Study on potential plants for use in constructed wetlands to strengthen phosphorus treatment performance from swine wastewater

Tissue and cellular phosphorus storage during development of phosphorus toxicity in Hakea prostrata (Proteaceae). Nutrient uptake by constructed floating wetland plants duri[r]

VIETNAM NATIONAL UNIVERSITY, HANOI VIETNAM JAPAN UNIVERSITY

VU THI THOM

STUDY ON POTENTIAL PLANTS FOR USE IN CONSTRUCTED WETLAND TO STRENGTHEN

PHOSPHORUS TREATMENT PERFORMANCE FROM SWINE WASTEWATER

MASTER'S THESIS

VIETNAM NATIONAL UNIVERSITY, HANOI VIETNAM JAPAN UNIVERSITY

VU THI THOM

STUDY ON POTENTIAL PLANTS FOR USE IN CONSTRUCTED WETLAND TO STRENGTHEN

PHOSPHORUS TREATMENT PERFORMANCE FROM SWINE WASTEWATER

MAJOR: ENVIRONMENTAL ENGINEERING CODE: PILOT

SUPERVISORS

DR NGUYEN THI AN HANG ASSOC PROF DR SATO KEISUKE

DR NGUYEN THI HOANG HA

i

ACKNOWLEDGMENTS

First of all, I would like to express the sincere gratitude to my principal supervisor, Dr Nguyen Thi An Hang at VNU Vietnam Japan University, for accepting me as her master student and continuous teaching and supporting me in the process of doing experiments as well as writing essays and making presentations She always encourages and is willing to help me when I have difficulties She is always beside me to teach me how to work effectively This helps me grow up in both personal and professional aspects A special thanks also go to Assoc Prof Dr Sato Keisuke for his recommendations to my research He provided me with the best conditions for implementing my experiments during my internship in Japan He is wholeheartedly devoted to his students I would like to express my deepest thanks to Dr Nguyen Thi Hoang Ha She gave me valuable supports in developing research methods, and enthusiastically guided me to fullfil my thesis I always feel grateful to her for accompanying me for such a long time The second, I would like to send my sciencere thanks to Prof Dr Jun Nakajima for supporting not only me but also all of memerbers in my class during Master course He cares for us like his children And he is our respected father

The third, I am grateful to Ms Nguyen Thi Xuyen, the project staff, for always supporting me in conducting experiments as well as analyzing environmental parameters

This research is funded by Vietnam National Foundation for Science and Technology Development (NAFOSTED) under grant number 105.99-2018.13, 2018 Asean Research Center (ARC) research grant of Vietnam National University, Hanoi (VNU), and Japan International Cooperation Agency (JICA) research grant for the academic year of 2018-2019

My heartfelt thanks and gratitudes to my family for their unconditional helps with plant sampling, the love and encouragement

Hanoi, June 9th, 2019

TABLE OF CONTENTS

ACKNOWLEDGMENTS i

LIST OF TABLES ii

LIST OF FIGURES iii

LIST OF ABBREVIATIONS v

INTRODUCTION vi

CHAPTER1:LITERATUREREVIEW

1.1 Overview of the situation of pig husbandry in Vietnam

1.1.1 Current status and development orientation of pig breeding

1.1.2 Environmental pollution due to swine wastewater

1.1.3 Technologies for the treatment of swine wastewater

1.2 Phosphorous pollution and treatment technologies

1.2.1 Phosphorus pollution and its consequences

1.2.2 Phosphorus treatment technologies

1.3 Constructed wetland for wastewater treatment

1.3.1 Definition and classification of CWs

1.3.2 Influential factors and treatment performance

1.4 Removal phosphorus by plants in the CWs 12

1.4.1 Classification of plants used in CWs 12

1.4.2 Removal P mechanisms by plants 14

CHAPTER2:MATERIALSANDRESEARCHMETHODOLOGY 18

2.1 Research object, scale, and scope 18

2.1.1 Research object 18

2.1.2 Research scale & scope 18

2.2 Materials, chemicals and equipment 19

2.2.1 Materials 19

2.2.2 Experimental design 20

2.2.4 Analysis of other water quality parameters 26

2.3 Data calculation 27

2.4 Data statistical analysis 28

CHAPTER3:RESULTSANDDISCUSSION 29

3.1 Screening potential plants for use in the CWs 29

3.1.1 Selection of potential plants based on their P content and biomass growth 29

3.1.2 Selection of CWs plants based on other growth characteristics 33

3.2 Factors influencing the growth and uptake of p of Cymbopogon citratus and Ubon paspalum 35

3.2.1 Effect of initial P concentration 35

3.2.2 Effect of pH 42

3.2.3 Effect of plant age 45

3.2.4 Effect of plant density 49

3.2.5 Effect of water level 52

3.3 Applicability of the selected plants in the constructed wetland 54

CHAPTER4:CONCLUSIONANDRECOMMENDATION 57

4.1 Conclusion 57

4.2 Recommendations 57

REFERENCE 58

ii

LIST OF TABLES

Table 1.1 Annual growth rate of culture sector (%)

Table 1.2 Composition and characteristics of swine wastewater

Table 1.3 Methods for handing and using liquids at systems

Table 1.4 main parameters of swine wastewater after biogas treatment

Table 1.5 P removal by Constructed Wetland 11

Table 1.6 Plant species are used to treat swine wastewater 13

Table 1.7 P removal efficiency by plants in CW (Jesus et al., 2017) 15

Table 2.1 The list of investigated plants 19

Table 2.2 Methods for examination of water quality parameters 27

Table 3.1 The P content in plants use for phytoremediation or CWs 31

Table 3.2 The P removal potential of the studied plants 33

Table 3.3 Growth characteristics of potential plants 34

Table 3.4 The P removal efficiency by different plant species 39

Table 3.5 Biomass growth rate of Ubon paspalum at different plant ages 48

iii

LIST OF FIGURES

Figure 1.1 Eutrophication: cause and effect

Figure 1.2 P removal in CWs

Figure 2.1 Scheme of horizontal constructed wetland 23

(at the start of experiment) 23

Figure 2.2 The structure of filter media in CWs and adsorption units 24

Figure 2.3 Plant sample preparation and analysis 25

Figure 2.4 Images of apparatus used in this study 26

Figure 3.1 The P content and its distribution in the studied plants 29

Figure 3.2 Images of the investigated plants in this study 34

Figure 3.3 Effect of initial P concentration on the removal efficiency of Cymbopogon citratus 37

Figure 3.4 Effect of initial P concentration on the removal efficiency of Ubon paspalum 37

Figure 3.5 P concentration left in solution plant with Ubon paspalum 38

Figure 3.6 P concentration left in solution planted with Cymbopogon citratus 38

Figure 3.7 Effect of intial P concentration on P removal rate of Ubon paspalum 40

Figure 3.8 Effect of intial P concentration on P removal rate of Cymbopogon citratus 40

Figure 3.9 Ubon paspalum died at the highest P concentration 41

Figure 3.10 Cymbopogon citratus could adapt with a wide range of initial P concentration 41

Figure 3.11 Normal growth of Cymbopogon citratus at all pH values 42

Figure 3.12 The death of Ubon paspalum at pH values of 9&11 43

Figure 3.13 Speciation of P in solution at various pH conditions 43

Figure 3.14 Effect of pH on P removal efficiency of Ubon paspalum and Cymbopogon citratus 44

iv

Figure 3.16 Effect of plant age on P removal efficiency and P removal rate of Ubon

paspalum (hydroponic experiment) 47

Figure 3.17 The effect of plant age on the growth of root system 47

Figure 3.18 Effect of plant age on the biomass growth of Ubon paspalum 48

(experiment with garden soil) 48

Figure 3.19 Effect of plant density on P removal rate of the investigated plants 50

Figure 3.20 Effect of plant density on the P removal efficiency of the investigated plants 50

Figure 3.21 The root growth of Ubon paspalum at different plant densities 52

Figure 3.22 Effect of Ph on P removal rate of plants 53

Figure 3.23 Effect of water level on root growth of Ubon paspalum 54

Figure 3.24 The change of phosphorus in the effluent over the time 54

Figure 3.25 P removal efficiency and Ph after treatment of HFCWs 55

v

LIST OF ABBREVIATIONS

COD Chemical oxygen demand CW Constructed Wetland BOD Biological oxygen demand

EBPR Enhanced biological phosphorus removal EPA Environmental Protection Agency

HF Horizontal flow

HLR Hydraulic loading rate HRT Hydraulic retention time SSF Subsurface water flow SF Surface flow

TN Total nitrogen TP Total phosphorus TSS Total suspended solids VF Vertical flow

vi

INTRODUCTION

Background

In Vietnam, in recent years, pig breeding industry has developed rapidly Since most of pig farms have not designed and operated appropriately, wastewater from pig farms cause serious environmental pollution, which poses a high risk to public health and surrounding ecosystems Therefore, the proper treatment of swine wastewater is urgent and necessary At present, swine wastewater in Vietnam is normally treated by biogas technology However, the concentration of pollutants in the effluent is still high, exceeding national discharge standards (QCVN 01-79: 2011/BNNPTNT) Thus, further processing after biogas treatment of swine wastewater is mandatory

