constructed wetlands for urban stormwater nutrient
TRANSCRIPT
Constructed wetlands for urban stormwater nutrient
attenuation: Diurnal to decadal dynamics
Tanveer Mehedi Adyel
B.Sc. (Hons.) in Environmental Sciences
M.S. in Environmental Sciences
This thesis is presented in fulfilment of the requirements for the degree of
Doctor of Philosophy of The University of Western Australia
School of Civil, Environmental and Mining Engineering
&
UWA School of Agriculture and Environment
May 2017
ii
[This page has left blank intentionally]
iii
Abstract
Stormwater runoff supplies organic and inorganic nutrients that have been
mobilised and redistributed from point and non-point sources within urban
catchments. Stormwater nutrient loads create challenges for the downstream receiving
waters, leading to water quality impairment, eutrophication and dissolved oxygen
(DO) depletion, and a reduction in recreational amenity. Therefore, strategies for
stormwater nutrient attenuation are necessary to maintain the ecological health and
amenities of such downstream waters. Constructed wetlands (CWs) as Water
Sensitive Urban Design options are widely used for stormwater nutrient attenuation
owing to their economic and environmental benefits. The surface flow (SF) and
subsurface flow (SSF) CWs are two of the most common types of CW. However,
optimisation of CW function remains challenging, particularly under Mediterranean
climates. CWs under Mediterranean climates receive nutrient-rich episodic
stormwater pulses during the wet winters. The same CWs experience prolonged low
to near-zero flow conditions and macrophyte senescence during the dry summers.
Furthermore, although CWs act as lentic systems during summer low flows, episodic
storm events during the winter convert the CWs into lotic systems, and frequent
lentic-lotic transitions make optimization of CW function difficult and prediction of
their nutrient reduction challenging.
This research aimed to assess stormwater nutrient attenuation in two CWs on
the Swan Coastal Plain, Western Australia, under different hydrological,
biogeochemical and climatic conditions. The first CW, a meandering SF CW, was
constructed in 2004 with a restoration effort after six years of operation. The second
CW, a hybrid system consisting of multiple alternating SF and laterite-based SSF
compartments, started operation in 2009. Baumea articulata and Schoenoplectus
validus were the dominant macrophytes in both CWs. The research adopted a holistic
approach to investigate nutrient attenuation at diurnal, event, seasonal, annual and
decadal scale through a synthesis of meteorological, water quality and quantity,
sediment quality and macrophyte data.
Nutrient species, except filtered total oxidised nitrogen (NOx), attenuated
immediately after the SF CW restoration. However, NOx attenuation started to
increase significantly three years after the restoration. Over decadal scales, attenuation
of inorganic and dissolved nitrogen, such as NOx and ammonia (NH3) was higher than
the attenuation of total nitrogen (TN) and dissolved organic nitrogen (DON). Three
iv
years after the restoration, the SF CW was consistently hypoxic, and started to release
phosphorus. The hybrid CW showed higher nutrient attenuation than the SF CW.
Although the SF and SSF compartments of the hybrid CW periodically released DON
and NOx, respectively, these nutrients attenuated in the subsequent alternating
compartments. A potential attenuation pattern between NH3 (highest under oxic
conditions) and NOx (highest under anoxic conditions) was observed in both CWs.
The sediments were a larger nutrient sink than macrophytes in both CWs. In addition,
during the summer low flows, senescent and floating macrophytes shaded the water
column and caused persistent low DO concentrations in the CW, and subsequently
suppressed nutrient attenuation. DO levels increased in the CWs after the removal of
the floating and senescent macrophytes.
A distinct diurnal DO signature was observed in the SF CW, with anoxia at
night, while groundwater was hypoxic throughout the year. The research developed a
low-cost proxy “metabolism” indicator using the diurnal change in DO. This proxy
“metabolism” was correlated with available solar radiation, water temperature, CW
volume and antecedent dry condition, and most importantly with the nutrient
attenuation capacity of the SF CW. This research specifically targeted systems under a
Mediterranean climate that experience large variability in hydrology, compared to
other regions of the world. This proxy indicator could be used to indicate CW function
when resources constrain the intensity of routine sampling. CWs also provide other
potential services and benefits for ensuring urban liveability and resilience.
v
Statement of originality
The research presented in this thesis is an original contribution to the field of
constructed wetland for urban stormwater management. All materials presented here
are original except where due acknowledgment is given, and has not been accepted for
any other degree or diploma. Besides, it has been substantially completed during my
enrolment at The University of Western Australia.
