the values of TSS and FC were considerably decreased after primary treatment and the percentages of reduction were 42.8% and 28.3% respectively. Similarly, 9.1% of BOD and 10.7% of COD were removed after primary treatment and the removal of BOD and COD was attributed mainly to TSS removal by sedimentation. Sedimentation was the most significant process to remove wastewater pollutants particularly TSS, as slow flow velocity of wastewater was maintained in the sedimentation tank. Following primary treatment, a slight increase in the concentration of PO4-3 and TP was observed in the wastewater. This might be due to the breaking down of high-molecular-weight polyphosphates to low-molecular-weight phosphates by the process of hydrolysis.
Alternatively, the effect of biological breakdown by the actions of microorganisms or chemical processes was insignificant since these processes require relatively longer period of time to play their role in wastewater treatment while the detention time of the wastewater in the sedimentation tank is short. As a result, the values of other pollutants of the raw domestic wastewater and the wastewater after primary treatment had only slight difference.
The characteristics of the domestic wastewater or the influent treated using CWs in various countries or locations gathered from different literatures are summarized in table 4.3. The wastewater parameter values determined for this study were also included in the table 4.3. The concentration of BOD5 in this study was found to be more or less similar with the concentration reported in most countries. But it was higher than the values reported in Mexico (Zurita et al, 2009) and Australia (Sleytre et al, 2007) and lower than the values in Brazil (Lana et al, 2013) and Ireland (Kayranli et al, 2010). Likewise, the domestic wastewater used in this study was relatively comparable in its COD concentration with the values reported in Egypt (Abdelhakeem et al, 2016) and Kenya (Mburu et al, 2013) although it was by far higher than Colombia (Caselles-Osorio et al, 2011) and Mexico (Zurita et al, 2009). In general, the ratio of BOD5: COD in this study was 0.45 and therefore it could be concluded that the domestic wastewater is convincingly biodegradable and it could be classified as “low strength” wastewater.

Table 4.3: Domestic wastewater characteristics of some other countries

Countries Wastewater parameters (mg/L)

References

BOD5
COD
TSS
NO3–N
NH4+-N
TN
PO4-3
TP FC (CFU/
100ml)
Austria 150 367 – – 42 – – 6.6 – Sleytr et al, 2007
Brazil 279 465 293 0.1 26.4 – – 3.9 – Lana et al, 2013
China 103-207 213-381 – – 48-112 71-104 – 4.8-12.1 – Lu et al, 2015
Colombia – 132 – 5.3 23 – 5 – 87677 Caselles-Osorio et al, 2011
Egypt 181-253 383-624 180-281 5.2-6.7 30-42 – 2.5-2.9 Abdelhakeem et al, 2016
Greece – 458 – 0.01 – 49 8.18 – – Prochaska et al, 2007
Ireland 761.1 1279.3 2183.8 4.8 32.1 – 3.7 – – Kayranli et al, 2010
Kenya 232 424 118 39 4 – – – – Mburu et al, 2013
Mexico 115.5 247.5 57.5 9.3 15.7 28.7 8.3 Zurita et al, 2009
Thailand – 93-136 47-65 1.1-1.4 11.4-21.1 17.4-27 4.3-7.1 6.7-9.8 – Konnerup et al, 2009
Turkey 200 343 333 – – 55.6 – 7.06 19823 Tuncsiper et al, 2012
UK 104-221 186-352 122-379 0.8-12 23-69.4 – 7.2-18.7 – – Al-Isawi et al, 2015
USA 230-392 – 418-2102 – – – – – 32900 -90400 Karathanasis et al, 2003
This study 186 417 206 7.51 50.8 76.6 5.84 18.27 95292

4.2 Meteorological data of the study area
Meteorological data such as daily average rainfall and temperature were obtained from the nearby office of the National Meteorological Agency (NMA). Therefore, the daily average values of rainfall and temperature for the sampling dates were presented in table 4.3. However, the amount of water loss from the CWs through evapo-transpiration was not available and therefore the water budget calculation for the system applied at Kotebe WWTP was not done.
Based on the meteorological data obtained from NMA, the daily average rainfall values of each season during the monitoring period were 0.75 ± 2.67 mm for winter (December, January and February), 4.31 ± 8.55 mm for autumn (March, April and May), 6.62 ± 8.18 mm for summer (June, July and August) and 1.64 ± 4.48 mm for spring (September, October and November). The highest daily average of rainfall was observed in summer and the lowest value was observed in winter. Although autumn is known to be light rainy season, the daily average value of rainfall was to some extent high in 2016.
Similarly, the daily average values of temperature were 24.2 ± 0.95 °C for winter, 24.3 ± 2.00 °C for autumn, 22.3 ± 1.95 °C for summer and 23.4 ± 1.36 °C for spring. The highest seasonal average value of temperature was seen in autumn while the lowest value was observed in spring.

Table 4.4: Daily average data of rainfall and ambient air temperature for the sampling dates.
Sr. No Sampling dates Average daily rainfall (mm) Average daily ambient air temperature (°C) Average daily evaporation (mm) Remark
1 15/12/15 0.0 24.3 NA
2 29/12/15 0.0 24.5 NA
3 13/01/16 1.0 23.4 NA
4 28/01/16 2.6 23.0 NA
5 11/02/16 0.0 25.2 NA
6 25/02/16 9.0 22.8 NA
7 10/03/16 0.0 24.6 NA
8 28/03/16 0.0 25.2 NA
9 12/04/16 2.1 26.4 NA
10 27/04/16 19.6 24.0 NA
11 12/05/16 0.0 23.6 NA
12 26/05/16 2.8 23.0 NA
13 09/06/16 0.0 25.2 NA
14 23/06/16 4.1 23.0 NA
15 11/07/16 2.5 20.6 NA
16 26/07/16 6.6 23.0 NA
17 10/08/16 8.2 20.4 NA
18 25/08/16 8.7 22.4 NA
19 08/09/16 0.0 21.5 NA
20 22/09/16 0.0 24.8 NA
21 10/10/16 1.3 25.6 NA
22 26/10/16 0.0 24.5 NA
23 10/11/16 0.0 23.0 NA
24 30/11/16 0.0 22.2 NA
NA: Not Available
Reference: National Meteorology Agency of Ethiopia (NMA), data delivery and dissemination case team, Addis Ababa.
4.3 Results of Performance Monitoring of the constructed wetlands
The performances of the horizontal, vertical and the hybrid of both horizontal and vertical subsurface flow constructed wetlands (CWs) were monitored from Dec. 15, 2015 to Nov. 30, 2016. Therefore, the data as well as their respective calculations and discussion for the most important wastewater parameters such as BOD5, COD, TSS, NO3-, NH4+, TN, PO43+, TP and TC were done and presented in the subsequent sections. Moreover, the result and discussion part included the data on plant tissue nutrient (N and P) content with the relevant calculations and discussion.
4.3.1 Temperature, pH, Electrical Conductivity and Dissolved Oxygen levels
In addition to the wastewater parameters listed above: temperature, pH, EC and DO of the influent and the effluent from each of the horizontal, vertical, and hybrid CWs were measured onsite to avoid change of concentration values due to time variability between sampling and laboratory analysis. Table 4.4 presented the mean values of these parameters in the influent and the effluents of the three constructed wetland cells. The measured average values of influent wastewater temperature in winter, spring, summer and autumn seasons were 21.6 ± 1.1°C, 22.4 ± 0.8 °C, 18.3 ± 1.4 °C and 16.1 ± 1.39 °C respectively. Similarly, the average pH values of the influent wastewater were 7.4 ± 0.31, 7.6 ± 0.28, 7.2 ± 0.31 and 7.3 ± 0.4 in winter, autumn, summer and spring respectively.
The average EC values in the four seasons were 959 ± 85 µs in winter, 920 ± 121 µs in autumn, 906 ± 66 µs in summer and 1012 ± 50 µs in spring. The onsite measurement of Dissolved Oxygen (DO) in this study revealed that the average concentrations of DO of the influent in winter, autumn, summer and spring were 0.38 ± 0.09 mg/L, 0.87 ± 0.51 mg/L, 1.2 ± 0.15 mg/L, and 0.56 ± 0.21 mg/L respectively. DO concentrations of the influent of the CWs system was observed to be higher in summer and lower in winter. This might be resulted due to the seasonal temperature difference that affects the solubility of oxygen in water.
Table 4.5: Mean influent values of temperature, pH, electrical conductivity and dissolved oxygen of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
Sr. No.
Influent Parameters Seasonal mean values
Winter Autumn Summer Spring
1 Temperature (°C) 21.6 (1.07) 22.4 (0.79) 18.3 (1.39) 16.1 (1.39)
2 pH 7.42 (0.31) 7.60 (0.28) 7.16 (0.31) 7.26 (0.40)
3 Electrical Conductivity (µs) 959 (85) 920 (121) 906 (66) 1012 (50)
4 Dissolved Oxygen (mg/L) 0.38 (0.09) 0.87 (0.51) 1.20 (0.15) 0.56 (0.21)

4.3.2 Removal of Biochemical Oxygen Demand (BOD5)
The concentrations of BOD5 in the influent of the horizontal, vertical and hybrid CWs applied at Kotebe WWTP in Addis Ababa varied between 132.7 and 224.7 mg/L for the period of performance monitoring and the average value was 186.4 ± 15.68 mg/L (Figure 4.1 and Table 4.5). The average BOD5 value of the influent used in this study was higher than the BOD5 value reported in Spain (Pedescoll et al, 2015), India (Mustafa, 2013) and China (Juan, 2012). But the values are still lower than the result found in Ireland and many African countries such as Kenya, Uganda, Cameroon and Nigeria (Khisa, 2011; Birol, 2010; Andualem, 2015).
During the period of wetland performance monitoring, the concentration of BOD5 turned of higher value during the dry season (winter) when the rainfall was extremely low. This could have resulted from the regular interruption of the municipal water supply which affected the water usage behavior of inhabitants in the area. In the contrary, the storm water infiltrated into the sewerage system and the use of rain water for different domestic purposes at least by some of the inhabitants during the rainy season might have caused dilution of biodegradable organic compounds. For instance, the lower concentrations of BOD5 were observed within the samples collected on 25th of February, 27th of April, 26th of July and 25th of August, 2016. The daily rainfalls recorded on these sampling dates were 9.0, 19.6, 6.6, 8.7 mm respectively, with the highest amount recorded in April. Moreover, the above assumption is evidenced by the observation that the sewerage system, from which the influent was piped to feed the pilot scale CWs system, was always full in rainy seasons/mostly in summer. Hence, in most cases, the decrease in BOD5 concentrations could be related with the dilution of the influent following precipitation.
During the monitoring period, the BOD5 concentrations in the effluent of the horizontal, vertical and hybrid wetlands system employed at Kotebe WWTP were found to be in the range of 6.3 – 45.7 mg/L (20.3 ± 9.02 mg/L), 4 – 32.3 mg/L (14.2 ± 7.00 mg/L), and 4.0 – 25.3 mg/L (12.0 ± 6.11 mg/L) respectively (table 4.5). The highest average BOD5 concentration (198.3 ± 21.08 mg/L) within the influent was recorded in winter season, while the lowest average concentration value (5.8 ± 2.17 mg/L) was found within the effluent of the hybrid CWs system in spring. The effluent concentration values of BOD5 from the three CWs were within the provisional discharge standards, 80 mg/L which is set by Ethiopian Environmental Protection Authority (EEPA, 2003).
Reduction of BOD, the organic matter which can be found in suspended or dissolved form in domestic wastewater is one of the primary objectives of wastewater treatment (Charles and Ian, 2009). Therefore, during the monitoring period, the performance of the pilot-scale constructed wetlands system in removing BOD5 was evaluated based on its decreasing concentrations, and in the removal efficiencies of the horizontal, vertical and hybrid flow systems ranged between 77.7 – 96.3% (89.1 ± 5.2 %), 84.2 – 97.3 % (92.2 ± 4.1 %) and 86.5-98.0 % (93.4 ±3.6%) respectively, with the average values given in brackets, with the highest removal percentage for the hybrid bed and the lowest removal percentage for the horizontal bed.
The result showed that the pilot scale CWs applied at Kotebe WWTP were efficient in removing BOD5 in domestic wastewater. The successful removal efficiency obtained in this study might be due to the fact that organic matter in domestic wastewater tends to have readily degradable compounds (Wallace, 2004) or the complex natural processes offered by the CWs system could be the other reason for its effectiveness (Kelvin and Tole, 2011).
The performances of the horizontal, vertical and hybrid flow systems in removing BOD5 were significantly different from one another statistically (one-way ANOVA; F0.95 (2, 69) = 5.843; P < 0.05). Hence, it could be described that the difference in the flow type of the influent into each of the three CWs beds resulted in significant effect on the performance of the wetlands for the removal of BOD5.
Azni et al (2010) described that the process of breaking down of organic carbon to CO2 by microorganisms to obtain energy for growth requires dissolved oxygen. In this study, the performance of hybrid/VF beds was better than HF for the removal of BOD5. This might be mainly occurred as the unsaturated flow condition in VF bed presents more oxygen for the oxidation-reduction potential to take place in vertical subsurface flow wetland (Pandey et al, 2013). In fact, both aerobic and anaerobic decomposition processes can take place in the removal of organic matter. However, the nature of biochemical reactions depends on the conditions created as a result of the rate of oxygen transfer in the wetland system. If there is adequate oxygen supply, aerobic decomposition is so rapid that the accumulation of organic matter in the wetland is small. But if the rate of oxygen transfer cannot meet the oxygen requirements, the removal process becomes anaerobic decomposition which results in the accumulation of organic matter in the wetland (Scott, 2004; Garcia et al, 2005). Hence, it could be explained that there was better oxygen transfer in the vertical flow and hybrid flow CWs system by which the higher BOD5 removal was observed in the study.
The CWs system in this study showed higher performance compared to the study conducted in Cairo, Egypt (Abdelhakeem et al, 2016), Nepal (Pandey et al, 2013), and El Salvador (Katsenovich et al, 2009). However, other reports from both temperate and tropical regions also show comparable results in removing BOD5 (Laaffat et al, 2013; Vymazal, 2014; Vymazal and Brezinova, 2014; Lu et al, 2015; Vergeles et al, 2015).
Although Vymazal (2014) reported that there was no significant difference between average outflow concentrations of wastewater pollutants in summer and winter periods, the performances of the horizontal, vertical and hybrid systems were differed significantly from season to season in removing BOD5; one-way ANOVA results of the horizontal, vertical and hybrid systems differed F0.95 (3, 20) = 13.595; P < 0.05, F0.95 (3, 20) = 12.496; P < 0.05 and F0.95 (3, 20) = 18.445; P 0.05). According to the result, the difference flow type did not pose any effect on the removal of COD although there was anticipation of different removal performance for each of the treatment bed.
The COD removal efficiency obtained in this study was lower compared to the study done in Greece (Prochaska et al, 2007) and China (Lu et al, 2015) while it showed better removal efficiency than the result reported in Egypt (Abdelhakeem et al, 2016), Ukraine (Vergeles et al, 2015) and china (Guo et al, 2015). The COD removal performance was however comparable with the result reported in Greece (Gikas and Tsihrintzis, 2010) and in India (Deeptha et al, 2015; Yadav et al, 2011).
On the other hand, the COD removal performances of the three CWs system were significantly different from season to season. The results of one-way ANOVA were F0.95 (3, 20) = 45.556; P < 0.05 for the horizontal system, F0.95 (3, 20) = 23.020; P < 0.05 for the vertical system and F0.95 (3, 20) = 19.294; P < 0.05 for the hybrid system. Valsero, Cardona and Becares (2012) revealed that pollutant removal efficiency of constructed wetlands is marked by seasonal difference. In addition to this, Wu et al (2014) described the influence of seasonal difference on the performance of CW system as the operation of CWs at cold climate is a big challenge.
Table 4.7: Average values of COD concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems applied at Kotebe WWTP, Addis Ababa, Ethiopia.

Sr. No. Seasons
of
the year Influent HFCW effluent VFCW effluent HyCW
Average COD
(St. dev) Average COD
(St. dev) COD
RE (%) Average COD
(St. dev) COD
RE (%) Average COD
(St. dev) COD
RE (%)

1
Winter 448.0
(65.06) 61.0
(14.29)
86.4 63.9
(15.83)
85.7 53.8
(12.64)
88.0

2
Spring 399.1
(39.27) 61.6
(14.40)
84.6 51.4
(10.76)
87.1 47.8
(10.14)
88.0

3
Summer 374.1
(44.08) 84.2
(13.53)
77.5 78.7
(11.53)
79.0 69.0
(12.72)
81.6

4
Autumn 390.1
(51.96) 102.1
(18.21)
73.8 91.3
(26.36)
76.6 84.3
(21.20)
78.4

5
Overall 402.8
(31.84) 77.2
(19.82) 80.6
(5.93) 71.3
(17.37) 82.1
(5.09) 63.7
(16.35) 84.0
(4.80)

Figure 4.4: COD Concentration of the influent and effluents of the horizontal, vertical and hybrid pilot scale CWs employed at Kotebe WWTP, Addis Ababa, Ethiopia.