Constructed wetlands (CWs) is a promising technology, which possesses many advantages, such as cost-effective, green technology (Wu et al., 2015; Yang et al, 2018;), low land, energy, and less-operational requirements (Wu et al., 2015); simple construction and operation (Bunce et al., 2018) However, the wide application of CWs is limited by intensive land requirement, long-term unsustainability (Bunce et al., 2018) Especially, although CWs can achieve high removal efficiency with TSS, COD, BOD, it is demonstrated to be inefficient in nutrient elimination It is well-known that the treatment performance of phosphorus by CWs is low and unstable Hence, the enhancement of phosphorus removal by CW is of great significance Since phosphorus is eliminated by CWs mainly via substrate adsorption, plant uptake, microbial degradation, selection and application of potential plants in CWs plays an important role

vii

Phragmites communis, Typha orientatlis and Sparganium stoloniferum was 72.67%, 73.39%, 71.54%, respectively (Liu et al., 2012) TP was considerably eliminated by Persicaria hydropiper, representing 97.63% (Zheng et al., 2013) However, these are wildlife grass type with no or less economic value There is still lack of information about potential plants with high economic value for phytoremediation and CWs to treat P- rich water and wastewater Also, very few studies on factors influencing the growth and utapke of phosphorus by plants

The objectives of this study comprise (1) to search for potential plants for phosphorus decontamination from wastewater, (2) to investigate five factors influencing the growth and phosphorus removal efficiency of Ubon paspalum and

Cymbopogon citratus, (3) to evaluate the applicapability of selected plant (Ubon paspalum) in the constructed wetlands for treatment of synthetic swine wastewater

Thesis’ outline

The thesis has been completed by chapters The main contents of each chapter are provided as below

Introduction: Introduce the research content, identify research issues, tasks,

purposes, research scope

Chapter 1: Literature review - Providing information on the characteristics of

P factor, the negative impact of P on the environment, technological solutions implemented to eliminate P, the removal efficiency of Contructed Wetland, roles and mechanisms P removal of plants

Chapter 2: Materials and methods - Describe potential plant selection

methods, methods of analyzing P content in plants and water, methods of setting up experiments to study absorption capacity of plants under the influencing factors

Chapter 3: Results - Focus on describing the results of plant uptake rates and

viii

Chapter 4: Discussion and recommendation - Summarizes major findings of

this study Additionally, the unique contributions of this study to the field of phosphorus removal by plants are provided Provide limitations and for future research direction

1

CHAPTER 1: LITERATURE REVIEW

1.1 Overview of the situation of pig husbandry in Vietnam

1.1.1 Current status and development orientation of pig breeding

According to FAO, Asia will become the region, which produces and consumes livestock products the most Like other countries in the region, Vietnam needs to maintain high growth level to meet the demand of domestic consumption and export In recent years, the livestock industry in Vietnam has developed rapidly The annual growth rate of livestock in the period of 2006-2010 was 8.5%

Table 1.1 Annual growth rate of culture sector (%) Year

Sectors

1986-1990 1990-1996 1997-2005 1986-2005 2006-2010 Other

agriculture 3.4 6.0 5.5 5.2 4.1

Cultivation 3.4 6.1 5.4 5.2 5.5

Livestock 3.4 5.8 6.7 5.6 8.5

Service 4.1 4.6 2.3 3.6 4.2

Source: Vietnam Agricultural Economics Institute, 2009

Recently, there is a trend to develop centralized animal husbandry As a result, the number of large-scale farms is increasing By 2006, Vietnam has 17,721 farms throughout the country, of which there is 7,415 pig farms, accounting for 42.18% The number of pig farms has increased from 3,293 to 14,481 in the period of 2011-2016

2

implementation of intensive animal husbandry Of which, pig raising plays an important role According to Decision No 10/2008/QĐ-TTg dated on 16 January 2008 by the Prime Minister on the approval of livestock development strategy to 2020, the proportion of livestock in the agriculture will increase from 32% (2010) to 38% (2015) and reach 42% (2020)

1.1.2 Environmental pollution due to swine wastewater

a Characteristics of swine wastewater

The swine wastewater can pose a high risk to the environment, due to high content of organic matter, TSS, N, P, and pathogen According to the report to assess the current status of the environment by Institute for Animal Husbandry (2006), in centralized pig farms in Ha Noi, Ninh Binh, Nam Dinh, Quang Nam, Binh Duong, Dong Nai, the swine wastewater is characterized as follows:

- Organic matter accounts for 70-80%, including cellulose, protein, acid amine, lipid, hydrate carbon, …

- Inorganic matter represents 20-30%, including, salts (chloride, sulfate, etc.) - Nutrients (N and P): The swine wastewater usually contains high levels of N and

P (e.g TN 200-350 mg/ L of which N-NH4 accounts for 80-90%, TP 60-100 mg/ L)

3

Table 1.2 Composition and characteristics of swine wastewater

No Parameters Unit Value Category A, QCVN 40:2011/BTNMT pH - 7.23 - 8.07 6-9

2 BOD5 mg/ L 1664 - 3268 30

3 COD mg/ L 2561 - 5028 75 SS mg/ L 1700 - 3218 50 N-NH4+ mg/ L 10 - 50

6 N total mg/ L 512 - 594 20 P total mg/ L 13.8 - 62

Source: Xuyen Viet Environment Company b Swine wastewater management

4

Table 1.3 Methods for handing and using liquids at systems

Swine wastewater with treatment

Biogas % 42.5 24.39 64.70 73.68

m3 3.87 ±5.43 4.41±1.28 3.73±1.83 3.98 ±2.98

Settling pond % 11.25 - - -

m3 5.59 ±2.86 - - -

Swine wastewater

without treatment

Pouring into fish pond

% 63.75 75.60 - -

m3 4.99 ±1.28 6,58±4.32 - -

Discharge into environment

% 11.25 12.19 57.14 63.15

m3 2.22 ±2.23 4.91±2.95 3.98 ±5.75 3.50±5.40

1.1.3 Technologies for the treatment of swine wastewater

Numerous technologies have been used for treatment of swine wastewater, which can be divided into (1) Mechanical treatment method, (2) Physio-chemical treatment method, and (3) Biological treatment method Of which, biological treatment method is mainly used after mechanical and physio-chemical treatment

5

Table 1.4 main parameters of swine wastewater after biogas treatment

No Parameter Unit Test method Results QCVN

40:2011/BTNMT Ammonium (NH4+) mg/L ISO 5664 (TCVN 5988) 180-340 10

2 BOD5 mg/L SMEWW 5210B 590-1260 50

3 Phosphorus (P) mg/L SMEWW 4500P B

SMEWW 4500P E 43-58

1.2 Phosphorous pollution and treatment technologies

1.2.1 Phosphorus pollution and its consequences

Phosphorous is an element essential for plant growth It is actively involved in main functions of plants, such as photosynthesis, energy conveyance, activation of protein, control of metabolic processes, etc (Vo et al., 2017) However, the excessive level of phosphorus may cause the eutrophication in natural water bodies

Figure 1.1 Eutrophication: cause and effect

This results in many water quality issues, including deterioration of water quality and algal bloom Degradation of algal reduces O2 but increase CO2 in water, thus influencing the life of aquatic organisms The lack of oxygen leads anaerobic decomposition of organic matter, producing foul smell (H2S, NH3, CH4) (Thongtha

6

activities Natural activities include earthquake, soil erosion, etc Human activities can release phosphorus into aquatic medium, such as agriculture, wastewater, storm water, etc The contribution of domestic sewage, agriculture and industry to the global phosphorous load is 54, 38, and 8% respectively The phosphorus load from agriculture in USA increased by 27%, from 579,000 to 734,000 tons in 2010

1.2.2 Phosphorus treatment technologies

Phosphorus decontamination from wastewater can be done via physical, chemical or biological methods Physical methods include membrane, magnetic separation, adsorption, ion exchange etc Chemical processes are precipitation, crystallization

7

requirement, complex configuration and operating regimes, more energy and space requirement (Nguyen et al., 2015) The utilization of biological process is not only limited to EBPR but also using algal or aquatic plants Sukacova et al (2015) stated that fixed growth algal bioreactor could eliminate up to 97% phosphorus in wastewater The ability of macrophyte in phosphorus decontamination from wastewater will be discussed thoroughly in other section of thesis

1.3 Constructed wetland for wastewater treatment

1.3.1 Definition and classification of CWs

a Definition

Artificial wetlands are human wetland areas designed to treat wastewater The system of artificial wetlands has low operating and maintenance costs, low energy consumption, does not require high operational and environmentally friendly techniques (Viet et al., 2019)

Components of CWs include substrate, plant, microorganism, animal (earthworm), water (Davis, 1985) The biomass of plants in the system can be used as animal feed, as a fiber material or as an organic fertilizer However, the land area for construction of artificial wetlands is relatively large, which is an obstacle to the selection of this treatment method, so artificial wetlands are suitable for crowded areas with wide and unfocused land area

b Advantages and limitations

CWs are inexpensive (building, maintenance), simple operation, tolerant to various flow, habitat of many submerged species, wild animals; improve the surroundings, get the community's approval (Davis, 1995)

CWs need to be a large land area, effectively processing inconsistently affected external environmental conditions such as rainfall, droughts, seasons, sensitive to toxic chemicals (Davis, 1995)

8

There are many ways to classify CWs which are based on water level, direction of water (Vymazal, 2008) Based on CWs water levels are classified into categories: Surface flow CWs, Subsurface flow CWs (VF-CWs, HF-CWs), Hybrid / integrated / combined CWs The water level in the

SF system is higher than the substrate surface while it is equal to or lower than the substrate surface in the SSF system (Davis, 1995)

Based on the flow direction of water, SSF is divided into types of horizontal flow (HF) and vertical flow (VF) Water in HFCWs flooded the substrate in the system before exiting through water level control While water in the VFCWs system drains with the intermittent application of water to the system (Stefanakis et al 2014)

d Mechanisms of CWs for P removal

9

Figure 1.2 P removal in CWs (Hristina Bodin, 2013)