The body of this thesis (Chapters 3 to 5) is presented as a series of self-
contained papers intended for journal publication, and some repetition of the literature
review, study site details and methodology has therefore been necessary. I have been
solely responsible for all data analysis and figures contained in this thesis, and have
led all of the writing of the text with feedback from my supervisors.
vi
Acknowledgements
My sincere thanks to my supervisors, Carolyn Oldham and Matthew Hipsey,
for their guidance, truthful criticism and sensible advice throughout my doctoral
journey. My thanks to Ana Deletic (Monash University) for her deep interest on the
progress of this research. A very big thanks to Peter Adkins and Christie Atkinson
(Department of Parks and Wildlife, WA) for taking care of the field sites over decade
and allowing me to work in these sites and their continuous assisting with historical
data, information and comments on writing. Thanks to Ryan Vogwill and Nicki
Mitchell for their instrumental support, and Anas Ghadouani and Jason Beringer for
making the official procedure smooth.
Thanks also to Ana Ruibal-Conti, Benya Wang, Carlos Ocampo, Gayan
Gunaratne and Jana Coletti for their assistance during the fieldwork, experiment and
data analysis. Special thanks to Hasnein Tareque for his support in GIS and
Mahmudur Rahman for giving me his inflatable boat during a field trip. My sincere
acknowledgement also to Taj Sarker for his help during fieldwork, particularly at
night in the cold, wind and rain. Thanks to Krish Seewraj (Department of Water, WA)
for his comments on early version of different chapters. Special thanks to Joanne
Edmondston for her constructive comments on academic writing during the doctoral
journey. My thanks to Chris Brouwer, Dave and James Hehre for their efforts and help
at field and lab.
Thanks go to my eldest brother Ferdous Sohel, for his deep interest in my
research. My sister-in-law, nephew, and niece have made my Ph.D. journey easy.
I acknowledge the help of the Department of Agriculture and Food, WA;
Department of Water, WA; the Department of Parks and Wildlife, WA; Water
Corporation, WA; City of Canning; the Bureau of Meteorology and the South East
Regional Centre for Urban Landcare (SERCUL) for their assistance with data and
information.
This study was funded by Cooperative Research Centre for Water Sensitive
Cities under Project C4.1 “Multi-functional urban water systems” and Project B2.4
“Hydrology and nutrient transport processes in groundwater/surface water systems”.
I was financially supported by UWA with Scholarship for International Research
Fees, University International Stipend, Safety Net Top-Up Scholarship and Ad Hoc
Scholarship.
vii
Dedication
To my beloved parents
viii
Potential publications arising from this thesis
Peer reviewed journal papers:
1. Tanveer M. Adyel, Matthew R. Hipsey and Carolyn E. Oldham (2017) Temporal
dynamics of stormwater nutrient attenuation by an urban constructed wetland
experiencing summer low flows and macrophytes senescence. Ecological
Engineering, 102, 641-661; [Chapter 3].
2. Tanveer M. Adyel, Carolyn E. Oldham and Matthew R. Hipsey (2016)
Stormwater nutrient attenuation in a constructed wetland with alternating surface
and subsurface flow pathways: event to annual dynamics. Water Research, 107,
66-82; [Chapter 4].
3. Tanveer M. Adyel, Matthew R. Hipsey and Carolyn E. Oldham (2017). Storm
event-scale nutrient attenuation in constructed wetlands experiencing a
Mediterranean climate: A comparison of surface flow and hybrid surface-
subsurface flow system. Science of the Total Environment, 598, 1001-1014;
[Chapter 5].
Technical reports:
4. Tanveer M. Adyel, Carlos Ocampo, Hasnein Tareque, Carolyn Oldham and
Matthew R. Hipsey (2015) Performance assessment of Wharf St Constructed
Wetland 2009-2014. Cooperative Research Centre for Water Sensitive Cities,
Australia, pp. 1-74.
5. Ana Ruibal-Conti, Carlos Ocampo, Tanveer M. Adyel, Matthew R. Hipsey and
Carolyn Oldham (2015) Performance assessment the of Anvil Way Compensation
Basin living stream: 2004-2013. Cooperative Research Centre for Water Sensitive
Cities, Australia, pp. 1-83.
Conference papers, abstracts and posters:
6. Tanveer M. Adyel, Matthew R. Hipsey, Carolyn E. Oldham (2017) Stormwater
nutrient dynamics during riparian zone saturation and lentic-lotic transition. 7th
International Multidisciplinary Conference on Hydrology and Ecology (HydroEco
2017), to be held in Birmingham, UK from 18-23 June 2017 (accepted).
7. Tanveer M. Adyel, Matthew R. Hipsey and Carolyn Oldham (2016) Comparative
assessment of nutrient attenuation in constructed wetlands during episodic
ix
stormwater pulses. 10th
INTECOL International Wetlands Conference, 19-24
September 2016, Changshu, China, pp. 71-72.
8. Tanveer M. Adyel, Matthew R. Hipsey and Carolyn Oldham (2016) Constructed
wetlands to support urban waterway protection from stormwater pollutants.