Figure 4.5 Seasonal COD removal efficiencies of the horizontal, vertical and hybrid pilot scale CWs employed at Kotebe WWTP, Addis Ababa, Ethiopia.
Based on the results of performance monitoring study of the three CWs system applied at Kotebe WWTP, the COD loading rate was ranged between 13.79 g/m2/d and 22.23g/m2.d (17.72 ± 2.433 g/m2.d) (Fig 4.6). The linear correlation between COD loading rate and COD removal rate was relatively similar among the three wetland systems. But the linear correlation for hybrid system (R2 = 0.872) was slightly better than the horizontal (R2 = 0.841) and vertical system (R2 = 0.838) while the horizontal system had to some extent better correlation than the vertical system. Even though, the correlation between COD loading rate and removal rate presented merely little differences among the three design types, the capacity to hold and remove COD was better in the hybrid system as the concentration increase.

COD loading rate (g/m2.d)
Figure 4.6: COD loading rate (g/m2.d) against the COD removal rate (g/m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
4.3.4 Removal of Total Suspended Solids (TSS)
Suspended solids are the typical contaminants in wastewater and are originated from either internal or external sources. The external sources are usually the influent and atmospheric inputs whereas planktons and animal and plant detritus are created within the wetland. Wastewater commonly constitutes 99.9% water and the rest 0.1% is solids. Hence, suspended solids are essential parameter in water quality monitoring and therefore applied to measure the quality of influent and effluent, and also to evaluate the performance of many processes (Thomas and William, 2001; Kayombo et al, 2003; Frank, 2003; Kadlec and Wallace, 2009).
As shown in Figure 4.5 and Table 4.7, during the monitoring period, the average concentrations of the TSS in the influent of the horizontal, vertical and hybrid pilot scale CWs applied at Kotebe WWTP were ranged between 94.0 – 290.7 mg/L (206.1 ± 21.44 mg/L). The inflow of TSS varies from season to season (Sani et al, 2013), and unlike the concentration of BOD5 and COD, the average TSS value was found to be higher in summer when there was peak rainfall and low in winter/the dry season. This might be occurred as a result of addition of suspended solids from surface runoff which could take place following the rainfall. The average TSS concentration of the influent was similar with the value reported in Greece (Gikas and Tsihrintzis, 2010) while its value was higher than the result reported in Mexico (Giacoman-Vallejos et al, 2015) and UK (Al-Isawi et al, 2015). But the value of TSS concentration found in this study was slightly lower than the value reported in Malaysia (Ling et al, 2012).
Similarly, TSS concentrations in the effluents of the pilot scale CW systems applied at Kotebe WWTP for the treatment of domestic wastewater were in the range of 10.7 – 41.0 mg/L (22.7 ± 9.38 mg/L), 18.3 – 51.7 mg/L (33.6 ± 8.47 mg/L) and 16.7 – 44.7 mg/L (30.7 ± 7.65 mg/L) for the horizontal, vertical and hybrid flow systems respectively. The lowest TSS concentration was observed in the effluent of the horizontal flow system in winter. TSS concentration values of the CWs system applied at Kotebe WWTP comply with the provisional discharge standards, 100 mg/L (EEPA, 2003).
In line with this, the concentration based removal percentage of TSS by the horizontal, vertical and hybrid wastewater flow constructed wetland system at Kotebe WWTP varied between 83.4 – 94.9% (89.1± 3.19%), 73.1 – 91.4% (83.8 ± 2.86%), and 75.1 – 91.3% (84.7 ± 1.65%) respectively (Figure 4.6 and Table 4.7). The efficiency of the horizontal, vertical and hybrid flow systems in removing TSS were significantly different from one another statistically (one-way ANOVA; F0.95 (2, 69) = 10.127; P < 0.05). The overall TSS removal efficiencies of each of the three wetland systems were comparable with the result obtained in Egypt (Abou-Elela et al, 2014; Mustafa, 2013).
Wetlands are able to remove suspended solids from wastewater efficiently (Thomas and William, 2001; Avsar et al, 2007; Kadlec and Wallace, 2009). This is because, the system has normally extended HRTs and low flow velocity, which create desirable conditions for easily removal of settleable solids by gravitational settlement. Alternatively, processes including biodegradation, adsorption on submerged parts of the plant and wetlands media, filtration, and flocculation/precipitation are involved in the removal of colloidal or non-settling solids. In the removal of suspended solids, the nature and size of contaminant solids and the type fill media are the major factors on which the practical action of each removal mechanism depends (Kayombo et al, 2003; Robert, 2004; Vymazal, 2008; Azni et al, 2010).
Although, there was anticipation that vertical flow wetlands are more successful in removing TSS (Kadlec and Wallace, 2009), the highest removal percentage was shown by the horizontal bed while the lowest removal percentage was observed in the hybrid bed system in this study. In a similar manner, Haghshenas-Adarmanadabi et al (2016) explained that the removal efficiencies of HFCWs were significantly higher than those of the VFCWs; and the higher efficiency of the HFCWs was due to higher retention time and loading rate of them. This could be occurred as a result of the more rapid biological processes taken place internally in the wetland system to add more solids into the effluents in the case of the vertical and hybrid flow beds. On the other hand, Thomas and William (2001) revealed that in addition to the settling rate of particles which in turn depends on a number of factors, the wetland length highly affects the efficiency in removing SSs. So, the slight increase in the TSS removal percentage of the horizontal flow CW might be related with the length, the wastewater flows through it.
But in the case of the VFCW the influent enters into the bed vertically all over the bed. The maximum average TSS concentration within the influent of the CW systems employed at Kotebe WWTP was 230.3 ± 20.523 mg/L and recorded in summer. But the lowest average TSS concentrations, 15.8 ± 3.764 mg/L was obtained within the effluent of the horizontal constructed wetlands in winter. Vymazalc (2014) pointed out that the removal of TSS and some other wastewater parameters is stable regardless of seasonal difference. However, in this study, the performance of the horizontal flow system during the monitoring period was different from season to season, F0.95 (3, 20) = 9.362; P 0.05 and F0.95 (3, 20) = 1.910; P > 0.05 respectively.
Table 4.8: Average values of TSS concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.

Sr.
No. Seasons of
the year Influent HFCW effluent VFCW effluent HyCW
Average
TSS (St. dev) Average
TSS (St. dev) TSS
RE (%) Average
TSS (St. dev) TSS
RE (%) Average
TSS (St. dev) TSS
RE (%)

1
Winter 186.9
(41.52) 15.8
(3.76)
91.5 28.4
(7.90)
84.8 23.5
(5.93)
86.6

2
Spring 189.1
(73.97) 18.6
(7.41)
90.0 30.9
(9.78)
83.7 25.4
(10.10)
85.3

3
Summer 230.3
(20.52) 36.6
(2.75)
84.1 46.2
(5.35)
80.0 39.8
(3.28)
82.7

4
Autumn 217.9
(43.06) 20.0
(4.05)
90.9 28.8
(7.52)
86.8 34.2
(8.29)
84.3

5
Overall 206.1
(21.44) 22.7
(9.38) 89.1
(3.41) 33.6
(8.47) 83.8
(2.86) 30.7
(7.65) 84.7
(1.65)

Figure 4.7: Concentration of TSS of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW system employed at Kotebe WWTP, Addis Ababa, Ethiopia.

Figure 4.8: Seasonal TSS removal efficiencies of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.

During the one year performance monitoring period, the loading rate of TSS were found to be in the range of 4.14 and 12.79 g/m2.d (9.1 ± 2.164 g/m2.d) (Fig 4.9). The linear correlation between loading rate and the removal rate by the horizontal, vertical and hybrid system showed that the TSS removal rate was dependent on the TSS loading rate and the removal rate for horizontal system (R2 = 0.973) was slightly more dependent than the hybrid system (R2 = 0.968), and the vertical flow system showed relatively lower correlation (R2 = 0.963) between the TSS loading rate and removal rate compared to both the horizontal and hybrid flow systems.

TSS loading rate (g/m2.d)
Figure 4.9: TSS loading rate (g/m2.d) against the TSS removal rate (g/m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
4.3.5 Removal of Ammonium (NH4+)
Nitrogen is among the most important constituents in wastewater since it causes eutrophication, toxicity to aquatic life and undesirable consequence on the level of oxygen in the water bodies (Kadlec and Wallace, 2009). It exists in the environment in different forms although the transformation from one form to another may occur quickly. The quantity of both organic and inorganic N forms that can be present in wastewater is high (Thomas and William, 2001). Proteinaceous matter and urea are the two principal components in which nitrogen can be available in domestic wastewater. If the condition in the pretreatment is kept anaerobic, ammonia, ammonium (NH4+) is produced as a result of the breaking down of protein and urea and the remaining organic N will be converted to NH4+ by the process of ammonification (Scott, 2004). To evaluate the efficiency of the pilot scale CWs system employed at Kotebe WWTP, in removing N, the influent and effluent NH4+, NO3- and TN were determined.
Based on the laboratory result, the average concentration of NH4+ in the influent varied between 32.6 – 69.6 mg/L (50.8 ± 8.39 mg/L). The highest average value of NH4+ (58.5 ± 6.07 mg/L) within the influent was obtained in winter while the influent had the lowest average concentration of NH4+ (39.6 ± 2.35 mg/L) in summer. As indicated in Table 4.3, the value was higher than the study done in Kenya (Mburu et al, 2013), Mexico (Zurita et al, 2009) and Brazil (Lana et al, 2013). However, Lu et al (2015) reported relatively higher NH4+ concentration in China and similar trends of NH4+ concentration were reported by Al-Isawi et al (2015).
The relatively high NH4+ concentration in the domestic wastewater used in this study might be related with the lower per capita water consumption by the residents. This could increase the concentration of animal and plant tissue and excreted urea in the wastewater. The assumption is strengthened by the fact that the concentration had lower values during the rainy season, when the water supply and consumption by the residents was increased causing dilution. NH4+ is formed, under aqueous conditions, by the rapid hydrolysis of ammonia formed from the breaking down of N containing organic compounds (Scott, 2004).
The average effluent concentrations of NH4+ were found to be in the range of 13.7 – 30.5 mg/L (20.7 ± 2.18 mg/L) for the horizontal flow, 10.8 – 27.2 mg/L (17 ± 1.98 mg/L) for the vertical flow and 11.9 – 26.7 mg/L (17.3 ± 2.57) for the hybrid flow type pilot-scale CWs employed at Kotebe WWTP (Figure 4.11 and Table 4.10). The lowest average concentration of NH4+ was obtained within the effluent of the vertical flow CW in summer. However, the concentration values of NH4+ within the effluents of all the three CWs were higher than the provisional discharge standard, which is 5 mg/L (EEPA, 2003).
Accordingly, the concentration based removal efficiencies of the pilot scale CWs employed at Kotebe WWTP were in the range of 43.0 – 74.2% (58.6 ± 6.21%) for the horizontal, 50.9 – 78.7% (66.2 ± 5.17) for the vertical and 51.8 – 77.6 (65.4 ± 16.68%) for the hybrid flow CW type (Figure 4.12 and Table 4.10). The removal efficiency showed relatively high variability among the three wetland beds and the lowest average removal percentage was seen in the horizontal bed and the highest average removal percentage was recorded in the vertical flow system. The performances of the horizontal, vertical and hybrid flow systems in removing NH4+ were significantly different from one another statistically (one-way ANOVA; F0.95 (2, 69) = 7.952; P < 0.05). Better NH4+ removal efficiency using pilot-scale CWs in treating mild domestic wastewater was reported in Malaysia (Katayon et al, 2008).
Tuncsiper et al (2009) pointed out that HSSF wetland system shows higher NO3- removal and lower NH4+ removal. In fact, the horizontal flow system gives suitable environmental conditions for denitrification though the condition for nitrification is limited. Contrasting to the HSSF, the VSSF has better aeration condition and hence it provides a higher NH4+ removal efficiency. All the three systems showed higher NH4+-N removal efficiency than the result reported in China (Zhiwen et al, 2006).
The maximum average NH4+ concentration within the influent of the CWs system employed at Kotebe WWTP was 58.5 ± 6.07 mg/L and recorded in winter. Meanwhile, the lowest average NH4+ concentrations, 14.6 ± 3.56 mg/L was obtained within the effluent of the vertical CW in summer. The performances of the horizontal, vertical and hybrid systems differed significantly from season to season in removing NH4+; one-way ANOVA results of the horizontal, vertical and hybrid systems were F0.95 (3, 20) = 8.095; P < 0.005, F0.95 (3, 20) = 5.104; P < 0.05 and F0.95 (3, 20) = 21.344; P 0.05).
Canga et al (2011) pointed out that the removal rate of N with horizontal flow CW system alone is low due to the deficiency of nitrification. On the other hand, single-stage VF CW cannot attain high N removal since environmental conditions that favor denitrification process lack. High removal of N can be realized in hybrid CWs system where the combination of HF and VF beds is applied.
Microbial metabolism affords removal of inorganic nitrogen, that is, nitrate and ammonium, in wetland soils. Certain bacteria (e.g., Pseudomonas spp.) metabolically transform nitrate into nitrogen gas (N2), a process known as denitrification. The N2 is subsequently lost to the atmosphere, thus denitrification represents a means for permanent removal, rather than storage, of nitrogen by the wetland. Ammonium removal in constructed wetlands can occur as a result of the sequential processes of nitrification and denitrification. Nitrification, the microbial (Nitrosomonas and Nitrobacter) transformation of ammonium to nitrate, takes place in aerobic regions of the soil and surface water. The newly formed nitrate can then undergo denitrification when it diffuses into the deeper, anaerobic regions of the soil. The coupled processes of nitrification and denitrification are universally important in the cycling and bioavailability of nitrogen in wetland and upland soils (Thomas and William, 2001).
In addition to microbial removal mechanisms, plant uptake and storage of nutrients in the sediment could be the chief N conversion and removal routes in CWs in treating wastewater (Wu et al, 2013). In planted CWs system, there is higher N (Wang et al, 2016), and therefore plant uptake is one of the major means to remove nitrate produced by the process of nitrification (Robert, 2004).
The maximum average NO3- concentration within the influent of the CWs system employed at Kotebe WWTP was 10.05 ± 2.878 mg/L and recorded in autumn. But the lowest average NO3- concentrations, 1.43 ± 0.611 mg/L was obtained within the effluent of the vertical CWs in winter. The performances of the horizontal, vertical and hybrid systems were not differed significantly from season to season in removing NO3-; one-way ANOVA results of the horizontal, vertical and hybrid systems were F0.95 (3, 20) = 0.772; P > 0.05, F0.95 (3, 20) = 0.580; P > 0.05 and F0.95 (3, 20) = 0.655; P > 0.05 respectively. However, environmental factors are known to influence denitrification rates and temperature is one of the key factors within the CWs system (Robert 2004).
As in other biological processes, growth rates in aquatic plant systems depend on temperature and the vegetated system show a much better performance during the warmer months of the year (Karathanasis, Potter and Coyne, 2003; Wang and Li, 2014). Lee et al (2013) also pointed out that the dependence of removal efficiency on temperature is significant due to plant uptake, which plays a significant role in nutrient removal. Greater bacterial activity is shown during the warmer season than the colder one (Chon and Chon, 2015). So, warmer climate improves performances, especially for nitrification (Masi, Caffaz and Ghrabi, 2013; Molle et al, 2015).
In general, it is stated that the minimum temperatures for nitrates production are 4–5 ?C (Prochaska et al, 2007). But NO3- removal was not significantly differed in pertinent to the change of seasons in this study. This could be related with the fact that the temperature in the study (tropical and sub-tropical areas) usually exceeds the minimum temperature required for nitrate production.
Table 4.10: Average values of NO3- concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
Sr. no Seasons of
the year Influent HFCW effluent VFCW effluent Hyb CW
Average
NO3-
(St. dev) Average
NO3-
(St. dev) NO3-
RE (%) Average
NO3-
(St. dev) NO3-
RE (%) Average
NO3-
(St. dev) NO3-
RE (%)
1 Winter 6.09
(2.20) 1.83
(0.71)
70.0 1.43
(0.61)
76.6 1.56
(0.56)
74.5
2 Spring 8.08
(2.90) 2.65
(0.83)
67.2 1.92
(0.84)
76.2 1.69
(0.73)
79.0
3 Summer 5.83
(2.60) 1.88
(1.13)
67.7 1.82
(1.01)
68.8 1.61
(0.58)
72.5
4 Autumn 10.05
(2.88) 4.90
(2.39)
51.2 3.57
(0.93)
64.5 3.21
(1.98)
68.1
5 Overall
7.51
(1.97) 2.82
(1.44) 64.0
(8.64) 2.19
(0.95) 71.5
(5.90) 2.02
(0.80) 73.5
(4.52)

Figure 4.13: Concentration of NO3- of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.

Figure 4.14: Seasonal NO3- removal efficiencies of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
Based on the monitoring result, the loading rate of NO3- was found to be in the range of 0.15 and 0.64 g/m2.d (0.33 ± 0.133 g/m2.d) (Fig 4.12). The correlation between loading rate and removal rate of the vertical system (R2 = 0.855) was slightly strong compared to the hybrid system (R2 = 0.826), while both the vertical and the hybrid system showed strong correlation than the horizontal system (R2 = 0.641). So, it is possible to conclude that the vertical and the hybrid CWs had the capacity to hold and treat nitrate as the concentration increases.