The most important way to remove phosphorus is adsorption and precipitation of matrix in constructed wetland system, there is less effect for plant adsorption of organic phosphorus, but the absorption of plant is given priority to remove inorganic phosphorus, which may be related to the large plants, like reed plants, need for inorganic phosphorus with a longterm growth.( LI jianbo, 2008) considered that: the adsorption by plants is a major way when at a low concentration of phosphorus, and the absorption by plants appear to be negligible when at a higher concentrations However, the adsorption of medium is limited, that is the absorption effect will be reduced after reaching saturation (Qin and Chen, 2016)

1.3.2 Influential factors and treatment performance

a Influence factors

Substrates (medium)

10

hydraulic conductivity would result in clogging of systems, severely decreasing the effectiveness of the system, and low adsorption by substrates could also affect the long-term removal performance of CWs (Wang et al., 2010) Many studies also suggest that substrates such as sand, gravel, and rock are the poor candidate for long-term phosphorus storage, but by contrast, artificial and industrial products with high hydraulic

HRT (Hydraulic retention time)

HRT determines the average contact time of microbial communities with pollutants (Lee et al., 2009) Furthermore, the effect of HRT may differ between CWs depending on the dominant plant species and temperature, as those factors can affect the hydraulic efficiency of wetlands

HLR (Hydrologic loading rate)

HLR is defined as following formular: 𝑞 = 𝑄/𝐴

100

Where q is defined as the volume per time per unit area (cm day -1); A is the wetland surface area (m2), Q is the flow rate (m3 day-1) Avila et al (2014) also studied the feasibility of hybrid CW systems used for removing emerging organic contaminants, and demonstrated that the removal efficiency for most compounds decreased as the HLR increased (Yan and Xu, 2014; Huang et al., 2000)

Feeding mode

11 b P removal efficiency

Table 1.5 P removal by Constructed Wetland

Types of plants Environments Types of wastewater Initial concentration (mgP/L) Removal (%) Foxtail Grass, Flax Lily, Banksia, and Bottlebrush Pea gravel, sand, and loamy sand Synthetic

strom water 4.51 mg/l

6-36 Typha Lockport dolomite Sewage wastewater 189

(mg/m2/d) 18 Queenston

shale

400-700

(mg/m2/d) 17-28 Fonthill

sand

105-331 (mg m−2

day−1)

5-58 P australis, T latifolia, P hydropiper, A sessilis, C esculenta and P stratoites Gravel

5.75 (mg/l) 76

J effuses

Sand and clay Agricultural

runoff 2.5 (mg/l)

77

C lurida 85

D acuminatum 74

Phragmites australis and cattails, Typha

latifolia

Gravel 10mg/l

(P- PO43-) 60 T.angustifolia,

P.australis, S pungens

Paxton fine

sandy loam soil Dairy cows 68

12

1.4 Removal phosphorus by plants in the CWs

1.4.1 Classification of plants used in CWs

a Role of plants in CWs

Plants is one of factors will affect the performance of CWs Plants provide an environment for microorganisms to attach and release oxygen from the root system which affect removal efficiency of plants (Jethwa and Bajpai, 2016) Using green plants to reduce pollutant concentration in soil and water was defined as phytoremediation (“Phyto” meaning plants, “remediation” meaning to restore and clean) (Cunningham et al., 1997)

Phytoremediation is more attractive than other technologies thanks to low maintenance, far-reaching, reducing pollution emissions, dust and by-products, preventing soil erosion, surface water flow, permeability, noise reduction, and increased aesthetics, carbon dioxide absorption, improved soil supply after treatment (Champagne, 2007)

In addition, phytoremediation (phytoremediation) is economically viable According to Champagne (2007), this method is at least 40% cheaper than other on-site processing methods and 90% less than ex situ technologies

b Classification of plants used in CWs

Wetland plants can be categorized under four main classes, namely, emergent plants, floating leave macrophytes, submerged plants, and freely floating macrophytes Wu et al (2014)

Macrophytes frequently used in CW treatments include emergent plants, submerged plants, floating leaved plants and free floating plants Although more than 150 macrophyte species have been used in CWs globally, only a limited number of these plant species are very often planted in CWs in reality Emergent species are

13

frequently used submerged plants are Hydrilla verticillata, Ceratophyllum demersum, Vallisneria natans, Myriophyllum verticillatum and Potamogeton crispus The floating leaved plants are mainly Nymphaea tetragona, Nymphoides peltata, Trapa bispinosa and Marsilea quadrifolia The free-floating plants are Eichhornia crassipes, Salvinia natans, Hydrocharis dubia and Lemna minor In addition, Ornamental flowering plants, especially Canna indica (Sandoval et al., 2019)

P.australis is the most common species in Asia and Europe while T latifolia is the most popular plant used in North America The most used plants in Africa are Cyperus papyrus L., P australis and Typha domingensis, Schoenoplectus tabernaemontani In Central and South Americas, Oceania, Palla was recored the most popular wetland plants Regarding types of the wetland plants used subsurface wetland, the second most common plant is Typha spp which is found in Australia, East Asia, North America, Africa In addition, P.australis is the most popular species globally (IWA Specialist Group 2000; Scholz 2006; Vymazal 2014)

Common species used in HFCWs are Scirpus (lacustri, acutus, californicus

and validus) Typha (domingensis, glauca, orientalis, latifolia and angustifolia),

Bulrush and comment reeds Phragmites spp is the most popular (Vymazal, 2011) And most of them are herbaceous plants (Vymazal, 2011; Jethwa and Bajpai, 2016)

Table 1.6 Plant species are used to treat swine wastewater

Species Common name Science name Submerged plant

Hydrilla Hydrilla verticilata

Water milfoil Myriophyllum spicatum

Blyxa Blyxa aubertii

Free floating plants

Water hyacinth Eichhornia crassipes

Rootless duckweed Wolfia arrhiga

Water lettuce Pistia stratiotes

Water fern Salvinia spp

14

1.4.2 Removal P mechanisms by plants

a Removal P mechanisms and P removal efficiency of plants

The roots use energy to get P into the tree through the cell membrane Other changes take place in rhizosphere affecting plant P uptake The roots secrete organic acids (citrate and oxalate) which increase the availability of P availability Amount of excreted organic acid, mycorrhizal fungi, root-zone microorganisms allow a plant to uptake P more from soil P is removed from the system by harvesting the plants (Brix 1997; Ma et al 2016)

In fact, the CWs with plants are more effective (Vymazal, 2011; Tanner, 2001) Depending on the stage of the system, plants will contribute to various removal effects For immature CWs, the role of plants in eliminating P will not be clearly shown However, the P removal efficiency of the system can still be enhanced by plants through its indirect impact on the treatment conditions of the system (Tanner, 2001)

In addition, phytoremediation (phytoremediation) is economically viable According to Champagne (2007), this method is at least 40% cheaper than other on-site processing methods and 90% less than ex situ technologies

The removal efficiency of P of Typha latifolia, Canna indica, Phragmites

australisdao is 0.06 -74.87%, 0.43 - 4.17, 0.56 - 36.7%, respectively under different

conditions In the same research conditions, the efficiency of removing P of Cladium

mariscus and Iris pseudacorus is 10% and 18% (Jesus et al., 2017) The treatment

15

Table 1.7 P removal efficiency by plants in CW (Jesus et al., 2017) CW

type Plant species

P uptake

(%) Notes Source

FWS

Phragmites australis 36.71 1st year harvested Zheng et al (2015) Phragmites australis 34.19 1st year

unharvested Zheng et al (2015) Phragmites australis 32.02 2nd year harvested Zheng et al (2015) Phragmites australis 35.93 2nd year

unharvested Zheng et al (2015)

VSSF

Typha latifolia 35.53 Tang et al (2008) Typha latifolia 42.54 Tang et al (2008) Typha latifolia 74.87 Tang et al (2008) Typha orientalis 14.31 Tang et al (2008) Phragmites australis 10.76

Scirpus validus 32.27

Iris pseudacorus 34.17 High nutrient Iris sibirica 13.19 Medium nutrient Iris sibirica 13.19 Low nutrient Iris sibirica 13.19

Phragmites

Australis 22

Sara G Abdelhakeem ,

Samir A Aboulroos

HSSF

Canna indica 0.7 Cui et al (2015) Canna indica 0.43 Cui et al (2015) Phragmites australis 0.56 Meng et al (2014)

Arundo donax 0.36 Meng et al (2014) Typha latifolia 0.06 Meng et al (2014)

SSF

Arachis duranensis 10.4 Van et al (2015) Cyperus

alternifolius 29.8 Van et al (2015) Philodendron

hastatum 29.8 Van et al (2015)

Phalaris

arundinacea 45.9 Lower input

Březinová and Vymazal (2015) Phalaris

arundinacea 3.1 Higher input

Březinová and Vymazal (2015) b Plant selection criteria

16

species, growth conditions, root surface area, oxidizing supply capacity, type of waste water and rate of loading, ability to withstand, absorb pollutants, be resistant to flooding, huge biomass (Jesus et al., 2017)

Fast growth

The rapid growth of plants corresponded with high level of P Therefore, they uptake a significant amounts of nutrient during the period of their growth Harvesting their above parts is the way to remove nutrients from wastewater (Vymazal, 2007)

Tolerant to continuous flooding

Beside the requirement of wetland plants with huge biomass and well developed root system, the tolerance ability to flood affects the nutrient removal efficiency by plants (Almuktar, 2018)