Stormwater Conference, 29 August-03 September 2016, Gold Coast, Qld,
Australia, pp. 1-8.
9. Tanveer M. Adyel, Matthew R. Hipsey and Carolyn Oldham (2015) Multi-
functional and multi-compartment constructed wetland for urban waterway
restoration. Singapore International Water Week, 10-16 July 2016, Singapore,
pp.1- 6.
10. Tanveer M. Adyel, Ana L. R. Conti, Carlos Ocampo, Jana Z. Coletti, Matthew
Hipsey and Carolyn Oldham (2015) Ecohydrology and stormwater nutrient
attenuation performance of constructed wetland in Western Australia. 34th
IAHR
World Congress, 28 June-03 July 2015, The Hague, the Netherlands. pp. 1-11.
11. Tanveer M. Adyel, Hasnein Tareque, Matthew R. Hipsey and Carolyn Oldham
(2015) A multi-functional and multi-compartment constructed wetland to support
urban waterway restoration. Poster presented at 2nd
Water Sensitive Cities
Conference, 8-9 September 2015, Brisbane, Australia.
12. Tanveer M. Adyel, Matthew Hipsey and Carolyn Oldham (2015) Stormwater
nutrient attenuation by constructed wetlands on the Swan Coastal Plain,
Hydropolis Conference, 21-22 April 2015, Perth, Australia.
13. Tanveer M. Adyel, Ana Ruibal, Carolyn Oldham and Matthew Hipsey (2014)
Stormwater nutrient attenuation by Anvil Way Compensation Basin. Poster
presented at 1st Water Sensitive Cities Conference, October 10 2014, Melbourne,
Australia.
x
List of Acronyms and Variables
General:
ADD: Antecedent dry day
AGB: Above ground biomass
AHD: Australian Height Datum
ANAMMOX: Anaerobic ammonium oxidation
ANZECC: Australian and New Zealand Environment and
Conservation Council
AWCB: Anvil Way Compensation Basin
BCD: Bebington Court Drain
BGB: Below ground biomass
BMPs: Best Management Practices
BOD: Biochemical oxygen demand
BoM: Bureau of Meteorology
CANON: Completely autotrophic nitrate removal over
the nitrate
CE: Cation exchange
COD: Chemical oxygen demand
CR: Community respiration
CW: Constructed wetland
DAFWA: Department of Agriculture and Food
DEM: Digital Elevation Model
DNRA: Dissimilatory nitrate reduction
DO: Dissolved oxygen
DoW: Department of Water
DWC: Dry weather concentration
EC: Electric conductivity
Eh: Oxidation-reduction potential
FF: First flush
FFM: Floating macrophytes
GPP: Gross primary production
HF: Horizontal flow
HRAP: Healthy River Action Plan
HRT: Hydraulic retention time
HSSF: Horizontal subsurface flow
ISWMS: Integrated Stormwater Management System
LID: Low Impact Design or Development
MSD: Mars Street Drain
MSMD: Mills Street Main Drain
MUSIC: Model for Urban Stormwater Improvement
Conceptualisation
NEP: Net ecosystem production
OL: Ornamental lake
PND: Partial nitrification and denitrification
R: Respiration
RCM: Recycled concrete material
SCP: Swan Coastal Plain
SDC: Standardised delta concentration
SDC_ave: Standardised delta concentration_averaged
SERCUL: South East Regional Centre for Urban
Landcare
SF: Surface flow
SOD: Sediment oxygen demand
Element, chemical, nutrient:
Al: Aluminium
Al3(OH)3(PO4)2·5H2O: Avellite
Al(PO4)·2H2O: Variscite
C: Carbon
Ca: Calcium
CaCO3: Calcite
Ca5(Cl,F)(PO4)3: Apatite
Ca5(OH)(PO4)3: Hydroxylapatite
CH4: Methane
CO2: Carbon dioxide
DIC: Dissolved inorganic carbon
DIN: Dissolved inorganic nitrogen
DOC: Dissolved organic carbon
DON: Dissolved organic nitrogen
DOP: Dissolved organic phosphorous
Fe: Iron
Fe(PO4)·2H2O: Strengite
Fe3(PO4)2·8H2O: Vivianite
FRP: Filtered reactive phosphorus
FTN: Filtered total nitrogen
FTP: Filtered total phosphorus
Mg: Magnesium
Mn: Manganese
N: Nitrogen
N2: Nitrogen gas
NH3: Filtered ammonia
NO: Nitric oxide or nitrogen monoxide
N2O: Nitrous oxide
NO2: Filtered nitrite
NO3: Filtered nitrate
NOx: Filtered total oxidised nitrogen
NUON: Filtered non-urea organic
nitrogen
P: Phosphorus
PH3: Phosphine gas
PN: Particulate nitrogen
PP: Particulate phosphorus
S: Sulphur
SS: Suspended solid
TDC: Total dissolved carbon
TDS: Total dissolved solid
TKN: Total Kjeldahl nitrogen
TN: Unfiltered total nitrogen
TOC: Total organic carbon
TP: Unfiltered total phosphorus
TSS: Total suspended solid
UreaN: Filtered urea nitrogen
xi
SSF: Subsurface flow
SUDS: Sustainable Urban Drainage Systems
VCUA: Volumetric contribution from ungauged area
VF: vertical flow
VSSF: Vertical subsurface flow
WA: Western Australia
WSC: Water Sensitive Cities
WSCW: Wharf Street Constructed Wetland
WSMD: Wharf Street Main Drain
WSUD: Water Sensitive Urban Design
Numerical:
A: Active surface of the catchment that effectively contributed to runoff (ha)
Ab: Biomass area (m2)
As: Sediment area (m2)
Bs: Sediment bulk density (kg/m3)
C*: Average nutrient concentration at the inlet and MSD (mg/L)
Ci: Nutrient concentration at ith timestep (mg/L)
Cinlet: Nutrient concentration at the inlet (mg/L)
Coutlet: Nutrient concentration at the outlet (mg/L)
DOmax: Maximum DO concentration (mg/L) or saturation (%) during the afternoon
𝐷𝑂𝑚𝑎𝑥𝑖: Average maximum DO saturation per day during the afternoon before the storm
perturbation over four consecutive days (%)
DOmin: Minimum DO concentration (mg/L) or saturation (%) during the morning
𝐷𝑂𝑚𝑖𝑛𝑖: Average minimum DO saturation per day during the morning before the storm perturbation
over four consecutive days (%)
Ds: Sediment depth (m)
DWb: Dry weight of the biomass (kg)
EC: Electric conductivity (μS/cm)
EF: Event flux (kg/ha)
EMC: Event mean concentration (mg/L)
L: Nutrient load (mg/s)
LA: Nutrient load attenuation (%)
Li: Load at the inlet (mg/s)
Lo: Load at the outlet (mg/s)
NCb: Biomass nutrient concentration (mg/kg)
NCs: Sediment nutrient concentration (mg/kg)
NMb: Biomass nutrient mass (kg)
NMs: Sediment nutrient mass (kg)
Qb: Quadrat area as 0.0625 m2
Qi: Flow rate at ith timestep (m3/s)
Qoutflow: Discharge rate at the outflow (m3/s)
SMC: Site mean concentration (mg/L)
V: Volume of the AWCB (m3)
VE: Total run-off volume across an event (m3)
Vi: Volume proportional to flow rate at ith timestep (m3)
ΔDO: Delta dissolved oxygen (mg/L or %)
ΔDOsat: Delta dissolved oxygen saturation (%)
∆𝐷𝑂̅̅ ̅̅ ̅̅𝑝𝑟𝑒: Average metabolism over four consecutive days before the storm perturbation (%)
∆𝐷𝑂̅̅ ̅̅ ̅̅𝑝𝑜𝑠𝑡: Average metabolism over four consecutive days after the storm perturbation (%)