NO3- loading rate (g/m2.d)
Figure 4.15: NO3- loading rate (g/m2.d) against the NO3- removal rate (g/m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
4.3.7 Removal of Total Nitrogen (TN)
Total nitrogen commonly comprises of different forms of nitrogen including organic, ammonia and oxidized nitrogen (nitrite and nitrate); and hence its removal is considered as the reduction of those compounds in wetlands. These forms of nitrogen undergo sequential changes, including primarily ammonificaiton which is again converted by the process of nitrification and finally by denitrification. The conversion from one form to another is carried out at varying rates depending on environmental conditions. Releases from decaying plants and microbial biomass also enhance TN in the water. As a result, the rate of TN reduction along the flow path in wetlands is expected to be decided partly by the speciation of the incoming nitrogen. The rates of TN removal for nitrified influents are expected to be considerable since because the antecedent alterations of ammonia and organic nitrogen have already taken place in pretreatment (Kadlec and Wallace, 2009).
As it is shown in Figure 4.13 and Table 4.11, the average TN values in the influent of the horizontal, vertical and hybrid CWs applied at Kotebe WWTP were found to be in the range of 57.7 – 92.7 mg/L and the average value during the performance monitoring period was 76.6 ± 9.06 mg/L. Although the value was comparable with the result reported by Lu et al (2015), In the contrary, it was lower than the values reported by (Prochaska et al, 2007; Konnerup et al, 2009; Zurita et al, 2009 and Tuncsiper et al, 2012) (Table 4.3).
Similarly, the concentrations of TN in the outlets of the horizontal, vertical and the hybrid constructed wetland systems varied between 28.7 – 46.0 mg/L (38.8 ± 3.36), 24.3 – 42.7 mg/L (34.2 ± 1.66 mg/L) and 21.3 – 38.7 mg/L (31.4 ± 1.52 mg/L), respectively. The TN concentration values within the effluents of the three CWs system comply with provisional discharge standards, 60 mg/L (EEPA).
In line with this, the performance of the horizontal, vertical and hybrid CWs system employed at Kotebe WWTP in removing TN was evaluated based on its concentration and the removal efficiencies were ranged between 27.3 – 63.0% (49.1 ± 3.68%) for the horizontal flow, 34.2 – 70.5% (54.9 ± 4.94%) for the vertical flow and 47.1 – 69.8% (58.7 ± 5.38%) for the hybrid flow type; with the lowest removal efficiency for the horizontal flow type and the highest removal efficiency for the hybrid system (Figure 4.14 and Table 4.11). The performances of the horizontal, vertical and hybrid flow constructed wetland systems in removing TN were significantly different from one another statistically (one-way ANOVA; F0.95 (2, 69) = 7.543; P 0.05, F0.95 (3, 20) = 0.580; P > 0.05 and F0.95 (3, 20) = 0.655; P > 0.05 respectively. But Seong-Heon et al (2016) pointed out that the TN removal efficiency in summer season was generally lower than that in winter, autumn, and spring.
In most CW types, denitrification plays the major role in the removal of nitrogen although the nitrate concentration in wastewater is usually low (Thomas and William, 2001). Nevertheless, environmental factors known to influence denitrification rates include the absence of O2, redox potential, substrate moisture, temperature, pH value, presence of denitrifiers, substrate type, organic matter, nitrate concentration and the presence of overlying water (Vymazal, 2007; Russo, 2008). Also, Robert (2004) explained that pH, temperature, organic carbon, nitrate levels, and the ecological interactions and exposure times of the denitrifying bacteria within the system are the key factors. In general, although the reaction dependent on a number of factors, denitrification is the permanent removal of Nitrogen from the system (WI, 2003).

Table 4.11: Average values of TN concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
Sr. no Seasons of the year Influent HFCW effluent VFCW effluent HyCW
Average
TN
(St. dev) Average
TN
(St. dev) TN
RE (%) Average
TN
(St. dev) TN
RE (%) Average TN
(St. dev) TN
RE (%)
1 Winter 87.2
(4.02) 43.2
(1.63)
50.5 36.2
(3.71)
58.5 32.2
(3.12)
63.1
2 spring 80.8
(12.01) 37.6
(5.27)
53.4 32.5
(5.04)
59.8 29.5
(2.22)
63.5
3 Summer 67.3
(2.96) 35.2
(6.15)
47.8 33.4
(5.51)
50.4 30.8
(6.79)
54.3
4 autumn 71.1
(4.44) 39.3
(4.33)
44.8 34.9
(4.47)
50.9 32.9
(3.64)
53.7
Overall
performance 76.6
(9.06) 38.8
(3.36) 49.1
(3.68) 34.2
(1.66) 54.9
(4.94) 31.4
(1.52) 58.7
(5.38)

Figure 4.16: Concentration of TN of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.

Figure 4.17: Seasonal TN removal efficiencies of the horizontal, vertical and hybrid CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
During the performance monitoring period, the loading rate of TN were found to be in the range of 2.54 and 4.08 g/m2.d (3.37 ± 0.452 g/m2.d) (Fig 4.18). The linear correlation of TN loading rate and TN removal rate for the hybrid system (R2 = 0.855) was better than the correlation by the vertical flow system (R2 = 0.823) and the vertical flow system was better than the correlation by the horizontal flow system (R2 = 0.752). So, as the loading rate increases, the hybrid flow system could have the capacity to hold and remove TN.

TN loading rate (g/m2.d)
Figure 4.18: TN loading rate (g/m2.d) against the TN removal rate (g.m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
4.3.8 Removal of Phosphate (PO43-)
Soluble reactive P is the analytical term given to biologically available orthophosphate, which is the primary inorganic form (Scholz, 2006). The most reactive forms are the dissolved phosphates, which change hydration in response to pH. The most common species are mono- and dibasic phosphates, which dominate at all typical wetland pH values (4 < pH 0.05). This might have occurred since all the three wetland systems use the same substrate/media. Whereas (Prachaska et al, 2007) pointed out that the main removal mechanism for the reduction of orthophosphate is chemical adsorption.
A combination of physical, chemical and biological processes is employed in treatment wetlands for removing P from wastewaters. Organic forms of P are much less biologically and chemically reactive in wetlands than orthophosphate. Settling of particulate organic P is important means for its removal. Both dissolved and particulate organic P ultimately may be biologically broken down to inorganic P (mineralization) and subsequently removed through different processes (Thomas and William, 2001; Kadlec and Wallance, 2009). Usually, bacteria removal and plant uptake are responsible for P-PO4-3 removal, while precipitation and adsorption are responsible for the removal of all P forms (Kadlec and Knight, 1996).
Iron and Aluminum oxides can precipitate phosphate to form new mineral compounds: Fe phosphate and Al phosphate. These new forms are likely very stable in the soil and create long-term storage of P. Thus, precipitation is governed by the availability of co-precipitating compounds such as Fe, Al, and dissolved oxygen and water chemistry, especially pH. Similarly, high concentration of calcium in wetlands can precipitate phosphate to form calcium phosphate, which is stable for long period of time (Thomas and William, 2001; UNHSC and NEIWPCC, 2010).
Additionally, Zheng et al (2016) reported that nutrient uptake of plants accounted for a higher proportion of P removal in SSF wetland system. Uptake of phosphates by microorganisms, such as bacteria, algae, and macrophytes, functions as a short-term, rapid-cycling means for soluble and insoluble forms. But most of the phosphate is returned back into the water column by cycling through the growth, death, and decomposition process. Some phosphate is lost in the process due to long-term accretion in newly formed sediments (USEPA, 2000).
The maximum average PO43- concentration within the influent of the CW systems employed at Kotebe WWTP was 7.7 ± 0.480 mg/L and recorded in spring; while the lowest average PO43- concentration, 1.81 ± 0.921 mg/L was obtained within the effluent of the hybrid CW in autumn. The performances of the horizontal, vertical and hybrid systems did not differ significantly from season to season in removing PO43-; one-way ANOVA results of the horizontal, vertical and hybrid systems were F0.95 (3, 20) = 2.069; P > 0.05, F0.95 (3, 20) = 1.746; P > 0.05 and F0.95 (3, 20) = 2.381; P > 0.05 respectively. Even though, the performance of the CWs system did not show significant difference with change of seasons, Prachaska et al (2007) revealed that seasons have an effect on orthophosphates removal.
Table 4.12: Average values of PO43- concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
Sr. no Seasons of the year Influent HFCW effluent VFCW effluent HyCW
Average PO43- (St. dev) Average PO43- (St. dev) PO43-
RE (%) Average PO43- (St. dev) PO43-
RE (%) Average PO43- (St. dev) PO43-
RE (%)
1 Winter 7.00
(0.90) 4.06
(0.94)
42.0 3.77
(0.78)
46.1 4.16
(0.86)
40.6
2 Spring 7.67
(0.48) 4.55
(0.65)
40.6 4.29
(0.37)
44.0 4.33
(0.64)
43.5
3 Summer 4.09
(1.58) 2.40
(0.48)
41.4 2.01
(0.68)
51.0 2.10
(0.37)
48.6
4 Autumn 4.60
(1.33) 1.96
(0.84)
57.4 1.84
(0.87)
59.9 1.81
(0.92)
60.8
5 Overall
5.84
(1.76) 3.24
(1.26) 45.4
(8.05) 2.98
(1.24) 50.3
(7.07) 3.10
(1.33) 48.4
(8.92)

Figure 4.19: Concentration of PO43- of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.

Figure 4.20: Seasonal PO43- removal efficiencies of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
The PO43- loading rate at the time of performance monitoring varied in the range between 1.31 and 0.364 g/m2.d (0.257 ± 0.083 g/m2.d) (Fig 4.21). The correlation between loading rate and removal rate by the horizontal, vertical and hybrid system showed that the PO43- removal rate was dependent on the PO43- loading rate and the removal rate for vertical flow system (R2 = 0.605) was somewhat more dependent on loading rate than the horizontal system (R2 = 0.543) and the hybrid system (R2 = 0.490); while the correlation between remval rate and loading rate in the horizontal system is slightly stronger than the hybrid system.

PO43- loading rate (g/m2.d)
Figure 4.21: PO43- loading rate (g/m2.d) against the PO43- removal rate (g/m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
4.3.9 Removal of Total Phosphorous (TP)
Phosphorus (P) is one of the essential macronutrients required by plants for growth, and is a limiting factor for the growth of vegetation (Kadlec and Wallace, 2009). Hence, addition of P to the environment often contributes to the occurrence of eutrophication of lakes and coastal waters (Thomas and William, 2001). A measure of relative ecosystem requirements is the proportion among the nutrient elements in the biomass, which is often represented as a molar proportion of C: N: P =106:16:1, or 41:7:1 on a mass basis (the Redfield ratio). Wastewaters do not have this ratio except by rare chance, and most often, there is excess P in domestic wastewater (Kadlec and Wallace, 2009).
During the monitoring period, the average TP concentrations of the influent which fed the pilot scale CWs system applied at Kotebe WWTP were found to be in the range of 3.30 – 13.03 mg/L (7.42 ± 2.46 mg/L). Even though, the value was comparable with the result obtained in a number of studies (Sleytr et al, 2007; Konnerup et al, 2009; Zurita et al, 2009; Tuncsiper et al, 2012 and Lu et al, 2015), it was higher than the influent concentrations reported in Brazil (Lana et al, 2015) and Ireland (Abdelhakeem et al, 2016).
Although none of the processes involved in phosphorous removal actually remove P from the wetland; they transform and/or store P in materials and compounds which may then re-release the P when conditions change (UNHSC and NEIWPCC, 2010). The principal P removal mechanisms in wetlands systems are sorption within the bottom soils, precipitation of phosphates under elevated pH conditions, uptake by the macrophytic plants, storage and fixation by algae and bacteria (Robert, 2010; UNHSC and NEIWPCC, 2010).
The average TP concentrations in the effluents were varied between 1.20 – 4.87 mg/L (3.06 ± 0.77 mg/L) for the horizontal flow system, 1.07 – 7.33 mg/L (3.52 ± 1.01 mg/L) for the vertical flow system and 1.03 – 5.73 mg/L (3.35 ± 1.02 mg/L) for the hybrid flow system. These values meet the provisional discharge standards, 10 mg/L which is set by Ethiopian Environmental Protection Authority (EEPA, 2003).
The concentration based removal efficiencies of the horizontal, vertical and hybrid CWs in removing TP varied from 29.3 – 75.5 % (58.0 ± 4.53 %), 22.7 – 72.9 % (51.8 ± 4.13%) and 28.7 – 72.1 % (54.5 ± 3.53 %), respectively. The slight increase in TP removal in the horizontal flow system could be attributed to the accumulation of undecomposed organic P in the system as the rate of microbial action is slow under anaerobic condition/limited aeration. This might prohibit the re-release of phosphorous into the effluent of the horizontal system which in turn decreases the concentration of TP in the outlet.
Nevertheless, the performance of the horizontal, vertical and hybrid flow systems in removing TP were not significantly different from one another (one-way ANOVA; F0.95 (2, 69) = 2.436; P > 0.05). Higher removal of TP was reported in a study conducted in small scale CWs for domestic wastewater treatment in China (Qiong et al, 2007).
The removal of TP varied between 40 and 60% in all types of CWs depending on the type of CWs and inflow loading and P removal is mainly influenced by wetland substrate (Vymazal, 2007; Li et al, 2010). Siti et al (2011) described that the use of specialized media in CWs to improve P removal should be developed and demonstrated since P removal always shows worse performance in the wetlands. The low removal efficiency in this study could be related with the application of gravel as fill media since gravel could not be considered as a good P-adsorption wetland media. Therefore, removal mechanisms such as filtration, plant uptake and biological assimilation are assumed to be the main ones in removing phosphorous in this study.
In many cases, wetlands do not provide the high level of efficient long-term removal for P than they provide for N. This is, partly, due to the lack of a gaseous sink, analogous to denitrification, for P removal (Haberl, Perfler and Mayer, 1995; Thomas and William, 2001; Siti et al, 2011; Rozema et al, 2016). In wetland soils, P occurs as soluble or insoluble, organic or inorganic complexes. Its cycle is sedimentary rather than gaseous and predominantly forms complexes within organic matter in peatlands or inorganic sediments in mineral soil wetlands. Over 90% of the P loads in streams and rivers may be present in particulate inorganic form (Scholz, 2006).
The maximum average TP concentration within the influent of the CWs system employed at Kotebe WWTP was 10.41 ± 2.23 mg/L and recorded in spring while the minimum concentration was recorded in summer. The lowest value during the rainy season might be attributed to the dilution effect caused by the infiltration of surface runoff into the sewerage system, from which the wastewater was piped to the pilot-scale CW systems. Meanwhile, the lowest average TP concentration, 2.07 ± 0.82 mg/L was obtained within the effluent of the horizontal CW in summer. But performances of the horizontal, vertical and hybrid systems did not differ significantly from season to season in removing TP; one-way ANOVA results of the horizontal, vertical and hybrid systems were F0.95 (3, 20) = 0.882; P > 0.05, F0.95 (3, 20) = 0.417; P > 0.05 and F0.95 (3, 20) = 0.657; P > 0.05 respectively.

Table 4.13: Average values of TP concentration (mg/L) and removal efficiencies (%) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
Sr. no Seasons of
the year Influent HFCW effluent VFCW effluent HyCW
Average
TP
(St. dev) Average
TP
(St. dev) TP
RE (%) Average
TP
(St. dev) TP
RE (%) Average TP
(St. dev) TP
RE (%)
1 Winter 8.28
(2.71) 3.26
(0.95)
60.6 3.59
(1.41)
56.6 3.49
(1.39)
58.0
2 Spring 10.41
(2.23) 3.92
(0.77)
62.4 4.83
(1.62)
53.6 4.63
(1.05)
55.5
3 Summer 4.77
(1.15) 2.07
(0.82)
56.7 2.39
(0.72)
49.9 2.16
(0.66)
54.7
4 Autumn 6.21
(1.64) 2.97
(1.15)
52.2 3.28
(1.68)
47.2 3.13
(1.42)
49.6
5 Overall
7.42
(2.46) 3.06
(0.77) 58.0
(4.53) 3.52
(1.01) 51.8
(4.13) 3.35
(1.02) 54.5
(3.53)

Figure 4.22: Concentration of TP of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.

Figure 4.23: Seasonal TP removal efficiencies of the horizontal, vertical and hybrid pilot scale CW systems employed in Addis Ababa, Ethiopia.
During the performance monitoring period, the loading rate of TP were ranged between 0.145 and 0.573 g/m2.d (0.326 ± 0.127 g/m2.d) (Fig 4.23). The correlation between loading rate and removal rate by the horizontal, vertical and hybrid system showed that the TP removal rate was dependent on the TP loading rate; and the correlation for horizontal system (R2 = 0.902) was slightly stronger than the hybrid system (R2 = 0.837). On the other hand, the correlation for both the horizontal and the hybrid system were stronger compared to the vertical system (R2 = 0.743). The difference might be attributed to the type of wastewater flow into each wetland cell and the hydraulic condition of the system as the pilot scale CWs were made to function at similar situations.