Tolerant to contaminant

Plants can be affected by environment stresses because many pollutants are present in CWs The concentration of influences in wastewater is too high to exceed the capacity of plants which reduces the growth and survival of plants (Surrency, 1993) In addition, high levels of pollutants directly affect the ecosystem of CWs causing inhibition of plant growth, even causing the disappearences of plants (Wu et al., 2015)

The high concentration of pollutants in water resulting in disadvantage of both treatment efficiency and plant survival Plant tolerance to the high concentration of pollutants is another important factor which is considered when selecting them for CWs (Almuktar, 2018)

Ability to accumulate contaminant

Wetlands plants are recognized as an important factor affecting water quality in CWs Absorption capacity of pollutants of plants contributes to CWs removal efficiency (Wu et al., 2015)

17

18

CHAPTER 2: MATERIALS AND RESEARCH METHODOLOGY

2.1 Research object, scale, and scope

2.1.1 Research objects

In this research, experiments to investigate influential factors on P removal by selected plants and those with CWs were conducted with synthetic wastewater The experiments to search for potential plants were carried out with soil

This study investigated 05 types of plants, including Colocasia gigantean, Piper lolot, Sauropus androgynous, Cymbopogon citratu, and Ubon paspalum

2.1.2 Research scale & scope

The experiments to evaluate influential factors and those with CWs were implemented at lab-scale The former was done at the laboratory of Master’s Program in Environmental Engineering (MEE), VNU Vietnam Japan University (VJU), whereas the latter was located on the roof of a residential building in Yen Hoa, Cau Giay, Hanoi

Concerning wastewater quality, this study focused on the removal of ortho phosphate (P-PO43-) of investigated plants Besides, other environmental parameters,

such as TSS, pH, BOD5, COD, TN, N-NH4+, TP, P-PO43- were measured to evaluate

the composition of real swine wastewater

19 2.2 Materials, chemicals and equipment

2.2.1 Materials

a Plants

The investigated plant Ubon paspalum was collected in a home garden in Dong Phong commune, Tien Hai district, Thai Binh province whereas Colocasia gigantean, Piper lolot, Sauropus androgynous, and Cymbopogon citrate were gathered in Quang Bi commune, Chuong My district, Hanoi City Depending on the purpose of experiments, the plants of different ages were utilized

Table 2.1 The list of investigated plants

Common name Science name Picture Location

Lemongrass Cymbopogon citratus

Quang Bi commune,

Thinh Da hamlet, Chuong

My district, Hanoi

city Piper lolot Piper sarmentosum

20

Star gooseberry Sauropus androgynus

Ubon paspalum Ubon paspalum

Dong Phong commune,

Tien Hai Distric, Thai Binh province

b Synthetic wastewater

Synthetic wastewater was prepared accordingly the composition and characteristics the real swine wastewater collected from a pig farm, which was located in Luong Xa village, Nam Dien commune, Chuong My district, Ha Noi c Chemicals

KH2PO4, NH4Cl used in this study were of analytical grade and purchased from

ESQ Co., Ltd (Ba Dinh, Hanoi)

2.2.2 Experimental design

a Screening potential plants

This experiment was to search for plants, which have potential for used in phytoremediation or CWs to eliminate phosphorus All investigated plants were grown in the soil They were harvested for determination of phosphorus content at the mature age In this experiment, phosphorus content and biomass growth rate were used for comparison purpose

b Investigating influential factors

21

P working solution (50 mg P/ L) After that, the certain amount of NH4Cl was added to make the nutrient solution A (50 mg P/ L and 500 mg N/ L) Finally, the A solution was diluted four times to get the background nutrient solution B (12.5 mg P/ L and 125 mg N/ L) The experiments to investigate influential factors were carried out by varying the influential factors while using the same background nutrient solution (except the experiment to investigate effect of initial P concentration)

Effect of initial phosphorus concentration: This experiment was designed to investigate how the plant can grow and uptake phosphorus in the solutions of different phosphorus concentrations This experiment was conducted with synthetic wastewater, which simulated 100%, 50 and 25% of real swine wastewater (P-PO4 50 mg/L, N-NH4-500 mg/L), in term of P and N This experiment included turns, each turn lasted for days, when the phosphorus in the solution was almost removed Cymbopogon citrate and 1.5-month Ubon paspalum were utilized for comparison 45 g of each plant was cultivated in a beaker containing 200 mL of nutrient solution The water sampling was done at the beginning of the experiment and after every days for determination of phosphorus concentration First, the water volume was measured Next, tape water was added for compensation of vaporization After that, mL of water was sampled for P analysis

22

Effect of plant age: The purpose of this test was to identify at which growth period, the plant is most efficient in phosphorus removal from wastewater This experiment was implemented with the background nutrient solution Ubon paspalum of kinds of age (1, 1.5, and months) and Cymbopogon citrate of kinds of age (baby and mature) were used This experiment lasted for turns (2 days/ turn) The water sampling was implemented the same as that for experiment to investigate the effect of pH

Effect of plant density: The experiment was to determine the best plant density for plant growth and phosphorus accumulation This experiment was carried out in both soil and auqeous solutions For hydroponic experiments, three kinds of plant density, such as 1, 3, and plant(s)/ beaker were applied The water sampling was done in a similar procedure to that of experiments for pH and plant age For soil experiment, it was conducted only with Ubon paspalum with types of plant density (1, 3, and plants/ trough) The soil was fertilized times/ week, months after planting, the Ubon paspalum was harvested for determination of both P content and biomass growth

Effect of water level: This experiment was to evaluate the ability of plant to adapt with different water levels It was done with kinds of water levels (2, 5, and cm) The water sampling frequency was the same as that of the experiment on pH, plant age, and density

c Trial application of the selected plant in CWs

23

Figure 2.1 Scheme of horizontal constructed wetland

(at the start of experiment)

Dimension and filter materials: The four CWs tanks were made of stainless steel, in rectangle shape and have the same dimension (L x W x H = 68.5 cm x 33 cm x 42 cm The efficient volume of each tank was 0.095 m3

24

Tank 1+ Tank Tank 3+ Tank Adsorption unit

Figure 2.2 The structure of filter media in CWs and adsorption units

Plants

The plant used in the CWs experiment was Ubon paspalum This plant was selected because of high biomass, fast growth, deep roots, good tolerance to flooding, high uptake of nutrient, high economic value To the best of our knowledge, this is the first time this plant has been used in CWs for phosphorous decontamination

The plant was grown in soil of a home garden It was cut off leaves to mitigate the water evaporation The stem with the height of 25 cm was put into a bucket of tap water for week in the cool place with daylight When the new leaves reached the height of 10 cm, the water in the bucket was replaced by the background nutrient solution and kept for another week for adaptation After that, the plants were transferred into CWs It took around weeks for CWs plants to stabilize and grow well

Wastewater characteristics

This experiment used the background nutrient solution, which simulated 25% real swine wastewater in terms of P and N contents (12.5 mg P/ L, and 125 mg N/ L) Every two days, waster sampling was implemented at the inlet and the outlet of all four CWs and after two adsorption units for P and pH measurement

25

The treatment systems were operated with hydraulic loading rate (HLR) of 0.032 and 0.027 m3/m2/d for the treatment systems 1&2 and the treatment system 3&4, respectively The hydraulic retention time (HRT) of the CWs, adsorption unit and adsorption unit was 5.4, 0.16 and 0.19 d, respectively The HRT of the hybrid treatment system (CW1 + adsorption unit 1) and the hybrid treatment system (CW3 + adsorption unit 2) was 5.5 and 5.59 d, respectively

2.2.3 Plant sample preparation and P analysis

Figure 2.3 Plant sample preparation and analysis

Plant sample preparation: First, plant samples were well rinsed with double distilled water, dried in the room with the air conditioner (27 0C and high fan speed) for 15 Next, the plant was cut into different parts (stem, leaves, and roots) and then measured the fresh weight Then, the cut plants were placed in the Thermo Scientific oven at 70 0C for 48 h, and cooled down to the room temperature After

that, the dried plant samples were measured to determine the dry weight The dried samples were ground into powder using the plant crusher and coffee grinder The powdered plants were kept in a tight glass bottle for P analysis

Plant sample digestion: The plant samples were digested according to Vietnam Standard TCVN 8551:2010 using heating digester (DK6) Accordingly, 0.5 g of the plant sample was mixed with 10 mL of the mixture of HClO4: HNO3 (1: in volume),

kept overnight and then digested in DK6 for h at different temperatures

26

Thermo Scientific

Oven Plant crusher

Kjeldahl Digestion Unit (DK6, Velp, EU)

UV-VIS Diode Array Spectrophotometer (S2100, Unico, USA)

SevenCompactTM pH meter (Mettler Toledo S220-Kit, China)

Peristaltic pump (HV-77200-50, Masterflex Cole-

Parmer, USA)

Double distilled water machine (A4000D, Bibby,

England)

Figure 2.4 Images of apparatus used in this study

2.2.4 Analysis of other water quality parameters

27

Table 2.2 Methods for examination of water quality parameters

Parameters Methods of analysis pH TCVN 6492:2011 DO TCVN 5499:1995 SS TCVN 6625:2000 BOD5 TCVN 6001:2008

COD TCVN 6491:1999 TP TCVN 6202:2008 N-NH4+ TCVN 5988:1995 2.3 Data calculation

Phosphorus removal efficiency by plants

Removal efficiency (%) = (C0 – Ci)/ C0 × 100%

C0: the nutrient concentration at the beginning of the experiment

Ci: the nutrient concentration on the day i with the plant growing

Phosphorus removal rate by plants

Removal rate (mg/ kg biomass/ d) = 𝑚𝑜−𝑚𝑖

𝑀 ×𝑑

mo: The mass of P in initial solution (mg)

mi: The mass of P in left solution (mg)