∆(∆𝐷𝑂̅̅ ̅̅ ̅̅ ): Average change in metabolism due to a storm perturbation (%)
xii
Table of Contents
Abstract .................................................................................................................................................. iii
Statement of originality .......................................................................................................................... v
Acknowledgements ................................................................................................................................. vi
Chapter 1 ................................................................................................................................................. 1
Introduction ............................................................................................................................................. 1
1.1 Background .................................................................................................................................... 2
1.2 Research aim and questions ........................................................................................................... 4
1.3 Research approach ......................................................................................................................... 5
1.4 Thesis outline .................................................................................................................................. 5
Chapter 2 ................................................................................................................................................. 7
Constructed wetlands for urban stormwater management: current research and future needs ..... 7
2.1 Background .................................................................................................................................... 8
2.2 Urban stormwater management ..................................................................................................... 9 2.2.1 Sources of stormwater nutrient ............................................................................................... 9 2.2.2 WSUD and WSC aspect for stormwater management ......................................................... 10
2.3 Constructed wetland and its biotic-abiotic component ................................................................ 12
2.4 Nutrient attenuation pathways...................................................................................................... 14 2.4.1 Attenuation of nitrogen ......................................................................................................... 15 2.4.2 Attenuation of phosphorus .................................................................................................... 18
2.5 Research needs and knowledge gaps ............................................................................................ 20 2.5.1 Performance variability due to CW design and type ............................................................ 20 2.5.2 Filter media as an effective nutrient sink .............................................................................. 25 2.5.3 Nutrient attenuation variability during flow dynamics ......................................................... 25 2.5.4 Role of macrophyte growth and decay in wetland performance ........................................... 27 2.5.5 Does the overall ‘metabolism’ of wetlands impact their attenuation ability ......................... 27 2.5.6 Effectiveness of CW restoration to improve nutrient attenuation ......................................... 29
Chapter 3 ............................................................................................................................................... 31
Nutrient attenuation in a surface flow constructed wetland experiencing summer low flow and
macrophyte senescence ......................................................................................................................... 31
3.1 Introduction .................................................................................................................................. 33
3.2 Materials and methods ................................................................................................................. 36 3.2.1 Site background and restoration works ................................................................................. 36 3.2.2 Data collection ...................................................................................................................... 36 3.2.3 Data analysis ......................................................................................................................... 41
3.3 Results .......................................................................................................................................... 44 3.3.1 Long term changes in wetland properties ............................................................................. 44 3.3.2 Nutrient accumulation or uptake in the sediment and macrophytes ..................................... 46 3.3.3 Hydrological variability on nutrients dynamics .................................................................... 53 3.3.4 Diurnal changes in wetland function .................................................................................... 58
3.4. Discussion ................................................................................................................................... 62 3.4.1 Restoration improves attenuation of specific nutrient species .............................................. 62 3.4.2 Hydrological and biogeochemical controls on nutrient attenuation ...................................... 66 3.4.3 DO as a proxy indicator of wetland metabolism ................................................................... 70
xiii
3.5 Conclusions .................................................................................................................................. 71
Chapter 4 ............................................................................................................................................... 73