TP loading rate (g/m2.d)
Figure 4.24: TP loading rate (g/m2.d) against the TP removal rate (g/m2.d) of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
4.3.10 Removal of Fecal Coliform (FC)
Although total coliforms have been used in older and monitoring studies, many of the organisms in this broad group are not limited to fecal sources. Accordingly, the subgroup of fecal coliform has become the preferred indicator (Kadlec and Wallace, 2009).
The average concentration of FC in the inlet of the CWs system used at Kotebe WWTP to treat domestic wastewater varied from 72,000 – 127,667 CFU/100 ml (95,292 ± 13,190 CFU/100 ml). The average concentration of FCs obtained in this study was noticeably lower compared to the value reported in Tanzania (Mahenge, 2014), while Mahauri et al (2000) reported lower concentration of fecal coliforms. Meanwhile, the values in this study were in agreement with the range of FC concentration reported by Caselles-Osorio et al (2011).
During the monitoring period, the average concentrations of the effluent from the horizontal, vertical and hybrid CWs system varied between 967 – 5,800 CFU/100 ml (2,010 ± 477 CFU/100 ml); 2,167 – 5,233 CFU/100 ml (3,250 ± 927 CFU/100 ml) and 1,967 – 5,400 CFU/100ml (3,353 ± 834 CFU/100ml) respectively.
Accordingly, the concentration based removal efficiency of the horizontal, vertical and hybrid flow CWs system applied at Kotebe WWTP was ranged between 96.1 – 98.8 %(97.9 ± 0.383%), 95.7 – 98.3% (96.6 ± 0.568 %) and 94.8 – 97.6% (96.5 ± 0.427 %) respectively. The performance of the horizontal, vertical and hybrid flow systems in removing FC were significantly different from one another (one-way ANOVA; F0.95 (2, 69) = 29.518; P < 0.05).
But the performances of the horizontal and vertical flow systems differed significantly from season to season while the hybrid flow system was not significantly different from season to season in removing TN; one-way ANOVA results of the horizontal, vertical and hybrid systems were F0.95 (3, 20) = 3.125; P < 0.05, F0.95 (3, 20) = 8.052; P 0.05, respectively.
Many of the enteric bacteria and viruses cannot survive long once they are out of the host organisms. As a result, they start to die out and are removed by natural processes/hostile environment in the treatment wetland systems (Thomas and William, 2001). In addition to this, competition, predation, sedimentation, filtration, adsorption, pH extremes and photolysis can play significant role in the removal of bacteria and viruses in natural systems such as CWs (Thomas and William, 2001). In tropical and subtropical climates, it is possible to remove harmful pathogenic organisms and to produce disinfected reclaimed wastewater without using expensive disinfectants, especially in poor areas where the reclamation of raw wastewater in agriculture is endangering human health. Hence, application of wetland system is especially suitable for small communities in developing countries, where the potential health benefits from pathogen removal are considerable (Shutes, 2001; Zurita and Carreon-Alvarez, 2015).
In recent times, CWs have been implemented to treat different types of municipal wastewaters (Desena, 1999) with examples that include effective secondary and tertiary applications for the removal of pathogens such as FCs (Raymundo, 2008). Mahenge (2014) pointed out that there is a significant reduction in concentration of FCs in treatment wetlands and the removal rate of FCs reaches 98%. But adequate time (> 5-15 days) is required to allow the system to operate more in a steady state condition for treatment of sewage to acceptable levels. In general, treatment wetlands show considerable potential for removing fecal bacteria from domestic wastewater (Fountoulakis et al, 2009; Vallejos, Caballero and Champagne, 2015; Sleytr et al, 2007).
Macrophytes-based systems turned out to be a good alternative for wastewater treatment concerning bacterial removal and water quality. In contrast, those systems without plants show lower efficiencies than their corresponding planted wetlands. It is also found that mean removal efficiencies and surface removal rates turn out to be significantly high in wetlands, and some increases in removal efficiencies are associated with warm season (Mercedes et al, 2008; Foladori, Bruni and Tamburini, 2015; Wu et al, 2016). Zurita and Carreon-Alvarez (2015) pointed out that a wetland tends toward a better performance during the maturity period reached by the system, noticeable through the presence of well-developed macrophytes.
Additionally, Tuncsiper, Ayaz and Akca (2012) revealed that the HRT and the loading rates are two of the most important factors in removing coliforms although the rate can be affected by a number of other conditions and environmental factors.
Table 4.14: Average values of FC concentration (CFU/100ml) and removal efficiencies (%) of the pilot scale CW systems applied at Kotebe WWTP, Addis Ababa, Ethiopia.
Sr. no Seasons of the year Influent HFCW effluent VFCW effluent HyCW
Average
FC
(St. dev) Average
FC
(St. dev) FC
RE (%) Average
FC
(St. dev) FC
RE (%) Average
FC
(St. dev) FC
RE (%)
1 Winter 101000
(26393) 2410
(1672)
97.6 3811
(1304)
96.2 3217
(2087)
96.8
2 Spring 110778
(21089) 2167
(494)
98.0 4183
(1091)
96.2 4545
(643)
95.9
3 Summer 81167
(4839) 1317
(260)
98.4 2133
(5834)
97.4 2611
(481)
96.8
4 Autumn 88222
(9708) 2145
(482)
97.6 2872
(581)
96.7 3039
(994)
96.6
Overall
performance 95292
(13190) 2010
(477) 97.9
(0.383) 3250
(927) 96.6
(0.5679) 3353
(834) 96.5
(0.4272)

Figure 4.25: Concentration of FC of the influent and effluents of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.

Figure 4.26: Seasonal FC removal efficiencies of the horizontal, vertical and hybrid pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.

According to the results of the study, the FC loading rate was ranged between 3.168 x 107 – 6.527 x 107 CFU/m2/d (4.193 x 107 ± 8.91 x 106 CFU/m2.d) (Fig 4.6). The linear correlation between FC loading rate and FC removal rate was almost similar among the three wetland systems. For that reason, the linear correlation were (R2 = 0.999) for the horizontal flow wetland, (R2 = 0.999) for the vertical flow wetland and (R2 = 0.999) for the hybrid flow bed. Although, all of the three systems show almost similar linear correlation between FC loading rate and FC removal rate, the capacity to hold and remove FC was slightly better in the horizontal and vertical systems as the concentration increases.

Figure 4.27: FC loading rate against FC removal rate of the horizontal, vertical and hybrid flow pilot scale CW systems applied at Kotebe WWTP, Addis Ababa, Ethiopia.
4.4 Kinetic parameters determination
The first order plug flow equation (Equation 2.12 to 2.14) which was discussed in detail in section 2.7 was used to calculate the areal removal rate constants. Therefore, the rate constants for each of the parameters were calculated using the annual and seasonal mean concentrations of the influent, effluent of the horizontal, vertical and hybrid flow systems and the loading rate of the respective pollutants. The results of the calculation were presented in Table 4.15 and Annex Table 10.
The values of the areal removal rate constants of BOD5, COD, TSS, NH4+, TN, PO43+, and TP were compared with literature values. In view of that, the values for BOD, TN, PO43- – P and TN were found to be higher than the literature values while the areal removal rates for COD, TSS, NH4+ – N of this study were lower than the literature values.
On the other hand, winter, spring, summer and autumn seasons were compared each other to understand the effects of seasons on the values of areal removal rate constants. The effect of seasons on pollutants removal was clearly indicated as virtually all pollutants showed higher areal removal rate constants in winter (Dec., 2015 – Feb., 2016) and spring (Mar. – May, 2016) than in summer and autumn seasons.
Table 4.15: Summary of areal removal rate constants, K (m/d) of the parameters considered in this study.
Parameters K, HF* K, VF** K, HybF*** K, literatures
BOD 0.098 0.113 0.121 0.077
0.27 Sheridan et al, 2014
Abdelhakeem et al, 2016
COD 0.073 0.076 0.081 0.283
0.44
0.2 Konnerup et al, 2009
Yan Zhang, 2014
Abdelhakeem et al, 2016
TSS 0.097 0.080 0.084 0.109
0.82
0.19 Korkusuz et al, 2004
Yan Zhang, 2014
Abdelhakeem et al, 2016
NH4+-N 0.040 0.048 0.048 0.085
0.05 Korkusuz et al, 2004
Abdelhakeem et al, 2016
TN 0.030 0.036 0.039 0.0158
0.73 Konnerup et al, 2009
Yan Zhang, 2014
PO43—P 0.026 0.030 0.028 0.002 Korkusuz et al, 2004
TP 0.039 0.033 0.035 0.0149
0.46
0.03 Konnerup et al, 2009
Yan Zhang, 2014
Abdelhakeem et al,2016
* Horizontal Flow ** Vertical Flow *** Hybrid Flow

4.5 Plant tissue nutrient (N and P) content
The nutrient contents of the above-ground (stem and leaves) and below-ground (root and rhizomes) were determined and expressed in terms of TN and TP percentages in dry weight of C. papyrus biomass (Table 4.16). Based on the result, the above-ground parts of C. papyrus had higher N content among the three CWs system. Below-ground and above-ground N contents of C. papyrus were 1.56 ± 0.26% and 2.27 ± 0.57% in horizontal bed, 1.75 ± 0.44 and 2.74 ± 0.52% in vertical bed, and 1.80 ± 0.45% and 2.63 ±0.53% in hybrid bed, respectively. This was in agreement with the results of several studies conducted by many authors who reported that the above-ground parts of wetland plants have in general higher N content (Korkusuz et al, 2004, Achyut et al, 2011, Yesbie and Mengistu, 2014; Yi Chen et al, 2014). This could be due to the fact that nitrogen is essential element to produce proteins and chlorophyll to make large leaves and thick stems and to carryout photosynthesis.
The N contents of the below-ground and above-ground parts of C. papyrus in the horizontal, vertical and hybrid type CWs differed significantly from each other statistically and the one-way ANOVA results were F0.95 (1, 10) = 7.866; P < 0.05, F0.95 (1, 10) = 12.717; P < 0.05 and F0.95 (1, 10) = 8.694; P < 0.05, respectively.
However, N uptake by wetland plants reduces while its concentration and load in wastewater rises up. This shows that plants capability for N uptake is limited and it can be considered as efficient method under situations where the load of N is minimal (Avsar et al, 2007). Zheng et al (2016) reported that plants nutrients uptake accounted for a higher proportion of the N removal in FWS, and higher proportion of P removal in SSF wetland system.
Zhang, Gersberg and Keat (2009) reported that the removal efficiency of planted CWs is higher than unplanted CWs for certain pollutants such as TN and NH4-N. Moreover, the performance of planted wetlands in removing nitrogen is usually found to be efficient and steady in all months of the year (Lee and Scholz, 2007; Fonkou et al, 2011; Abou-Elela and Hellal, 2012; Mesquita, Albuquerque and Nogueira, 2012).
In the process of assimilation, the plants reduce inorganic N to organic N compounds, plant structure. There is significantly high rate of N uptake by wetland plants from water and sediments during the growing season. Increased immobilization of nutrients by microbes and uptake by algae and epiphytes also lead to retention of inorganic N. The net annual uptake of N by macrophytes approximately ranges between 0.5 to 3.3 g N/m2/yr (USEPA, 2000).
But in the case of phosphorous, below-ground parts of C. papyrus were found to have higher percentage than above-ground parts. The P percentage in below-ground and above-ground C. papyrus parts were 0.139 ±0.43% and 0.064 ± 0.033% in horizontal flow, 0.167 ± 0.063% and 0.067 ±0.029% in vertical flow, and 0.115 ± 0.026% and 0.065 ± 0.031% in hybrid flow type, respectively. The higher P content in belowground C. papyrus might be related with the fact that P is used by plants to develop strong root system.
The percentage of P in below-ground and above-ground parts of C. papyrus in the horizontal, vertical and hybrid type CWs differed significantly from each other statistically and the one-way ANOVA results were F0.95 (1, 10) = 11.563; P < 0.05, F0.95 (1, 10) = 12.482; P < 0.05 and F0.95 (1, 10) = 9.082; P < 0.05, respectively.
The uptake and release of P occur similarly as that of the microbes, but the reactions require long period of time, possibly months to years. Uptake occurs during the growth phase of the plant and release occurs during plant senescence and death, followed by decomposition (USEPA, 2000). Adhikari et al (2011) states that P concentration was higher in the belowground parts of wetland plants, which suggests that harvest of the root system would be necessary for achieving maximum P removal. Assimilation of P in vegetation is usually short-term and decomposition of detrital plant tissue is usually rapid resulting in release of P. However, the undecomposed organic P accumulates in the system and becomes an integral part of the soil/sediment P pool (Reddy et al, 1995).
Generally, the nutrient removal capacity of a wetland system was more dependent on individual plant biomass irrespective of plant type, i.e., on the size of individual plants or plant density. The nitrogen concentration was higher in aboveground plant parts but the phosphorus was higher in the belowground parts, which suggests that harvest of the root system would be necessary for maximum phosphorus removal, but an aboveground harvest would be sufficient for nitrogen removal from our wetlands systems. Plant nutrients in the four wetland sites correlated well with ambient nutrient concentrations in the sediments and water columns, irrespective of the type of plants present (Adhikari et al, 2011).
Table 4.16: Average values of plant tissue nutrient content (%) of the pilot scale CW systems applied at Kotebe WWTP, Addis Ababa, Ethiopia.

Nutrient C. papyrus
in horizontal cell C. papyrus in vertical cell C. papyrus in hybrid cell
Root Stem Root Stem Root Stem

% of TN (May) 1.59
(±0.17) 2.18
(±0.81) 2.08 (±0.19) 2.39 (±0.35) 1.97 (±0.619) 2.61 (±0.64)

% of TN (November) 1.52
(±0.37) 2.36
(±0.40) 1.43 (±0.35) 3.08 (±0.44) 1.62 (±0.14) 2.64 (±0.53)

TN average 1.56
(±0.26) 2.27
(±0.57) 1.75 (±0.44) 2.74 (±0.52) 1.80 (±0.45) 2.63 (±0.53)

% of TP (May) 0.139 (0.045) 0.057 (±0.022) 0.211 (±0.038) 0.081 (±0.027) 0.118 (±0.02) 0.058 (±0.011)

% of TP (November) 0.138 (±0.050) 0.071 (±0.045) 0.123 (0.051) 0.054 (±0.028) 0.111 (±0.036) 0.071 (±0.047)

% TP average 0.139 (±0.43) 0.064 (±0.033) 0.167 (±0.063) 0.067 (±0.029) 0.115 (±0.026) 0.065 (±0.031)

Figure 4.28: Percentage of TN in plant tissue of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.

Figure 4.29: Percentage of TP in plant tissue of the pilot scale CW systems employed at Kotebe WWTP, Addis Ababa, Ethiopia.
CHAPTER FIVE
CONCLUSION AND RECOMMENDATION
5.1 Conclusion
The application of effective, low-cost, less-energy intensive, and easily operated secondary wastewater treatment methods is a very demanding issue in Ethiopia as discharging of untreated wastewater into water bodies is the common practice in most parts of the country. It is crucial to protect the environment and public health, and beyond this it can create further opportunity to re-use the treated wastewater for non-domestic purposes such as irrigation, construction works and ground water recharging.
The findings of this study give a snapshot of the performance of CW systems in tropical areas like Ethiopia. It is assumed to contribute to the understanding of how the horizontal, vertical and hybrid subsurface flow constructed wetland systems with similar wetland plants and fill media worked in the prevailing climatic conditions in Addis Ababa. It also helps to weigh up the impact of different seasons in the removal of pollutants in domestic wastewater. Based on the results presented, the HFCW, VFCW and hybrid of the two systems showed good BOD, COD, TSS, NH4+, NO3-, TN, TP and FC removal and the effluent concentrations of all wastewater pollutants except NH4+ comply with the discharge standard set by Ethiopian Environmental Protection Authority (EEPA, 2003). However, the annual average NH4+ concentrations of the effluents were higher than the provisional discharge standards. The annual average removal of PO43- was low during the monitoring period although the effluent concentrations comply with the discharge standard.
In line with this, the results of one way ANOVA showed that the performances of the HFCW, VFCW and the hybrid wetland systems were different from one another statistically for BOD5, TSS, NH4+, TN and FC; while the efficiency of the three systems did not differ from one another for the parameters: COD, NO3-, PO44+, and TN. Similarly, one way ANOVA to compare the performances of the CWs in different seasons revealed that the performances of all CW systems were differed significantly from season to season for BOD5, COD, and NH4+.
Concerning the nutrient content of the wetland plant /Cyprus papyrus/, it was observed that the TN content of the above-ground part of the wetland plant was higher than the TN content of below-ground plant part. On the contrary, the TP content of the below-ground plant part was higher than the TP content of the above-ground part.
In general, it can be concluded that the treatment performance of the pilot-scale CW systems applied At Kotebe WWTP was very promising for the promotion and application of CWs as an alternative wastewater treatment system to protect the environment and public health. Ethiopia has favorable climatic conditions for the implementation of CW systems and hence, the technology can continue as competent solution to alleviate the inherent environmental problems associated with discharging of untreated wastewaters. It is possible to use CW systems for wider application in different towns for the treatment of municipal wastewater or in small communities/institutions such as universities, colleges, military camps, farms, factories and hospitals.