M: Biomass of plants (g)

d: Time for each turn of experiment (day) Biomass growth rate

Biomass growth rate (g/ d) = 𝑚𝑖−𝑚𝑜

𝑑

28

mi: The mass of harvested plants (g)

d: Time for experiment (day)

2.4 Data statistical analysis

29

CHAPTER 3: RESULTS AND DISCUSSION

3.1 Screening potential plants for use in the CWs

3.1.1 Selection of potential plants based on their P content and biomass growth

Figure 3.1 The P percentage of studied plants

The P removal capacity of plants depends on not only the P percentage in the plant but also the plant biomass (Greenway, 1997) Thus, this study conducted a comparative study with different plants on both their P content in plant and the biomass growth The experiment was performed with mature plants grown in the P-rich soil in Chuong My district, Ha Noi City In this study, P content was determined in the different parts of plants Based on that, the P percentage in the whole plant was calculated The obtained results were shown in Figure 3.1 The P percentage in the whole plants of different plants tudy was decreased in the following order Ubon

paspalum (0.49%) > Piper lolot (0.39%) > Cymbopogon citratus (0.36%) = Colocasia gigantean (0.36%) > Sauropus androgynous (0.23%) The P content in the

investigated plants in this study was in the range of 0.23-0.49% The results obtained in this study agree well with the finding of Mcjannet et al (1995) and Bodin (2013),

0 0.2 0.4 0.6

Sauropus androgynus

Cymbopogon citratus

Colocasia gigantean

Piper sarmentosum

Ubon paspalum

P

percentage

of

plants

30

who reported that the P content in emergent macrophyte plants varied in the range of 0.13-1.07% or 0.05-0.5% dry weight, respectively Specifically, the P content in investigated plants (0.23-0.49%) were higher than that in Brophytes, Helophytes,

Hydrophytes (0.1-0.3%) (Demars and Edwards, 2008) However, it was found to be

equivalent to that in water hyacinth and water lettuce (0.3-0.5%) (Bhole, 2013; Lu et al., 2010) In constrast, it was found to be lower than that in such emergent plants as

Eleocharis sphacelata, Baumea aniculata, Typha domingensis, and Cyperus involucratus from 1.9 to times Especially, in comparison with such floating plant

as duckweek, Ceratophyllum, Water hyacinth, Nymphoides indica, Aquatica,

Ludwigia peploides, Ceratopteris thalictroides, Marsilea, plants in this study

exhibited the significantly lower P content (3.3-7 times) (Greenway, 1997) The lower P content of the investigated plants compared with floating plants can be explained by the fact that the plants in this study were emergent plants, which were reported to be inferior to floating plants in P accumulation (Greenway, 1997) This remark was strongly supported by Green et al (2003), who reported that the P content in floating plants was higher than that in emergent plants (0.2-0.4%) It was worth noting that two emergent plants, namely Phragmites australis and Typha

31

Table 3.1 The P content in plants use for phytoremediation or CWs

Plants Parts of plants* P content (mg/g) P content

(%) Reference FLOATING PLANTS

Water lettuce W 0.3 Lu et al (2010)

Aquatica W 0.9 Greenway (1997)

Ipomoea diamantinensis W 10 Greenway (1997)

Ludwigia peploides W 10 Greenway (1997)

Ceratopteris thalictroides W 10 Greenway (1997)

Marsilea W 10 Greenway (1997)

Monochoria cyanea W 13 1.3 Greenway (1997)

Ceratophyllum W 14 1.4 Greenway (1997)

Nymphoides indica W 16 1.6 Greenway (1997)

Duckweek W 18 1.8 Greenway (1997)

EMERGENT

MACROPHYTE PLANTS

Reaumuria soongorica L 0.73 0.073 He et al (2015)

Brophytes W 0.1 Demars and

Edwards (2008)

Helophytes W 1.5 0.15 Demars and

Edwards (2008)

Phragmites australis W 0.2 Greenway et al (2003)

Typha domingensis W 2.3 0.23 Greenway et al (2003)

Schoenoplectus validus W 2.6 0.26 Greenway et al (2003)

Eleocharis sphacelata W 2.7 0.27 Greenway et al (2003)

Eleocharis acuta W 3.4 0.34 Greenway et al (2003)

Cymbopogon citratus W 3.59 0.359 THIS STUDY

Baumea articulata W 3.7 0.37 Greenway et al (2003)

Persicaria orientalis W 4.4 0.44 Greenway et al (2003)

32

The P distribution in plants may affect the harvest strategy of plants from CWs The P distribution in five investigated plants was shown in Figure 3.1 It was found that all investigated plants showed higher P content in the aboveground parts (leaf and stem) For Ubon paspalum, the P content in leaf and root was 4.4 and 3.0 mg/g For Cymbopogon citratus, the P content in leaf and root was 3.89 and 1.82 mg/g This can be attributed to high content of P in soil According to previous study, P tended to be concentrated in the root if the P centration in the soil was low In other words, in the soil rich in P, higher P percentage could be found in the leaf This feature favors the harvest strategy to collect aboveground part rather than whole plant This is especially significant since the investigated plants are mainly perennial This is in the contraty to the results obtained by Greenway et al (2003) indicating that for many types of emergent plants, higher P content was detected in the root/rhizome It is interesting that the list included Cuperus spp., Phragmites australis, and Typha sp., which were the common CWs plants Thus, the search for CWs plants, which both have high P content and distribution in the aboveground part is necessary

As mentioned above, the biomass growth is one of main factors influencing the P removal potential fo the plant The P removal potential of the studied plants were estimated based on P content and biomass growth The results were shown in Table 3.2 It can be seen that among investigated plants, the Ubon paspalum

33

Table 3.2 The P removal potential of the studied plants

Plant

P content in the whole plants (%

DW)

Biomass growth (ton FW/ha/year)

Biomass growth (ton DW/ha/year)

P removal potential (kg/ha/year)

Ubon paspalum 0.49 35 5.03 24.65

Cymbopogon citratus 0.36 27 6.05 21.80

Sauropus androgynus 0.23 28 7.39 17.00

Piper sarmentosum 0.39 17 3.00 11.69

Colocasia gigantean 0.36 18 1.40 5.04

3.1.2 Selection of potential CWs plants based on other growth characteristics

In addition to P content and biomass growth, the applicability of plants in CWs depends on other factors, such as flood tolerance, root system, long lasting, economic values, etc The information about growth characterists of five investigated plants were summarized in Table 3.3

34

Table 3.3 Growth characteristics of potential plants

Plant species Perennial plants

Harvest time

Tolerant to flood

Cropping season First

harvesting

Next harvesting Colocasia

gigantean  months 50 days  All year round Piper lolot  month 30 days  All year round

Sauropus

androgynus  months 25-30 days  All year round Cymbopogon

citratus  months 45-50 days  All year round Ubon paspalum  1,5 months 21 - 30 days  All year round

Cymbopogon citratus Piper lolot Sauropus androgynus

Colocasia gigantean Ubon paspalum

35

In short, taking into consideration of both P content, biomass growth, and other growth characteristics, among investigated plants, Ubon paspalum and

Cymbopogon citratus demonstrated the highest potential for use as CWs plants

However, Ubon paspalum has never been used in CWs for P decontamination That is the reason why it was selected to be futher investigated in the hybrid CWs and adsorption treatment system of this study

3.2 Factors influencing the growth and uptake of P of Cymbopogon

citratus and Ubon paspalum

3.2.1 Effect of initial P concentration on the growth and P uptake by plants

This experiment aims at evaluating the tolerance and P uptake capacity of two investigated plants Based on that, the optimum concentration of P in the simulated wastewater used as the influent for CWs will be proposed This experiment was carried out with three kinds of the initial P concentrations (12.5, 25, and 50 mg P/L), which are equivalent to those in the real swine wastewater with three levels of dilution (25, 50, and 100% real swine wastewater) For comparison purpose, this experiment was done in parallel with two plants (Colocasia gigantean and Ubon paspalum) The experiment lasted for days until the left P concentration in the solution was exhausted The left P concentration in the solution, the P removal rate, and P removal efficiency were utilized for assessment

The effect of initial P concentration on P removal efficiency of investigated plants was illustrated in Figures 3.3; 3.4 It is clear from Figure 3.3 that for

Cymbopogon citratus, the lower initial P concentration it was, the higher P removal

efficiency could be obtained After days, the P removal efficiency was 97.85, 89.85, and 64.13 % at the P initial concentration of 12.5, 25, and 50 mg/ L, respectively In contrast, the difference in P removal efficiency caused by the change in P initial concentration was negligible

36

mg/ L), the P removal efficiency (97.85 %) of Cymbopogon citratus was significantly higher than that of Ubon paspalum (64.20 %) The P removal efficiency in the constructed wetland is normally not high As can be seen from Table 3.2, it fluctuated in the range of 6-90 % It is worthy to note that, the extremely high P removal efficiency (90 and 85%) is usually achieved at very low P initial concentration (0.33 and 2.5 mg/L) (Jethwa and Bajpai (2016) In this study, although the initial P concentration (12.5-50 mg/L) was higher than that in previous studies (0.33-10 mg/L), the P removal efficiency was still high (64.13-97.85% for Cymbopogon

citratus, 41.29-64.20% for Ubon paspalum) This can be explained by the fact that

this study used ortho-phosphate as the nutrient source for the plants, which is recognized to be readily available to plants (Chen et al., 2017)