Stormwater nutrient attenuation in a multi-compartment constructed wetland ............................ 73
4.1 Introduction .................................................................................................................................. 75
4.2 Materials and methods ................................................................................................................. 77 4.2.1 Site background .................................................................................................................... 77 4.2.2 Data collection ...................................................................................................................... 80 4.2.3 Data analysis ......................................................................................................................... 82
4.3 Results .......................................................................................................................................... 85 4.3.1 Long term dynamics of nutrient quality ................................................................................ 85 4.3.2 Nutrient attenuation across compartments ............................................................................ 88 4.3.3 Impact of hydrological variability on nutrient attenuation ................................................... 88 4.3.4 Nutrients storage within sediments and macrophytes ........................................................... 95 4.3.5 DO variability ....................................................................................................................... 97
4.4 Discussion .................................................................................................................................... 98 4.4.1 Variability in performance within the system and over time ................................................ 98 4.4.2 ΔDO as an indicator of wetland metabolism ...................................................................... 106 4.4.3 Benefits of multi-functional system .................................................................................... 107
4.5 Conclusions ................................................................................................................................ 109
Chapter 5 ............................................................................................................................................. 111
Storm event-scale nutrient attenuation and metabolism in constructed wetlands ........................ 111
5.1 Introduction ................................................................................................................................ 113
5.2 Materials and methods ............................................................................................................... 114 5.2.1 Study sites ........................................................................................................................... 114 5.2.2 Data and field campaign ..................................................................................................... 115 5.2.3 Water balance ..................................................................................................................... 120 5.2.4 Nutrient attenuation estimation during events .................................................................... 120 5.2.5 Metabolism estimation ........................................................................................................ 122
5.3 Results ........................................................................................................................................ 123 5.3.1 Water balance during events ............................................................................................... 123 5.3.2 Water quality and nutrient attenuation during events ......................................................... 123 5.3.3 Changes of soil porewater properties due to riparian saturation ......................................... 126 5.3.4 DO dynamics during events ................................................................................................ 130
5.4 Discussion .................................................................................................................................. 131 5.4.1 Hydro-climate and catchment character determine nutrient export .................................... 131 5.4.2 CWs design and hydrological variability control nutrient attenuation................................ 135 5.4.3 Diurnal DO and metabolism shape nutrient dynamics ....................................................... 138
5.5 Conclusions ................................................................................................................................ 139
Chapter 6 ............................................................................................................................................. 141
Conclusions .......................................................................................................................................... 141
References ............................................................................................................................................ 145
xiv
List of Figures
Figure 2. 1 Conceptual model indicating how stormwater flows in undeveloped or natural area
and urbanized area .............................................................................................................. 10
Figure 2. 2 Cross section of a surface flow constructed wetland with groundwater
connectivity.. ....................................................................................................................... 16
Figure 2. 3 Schematic representation of the main phase of above ground biomass of
macrophytes including leaf expansion, maturity and senescence. ...................................... 29
Figure 2. 3 Schematic representation of the main phase of above ground biomass of