5.2 Recommendations
Based on the results obtained from this study, the following recommendations are forwarded for further studies and wider application of constructed wetlands as alternative wastewater treatment methods.
The removal efficiency of the constructed wetland systems for the removal of most wastewater pollutants considered in this study was good to use the method particularly in small towns and various institutions. However, the selection of wetland plants and fill media should be based on adequate knowledge and experience to optimize the efficiency of constructed wetlands in removing pollutants particularly nitrogen and phosphorous.
Integrated constructed wetland systems which comprise of two or more cells connected in series should be employed to improve the quality of effluents.
The performance of the pilot scale constructed wetland systems under similar environmental conditions should be evaluated before their implementation at full scale level and also the pilot scale CW systems should be monitored for at least one year to properly address the effect of different seasons.
The characteristics of raw wastewaters in wet and dry seasons should be determined and taken in to account during designing and operation of constructed wetland systems.
More detailed research works should be conducted to understand the different processes carried out in constructed wetlands.

References:
Abdelhakeem, S. G., Aboulroos, S. A. and Kamel, M. M. (2016). Performance of a vertical subsurface flow constructed wetland under different operational conditions. Journal of Advanced Research, 1-12.
Abdel-Halim, W. and Rosenwinkel, K. H. (2005). Domestic wastewater treatment and re-use. Closing the loop. EMPOWERS Regional Symposium: End-user ownership and involvement in IWRM, Cairo, Egypt.
Abidi, S., Kallali, H., Jedidi, N., Bouzaiane, O. and Hassen, A. (2009). Comparative pilot study of the performances of two constructed wetlands wastewater treatment hybrid systems. Desalination, 246, 370-377.
Abou-Elela, S. I. and Hellal, M. S. (2012). Municipal wastewater treatment using vertical flow constructed wetlands planted with Canna, Phragmites and Cyprus. Ecological Engineering, 47, 209-213.
Abou-Elela, S. I., Golinelli, G., El-Tabl, A. S. and Hellal, M. S. (2014). Treatment of municipal wastewater using horizontal flow constructed wetlands in Egypt. Water Science and Technology, 69 (1), 38-47.
Adhikari, A. R., Acharya, K., Shanahan, S. A. and Zhou, X. (2011). Removal of nutrients and metals by constructed and naturally created wetlands in the Las Vegas valley, Nevada. Environ Monit Assess, 180, 97-113.
Albalawneh, A., Chang, T. K., Chou, C. and Naoum, S. (2016). Efficiency of a horizontal su-bsurface flow constructed wetland treatment system in an arid area. Water, 8, 51.
Al-Baldawi, I. A., Abdullah, S. R. S., Anuar, N., Suja, F. and Idris, M. (2013). Performance assessment of pilot horizontal sub-surface flow constructed wetlands for removal of diesel from wastewater by Scirpus grossu. Water Science and Technology, 68 (10), 2271-2278.
Alebel, B. W., Oliver, F., Eline, B. and Stephan, D. (2009). The economic value of improved wastewater irrigation: A contingent valuation study in Addis Ababa, Ethiopia. Journal of Agriculture and Resource Economics, 34(3), 428-449.
Alley, B. L., Willis, B., Rodgers, J. and Castle, J. W. (2013). Seasonal performance of a hybrid pilot-scale constructed wetland treatment system for simulated fresh oil field-produced water. Water Air Soil Pollut, 244, 1639.
Amaral, R., Ferreira, F., Galvao, A. and Matos, J. S. (2013). Constructed wetlands for combined sewer overflow treatment in a Mediterranean country, Portugal. Water Science and Technology, 67 (12), 2739 – 2746.
Anish, G., Ajay, K. K. C. and Bijay, T. (2012). Design approach for sub-surface flow constructed wetlands. Hydro Nepal, 10, 42-47.
Andualem, M., Seyoum, Leta. and Karoli, N. N. (2015). Wastewater treatment performance efficiency of constructed wetlands in African countries: A review. Water Science and Technology, 71 (1), 1-8.
Arias, C. A., Bubba, M. D. and Brix, H. (2001). Phosphorous removal by sands for use as a media in subsurface flow constructed reed beds. Wat. Res., 35 (5), 1159-1168.
Asaye, K. (2009). Evaluation of selected plant species for the treatment of tannery effluent in constructed wetland system: Unpublished.
Avila, C., Bayona, J. M., Martin, I., Salas, J. J. and Garcia, J. (2015). Emerging organic contaminat removal in a full-scale hybrid constructed wetland system for wastewater treatment and reuse. Ecological Engineering, 80, 108-116.
Avsar, Y., Tarabeah, H., Kimchie, S. and Ozturk, I. (2007). Rehabilitation by constructed wetlands of available wastewater treatment plant in Sakhnin. Ecological Engineering, 29, 27-32.
Azni, I., Abdul, G. L. A., Yung-Tse, H. and Lawrence, K. W. (2010). Wetlands for wastewater treatment. In: Lawrence, K. W., Joo-Hwa, T., Stephen, T. T. and Yung-Tse, H. (eds.) Handbook of Environmental Engineering: Environmental Bioengineering, Springer science+business media, LLC.
Babatunde, A. O., Zhoa, Y. Q., Doyle, R. J., Rackard, S. M., Kumar, J. L. G. and Hu, Y. S. (2011). Performance evaluation and prediction for a pilot two-stage on-site constructed wetland system employing dewatered alum sludge as main substrate. Bioresource Technology, 102, 5645-5652.
Baddam, R., Reddy, G. B., Raczkowski, C. and Cyrus, J. S. (2016). Activity of soil enzymes in constructed wetlands treated with swine wastewater. Ecological Engineering, 91, 24-30.
Bakhshoodeh, R., Alavi, N. Mohammadi, A. S. and Ghanavati, H. (2016). Removing heavy metals from Isfahan composting leachate by horizontal subsurface flow constructed wetland. Environ Sci Pollut Res, 23, 12384-12391.
Beharrell, M. (2004). Planting, selection and plant establishment in constructed wetlands in a tropical environment. In: Ming, H. W. (ed.) Constructed wetlands in Asia.
Belmont, M. A., Cantellano, E., Thompson, S., Williamson, M., Sanchez, A. and Metcalfe, C. D. (2004). Treatment of domestic wastewater in a pilot-scale natural treatment system in Central Mexico. Ecological Engineering, (23), 299 – 311.
Bialowiec, A., Albuquerque, A. & Randerson, P. F. (2014). The influence of evapotranspiration on vertical flow subsurface constructed wetland performance. Ecological Engineering, 67, 89-94.
Birhanu, G. F. (2007). Constructed wetland system for domestic wastewater treatment: A case study in Addis Ababa, Ethiopia; Unpublished.
Biswas, A. K. (2010). Water for a thirsty urban world. Brown Journal of World Affairs, 147-162.
Bixio, D., De hyder, B., Cikurel, H., Muston, M., Miska, V., Joksimovic, D., Schafer, A. I., Ravazzin, A., Aharoni, A., Savic, D. and Thoeye, C. ( 2005). Municipal wastewater reclamation: Where do we stand? An overview of treatment technology and management practice. Water Science and Technology: Water Supply, 5 (1), 77 – 85.
Borling, K. (2003). Phosphorous sorption, accumulation and leaching: effects of long-term inorganic fertilization of cultivated soils. Doctoral thesis. ISSN 1401-6249. ISBN 91-576-6466-8.
Brix, H. (1995). Use of constructed in water pollution control: historical development, present status, and future perspectives. Wat. Sci. Tech., 30 (8), 209-223.
Brix, H., Arias, C. A. and Bubba, M. (2001). Media selection for sustainable phosphorous removal in subsurface flow constructed wetlands. Water Science and Technology, 44 (11-12), 47-54.
Brix, H. (2004). Danish guidelines for small-scale constructed wetland system for onsite treatment of domestic sewage. Proceedings of the 9th international conference on wetland systems for water pollution control, Avignon, France, pp. 1-8.
Cakir, R., Gidirislioglu, A. and Cebi, U. (2015). A study on the effects of different hydraulic loading rates (HLR) on pollutant removal efficiency of subsurface horizontal-flow constructed wetlands used for treatment of domestic wastewaters. Journal of Environmental Management, 164, 121-128.
Canga, E., Santo, S. D., Pressl, A., Borin, M. and Langergraber, G. (2011). Comparison of nitrogen elimination rates of different constructed wetland designs. Water Science and Technology, 64 (5), 1122-1129.
Charles, P.G. and Ian, L.P. (2009). Wastewater treatment and biosolids reuse. In: Raina, M.M., Ian, L.P. & Charles, P.G. (eds.) Environmental microbiology. Elsevier ink, pp. 503-530.
Chazarenc, F., Merlin, G. and Gonthier, Y. (2004). Hydrodynamics of horizontal subsurface flow constructed wetlands. Ecological Engineering, 21, 165-173.
Chen, T. Y., Kao, C. M., Yeh, T. Y., Chien, H. Y. and Chao, A. C. (2006). Application of a constructed wetland for industrial wastewater treatment: A pilot-scale study. Chemosphere, 64, 497 – 502.
Choi, J., Maniquiz, M. C., Lee, B., Jeong, S. and Kim, L. H. (2012). Characteristics of contaminant and phosphorous existence types in sediment of a constructed wetland. Desalination and Water Treatment, 38, 342-348.
Chon, K. and Cho, J. (2015). Abundance and expression of denitrifying genes (narG, nirS, norB and nosZ) in sediments of wastewater stabilizing constructed wetlands. Environ. Eng. Res., 20 (1), 51-57.
Choudhary, A. K., Kumar, S. and Sharm, C. (2011). Constructed Wetlands: An option for pulp and paper mill wastewater treatment. Electronic Journal of Environmental, Agriculture and Food Chemistry, 10 (10), 3023-3037.
Corcoran, E., Nellemann, C., Baker, E., Bos, R., Osborn, D. and Savellie, H. (2010). Sick Water? The central role of wastewater management in sustainable development. Birkeland Trykkeri AS, Norway.
Cortes-Esquivel, J. A., Giacoman-Vallejos, G., Barcelo-Quintal, I. D., Mendez-Novelo, R. and Ponce-Caballero, M. C. (2012). Heavy metals removal from swine wastewater using constructed wetlands with horizontal subsurface flow. Journal of Environmental Protection, 3, 871-877.
Costa, J. F., Martins, W. L. P., Seidl, M. and Sperling, M. V. (2015). Role of vegetation (Typha Latifolia) on nutrient removal in a horizontal subsurface-flow constructed wetland treating UASB reactor-trickling filter reactor. Water Science and Technology, 71 (7), 1004-1010.
Deeptha, V. T., Sudarsan, J. S. and Baskar, G. (2015). Performance and cost evaluation of constructed wetland for domestic wastewater treatment. Journal of Environmental Biology, (36), 1071 – 1074.
De Paoli, A. C. and Sperling, M. V. (2013). Evaluation of clogging in planted and unplanted horizontal subsurface flow constructed wetlands: solids accumulation and hydraulic conductivity reduction. Water Science and Technology, 67 (6), 1345-1352.
Despland, L. M., Clark, M. W., Vancov, T. and Aragno, M. (2014). Nutrient removal and microbial communities’ development in a young unplanted constructed wetland using BauxsolTM pellets to treat wastewater. Science of the Total Environment, 484, 167-175.
Dimuro, J. L., Guertin, F. M., Helling, R. K., Perkins, J. L. and Romer, S. (2014). A financial and environmental analysis of constructed wetlands for industrial wastewater treatment. Journal of Industrial Ecology, 18 (5), 631 – 641.
Dong, C., Huang, Y. H., Wang, S. C. and Wang, X. H. (2016). Oxygen supply and wastewater treatment in subsurface-flow constructed wetland mesocosm: role of plant presence. Pol. J. Environ. Stud., 25 (2), 573-579.
Dordio, A. V. and Carvalho, A. J. P. (2013). Organic xenobiotics removal in constructed wetlands, with emphasis on the importance of the support matrix: review. Journal of Hazardous Materials, 252 (253), 272-292.
Duncan, M. (2003). Domestic wastewater treatment in developing countries. Earthscan, UK, Pp. 211.
Dzakpasu, M., Scholz, M., MCIWEM, McCharthy, V. and Jordan, S. (2015). Phosphorous retention and mass balance in an integrated constructed wetland treating constructed wetland. Water and Environment Journal, 29, 298-306.
Erakhrumen and Agbontanor, A. (2007). Phytoremediation: An environmentally sound technology for pollution prevention, control and remediation in developing countries. Educational research and review, 2 (7), 151-156.
Foladori, P., Bruni, L. and Tamburini, S. (2015). Bacterial viability and decay in water and soil of vertical subsurface flow constructed wetlands. Ecological Engineering, 82, 49-56.
Fonkou, T., Ivo, B. S., Lekeufack, M., Mekontso, T. F. and Amougou, A. (2011). Potential of Cyprus papyrus in yard-scale horizontal flow constructed wetlands for wastewater treatment in Cameroon. Universal Journal of Environmental Research and Technology, 1 (2), 160-168.
Fountoulakis, M. S., Terzakis, S., Chatzinotas, A., Brix, H., Kalogerkis, N. and Manios, T. (2009). Pilot-scale comparison of constructed wetlands operated under high hydraulic loading rates and attached biofilm reactors for domestic wastewater treatment. Science of the Total Environment, 407, 2996-3003.
Fu, G., Guo, Z., Zhang, J., Chen, Z. and Wong, M. H. (2015). Organic matter transplant improved purification performance of newly built constructed wetlands. Ecological Engineering, 83, 338-342.
Gaboutloeloe, G. K., Chen, S., Barber, M. E. and Stockle, C. O. (2009). Combinations of horizontal and vertical flow constructed wetlands to improve Nitrogen removal. Water, Air, Soil pollun. 9, 279-286.
Garcia, J., Aguirre, P., Mujeriego, R., Huang, Y., Orgiz, L. and Bayona, J. M. (2004). Initial contaminant removal performance factors in horizontal flow reed beds used for treating urban wastewater. Water Research, 38, 1669-1678.
Garcia, J., Aguirre, P., Barragan, J., Mujeriego, R., Matamoros, V. and Bayona, J. M. (2005). Effect of key design parameters on the efficiency of horizontal subsurface flow constructed wetlands. Ecological Engineering, 25, 405-418.
Garcia, M., Soto, F., Gonzalez, J. M. and Becares, E. (2008). A comparison of bacterial removal efficiencies in constructed wetlands and Algae-based systems. Ecological Engineering, 32, 238-243.
Georgia Environmental Protection Division (2010). Guidelines for constructed wetlands for municipal wastewater facilities. (unpublished).
Getahun, W. and Adinew, A. (1999). Wastewater Management in Addis Ababa: Integrated Development for Water Supply and Sanitation.
Gikas, G. D. and Tsihrintzis, V. A. (2010). On-site treatment of domestic wastewater using a small-scale horizontal subsurface flow constructed wetland. Water Science and Technology, 62 (3), 603-614.
Gill, L. W., Ring, P., Higgins, N. M. P. and Johnston, P. M. (2014). Accumulation of heavy metals in a constructed wetland treating road runoff. Ecological Engineering, 70, 133-139.
Global Water Partnership (2009). Managing the other side of the water cycle: Making wastewater an asset. Elanders in Molnlycke, Sweden.
Gokalp, Z., Karaman, S., Tas, I. and Kirnak, H. (2016). Constructed wetland technology to prevent water resources pollution. Current Trends in Natural Sciences, 5 (9), 125-132.
Haberl, R., Perfler, R. and Mayer, H. (1995). Constructed wetlands in Europe. Water Science Technology, 32 (3), 305-315.
Haberl, R., Grego, S., Langergraber, G., Kadlec, R. H., Cicalini, A. R., Dias, S. M., Novais, J. M., Aubert, S., Gerth, A., Thomas, H. and Habner, A. (2003). Constructed wetlands for the treatment of organic pollutants. Journal of soil and sediments, 3(2), 109-124.
Han, S. S., Chen, Z., Zhou, F. Y. and Lu, X. Q. (2014). Assessment of suspended solid removal in a surface flow constructed wetland using a three-dimensional numerical model. Water Resource Management, 28, 3111-3125.
Hench, K. R., Bissonnette, G. K., Sexstone, A. J., Coleman, J. G., Garbutt, K. and Skousen, J. G. (2003). Fate of physical, chemical and microbial contaminants in domestic wastewater following treatment by small constructed wetlands. Water Research, 37, 921-927.
Hsueh, M. L., Yang, L., Hsieh, L. Y. and Lin, H. J. (2014). Nitrogen removal along the treatment cells of a free-water surface constructed wetland in subtropical Taiwan. Ecological Engineering, 73, 579-587.
Hu, Y. S., Kumar, J. L. G., Akintunde, A. O., Zhao, X. H. and Zhao, Y. Q. (2011). Effects of livestock wastewater variety and disinfectants on the performance of constructed wetlands in organic matters and nitrogen removal. Environ Sci Pollut Res, 18, 1414-1421.
Hussein, I. A. and Ahmed, D. (2012). Constructed wetlands for urban wastewater treatment in Egypt: Sanitation sustainable practice, (12), 27-32.
Jern, N. G. W. (2006). Industrial wastewater treatment. Imperial College Press, pp. 2-3.
Jimenez, B. (2006). Irrigation in developing countries using wastewater. International Review for International Strategies, 6 (2), 229-250.
Johannesson, K., Andersson, J. L. and Tonderski, K. S. (2011). Efficiency of a constructed wetland for retention of sediment-associated phosphorous. Hydrobiologia, 674, 179-190.
John, R. K. (2009). Wastewater treatment and disposal. In: Nelson, L.N., Franklin, J. A., Patrick S. ; Joseph, A. S. (eds.) Environmental engineering: Water, wastewater, soil and groundwater treatment and remediation. sixth edition, John Wiley and Sons, pp. 283-370.
Johnson, D. B. and Hallberg, K. B. (2005). Acid mine drainage remediation options: a review. Science of the Total Environment, 338, 3-14.
Kadlec R. H. and Knight R. L. (1996). Treatment wetlands. CRC Press, Boca Raton, New York, London,Tokyo.
Kadlec, R. H. and Wallance, S. D. (2009). Treatment wetlands (sec. ed.). CRC press, USA.
Karathanasis, A. D., Potter, C. L. and Coyne, M. S. (2003). Vegetation effects on fecal bacteria, BOD, and suspended solids removal in constructed wetlands treating domestic wastewater. Ecological Engineering, 20, 157 – 169.
Katayon, S., Fiona, Z., Noor, M. J. M., Abdulhalim, G. and Ahmad, J. (2008). Treatment of mild domestic wastewater using subsurface constructed wetlands in Malaysia. International Journal of Environmental Studies, 65(1), 87 – 102.
Kayombo, S., Mbwette, T.S. A., Katima, J.H. Y., Ladegaard, N. ; Jørgensen, S.E. (2003). Waste stabilization ponds and constructed wetlands design manual. WSP ; CW Research Project, College of Engineering and Technology, University of Dar es Salaam.
Kelvin, K. and Tole, M. (2011). The efficacy of a tropical constructed wetland for treating wastewater during the dry season: The Kenyan experience. Water Air Soil Pollut, 215, 137-143.
Kenatu, A.W. (2011). Evaluation of the performance of constructed wetland system for the treatment of brewery wastewater. Unpublished.
Kent, D. M. (2001). Applied wetlands science and technology (second editon). Boca Ration, CRC press LLC, Florida.
Khisa, K. and Mwakio, T. (2010). The efficiency of the tropical constructed wetland for treating wastewater during the dry season: the Kenyan experience. Water Air Soil Pollution, 215, 137-145.
Kivaisi, A. K. (2001). The potential for constructed wetlands for wastewater treatment and reuse in developing countries: a review. Ecological Engineering, 16, 545-560.
Knox, A. S., Nelson, E. A., Halverson, N. V. and Gladden, J. B. (2010). Long-term performance of a constructed wetland for metal removal. Soil and Sediment Contamination, 19, 667-685.
Konnerup, D., Koottatep, T. and Brix, H. (2009). Treatment of domestic wastewater in tropical, subsurface flow constructed wetlands planted with Canna and Heliconia. Ecological Engineering, 35, 248-257.
Korkusuz, A. E., Beklioglu, M. and Demirer, G. N. (2004). Treatment efficiency of the vertical flow pilot-scale constructed wetlands for domestic wastewater treatment. Turkish Journal of Engineering and Environmental Science, 28, 333-344.
Kurniadie, D. (2011). Wastewater treatment using vertical subsurface flow constructed wetland in Indonesia. American Journal of Environmental Sciences, 7 (1): 15-19.
Kyambadde, J., Kansiime, F., Gumaelius, L. and Dalhammar, G. (2004). A comparative study of Cyprus papyrus and Miscanthidium Violaceum-based constructed wetland for wastewater treatment in a tropical climate. Water Research, (38) 475-485.
Kyambadde, J, Kansiime, F. and Dalhammar, G. (2005). Nitrogen and phosphorous in substrate-free pilot constructed wetlands with horizontal surface flow in Uganda. Water, Air and Soil Pollution, 165, 37-59.
Laffat, J., Ouazzani, N. and Mandi, L. (2015). The evaluation of potential purification of a horizontal subsurface flow constructed wetland treating greywater in semi-arid environment. Process Safety and Environmental Protection, 95, 86-92.
Lana, L. C. O., Moraes, D.C., Sperling, M. V., Morato, M. L. N., Vasconcellos, G. R., Paraense, M. O. and Moreira, T. P. A. (2013). Performance of a single stage vertical flow constructed wetland system treating raw domestic sewage in Brazil. Water Science and Technology, 68 (7), 1599-1606.
Landry, G. M., Maranger, R. and Brisson, J. (2009). Effect of artificial aeration and macrophyte species on nitrogen cycling and gas flux in constructed wetlands. Ecological Engineering, 35, 221-229.
Langergraber, G. (2013). Are constructed treatment wetlands sustainable sanitation solutions. Water Science and Technology, 67 (10), 2133-2140.
Langergraber, G., Pressl, A. and Haberl, R. (2014). Experiences from the full-scale implementation of a new two-stage vertical flow constructed wetland design. Water Science and Technology, 69 (2), 335-342.
Lavrnic, S. and Mancini, M. L. (2016). Can constructed wetlands treat wastewater for use in agriculture? Review of guidelines and examples in South Europe. Water Science and Technology, (73.11).
Lee, B. H. and Scholz, M. (2006). What is the role of Phragmites australis in experimental constructed wetland filters treating urban runoff. Ecological Engineering, 29, 87-95.
Lee, S. Y., Manquiz, M. C., Choi, J. Y., Kang, J. H. and Kim, L. H. (2012). Phosphorous mass balance in a surface flow constructed wetland receiving piggery effluent. Water Science and Technology, 66 (4), 712-718.
Lee, S.Y., Maniquiz, M. C., Choi, J. Y., Jeong, S. M. and Kim, L. H. (2013). Seasonal nutrient uptake of plant biomass in a constructed wetland treating piggery wastewater effluent. Water Science and Technology, 67 (6), 1317-1323.
Li, L., Li, Y., Biswas, D. K., Nian, Y. and Jiang, G. (2007). Potential of constructed wetlands in treating the eutrophic water: evidence from Taihu lake of China. Bioresource Technology, 99, 1656-1663.
Li, D. W., Zhang, W. S., Yin, W. and Lei, A. L. (2010). Combined pond-wetland systems for treatment of urban surface runoff and lake water. Environmental Engineering Science, 27 (12), 1027-1034.
Liu, D., Ge, Y., Chang, J., Peng, C., Gu, B., Chan, G. Y. S. and Wu, X. (2008). Constructed Wetlands in China. Recent development and future challenges. Front Ecol environ, (6), 7-14.
Lu, S., Zhang, X., Wang, J. and Pei, L. (2016). Impacts of different media on constructed wetlands for rural household sewage treatment. Journal of Cleaner Production, 127, 325-330.
Maderra-Parra, C. A., Pena, M. R., Pena, E. J. and Lens, P. N. L. (2015). Cr (VI) and COD removal from landfill leachate by polyculture constructed wetland at a pilot scale. Environ Sci Pollut Res, 22, 12804-12815.
Mahenge, A. S. (2014). Performance of horizontal surface flow constructed mangroves wetland in faecal coliform removal. Journal of Applied Phytotechnology in Environmental Sanitation, 3 (2), 35-44.
Mangangka, I. R., Liu, A., Egodawatta, P. and Goonetilleke, A. (2015). Sectional analysis of stormwater treatment performance of a constructed wetland. Ecological Engineering, 77, 172-179.
Manahan, S. E. (2000). Environmental chemistry. Boca Raton, CRC press LLC.
Maniquiz, M. C., Lee, S. Y., Choi, J. Y., Jeong, S. M. and Kim, L. H. (2012). Treatment performance of a constructed wetland during storm and non-storm events in Korea. Water Science and Technology, 65 (1), 119-126.
Mara, D., Drangert, J. O. Anh, N. V., Tonderski, A., Gulyas, H. and Tonderski, K. (2007). Selection of sustainable sanitation arrangements. Water Policy, (9), 305 – 318.
Mairi, J. P., Lyimo, T. J. and Njau, K. N. (2012). Performance of subsurface flow constructed wetland for domestic wastewater treatment. Tanz. J. Sci., 38(2), 66-79.
Mashauri, D.A., Mulungu, D. M. and Abdulhussein. B. S. (2000). Constructed wetland at the university of Dar Es Salam. Water Resource, 34(4), 1135-1144.
Masi, F. and Martinuzzi, N. (2007). Constructed wetlands for the Mediterranean countries: hybrid systems for water reuse and sustainable sanitation. Desalination, 215, 44-55.