The Figure 3.3 & 3.4 also demonstrated that at the same initial P concentration, the P removal efficiency increased at the longer experiment period for both studied plants Specifically, at the initial P concentration of 12.5 mg/L, for Cymbopogon

citratus, the P removal efficiency increased by 83.12% when the experiment period

37

Figure 3.3 Effect of initial P concentration on the removal efficiency of Cymbopogon citratus

Figure 3.4 Effect of initial P concentration on the removal efficiency of Ubon paspalum

From Figures 3.5 and Figure 3.6, it is clear that the P concentration was decreased significantly with the cultivation time This is supported by Zheng et al

0 20 40 60 80 100

0 10 11

P

rem

ov

al

ef

ficie

nc

y

(%)

Samling date

Ci=12.5 mgP/L Ci=25 mgP/L Ci=50 mgP/L

0 20 40 60 80 100

0 10 11

P

rem

oval

ef

ficiency

(%)

Sampling date

38

(2013), who studied on P accumulation and removal from livestock wastewater using

Polygonum hydropiper It is obvious that the increase in P removal efficiency over

time was accompanied by the decline in the P concentration in the solution This trend is consistent with the finding by the Chen et al (2017)

Figure 3.5 P concentration left in solution plant with Ubon paspalum

Figure 3.6 P concentration left in solution planted with Cymbopogon citratus

0 10 20 30 40 50 60 70

0 10 11

P conc entrat ion in so luti on (m g P/ L ) Sampling date

Ci=12.5 mgP/L Ci=25 mgP/L Ci=50 mgP/L

0 10 20 30 40 50 60 70

0 10 11

P c onc entra ti on in sol uti on (mg P /L ) Sampling date

39

Table 3.4 The P removal efficiency by different plant species

Plant Media Wastewater

Initial P concentration (mg/ L) P removal efficiency (%) Reference Foxtail Grass, Flax Lily, Banksia, and Bottlebrush Pea gravel, sand, and loamy sand Synthetic storm water

4.51 (TP) 6-36 Jethwa and Bajpai (2016) P australis, T latifolia, P hydropiper, A sessilis, C esculenta and P stratoites

Gravel Domestic

wastewater 5.75 (PO4-P)

76 Jethwa

and Bajpai (2016) J effuses, Sand and clay Agricultural

runoff 2.5 (TP)

77

Jethwa and Bajpai

(2016) C lurida,

and 85

D acuminatum

74

Mixed 82

Typha latifolia and

Phragmites australis

Gravel Synthetic domestic wastewater

10 (PO4-P) 60

Jethwa and Bajpai (2016) Typha latifolia and Scirpus acutis

Gravel Municipal

wastewater 0.33 (PO4-P) 90

Jethwa and Bajpai

(2016)

40

trend occurred at the low and medium initial P concentrations However, at the high initial P concentration (50 mg/ L), the plants could not survive, thus affecting the P removal uptake of the plant The similar effect of initial P concentration on P removal rate was reported by Fu and He (2015)

Figure 3.7 Effect of intial P concentration on P removal rate of Ubon paspalum

Figure 3.8 Effect of intial P concentration on P removal rate of Cymbopogon citratus

-10 10 20 30 40

0 10 11

P

rem

oval

rate

(m

g

P/

kg/ d)

Sampling date

Ci=12.5 mgP/ L Ci=25 mgP/ L Ci=50 mgP/ L

0 10 20 30 40

0 10 11

P

rem

oval

rate

(m

g

P/

kg/ d)

Sampling date

41

Regarding the toxic P concentration, whilst Cymbopogon citratus could live with all initial P concentrations, most of Ubon paspalum died at the highest initial P concentration (50 mg/ L) This indicated that Ubon paspalum was more sensible to high P concentration This should be taken into consideration in the case of applying

Ubon paspalum in the constructed wetlands treatment system later The finding of

this study was in harmony with that of a previous study conducted by Haller and Sutton (1973), which released that the P concentration of 40 ppm (~40 mg/ L) posed a threat to Water hyacinth According to Abdolzadeh et al (2009), the P concentration higher than 10 µM (~ 0.31 mg/ L) severely affected the formation of root cluster of Lupinus atlanticus and Lupinus albus In this study, both Ubon

paspalum and Cymbopogon citratus grew well at much higher P concentrations (25

mg/ L) This suggests that various plant species may have different sensitivity to P level, thus may be subjected to P-toxicity or not

Figure 3.9 Ubon paspalum died at the highest P concentration

Figure 3.10 Cymbopogon citratus could adapt with a wide range of initial P

42

3.2.2 Effect of pH on the growth and P uptake by plants

pH plays a key role to the growth of plant because pH can affect the speciation, availability and toxicity of phosphorus, the growth of root system, the abundance and activity of microorganisms (Soti et al., 2015) Therefore, this experiment was carried out to determine the optimum pH for the growth as well as P uptake of Cymbopogon

citratus and Ubon paspalum This will provide useful information for preparing the

simulated wastewater for CWs In order to that, the experiment was performed at five pH values (3, 5, 7, 9, and 11) The experiment including turns lasted for 14 days The obtained results were shown in Figures 3.14 and Figure 3.15

The experiment at pH values of 3, 5, 7, the P removal efficiency, P removal rate and plant growth were recorded In the contrary, this experiment at pH values of 9&11, only growth of plant was recorded This is because of two reasons: (1) At pH values of 9&11, the white precipitate was formed when the pH was adjusted using NaOH and H2SO4 This was probably because in the strong alkaline medium,

phosphorus existed mainly in the form of PO43- rather than HPO42- or H2PO4- (Cerozi

and Fitzsim, 2019) Meanwhile, heavy metals, as the contaminant of NaOH and H2SO4 chemicals, existed in the form of metal hydroxides The reaction between

PO43- and metal ions resulted in precipitates This led to the reduction in the initial P

concentration of solutions used for phytoremediation experiments (2) pH values of 9&11 severely hindered the growth of Ubon paspalum

43

Figure 3.12 The death of Ubon paspalum at pH values of 9&11

Figure 3.13 Speciation of P in solution at various pH value

(Cerozi and Fitzsim, 2019)

The Figure 3.14 showed that the P removal efficiency of Ubon paspalum and

Cymbopogon citratus was most favored at pH of and 5, respectively In the pH

44

Figure 3.14 Effect of pH on P removal efficiency of Ubon paspalum and Cymbopogon citratus

Similarly, the P removal rate of Ubon paspalum and Cymbopogon citratus was highest at pH value of and 5, respectively For Ubon paspalum, the P removal rate was 10.96, 7.79, and 7.69 mg P/ kg biomass/ d at pH of 7, 3, and 5, respectively For

Cymbopogon citratus, the P removal rate was 11.56, 10.79, and 7.69 mg P/ kg

biomass/ d at pH of 5, 3, and 7, respectively

Considering the effect of pH on both P removal efficiency and P removal rate, it can be seen that the pH range of 5-7 was most suitable for both Ubon paspalum and

Cymbopogon citratus to grow and uptake phosphorus This agrees well with the

finding of Fitzsimmons (2019) that the solution pH of 5.5-7.2 was the best for the nutrient availability and plant uptake Similarly, Soti et al (2015) reported that the soil pH of 5.5–6.5 favored the growth and uptake of nutrient of Lygodium

microphyllum The optimum pH range for P uptake (5-7) can be attributed to the

formation of less precipitates At pH higher than 7, phosphate anions tended to react with calcium (Ca) and magnesium (Mg) cations to form insoluble compounds At pH

0 10 20 30 40 50

pH=3 pH=5 pH=7

P

rem

oval

ef

ficiency

(%)

45

lower than 5, phosphate anions reacted with aluminum (Al) and iron (Fe) cations to produce less soluble compounds (Haller and Sutton)

This study demonstrated that plants tend to adjust pH of the living medium to their favorite values It was observed that in the low pH medium (3-5), Ubon

paspalum seemed to have better adjustment ability than Cymbopogon citratus It is

evidenced that pH of the solution with Ubon paspalum was always higher than that with Cymbopogon citratus In the high pH medium (9-11), Cymbopogon citratus exhibited to be superior to Ubon paspalum in reducing from alkaline to neutral medium The buffer capacity of the plant can be explained by release of H+ or OH- as

plants balance their uptake of excess cations or anions (Mikkelsen, 2013)

Figure 3.15 Effect of pH on P removal rate of Ubon paspalum and

Cymbopogon citratus

3.2.3 Effect of plant age on its growth and P uptake

The plant age may affect the growth, uptake and thus accumulation of nutrient in the plant (Duran et al., 2008; Jie et al., 2019) Depending on each stage of growth, plants have different nutrient needs, thus affecting the nutrient uptake of plants Understanding the effect of plant age helps to determine (1) the suitable age of plant for transferring into CWs, (2) the best harvest time of the CWs plant In this study,

0 10 12 14

pH=3 pH=5 pH=7

P

rem

oval

rate

(m

g

P/

kg/ d

)

46

this experiment was conducted with Ubon paspalum in both aquaneous solutions and natural soil The experiments with hydroponic solutions were performed with plant ages (1, 1.5, and month) to determine P removal rate and P removal efficiency of the plant It lasted for 14 days with turns The experiment with natural soil was implemented with plant ages (1, 1.5, and months) to evaluate the biomass growth rate and P content in different parts of plants It has been conducted for months (from March to May, 2019) in a home garden in Thai Binh province