macrophytes including leaf expansion, maturity and senescence. ...................................... 29
Figure 3. 1 Map of the Anvil Way Compensation Basin showing monitoring and sampling
points ................................................................................................................................... 38
Figure 3. 2 Nutrient attenuation (as SDC) in the Anvil Way Compensation Basin over pre- and
post-restoration regimes ...................................................................................................... 48
Figure 3. 3 Nutrient attenuation (as SDC) in the Anvil Way Compensation Basin over pre- and
post-restoration regimes ...................................................................................................... 49
Figure 3. 4 Change of attenuation (as SDC_ave) of N species by the Anvil Way Compensation
Basin over the pre- and post-restoration regimes ............................................................... 51
Figure 3. 5 Change of attenuation (as SDC_ave) of P species by the Anvil Way Compensation
Basin over the pre- and post-restoration regimes. .............................................................. 52
Figure 3. 6 Nutrient pools of TN, TP and TOC at the in-stream and bench sediments of the
Anvil Way Compensation Basin from November 2011 to May 2015. ............................... 53
Figure 3. 7 Nutrient pools in AGB and BGB of the Anvil Way Compensation Basin from
October 2012 to October 2013. ........................................................................................... 54
Figure 3. 8 Monthly total rainfall, total discharge and average flow rate at and around the
Anvil Water Compensation Basin from 2012 to 2014. ....................................................... 56
Figure 3. 9 Selected rainfall amount and rainfall duration at and around the Anvil Way
Compensation Basin along with flow discharge in 2012 and 2013. ................................... 56
Figure 3. 10 Spatial and temporal trend of water level and porewater oxidation-reduction
potential within a riparian zone transect, and nutrients from 11 to 17 March 2015. .......... 60
Figure 3. 11 Changes of delta DO saturation in response of solar exposure and stream
coverage with floating macrophytes in the Anvil Way Compensation Basin .................... 61
Figure 3. 12 Changes of delta DO saturation in response of solar exposure and stream
coverage with floating macrophytes in the Anvil Way Compensation Basin .................... 62
Figure 3. 13 Diurnal pattern of meteorology and water quality in the Anvil Way
Compensation Basin from 12 to 13 March 2015.. .............................................................. 63
Figure 4. 1 Map of the Wharf Street Constructed Wetland indicating sampling locations of
surface water, flow, sediment and macrophyte and DO. .................................................... 79
Figure 4. 2 Trends of NH3, NOx, TKN, DON, TN, FRP and TP nutrient concentrations along
the Wharf Street Constructed Wetland from 2009 to 2015.. .............................................. 86
Figure 4. 3 Nutrient attenuation in the Wharf Street Constructed Wetland from 2009 to 2015.
............................................................................................................................................ 89
Figure 4. 4 Nutrient attenuation in the SF1 and SSF1 compartment of the Wharf Street
Constructed Wetland from 2009 to 2015 ............................................................................ 90
Figure 4. 5 Nutrient attenuation within different compartments of the Wharf Street
Constructed Wetland ........................................................................................................... 91
Figure 4. 6 Hydrograph at the inlet and outlet of the Wharf Street Constructed Wetland during
different hydro-climatological conditions ........................................................................... 93
xv
Figure 4. 7 Nutrient concentration and flow patterns at the main inlet and main outlet of the
Wharf Street Constructed Wetland during a storm event (17–21 November 2009).. ......... 94
Figure 4. 8 Nutrient pools of TN, TP and TOC from June 2010 to May 2014 at three sediment
sampling sites of the Wharf Street Constructed Wetland. .................................................. 96
Figure 4. 9 Dry biomass and nutrient pools in AGB and BGB of B. articulata from November
2010 to May 2012 at three sampling sites of the Wharf Street Constructed Wetland. ....... 96
Figure 4. 10 Spatial and temporal trend of DO concentrations in water samples of different
compartments of the Wharf Street Constructed Wetland from 2009 to 2015..................... 99
Figure 4. 11 Relationship between DO and nutrient concentrations in different compartments
of the Wharf Street Constructed Wetland ......................................................................... 100
Figure 4. 12 Relationship between DO and nutrient concentrations in different compartments
of the Wharf Street Constructed Wetland ......................................................................... 101
Figure 4. 13 Relationship between inlet nutrient concentrations and nutrient attenuation in
different compartments of the Wharf Street Constructed Wetland. .................................. 103
Figure 4. 14 Conceptual relationships between different attenuation pathways of NH3 and
NOX ................................................................................................................................... 104
Figure 4. 15 Monthly ΔDO saturation as a function of solar exposure during the wet and dry
periods in the Wharf Street Constructed Wetland ............................................................ 107
Figure 4. 16 Some multi-functional aspects of the Wharf Street Constructed Wetland. ....... 108
Figure 5. 1 Map of the Wharf Street Constructed Wetland indicating different sampling points
.......................................................................................................................................... 116
Figure 5. 2 Map of the Anvil Way Compensation Basin indicating different sampling points...