Masi, F., Caffaz, S. and Ghrabi, A. (2013). Multi-stage constructed wetland systems for municipal wastewater treatment. Water Science and Technology, 67 (7), 1590-1598.
Massoud, M. A., Tarhini, A. Nasr, J. A. (2009). Decentralized approaches to wastewater treatment and management: Journal of Environmental Management, 90, 652 – 659.
Mekala, G. D., Davidson, B., Samad, M. and Boland, A. M. (2008). A framework for efficient wastewater treatment and recycling systems. Colombo: IWMI and IF technologies.
Menon, R. and Holland, M. M. (2013). Phosphorous retention in constructed wetlands vegetated with Juncus effusus, Carex lurida, and Dichanthelium acuminatum var. Acuminatum. Water Air Soil Pollut, 224, 1602.
Merchand, L., Mench, M., Jacob, D. L. and Otte, M. L. (2010). Metal and metalloid removal in constructed wetlands, with emphasis on the importance of plants and standardized measurements: A review. Environmental Pollution, 158, 3447-3461.
Merlin, G., Pajean, J. L. and Lisollo, T. (2002). Performance of constructed wetlands for municipal wastewater treatment in rural mountainous area. Hydrobiologia, 469, 87-98.
Merlin, G. and Lisollo, T. (2010). Energy and emergy analysis to evaluate sustainability of small wastewater treatment plants: application to a constructed wetland and a sequencing batch reactor. J. Water Resource and Protection, 2, 997-1009.
Mesquita, M. C., Albuquerque, A., Amaral, L. and Nogueira, R. (2012). Effect of vegetation on the performance of horizontal subsurface flow constructed wetlands with lightweight expanded clay aggregates. Int. J. Environ. Sci. Technol., 10, 433-442.
Meyer, D., Molle, P., Esser, D., Troesch, S., Masi, F. and Dittmer, U. (2013). Constructed wetlands for combined sewer overflow treatment-comparison of German, French and Italian approaches. Water, 5, 1-12.
Mietto, A., Politeo, M., Breschigliaro, S. and Borin, M. (2015). Temperature influence on nitrogen removal in a hybrid constructed wetland system in Northern Italy. Ecological Engineering, 75, 291-302.
Millennium Ecosystem Assessment (2005). Ecosystems and human wellbeing: biodiversity synthesis. World Resources Institute, Washington DC, USA.
Milne, N. A. and Gray, S. R. (2013). General application for the treatment of effluent and reuse of wastewater. In: Tamime, A. Y. (eds) Membrane processing: Dairy and beverage applications. Blackwell Publishing Ltd. Pp. 107.
Mohammadpour, R., Shaharuddin, S., Chang, C. K., Zakari, N. K. and Ghani, A. A. (2014). Spatial pattern analysis for water quality in free-surface constructed wetland. Water Science and Technology, 70 (7), 1161-1167.
Mojiri, A., Ziyang, L., Tajuddin, R. M., Farraji, H. and Alifar, N. (2016). Co-treatment of landfill leachate and municipal wastewater using the ZELIAC/zeolite constructed wetland system. Journal of Environmental Management, 166, 124-130.
Molle, P., Latune, R. L., Riegel, C., Lacombe, G., Esser, D. and Mangeot, L. (2015). French vertical-flow constructed wetland design: adaptations for tropical climates. Water Science and Technology, 71 (10), 1516-1523.
Muresan, M. V. (2012). Optimal treatment technology for rural wastewater: constructed wetlands. Scientific Bulletin Series, 26 (2), 101-109.
Murray, A. and Ray, I. (2010). Commentary: Back-end users: the unrecognized stakeholders in demand-driven sanitation. Journal of planning education and research, 20 (10), 1-9.
Mustafa, A., Scholz, M., Harrington, R. and Carroll, P. (2009). Long-term performance of a representative integrated constructed wetland treating farm yard runoff. Ecological Engineering, 35, 779-790.
Mustafa, A. (2013). Constructed wetland for wastewater treatment and use: A case study of developing country. International Journal of Environmental Science and Development, 4 (1), 20-24.
Naz, M., Uyanik, S., Yesilnacar, M. I. and Sahinkaya, E. (2009). Side-by-side comparison of horizontal subsurface flow and free water surface flow constructed wetlands and artificial neutral network (ANN) modelling approach. Ecological Engineering, 35, 1255-1263.
Nelson, L. N., Franklin, J. A., Patrick, S. and Salvato, J. A. (2009). Environmental engineering: Water, wastewater, soil and groundwater treatment and remediation. (6th ed.), John Wiley ; Sons, USA.
Nhapi, I. and Gijzen, H. G. (2005). A 3-Step strategic approach to sustainable wastewater management. Water SA, 31 (1), 133 – 140.
Nicholas, P. C. (2002). Handbook of water and wastewater treatment technologies. Butterworth-Heinemann publication, USA.
Odinga, C., Otieno, F. and Adeyemo, J. (2011). Investigating the effectiveness of aquatic plants (Echinoclao L and Cyprus L) in removing nutrients from wastewater: the case of Chemelil constructed wetland-Kenya. International Journal of the Physical Sciences, 6 (16), 3960-3970.
Paing, J., Guilbert, A., Gagnon, V. and Chazarec, F. (2015). Effect of climate, wastewater composition, loading rates, system age, and design on performances of French vertical flow constructed wetlands: A survey based on 169 full scale systems. Ecological Engineering, 80, 46-52.
Paing, J., Serdobbel, V., Welschbillig, M., Calvez, M., Gagnon, V. and Chazarenc, F. (2015). Treatment of high organic content wastewater from food-processing industry with the French vertical flow constructed wetland system. Water Science and Technology, 72 (1), 70-76.
Pandey, M.K., Jenssen, P.D., Krogstad, T. and Jonasson, S. (2013). Comparison of vertical and horizontal flow planted and unplanted subsurface flow wetlands treating municipal wastewater. Water Science and Technology, 68 (1), 117-123.
Pietro, K. C. and Ivanoff, D. (2015). Comparison of long-term phosphorous removal performance of two large-scale constructed wetlands in south Florida, USA. Ecological Engineering, 79, 143-157.
Polprasert, C. (2004). Constructed wetlands for wastewater treatment: principles and practice. In: Ming, H. W. (ed.) Constructed wetlands in Asia.
Prochaska, C. A., Zouboulis, A. I. and Eskridge, K. M. (2007). Performance of pilot-scale vertical-flow constructed wetland as affected by season, substrate, hydraulic load and frequency of application of simulated urban sewage. Ecological Engineering, 31, 57-66.
Prashant, Billore, S. K., Sharma, J. K., Singh, N. and Ram, H. (2013). Treatment of wastewater and restoration of aquatic systems through an echo-technology based constructed treatment wetlands- a successful experience in central India. Water Science and Technology, 68 (7), 1566-1574.
Qasaimeh, A. Alsharie, H. and Masoud, T. (2015). A review on constructed wetlands components and heavy metal removal from wastewater. Journal of Environmental Protection, 6, 710-718.
Rai, U. N., Upadhyay, A. K., Singh, N. K., Dwivedi, S. and Tripathi, R. D. (2015). Seasonal applicability of horizontal sub-surface flow constructed wetland for trace elements and nutrient removal from urban wastes to conserve Ganga river water quality at Haridwar, India. Ecological Engineering, 81, 115-122.
Ramprasad, C. and Philip, L. (2016). Surfactants and personal care products removal in pilot-scale horizontal and vertical flow constructed wetlands while treating grey water. Chemical Engineering Journal, 284, 458-468.
Ramsar Convention Secretariat (2013). The ramsar convention manual: A guide to the convention on wetlands (6th ed.). Gland, Switzerland: Ramsar convention secretariat.
Redder, A., Durr, M., Daeschlein, G., Baeder- Bederski, O., Koch, C., Muller, R., Exner, M. and Borneff-Lipp, M. (2009). Constructed wetlands – are they safe in reducing protozoan parasites? International journal of hygiene environment health, 213, 72-77.
Reddy, K. R., Diaz, O. A., Scinto, L. J. and Agami, M. (1995). Phosphorous dynamics in selected wetlands and streams of the lake Okeechobee basin. Ecological Engineering, 5, 183-207.
Robert, A. C. (2004). Standard handbook of environmental engineering. (2nd ed.) McGraw-Hill.
Romain, V., Sylvie, D. and David, B. (2015). Water residence time and pesticide removal in pilot-scale wetlands. Ecological Engineering, 85, 76-84.
Rousseau, D. P. L., Vanrolleghem, P. A. and Pauw, N. D. (2004). Constructed wetlands in fladers: a performance analysis. Ecological Engineering, 23, 151-163.
Rousseau, D. P. L., Lesage, E., Story, A., Vanrolleghem, P. A. and Pauw, N. D. (2008). Constructed wetlands for water reclamation. Desalination, 218, 181-189.
Rozema, E. R., Rozema, L. R. and Zheng, Y. (2016). A vertical flow constructed wetland for the treatment of winery process water and domestic sewage in Ontario, Canada: six years of performance data. Ecological Engineering, 86, 262-268.
Rozema, E. R., VanderZaag, A. C., Wood, J. D., Drizo, A., Zheng, Y., Madani, A. and Gordon, R. J. (2016). Constructed wetlands for agricultural wastewater treatment in northeastern north America. Water, 8, 173.
Russo, R. E. (2008). Wetlands: Ecology, conservation and restoration. New York: Nova science publishers.
Saeed, T., Afrin, R., Al Muyed, A. and Sun, G. (2012). Treatment of tannery wastewater in a pilot-scale hybrid constructed wetland system in Bangladesh. Chemosphere, 88, 1065-1073.
Saravanan, V. S., Peter, P. M. ; Janos, J. B. (2011). Global change, wastewater and health in fast growing economies. Current Opinion in Environmental Sustainability, 3: 461-466.
Sarmento, A. P., Borges, A. C. and Matos, A. T. (2012). Evaluation of vertical-flow constructed wetlands for Swine wastewater treatment. Water Air Soil Pollut, 223, 1065-1071.
Scheierling, S., Bartone, C. R., Mara, D. D. and Drechsel, P. (2011). Towards an agenda for improving wastewater use in agriculture. Water International, 36(4), 420 – 440.
Schertenleib, R. (2005). From conventional to advanced environmental sanitation. Water Science and Technology, 51 (10), 7 – 14.
Scholz, M. (2006). Wetland systems to control urban runoff. Edinburgh.
Scholz, M., Harrington, R., Carroll, P. and Mustafa, A. (2010). Monitoring of nutrient removal within integrated constructed wetlands (ICW). Desalination, 250, 356-360.
Schulz, R. and Peall, S.K.C. (2001). Effectiveness of a constructed wetland for retention of nonpoint source pesticide pollution in the Lourens river catchment, South Africa. Environ Sci Technol, 35, 422-426.
Schulz, R., Hahn, C., Bennett, E. R., Dabrowski, J. M., Thiere, G. and Peall, S. K. C. (2003). Fate and effects of Azinphos-Methyl in a flow-through wetland in South Africa. Environ Sci Technol, 37, 2139-2144.
Seadira, T., Baloyi, J., Raphulu, M., Moutloali, R. and Ochieng, A. (2014). Acid mine drainage treatment using constructed wetland. International conference on chemical, integrated waste management and environmental engineering, Johannesburg.
Sehar, S., Sumera, Naeem, S.,Perveen, I., Ali, N. and Ahmed, S. (2015). A comparative study of macrophytes influence on wastewater treatment through subsurface flow hybrid constructed wetland. Ecological Engineering, 81, 62-69.
Sharma, G. and Brighu, U. (2016). Selection of suitable plant species in semi arid climatic conditions for quality improvement of secondary treated effluent by using vertical constructed wetland. Nature Environment and Pollution Technology, 15 (1), 325-329.
Shutes, R. B. E. (2001). Artificial wetlands and water quality improvement. Environment International, 26, 441-447.
Si, W., Zhang, W., Lv, Y., Yang, F., Liu, J. and Zhang, Y. (2014). Heavy metal removal in a constructed wetland and benefits for the development of the Toad Bufo Raddei. Pol. J. Environ. Stud., 23 (6), 2207-2215.
Siracusa, G. and Rosa, A. D. L. (2006). Design of a constructed wetland for wastewater treatment in a Sicilian town and environmental evaluation using the emergy analysis. Ecological Modelling, 197, 490497.
Siti, H. C. R., Mohd, F. M. D., Mohd, B. M. Y. and Shreeshivadasan, C. (2011). Overview of subsurface constructed wetlands application in tropical climates. Universal Journal of Environmental Research and Technology, 1 (2), 103-114.
Sivakumer, D., Shankar, D., Vijaya, P. A. J. and Valarmathi, M. (2013). Constructed wetland treatment of textile industry wastewater using aquatic macrophytes. International Journal of Environmental Sciences, 3 (4), 1223-1232.
Sleytr, K., Tietz, A., Langergraber, G. and Haberl, R. (2007). Investigation of bacterial removal during the filtration process in constructed wetlands. Science of the Total Environment, 380, 173 180.
Song, Z., Zheng, Z., Li, J., Sun, X., Han, X., Wang, W. and Xu, M. (2006). Seasonal and annual performance of a full-scale constructed wetland system for sewage treatment in China. Ecological Engineering, 26, 272-282.
Stottmeister, U., Wiebner, A., Kuschk, P., Kappelmeyer, U., Kastner, M., Bederski, O., Muller, R. A. and Moormann, H. (2003). Effects of plants and microorganisms in constructed wetlands for wastewater treatment. Biotechnology Advances, 22, 93-117.
Su, T. M., Yang, S. C., Shih, S. S. and Lee, H. Y. (2009). Optimal design for hydraulic efficiency performance of free-water-surface constructed wetlands. Ecological Engineering, 35, 1200-1207.
Sudarsan, J. S., Deeptha, V. T., Maurya, D., Goel, M., Kumar, K. R. and Das, A. (2015). Study on treatment of electroplating wastewater using constructed wetland. Nature Environment and Pollution Technology, 14 (1), 95-100.
Tadesse, A. (2010). Evaluation of selected plant species for the removal efficiency of Chromium and organic pollutants from tannery wastewater in constructed wetlands at Modjo tannery. Unpublished.
Tao, W., Bays, J. S., Meyer, D., Smardon, R. C. and Levy, Z. F. (2014). Constructed wetlands for treatment of combined sewer overflow in the US: a review of design challenges and application status. Water, 6, 3362-3385.
Thomas, A. D. And William, F. D. (2001). Wetlands for water treatment. In: Donald, M. K. (ed.) Applied Wetlands Science and Technology. Boca Raton: CRC press LLC.
Thullen, J. S., Sartoris, J. J. and Walson, W. E. (2002). Effects of vegetation management in constructed wetland treatment cells on water quality and mosquito production. Ecological Engineering, 18, 441-457.
Tietz, A., Langergraber, G., Watzinger, A., Haberl, R. and Kirschner, A. K. T. (2008). Bacterial carbon utilization in vertical subsurface flow constructed wetlands. Water Research, 42, 1622-1634.
Toscano, A., Marzo, A., Milani, M., Cirelli, G. L. and Barbagallo, S. (2015). Comparison of removal efficiencies in Mediterranean pilot-constructed wetlands vegetated with different plant species. Ecological Engineering, 75, 155-160.
Trivedy, R. K. and Siddharth, K. (2010). Low cost wastewater treatment technologies. New Delhi: ABD publisher.
Truu, J., Nurk, K., Juhanson, J. and Mander, U. (2005). Variation of microbiological parameters within planted soil filter for domestic wastewater treatment. Journal of Environmental Science and Health, 40, 1191-1200.
Tuncseper, B. (2009). Nitrogen removal in a combined vertical and horizontal sub-surface flow constructed wetland system. Desalination, 247, 466-475.
Tuncsiper, B., Ayaz, S. C. and Akca, L. (2012). Coliform bacteria removal from septic wastewater in a pilot-scale combined constructed wetland system. Environmental Engineering and management Journal, 11 (10), 1873-1879.
Tuncsiper, B., Drizo, A. and Twohig, E. (2015). Constructed wetlands as a potential management practice for cold climate dairy effluent treatment-VT, USA. Catena, 135, 184-192.
United Nations (2006). Global atlas of excreta, wastewater, sludge and biosolids management: moving forward the sustainable and welcome uses of a global resource.
United Nations Human Settlements Programme (2008). Constructed wetlands manual. Nairobi.
United States Environmental Protection Agency (1993). Constructed wetlands for wastewater treatment and wildlife habitat.
United States Environmental Protection Agency (2000). Manual; Constructed wetlands treatment of municipal wastewaters. Office of Research and Development, Ohio.
Upadhyay, A.K., Bankoti, N. S. and Rai, U. N. (2016). Studies on sustainability of simulated constructed wetland system for treatment of urban waste: design and operation. Journal of Environmental Management, 169, 285-292.
Vallejos, G. G., Caballero, C. P. and Champagne, P. (2015). Pathogen removal from domestic and swine wastewater by experimental constructed wetlands. Water Science and Technology, 71 (8), 1263-1270.
Valsero, M. H., Cardona, R. S. and Becares, E. (2012). Comparison of interannual removal variation of various constructed wetland types. Science of the Total Environment, 430, 174-183.
Van, V. (2010). Constructed wetlands for wastewater treatment. Water, 2, 530 – 549.
Varga, D., Ruiz, I. and Soto, M. (2013). Winery wastewater treatment in subsurface constructed wetlands with different bed depths. Water Air Soil Pollut, 224, 1485.
Vergeles, Y., Vystavna, Y., Ishchenko, A., Rybalka, I., Marchand, L. and Stolberg, F. (2015). Assessment of treatment efficiency of constructed wetlands in East Ukraine. Ecological Engineering, 83, 159- 168.
Vymazal, J. (2005). Horizontal subsurface-flow and hybrid constructed wetlands systems for wastewater treatment: a review. Ecological Engineering, 25, 478-490.
Vymazal, J. (2007). Removal of nutrients in various types of constructed wetlands. Science of the Total Environment, 380, 48-65.
Vymazal, J. and Lenk, K. (2008). Wastewater treatment in constructed wetlands with horizontal sub-surface flow. Springer Science and Business Media, Czech Republic.
Vymazal, J. (2009). The use of constructed wetlands with horizontal sub-surface flow for various types of wastewater. Ecological Engineering, 35, 1-17.
Vymazal, J. and Kropfelova, L. (2009). Removal of organics in constructed wetlands with horizontal sub-surface flow: A review of the field experience. Science of the Total Environment, 407, 3911-3922.
Vymazal, J. (2010). Constructed wetlands for wastewater treatment. Water, 2, 530-549.
Vymazal, J. (2014). Constructed wetlands for treatment of Industrial wastewater: A review. Ecological Engineering, 73, 724-751.
Vymazal, J. (2014). Long-term treatment efficiency of a horizontal subsurface flow constructed wetlands at Jimlikov Czech republic. Environmental Engineering and Management Journal, 13 (1), 73-80.
Vymazal, J. and Brezinova, T. (2015). The use of constructed wetlands for removal of pesticides from agricultural runoff and drainage: A review. Environmental International, 75, 11-20.
Vymazal, J. and Kropfelova, L. (2015). Multistage hybrid constructed wetland for enhanced removal of nitrogen. Ecological Engineering, 84, 202-208.
Wallace, S. D. (2008). Constructed wetlands-design approaches. University curriculum development for decentralized wastewater treatment (unpublished).
Wang, L. and Li, T. (2014). Vegetation effects on anammox spatial distribution and nitrogen removal in constructed wetlands treated with domestic sewage. Water Science and Technology, 70 (8), 1370-1375.
Wang, Y., Wang, J., Zhao, X., Song, X. and Gong, J. (2016). The inhibition and adaptability of four wetland plant species to high concentration of ammonia wastewater and nitrogen removal efficiency in constructed wetlands. Bioresource Technology, 202, 198-205.
World Health Organization/United Nations Environment Programme (1997). Wastewater as a resource.
Wu, H., Fan, J., Zhang, J., Ngo, H. H., Guo, W., Liang, S., Hu, Z. and Liu, H. (2015). Strategies and techniques to enhance constructed wetland performance for sustainable wastewater treatment. Environ Sci Pollun Res, 22: 14637 – 14650.
Wu, H., Zhang, J., Ngo, H. H., Guo, W., Hu, Z., Liang, S., Fan, J. and Liu, H. (2015). A review on the sustainability of constructed wetlands for wastewater treatment: design and operation. Bioresource Technology, 175, 594-601.
Wu, H., Zhang, J., Wei, R., Liang, S., Li, C. and Cie, H. (2013). Nitrogen transformations and balance in constructed wetlands for slightly polluted river water treatment using different macrophytes. Environ Sci Pollut Res, 20,443-451.
Wu, S., Kuschk, P., Brix, H., Vymazal, J. and Dong, R. (2014). Development of constructed wetlands in performance intensifications for wastewater treatment: A nitrogen and organic matter targeted review. Water Research, 57, 40-55.
Wu, S., Wallace, S., Brix, H., Kuschk, P., Kirui, W. K., Masi, F. and Dong, R. (2015). Treatment of industrial effluents in constructed wetlands: Challenges, operational strategies and overall performance. Environmental Pollution, 201, 107-120.
Wu, S., Carvalho, P. N., Muller, J. A., Manoj, V. R. and Dong, R. (2016). Sanitation in constructed wetlands: A review on the removal of human pathogens and faecal indicators. Science of the Total Environment, 541, 8 -22.
Xie, X. L., He, F., Xu, D., Dong, J. K., Cheng, S. P. and Wu, Z. B. (2011). Application of large-scale integrated vertical-flow constructed wetland in Beijing Olympic forest park: design, operation and performance. Water and Environment Journal, 26, 100-107.
Yang, Q., Chen, Z. H., Zhao, J. G. and Gu, B. H. (2007). Contaminant removal of domestic wastewater by constructed wetlands: effects of plant species. J. Integr. Plant Biol. 49 (4), 437-446.
Yang, X., Zhang, X., Wang, J., Zhao, G. and Wang, B. (2014). Performance of the subsurface flow constructed wetlands for pretreatment of slightly polluted source water. Ecotoxicology, 23, 699-706.
Ye, F. and Li, Y. (2009). Enhancement of nitrogen removal in towery hybrid constructed wetland to treat domestic wastewater for small rural communities. Ecological Engineering, 35, 1043-1050.
Ye, J., Wang, L., Li, D., Han, W. and Ye, C. (2012). Vertical oxygen distribution trend and oxygen source analysis for vertical-flow constructed wetlands treating domestic wastewater. Ecological Engineering, 41, 8-12.
Zapater-Pereyra, M., Dien, F. V., Bruggen, J. J. A. and Lens, P. N. L. (2013). Material selection for a constructed wetroof receiving pre-treated high strength domestic wastewater. Water science and technology, 68 (10), 2264-2271.
Zhang, D., Gersberg, R. M. and Keat, T. S. (2009). Constructed wetlands in China. Ecological engineering, 35, 1367 -1378.
Zhang, D. Q., Tan, S. K., Gersberg, R. M., Zhu, J., Sadreddini, S. and Li, Y. (2012). Nutrient removal in tropical subsurface flow constructed wetlands under batch and continuous flow conditions. Journal of Environmental Management, 96, 1-6.
Zhang, D. Q., Jinadasa, K. B. S. N., Gersberg, R. M., Liu, Y., Tan, S. K. and Wun, J. N. (2015). Application of constructed wetlands for wastewater treatment in tropical and subtropical regions (2000-2013). Journal of Environmental Sciences, 30, 30-46.
Zhang, X., Inoue, T., Kato, K., Harada, J., Izumoto, H., Wu, D., Sakuragi, H., Letsugu, H. and Sugawara, Y. (2016). Performance of hybrid subsurface constructed wetland system for piggery wastewater treatment. Water Science and Technology, 73 (1), 13-20.
Zheng, Y., Wang, X. C., Dzakpasu, M., Zhao, Y., Ngo, H. H., Guo, W., Ge, Y. and Xiong, J. (2016). Effects of interspecific competition on the growth of macrophytes and nutrient removal in constructed wetlands: A comparative assessment of free water surface and horizontal subsurface flow systems. Bioresource Technology, 207, 134-141.
Zhou, J. B., Jiang, M. M., Chen, B. and Chen, G. Q. (2007). Energy evaluations for constructed wetland and conventional wastewater treatments. Communications in Nonlinear Science and Numerical Simulation, 14, 1781-1789.
Zidan, A. R. A., El-Gamal, M. M., Rashid, A. A. and El-Hady Eid, M. A. A. (2015). Wastewater treatment in horizontal subsurface flow constructed wetlands using different media (setup stage). Water Science, 29, 26-35.
Zou, J., Guo, X., Han, Y., Liu, J. and Liang, H. (2012). Study of a noble vertical flow constructed wetland system with drop aeration for rural wastewater treatment. Water air soil pollut, 223, 889-900.
Zurita, F., Anda, J. D. and Belmount, M. A. (2009). Treatment of domestic wastewater and production of commercial flowers in vertical and horizontal subsurface-flow constructed wetlands. Ecological Engineering, 35, 861-869.
Zurita, F. and Carreon-Alvarez, A. (2015). Performance of three pilot-scale hybrid constructed wetlands for total coliforms and Escherchia coli removal from primary effluent- a 2-year study in a subtropical climate. Journal of Water and Health, 13 (2), 446-458.