Regarding the hydroponic experiment, the results on P removal rate and P removal efficiency were shown in Figure 3.16 It was found that the younger the plant was, the greater the P removal rate was The highest P removal rate was obtained with the 1-month plant (52.87 mg P/ kg/ d), which was 4-fold higher than that achieved with 3-month plant (12.94 mg P/ kg/ d) This can be explained by the observation that for 1-month and 1.5-month plants, many new roots appeared In contrast, this did not happen with the 3-month plant (Figure 3.17) This is strongly supported by Mikkelsen (2013), who claimed that the growth of root systems was one of main factors influencing the P uptake From the view point of P decontamination, the young Ubon

paspalum should be used to transfer into CWs instead of old plant The P removal

47

Figure 3.16 Effect of plant age on P removal efficiency and P removal rate of Ubon paspalum (hydroponic experiment)

Figure 3.17 The effect of plant age on the growth of root system

Concerning the experiment with the natural soil, the biomass, growth rate and P content in the plant were investigated The relationship between the plant age and the biomass growth of Ubon paspalum was shown in Figure 3.18 It is clear that the older the plant was, the higher the biomass could be obtained Specifically, the average weight of a plant was increased from 35.34 to 2240 g when the age of plant elevated from to months

0 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60

1 1.5

P rem oval ef ficiency (%) P rem oval rate (m g P/ kg/ d)

48

Figure 3.18 Effect of plant age on the biomass growth of Ubon paspalum

(experiment with garden soil)

The influence of plant age on the biomass growth rate of Ubon paspalum was showed in Table 3.5 It is evident that the biomass growth rate of Ubon paspalum was highest at the age of month This is followed by the P removal rate achieved at the ages of 1.5 months (13.87 g/ d) and months ( 9.42 g/ d) The results implies that the in the period of 3-5 moths, Ubon paspalum grew the best Based on this, it is highly recommeded that the best harvest time should be months after planting

Table 3.5 Biomass growth rate of Ubon paspalum at different plant ages

Initial plant age (month)

Initial weight (g)

Harvest plant age (month)

Harvest weight (g)

Growth rate (g/d) 1-month 35 months 600 9.42 1.5-month 68 3.5 months 900 13.87

3-month 540 months 4200 61

0 500 1000 1500 2000 2500

1 1.5

Leave 1.44 3.42 150 412

Stem 12.5 20.93 200 930

Root 21.4 44.98 190 925

W

eig

ht of

plant (

g

)

Plant age (month)

49

3.2.4 Effect of plant density on its growth and P uptake

The total amount of removed P by a plant depends not only on the P content in the plant but also on the biomass growth (Schwammberger et al, 2019) However, biomass growth is reporeted to be affected by the plant density (Jethwa and Bajpai, 2016) This experiment was designed to identify the optimum plant density for the best growth and P uptake by Ubon paspalum and Cymbopogon citratus The experiments with hydroponic solutions were conducted with kinds of plant densities (1, 3, and plants per beaker) for both Ubon paspalum and Cymbopogon citratus It was lasted for 21 days (7 turns) Every days, hydroponic solutions were changed and wastewater samples were collected for P analysis to evaluate the P removal rate and P removal efficiency The experiments with soil was performed with categories of plant densities (1, 2, 3, and plants per trough – 0.5 m2) for only Ubon paspalum

It lasted for months (March to May 2019) At the start and the end of experiment, the biomass was measured to evaluate the biomass growth rate and P content in different parts of plant

Relating the hydroponic experiments, the effect of plant density on the P removal rate and P removal efficiency of the investigated plants were shown in Figures 3.19 & 3.20 It was found that the higher the plant density was, the slower the P removal rate was oberverd for both Ubon paspalum and Cymbopogon citratus Specifically, the P removal rate of Cymbopogon citratus was decreased from 19.63 to 11.22 mg P/ kg biomass/ d when the plant density increased from to plant/ beaker The higher P removal rate at lower plant density can be attributed to less competition between plants when there were fewer plants in the beaker The similar results were released when Webb et al (2013) Accordingly, the P removal rate of

Salicornia europaea at high and low plant densities were found to be 19.96 26.09

50

nutrient demand was higher for a greater number of the plant in the beaker The same trend happened to Ubon paspalum However, the reduction in the P removal efficiency as the result of increase in plant density for Cymbopogon citratus (28.61%) was higher than that for Ubon paspalum (15.84%) This indicates that more attention should be paid to the plant desnity if Cymbopogon citratus was planted in the CWs

Figure 3.19 Effect of plant density on P removal rate of the investigated plants

Figure 3.20 Effect of plant density on the P removal efficiency

of the investigated plants

0 10 15 20 25

1 plant plants plants

P rem oval rate (m g P/ kg/ d)

Plant density (plant/ beaker)

Ubon paspalum Cymbopogon citratus

0 10 20 30 40 50 60

1 plant plants plants

P rem oval ef ficiency (%)

Plant density (plant/ beaker)

51

Concerning the soil tests with Ubon paspalum, the results on the biomass and biomass growth rate were summarized in Table 3.6 The biomass growth rate of

Ubon paspalum was decreased from 10.12 to 3.65 g/d when the plant density

increased from 10 to 40 plant/ m2 In contrast, the higher accumulative harvested

biomass was obtained at the higher plant density from 10 to 30 plants/ m2 The

biomass growth was increased from 6070 to 10290 g when the plant density elevated from 10 to 30 plants/ m2 The highest biomass (10290 g) was obtained at the plant

density of 30 plant/m2 At the plant density of 10, 20 or 40 plant/m2, the biomass was

low (6070, 7735 and 8748 g) compared with the max biomass Thus, these plant densities should not be used Taking into account both biomass and biomass growth rate, the best plant density was 30 plant/m2 This finding was validated by the

observation that the root system of Ubon paspalum was longest (up to 0.8 m) at this plant density (Figure 3.21)

Table 3.6 Effect plant density on the biomass growth rate of Ubon paspalum

Density (plant/ m2)

Initial biomass

(g)

Accumulative harvested biomass

(g)

Biomass growth (g)

Biomass growth rate

52 Density

(1 plants/ 0.1 m2)

Density (1 plants/ 0.1 m2)

Density (3 plants/ 0.1 m2)

Density (5 plants/ 0.125 m2) Figure 3.21 The root growth of Ubon paspalum at different plant densities

3.2.5 Effect of water level on its growth and P uptake

Water levels significantly affected plant growth rate and nutrient removal capacity (Ngo et al, 2018) Study the effect of flooding on the rate of P uptake and plant growth to determine the appropriate water level in CWs construction This experiment was conducted with three water levels (2, 5, and cm in 500 mL beaker) The experiment results were illustrated in Figure 3.22

It can be seen from Figure 3.22 that the P removal rate was highest at the water level of cm, which was 17.74 and 10.32 mg P/kg biomass/d for both Cymbopogon

citratus and Ubon paspalum, respectively The P removal rate was lowest at the water

level of cm This is probably because the flooding significantly affected the growth of plants, thus reducing the P uptake ability of the plants The P removal rate at the water level of cm was lower than that at the water level of cm This is because the less P content could be found in the P solution with lower volume This study indicates that the superior P removal rate could be obtained with lower water levels Similarly, Ngo et al (2013) found that Typha orientaliscould grow and uptake P better at lower water levels In the contrary, the better P removal rate of Spirodela

53

than emergent plants Therefore, higher water depths could not hamper its growth and P uptake

Figure 3.22 Effect of water level on P removal rate of plants

The removal P rate of Cymbopogon citratus varied from 11.58 to 17.74 mg/kg/d, which was 1.57-1.88 times higher than that of Ubon paspalum The result exhibited that Cymbopogon citratus could survive at all water levels In contrast,

Ubon paspalum was dramatically affected by water level of cm For two out of five

turns, the stem of plants turned black and the ruined At the water level of cm, new roots did not appear for all the experiment period In contrast, the growth of new roots was enhanced at two lower water level, especially at the water level of cm The effect of water level on the root growth was quite similar to that on P removal rate These results indicated that the water level of cm was the optimal for both investigated plants and Ubon paspalum was more sensitive to flooding than

Cymbopogon citratus

0 12 16 20

2

P

rem

oval

rate

(m

g

P/

kg/ d

)

Water level (cm)

54

Figure 3.23 Effect of water level on root growth of Ubon paspalum 3.3 Applicability of the selected plants in the Constructed Wetland

Figure 3.24 The change of phosphorus in the effluent over the time

This experiment was conducted to assess the role of removing P of plants in CWs tanks for Ubon paspalum and tanks for control were established Besides, the role of CW unit and adsoption unit is also assessed

The plants before being tranfered to the system is cut and leave 25 cm of stem with a volume of 400 g per tank In the beginning, the plants grew slowly However, after weeks of adapting the tree the growth rate of the plant was faster, the new

0

10-May 15-May 20-May 25-May 30-May 4-Jun 9-Jun

P

c

onc

entr

ation

(mg

P/

L

)

Date of sampling

Tank2 Tank1 Tank3

55

branch appeared and many new roots, the height of the plants from 25 cm initially increased to m

Figure 3.25 P removal efficiency and pH after treatment of HFCWs

The difference in the efficiency of treatment performance between the planted CWs system and without plant CWs has obtained some initial results For CWs with substrate as WHC, the role of the plants occupies 1.9% while the role of the plants in CWs with yellow sand as substarte was 3.5% The role of plants in the yellow sand system is slightly higher than that of a WHC substrate The P removal efficiency of CWs (containing WHC) is lower than that of the sand system, which may be related to differences in Ph in these two systems pH in tank - WHC (8.1) is higher than tank –yellow sand (7.7) According to previous studies, pH = 5.5 - is suitable for plant growth and absorption In addition to, the difference is not much Because the system's operating time is quite short and plants has not enough time to adapt When the CWs tank with sand is soon saturated, the difference of the role of plants in system CWs will be clearer Curently, the contribution of plants in the P removal efficiency is very small It need to takes longer observation time