.......................................................................................................................................... 117
Figure 5. 3 TN and TP concentrations at the inlet and outlet of the Wharf Street Constructed
Wetland and Anvil Way Compensation Basin during different storm events.. ................ 124
Figure 5. 4 Nutrient concentrations at the inlet and outlet of the Anvil Way Compensation
Basin during three storm events ........................................................................................ 125
Figure 5. 5 Stormwater hydrographs, pollutographs and nutrients dynamics in the Wharf
Street Constructed Wetland and the Anvil Way Compensation Basin ............................. 129
Figure 5. 6 Temporal variability of soil porewater properties in the transect of the Anvil Way
Compensation Basin ......................................................................................................... 129
Figure 5. 7 Nutrient concentrations in porewater samples of the Anvil Way Compensation
Basin ................................................................................................................................. 130
Figure 5. 8 Temporal variability of metabolism in the SF1 compartment of the Wharf Street
Constructed Wetland and the Anvil Way Compensation Basin during four consecutive
days of pre-storm events and post-storm events ............................................................... 132
Figure 5. 9 Principal component analysis of different parameters (average over four
consecutive days) during the pre-event and post-event in the Anvil Way Compensation
Basin.. ............................................................................................................................... 133
Figure 5. 10 Intra-storm variability flow and EMC of TN and TP in the Wharf Street
Constructed Wetland and Anvil Way Compensation Basin ............................................. 134
Figure 5. 11 Nutrient loads attenuation as a function of travel time of water flow in the Wharf
Street Constructed Wetland and Anvil Way Compensation Basin. Relationship between
nutrient load attenuation and changes in metabolism due to storm perturbation in the
AWCB .............................................................................................................................. 137
xvi
List of Tables
Table 2. 1 Traits to categorize constructed wetlands ............................................................... 14
Table 2. 2 A comparison of surface, subsurface, vertical flow and horizontal flow CWs. ...... 15
Table 2. 3 Nitrogen and phosphorus attenuation and transformation dynamics within and
around constructed wetlands ............................................................................................... 21
Table 2. 4 Attenuation of nutrients and physicochemical parameters from urban stormwater
runoff from some site-specific investigations. .................................................................... 23
Table 3. 1 Summary table of spatial and temporal dynamics of studied nutrient parameters,
data type, sampling frequency and data source for the Anvil Way Compensation Basin
assessment ........................................................................................................................... 40
Table 3. 2 Spatial and temporal variability of nutrient concentrations in the Anvil Way
Compensation Basin. The dry and wet periods represent December to May and June to
November, respectively ...................................................................................................... 47
Table 3. 3 Nutrient attenuation as SDC and SDC_ave during the dry and wet periods over pre-
and post-restoration regimes of the Anvil Way Compensation Basin.. .............................. 50
Table 3. 4 Spatial and temporal changes of physical properties of above ground biomass of
aquatic macrophytes per quadrate in the Anvil Way Compensation Basin. ....................... 55
Table 3. 5 Spatial and temporal variation of groundwater nutrient concentrations at four bores
in the Anvil Way Compensation Basin from 2010 to April 2015. ..................................... 57
Table 3. 6 Conceptual consequences of the Anvil Way Compensation Basin restoration
initiatives on nutrient attenuation and transformation pathways. ....................................... 64
Table 4. 1 Summary table of spatial and temporal dynamics of studied parameters, data type,
sampling frequency and data source for the Wharf Street Constructed Wetland assessment
............................................................................................................................................ 81
Table 4. 2 Spatial and temporal variability of nutrient concentration in the WSCW .............. 87
Table 4. 3 Nutrient attenuation during the dry and wet period in the Wharf Street Constructed
Wetland ............................................................................................................................... 92
Table 4. 4 Nutrient attenuation during the dry and wet period in the Wharf Street Constructed
Wetland ............................................................................................................................... 92
Table 4. 5 Estimation of nutrient load attenuation in the Wharf Street Constructed Wetland
during the base and high flow conditions. .......................................................................... 95
Table 4. 6 Spatial and temporal variability of physical properties of above ground biomass of
B. articulata per quadrate in the Wharf Street Constructed Wetland. ................................ 97
Table 5. 1 Comparative description of the Wharf Street Constructed Wetland and the Anvil
way Compansatin Basin .................................................................................................... 118
Table 5. 2 Event campaigns and associated hydro-meteorology for the assessment of the
Wharf Street Constructed Wetland and the Anvil Way Compensation Basin. ................. 119
Table 5. 3 Variation of EMC, SMC, EF, load and load attenuation of TN and TP during storm
events in the Wharf Street Constructed Wetland and the Anvil Way Compensation Basin
.......................................................................................................................................... 126
Table 5. 4 EMC and load variability at the inlet and outlet for nutrient species in the Anvil
Way Compensation Basin during different storm events. ................................................ 127
List of Box
Box 2. 1 Main objectives and principles of WSUD for stormwater management. .................. 12
Chapter 1
Introduction
Chapter 1
2
1.1 Background
Stormwater runoff carries nutrients from point and non-point sources of urban
catchment (Hatt et al. 2009; Gasperi et al. 2011; Burns et al. 2012; Zgheib et al.