Annexes
Table 1: Mean, standard error, minimum and maximum values of BOD5 within the influent and effluents of the HFCW, VFCW and hybrid of the HFCW and VFCW employed at Kotebe WWTP, Addis Ababa-Ethiopia.
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
Winter (Influent) 6 162.33 224.67 198.3333 21.08179
Spring (Influent) 6 146.67 217.67 194.3900 25.55951
Summer (Influent) 6 132.67 205.67 163.4450 25.09931
Autumn (Influent) 6 167.00 204.33 189.1650 14.95608
Winter (HFCW-effluent) 6 10.00 27.67 15.5567 6.53263
Spring (HFCW-effluent) 6 6.33 16.33 10.8867 3.94807
Summer (HFCW-effluent) 6 13.33 32.33 23.1650 8.36807
Autumn (HFCW-effluent) 6 18.00 45.67 31.4450 10.52766
Winter (VFCW-effluent) 6 5.00 15.00 9.5000 3.90788
Spring (VFCW-effluent) 6 4.00 11.67 7.3350 2.55708
Summer (VFCW-effluent) 6 9.00 27.67 17.5550 7.62098
Autumn (VFCW-effluent) 6 11.33 32.33 22.3300 8.14862
Winter (Hyb CW-effluent) 6 6.00 10.00 8.1117 1.96326
Spring (Hyb CW-effluent) 6 4.00 9.33 5.7767 2.16593
Summer (Hyb CW-effluent) 6 8.67 23.00 15.7800 5.72827
Autumn (Hyb CW-effluent) 6 10.67 25.33 18.5000 6.55096
Valid N (listwise) 6