0 10 12 14 20 40 60 80 100 120

Tank Tank Tank Tank Adsorption Adsorption pH P rem oval ef ficiency (%)

56

After 29 days of operation, the treatment efficiency of the system reaches 95-98.7% The removal efficiency of P for WHC system (98%) is 21% higher than the average yellow sand system

Summary, the removal efficiency of P of the system still follows the good trend in WHC while becoming more saturated in the sand system To improve the effectiveness of the plant's P removal, pay attention to selecting the young plants and putting them in the background soulution after week before transferring them to the

The first days of experiment After two weeks operation of experimetn

57

CHAPTER 4: CONCLUSION AND RECOMMENDATION

4.1 Conclusion

• Among investigated plants, Cymbopogon citratus and Ubon paspalum were most potential for use as CWs plants, because of high biomass, fast growth, high P uptake, good tolerance to flooding and perennial and uncommon CWs plants

• Concerning influential factors, Cymbopogon citratus grew the best at pH and was tolerant with high P concentration (up to 50 mg/L) and water level (8 cm for week) Ubon paspalum grew the best at pH 7, was less tolerant with high P concentration (25 mg/L) and water level (5 cm for week) than

Cymbopogon citratus The younger the Ubon paspalum was, the higher the P

removal rate was The best harvest time for Ubon paspalum was months after planting

• In the WHC-based CWs, Ubon paspalum adapted well In the start-up period, this plant contributed to 1.9 – 3.5 % to the P removal of CWs

4.2 Recommendations

• This study was implemented in a short period of time, which was inadequate for the complete assessment the P accumulation in the plant A longer study with the same plant should be done in the future

• This study was conducted with the simulated wastewater containing only P It is highly recommended to conduct further research with the real wastewater and multiple pollutants

58

REFERENCE

Almuktar, S A., Abed, S N., & Scholz, M (2018) Wetlands for wastewater treatment and subsequent recycling of treated effluent: a review Environmental

Science and Pollution Research, 25(24), 23595-23623

Bhandari, A., Surampalli, R Y., Champagne, P., Ong, S K., Tyagi, R D., & Lo, I M (2007) Remediation technologies for soils and groundwater Reston, Virginia: ASCE

Bodin, H (2013) Wastewater treatment in constructed wetlands: effects of

vegetation, hydraulics and data analysis methods (Doctoral dissertation, Linköping

University Electronic Press)

Bunce, J T., Ndam, E., Ofiteru, I D., Moore, A., & Graham, D W (2018) A review of phosphorus removal technologies and their applicability to small-scale domestic wastewater treatment systems Frontiers in Environmental Science, 6,

Chandekar, N., & Godboley, B J (2017) A review on phytoremediation a sustainable solution for treatment of kitchen wastewater Int J Sci Res (IJSR), 6(2), 1850-1855

Chen, C., Zhao, T., Liu, R., & Luo, L (2017) Performance of five plant species in removal of nitrogen and phosphorus from an experimental phytoremediation system in the Ningxia irrigation area Environmental monitoring

and assessment, 189(10), 497

Cunningham, S D., & Ow, D W (1996) Promises and prospects of phytoremediation Plant physiology, 110(3), 715

da Silva Cerozi, B., & Fitzsimmons, K (2016) The effect of pH on phosphorus availability and speciation in an aquaponics nutrient solution Bioresource

technology, 219, 778-781

Davis, L (1995) A guide to creating wetlands for agricultural wastewater, domestic wastewater, coal mine drainage and stormwater in the Mid-Atlantic Region 1, general considerations USDA-NRCS, US EPA-Region III, and PA Department of

Environmental Resources

59

Fu, X., & He, X (2015, April) Nitrogen and phosphorus removal from contaminated water by five aquatic plants In 2015 International Conference on

Mechatronics, Electronic, Industrial and Control Engineering (MEIC-15) Atlantis

Press

Nguyễn Sáng (2016) Nghiên cứu xử lý nước thải chăn nuoi phương pháp sinh học kết hợp lọc màng Luận án tiến sĩ, Đại học Khoa học Tự nhiên, ĐHQGHN Gao, J., Song, Z., & Liu, Y (2019) Response mechanisms of leaf nutrients of endangered plant (Acer catalpifolium) to environmental factors varied at different growth stages Global Ecology and Conservation, 17, e00521

Greenway, M (1997) Nutrient content of wetland plants in constructed wetlands receiving municipal effluent in tropical Australia Water Science and

Technology, 35(5), 135-142

Greenway, M (2003) Suitability of macrophytes for nutrient removal from surface flow constructed wetlands receiving secondary treated sewage effluent in Queensland, Australia Water Science and Technology, 48(2), 121-128

Haller, W T., & Sutton, D L (1973) Effect of pH and high phosphorus concentrations on growth of water hyacinth Hyacinth Control Journal, 11, 59-61

Jenkins, D., & Hermanowicz, S W (1991) Principles of chemical phosphate removal IN: Phosphorous and Nitrogen Removal from Municipal Wastewater:

Principles and Practice nd ed Lewis Publishers, Boca Raton, Florida 1991 p 91-110 15 fig, tab, 23 ref

Jethwa, K., & Bajpai, S (2016) Role of plants in constructed wetlands (CWS): A review J Chem Pharm Sci, 2, 4-10

Jesus, J M., Danko, A S., Fiúza, A., & Borges, M T (2018) Effect of plants in constructed wetlands for organic carbon and nutrient removal: a review of experimental factors contributing to higher impact and suggestions for future guidelines Environmental Science and Pollution Research, 25(5), 4149-4164

Kamran, M A., Mufti, R., Mubariz, N., Syed, J H., Bano, A., Javed, M T., & Chaudhary, H J (2014) The potential of the flora from different regions of Pakistan in phytoremediation: a review Environmental Science and Pollution

Research, 21(2), 801-812

Pasek, M (2019) A role for phosphorus redox in emerging and modern biochemistry Current opinion in chemical biology, 49, 53-58

60

phytoremediation of treated urban wastewater with Cyperus alternifolius L and Typha latifolia L in the West of Sicily (Italy) Ecological engineering, 61, 282-291 Lou, Y., Lu, X., Wang, G., & Jiang, M (2012) Nitrogen and phosphorus contents of wetland plant tissues in sanjiang plain, northeast china Fresenius

Environmental Bulletin, 21(2 A), 503-508

Qin, R., & Chen, H (2016, October) The procession of constructed wetland removal mechanism of pollutants In 2016 4th International Conference on

Mechanical Materials and Manufacturing Engineering Atlantis Press

Rezakhani, L., Motesharezadeh, B., Tehrani, M M., Etesami, H., & Hosseini, H M (2019) Phosphate–solubilizing bacteria and silicon synergistically augment phosphorus (P) uptake by wheat (Triticum aestivum L.) plant fertilized with soluble or insoluble P source Ecotoxicology and environmental safety, 173, 504-513

Shane, M W., McCully, M E., & Lambers, H (2004) Tissue and cellular phosphorus storage during development of phosphorus toxicity in Hakea prostrata (Proteaceae) Journal of Experimental Botany, 55(399), 1033-1044

Schwammberger, P F., Lucke, T., Walker, C., & Trueman, S J (2019) Nutrient uptake by constructed floating wetland plants during the construction phase of an urban residential development Science of The Total Environment, 677, 390-403

Soti, P G., Jayachandran, K., Koptur, S., & Volin, J C (2015) Effect of soil pH on growth, nutrient uptake, and mycorrhizal colonization in exotic invasive Lygodium microphyllum Plant ecology, 216(7), 989-998

Wu, H., Zhang, J., Ngo, H H., Guo, W., Hu, Z., Liang, S., & Liu, H (2015) A review on the sustainability of constructed wetlands for wastewater treatment: design and operation Bioresource technology, 175, 594-601

Xu, J., & Shen, G (2011) Effects of harvest regime and water depth on nutrient recovery from swine wastewater by growing Spirodela oligorrhiza Water

Environment Research, 83(11), 2049-2056

Yan, T., Zhu, J., Yang, K., Yu, L., & Zhang, J (2017) Nutrient removal under different harvesting scenarios for larch plantations in northeast China: Implications for nutrient conservation and management Forest Ecology and Management, 400, 150-158

61

Zheng, Z C., Li, T X., Zeng, F F., Zhang, X Z., Yu, H Y., Wang, Y D., & Liu, T (2013) Accumulation characteristics of and removal of nitrogen and phosphorus from livestock wastewater by Polygonum hydropiper Agricultural water

management, 117, 19-25

Zhou, Q., Zhu, H., Bañuelos, G., Yan, B., Liang, Y., Yu, X., & Chen, L (2017) Effects of vegetation and temperature on nutrient removal and microbiology in horizontal subsurface flow constructed wetlands for treatment of domestic sewage Water, Air, & Soil Pollution, 228(3), 95

62 APPENDIX

a) CW system in Koka Shiga

b) CW system in Mombetsu, Hokkaido

c) Material filter used in CWs

d) CW system in Higashikagura, Hokkaido

e) Doing experiment in Ritsumeikan University

Appendix 1: Visiting CWs and doing experiment during internship in Japan

63

a) Swine wastewater sampling b) Ubon paspalum sampling

c) Synthetic wastewater sampling d) Ubon paspalum in home garden

Appendix 2: Sampling and doing experiment in Vietnam