Table 2: Mean, standard error, minimum and maximum values of COD within the influent and effluents of the HFCW, VFCW and hybrid of the HFCW and VFCW employed at Kotebe WWTP, Addis Ababa-Ethiopia.
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
Winter (Influent) 6 344.00 505.33 447.9983 65.06278
Spring (Influent) 6 357.00 465.00 399.0567 39.27258
Summer (Influent) 6 323.00 445.00 374.0550 44.07695
Autumn (Influent) 6 313.33 445.33 390.1083 51.96378
Winter (HFCW-effluent) 6 38.00 78.67 60.9450 14.28724
Spring (HFCW-effluent) 6 46.00 83.00 61.6117 14.40239
Summer (HFCW-effluent) 6 67.67 103.00 84.2217 13.53198
Autumn (HFCW-effluent) 6 72.33 125.00 102.0550 18.21191
Winter (VFCW-effluent) 6 41.00 83.33 63.9433 15.82704
Spring (VFCW-effluent) 6 36.67 64.33 51.3883 10.76329
Summer (VFCW-effluent) 6 65.00 93.67 78.6683 11.53208
Autumn (VFCW-effluent) 6 56.67 123.33 91.2767 26.35577
Winter (Hyb CW-effluent) 6 38.67 72.00 53.7767 12.64386
Spring (Hyb CW-effluent) 6 33.00 61.00 47.7783 10.14388
Summer (Hyb CW-effluent) 6 56.00 85.33 68.9450 12.71809
Autumn (Hyb CW-effluent) 6 56.00 116.00 84.3333 21.20147
Valid N (listwise) 6

Table 3: Mean, standard error, minimum and maximum values of TSS within the influent and effluents of the HFCW, VFCW and hybrid of the HFCW and VFCW employed at Kotebe WWTP, Addis Ababa-Ethiopia.
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
Winter (Influent) 6 129.00 240.33 186.9433 41.52406
Spring (Influent) 6 94.00 290.67 189.1117 73.96847
Summer (Influent) 6 207.67 262.33 230.2783 20.52277
Autumn (Influent) 6 160.33 282.00 217.8883 43.06007
Winter (HFCW-effluent) 6 10.67 21.00 15.8350 3.76360
Spring (HFCW-effluent) 6 12.00 32.33 18.5567 7.40859
Summer (HFCW-effluent) 6 33.67 41.00 36.5567 2.75305
Autumn (HFCW-effluent) 6 13.67 25.67 19.9467 4.05212
Winter (VFCW-effluent) 6 18.33 39.00 28.4433 7.89983
Spring (VFCW-effluent) 6 19.67 49.00 30.8883 9.77642
Summer (VFCW-effluent) 6 36.67 51.67 46.1667 5.35218
Autumn (VFCW-effluent) 6 20.33 40.00 28.7767 7.51902
Winter (Hyb CW-effluent) 6 16.67 30.67 23.4450 5.92649
Spring (Hyb CW-effluent) 6 16.67 44.67 25.3917 10.10311
Summer (Hyb CW-effluent) 6 36.00 44.67 39.7783 3.27756
Autumn (Hyb CW-effluent) 6 21.67 44.33 34.1667 8.29155
Valid N (listwise) 6

Table 4: Mean, standard error, minimum and maximum values of NH4+ within the influent and effluents of the HFCW, VFCW and hybrid of the HFCW and VFCW employed at Kotebe WWTP, Addis Ababa-Ethiopia.
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
Winter (Influent) 6 53.17 69.58 58.4950 6.06465
Spring (Influent) 6 43.92 64.68 55.7450 7.57025
Summer (Influent) 6 37.30 43.67 39.6133 2.34730
Autumn (Influent) 6 32.63 64.83 49.2150 13.28853
Winter (HFCW-effluent) 6 13.73 26.33 20.2817 5.21357
Spring (HFCW-effluent) 6 17.10 24.70 20.9933 2.71070
Summer (HFCW-effluent) 6 14.67 22.97 18.1117 3.16745
Autumn (HFCW-effluent) 6 17.42 30.53 23.4217 4.98532
Winter (VFCW-effluent) 6 11.40 23.50 16.5650 4.20873
Spring (VFCW-effluent) 6 10.75 21.53 17.1750 4.50627
Summer (VFCW-effluent) 6 10.90 21.10 14.6117 3.55482
Autumn (VFCW-effluent) 6 14.63 27.23 19.4133 5.81890
Winter (Hyb CW-effluent) 6 11.93 17.03 15.1933 1.92236
Spring (Hyb CW-effluent) 6 14.43 24.90 18.3317 3.73288
Summer (Hyb CW-effluent) 6 12.83 19.27 15.1117 2.37746
Autumn (Hyb CW-effluent) 6 14.87 26.70 20.3500 5.14674
Valid N (listwise) 6

Table 5: Mean, standard error, minimum and maximum values of NO3- within the influent and effluents of the HFCW, VFCW and hybrid of the HFCW and VFCW employed at Kotebe WWTP, Addis Ababa-Ethiopia.
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
Autumn (Influent) 6 5.90 14.50 10.0500 2.87836
Winter (Influent) 6 3.47 9.30 6.0900 2.19695
Spring (Influent) 6 4.47 11.60 8.0783 2.90027
Summer (Influent) 6 3.33 9.63 5.8267 2.59512
Autumn (HFCW-effluent) 6 1.80 7.57 4.9017 2.38606
Winter (HFCW-effluent) 6 1.17 3.07 1.8283 .70822
Spring (HFCW-effluent) 6 .97 3.10 2.6517 .83219
Summer (HFCW-effluent) 6 .80 3.53 1.8833 1.12552
Autumn (VFCW-effluent) 6 2.60 4.70 3.5700 .92566
Winter (VFCW-effluent) 6 .73 2.43 1.4267 .61079
Spring (VFCW-effluent) 6 .83 3.37 1.9233 .83930
Summer (VFCW-effluent) 6 .63 3.57 1.8167 1.01348
Autumn (Hyb CW-effluent) 6 .93 6.07 3.2117 1.98321
Winter (Hyb CW-effluent) 6 1.07 2.60 1.5550 .56362
Spring (Hyb CW-effluent) 6 .83 2.63 1.6933 .72627
Summer (Hyb CW-effluent) 6 1.07 2.63 1.6067 .57712
Valid N (listwise) 6

Table 6: Mean, standard error, minimum and maximum values of TN within the influent and effluents of the HFCW, VFCW and hybrid of the HFCW and VFCW employed at Kotebe WWTP, Addis Ababa-Ethiopia.
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
Autumn (Influent) 6 65.00 77.33 71.1117 4.43919
Winter (Influent) 6 82.33 92.67 87.2233 4.02185
Spring (Influent) 6 57.67 90.33 80.7783 12.01259
Summer (Influent) 6 62.33 70.67 67.3333 2.96003
Autumn (HFCW-effluent) 6 32.33 45.33 39.2767 4.32843
Winter (HFCW-effluent) 6 41.00 46.00 43.1667 1.62904
Spring (HFCW-effluent) 6 31.33 45.67 37.6117 5.27408
Summer (HFCW-effluent) 6 28.67 45.33 35.1667 6.15381
Autumn (VFCW-effluent) 6 28.67 42.33 34.8883 4.47325
Winter (VFCW-effluent) 6 32.00 42.67 36.2233 3.71150
Spring (VFCW-effluent) 6 25.00 38.67 32.4450 5.03743
Summer (VFCW-effluent) 6 24.33 41.00 33.3867 5.51214
Autumn (Hyb CW-effluent) 6 27.00 36.67 32.9433 3.63583
Winter (Hyb CW-effluent) 6 30.00 38.33 32.1650 3.12421
Spring (Hyb CW-effluent) 6 26.00 32.67 29.5000 2.21971
Summer (Hyb CW-effluent) 6 21.33 38.67 30.7783 6.78958
Valid N (listwise) 6

Table 7: Mean, standard error, minimum and maximum values of PO43- within the influent and effluents of the HFCW, VFCW and hybrid of the HFCW and VFCW employed at Kotebe WWTP, Addis Ababa-Ethiopia.
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
Autumn (Influent) 6 3.00 6.73 4.6000 1.32828
Winter (Influent) 6 5.80 8.23 6.9950 .89717
Spring (Influent) 6 7.07 8.27 7.6717 .47993
Summer (Influent) 6 2.97 7.27 4.0917 1.58178
Autumn (HFCW-effluent) 6 .90 3.30 1.9617 .83712
Winter (HFCW-effluent) 6 2.60 5.37 4.0567 .94046
Spring (HFCW-effluent) 6 3.53 5.13 4.5533 .64537
Summer (HFCW-effluent) 6 1.97 3.27 2.3950 .48344
Autumn (VFCW-effluent) 6 .90 3.07 1.8433 .86929
Winter (VFCW-effluent) 6 2.70 4.90 3.7717 .78306
Spring (VFCW-effluent) 6 3.93 4.87 4.2933 .36909
Summer (VFCW-effluent) 6 1.20 2.97 2.0067 .67846
Autumn (Hyb CW-effluent) 6 .90 3.00 1.8050 .92115
Winter (Hyb CW-effluent) 6 2.80 5.27 4.1583 .85861
Spring (Hyb CW-effluent) 6 3.33 4.97 4.3333 .64357
Summer (Hyb CW-effluent) 6 1.67 2.53 2.1000 .36540
Valid N (listwise) 6

Table 8: Mean, standard error, minimum and maximum values of TP within the influent and effluents of the HFCW, VFCW and hybrid of the HFCW and VFCW employed at Kotebe WWTP, Addis Ababa-Ethiopia.
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
Autumn (Influent) 6 3.93 8.63 6.2100 1.64140
Winter (Influent) 6 4.37 11.33 8.2833 2.70450
Spring (Influent) 6 7.27 13.03 10.4067 2.22662
Summer (Influent) 6 3.30 6.73 4.7700 1.14666
Autumn (HFCW-effluent) 6 1.80 4.73 2.9700 1.15310
Winter (HFCW-effluent) 6 2.10 4.47 3.2617 .95120
Spring (HFCW-effluent) 6 2.93 4.87 3.9167 .76521
Summer (HFCW-effluent) 6 1.20 3.27 2.0683 .81835
Autumn (VFCW-effluent) 6 1.07 5.87 3.2783 1.68343
Winter (VFCW-effluent) 6 2.07 5.43 3.5933 1.40493
Spring (VFCW-effluent) 6 2.63 7.33 4.8267 1.62267
Summer (VFCW-effluent) 6 1.30 3.27 2.3900 .72014
Autumn (Hyb CW-effluent) 6 1.67 5.47 3.1283 1.41843
Winter (Hyb CW-effluent) 6 1.87 5.50 3.4850 1.38587
Spring (Hyb CW-effluent) 6 3.40 5.73 4.6267 1.04951
Summer (Hyb CW-effluent) 6 1.03 2.83 2.1600 .65663
Valid N (listwise) 6

Table 9: Mean, standard error, minimum and maximum values of FC within the influent and effluents of the HFCW, VFCW and hybrid of the HFCW and VFCW employed at Kotebe WWTP, Addis Ababa-Ethiopia.
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
Autumn (Influent) 6 74666.67 97666.67 88222.2250 9708.33869
Winter (Influent) 6 79000.00 148333.33 101000.0000 26393.59959
Spring (Influent) 6 72000.00 127666.67 110777.7783 21088.87845
Summer (Influent) 6 73666.67 88000.00 81166.6683 4838.50172
Autumn (HFCW-effluent) 6 1666.67 2766.67 2144.4467 481.51106
Winter (HFCW-effluent) 6 1466.67 5800.00 2410.0000 1671.61237
Spring (HFCW-effluent) 6 1266.67 2633.33 2166.6667 493.96248
Summer (HFCW-effluent) 6 966.67 1700.00 1316.6667 259.70025
Autumn (VFCW-effluent) 6 2166.67 3700.00 2872.2233 581.34470
Winter (VFCW-effluent) 6 2766.67 6333.33 3811.1117 1303.95258
Spring (VFCW-effluent) 6 2533.33 5233.33 4183.3333 1091.12556
Summer (VFCW-effluent) 6 1366.67 3000.00 2133.3333 583.85698
Autumn (Hyb CW-effluent) 6 2366.67 5033.33 3038.8900 993.84918
Winter (Hyb CW-effluent) 6 2000.00 7433.33 3216.6667 2087.39469
Spring (Hyb CW-effluent) 6 3700.00 5400.00 4544.4450 643.13957
Summer (Hyb CW-effluent) 6 1966.67 3233.33 2611.1117 480.58498
Valid N (listwise) 6

Table 10: Areal removal rate constants, K (m/d) of the parameters considered in this study.
BOD
Reference
Seasons K in HSSF type K in VF type K in Hybrid type K in literatures
Winter 0.112 0.134 0.141
Autumn 0.127 0.144 0.155
Summer 0.086 0.098 0.103
Spring 0.079 0.094 0.102
Annual 0.098 0.113 0.121 0.077
0.27 Sheridan et al, 2014
Abdelhakeem et al, 2016
COD
Seasons K in HSSF K in VF K in Hybrid K in literatures
Winter 0.088 0.086 0.093
Autumn 0.082 0.090 0.093
Summer 0.066 0.069 0.075
Spring 0.059 0.064 0.067
Annual 0.073 0.076 0.081 0.283
0.44
0.2 Konnerup et al, 2009
Yan Zhang, 2014
Abdelhakeem et al, 2016
TSS
Seasons K in HSSF K in VF K in Hybrid K in literatures
Winter 0.109 0.083 0.091
Autumn 0.102 0.080 0.088
Summer 0.081 0.071 0.077
Spring 0.105 0.089 0.082
Annual 0.097 0.080 0.084 0.109
0.82
0.19 Korkusuz et al, 2004
Yan Zhang, 2014
Abdelhakeem et al, 2016
NH4+-N
Seasons K in HSSF K in VF K in Hybrid K in literatures
Winter 0.047 0.055 0.059
Autumn 0.043 0.052 0.049
Summer 0.034 0.044 0.042
Spring 0.033 0.041 0.039
Annual 0.040 0.048 0.048 0.085
0.05 Korkusuz et al, 2004
Abdelhakeem et al, 2016
TN
Seasons K in HSSF K in VF K in Hybrid K in literatures
Winter 0.031 0.039 0.044
Autumn 0.034 0.040 0.044
Summer 0.029 0.031 0.034
Spring 0.026 0.031 0.034
Annual 0.030 0.036 0.039 0.0158
0.73 Konnerup et al, 2009
Yan Zhang, 2014
PO43-P
Seasons K in HSSF K in VF K in Hybrid K in literatures
Winter 0.024 0.027 0.023
Autumn 0.023 0.026 0.025
Summer 0.023 0.031 0.029
Spring 0.038 0.040 0.041
Annual 0.026 0.030 0.028 0.002 Korkusuz et al, 2004
TP
Seasons K in HSSF K in VF K in Hybrid K in literatures
Winter 0.042 0.038 0.039
Autumn 0.033 0.024 0.026
Summer 0.037 0.030 0.035
Spring 0.032 0.028 0.030
Annual 0.039 0.033 0.035 0.0149
0.46
0.03 Konnerup et al, 2009
Yan Zhang, 2014
Abdelhakeem et al,2016

Figure 1: Site clearing, excavation work and preparation for the construction of the pilot scale constructed wetlands at Kotebe WWTP, Addis Ababa-Ethiopia

Figure 2: Construction process of the pilot scale constructed wetland systems employed at Kotebe WWTP, Addis Ababa-Ethiopia.

Fig 3: The process of pipe installation, sealing of the wetland systems bed and application of the fill media at Kotebe WWTP, Addis Ababa-Ethiopia.

Fig 4: Source, transportation and plantation of the wetland plant/Cyprus papyrus/ to the pilot scale constructed wetlands system applied at Kotebe WWTP, Addis Ababa-Ethiopia.

Figure 6: Adaptation and growth of the wetland plants of the pilot scale constructed wetland systems employed at Kotebe WWTP, Addis Ababa-Ethiopia.