2012). Due to the rapid expansion of impervious areas in cities around the globe, the
magnitude of nutrient loads associated with untreated stormwater is increasing, and
subsequently creating significant challenges for the downstream receiving waterways
(Walsh et al. 2005b; Brown et al. 2009; Burns et al. 2012; Fletcher et al. 2013). These
challenges include issues associated with water quality impairment due to the
formation of algal blooms and eutrophication, the depletion of dissolved oxygen (DO)
and generation of anoxia, the decline in aquatic biodiversity and in extreme cases, the
mortality of fish; all of these challenges can reduce the ecological and recreational
amenity of watercourses (Taylor et al. 2004; Miller & Boulton 2005; Kaushal et al.
2008; Cuffney et al. 2010; Hathaway et al. 2012). A holistic management plan to
attenuate excess stormwater nutrient is therefore needed to restore ecological services
and mitigate adverse effects within and around the receiving watercourses, and to
improve urban liveability (Hatt et al. 2007; Wong et al. 2012; Deletic et al. 2013;
Managi et al. 2016). A paradigm shift in stormwater management has been taking
place in many cities based on Integrated Stormwater Management System (ISWMS)
and through the adoption of Best Management Practices (BMPs) (Fletcher et al.
2014). Water Sensitive Urban Design (WSUD) is such an approach that is widely
practiced in Australia (Lloyd 2001). WSUD is based on the idea that the management,
protection and conservation of the urban water cycle must take place to ensure that the
urban water environment is sensitive to natural hydrological and ecological processes,
and that the ecological services of urban waterways are maintained (Deletic et al.
2013). A commonly used WSUD element is the constructed wetland (CW) often
implemented prior to a watercourse, to provide stormwater retention and nutrient
attenuation (Hatt et al. 2006; Burns et al. 2012; Fletcher et al. 2013).
CWs are engineered green bioreactors (Mitsch & Jorgensen 2003), that can
attenuate stormwater nutrients through different hydrological and biogeochemical
processes within the water, soils/sediments, aquatic macrophytes and associated
microbial communities (Malaviya & Singh 2012; Saeed & Sun 2012; Vymazal &
Kröpfelová 2015). These engineered wetlands are economically viable and
environmentally sustainable in comparison with other conventional stormwater
treatment systems, due to low construction and maintenance costs, relatively simple
Chapter 1
3
operation, and low energy requirements (Kadlec 2000; Puigagut et al. 2008; Kadlec &
Wallace 2009). Individual CWs can be classified into different groups (Fonder &
Headley 2013), where surface flow (SF) and subsurface flow (SSF) CWs are the two
most common types. In isolation, a SF or SSF CW may attenuate a particular nutrient
more than another, while multi-compartment or hybrid CW can combine the
advantages of individual SF and SSF CW into a single system, and provide better
overall nutrient attenuation (Ávila et al. 2014; Ayaz et al. 2015; Vymazal &
Kröpfelová 2015). However, optimization of nutrient attenuation in a CW over diurnal
to decadal timescales is challenging. Moreover, the extent of nutrient attenuation
within different compartments of a CW needs to be investigated under different
hydro-climatological and biogeochemical regimes.
The performance of a CW varies depending on its geological settings and
available hydro-biological conditions. CWs in Mediterranean climates such as on the
Swan Coastal Plain (SCP) of Western Australia (WA), experience prolonged periods
of low or near zero surface flows during the dry summer and therefore CW can
become stagnant or lentic system. However frequent rainfall events during the wet
winter and possible groundwater inputs increase water levels, generate a continuous
flow and convert CW as a lotic system. Periodic lentic-lotic transition of CW across
events, days and seasons makes the nutrient attenuation estimation challenging.
Moreover, CWs of the Mediterranean climates also experience macrophytes
senescence during the summer. Therefore, nutrient attenuation varies over time, and
how to optimization nutrient attenuation within this context is not well understood.
Furthermore, CWs function most effectively in the first few years after the
construction (Kadlec & Wallace 2009; Mustafa et al. 2009; Vymazal 2013b), however
nutrient attenuation can decrease with time, due to the senescence and accumulation
of plant material that subsequently releases nutrients back into the system (Hancock
2002; Kröger et al. 2007), the accumulation of nutrients and organic matter in the
sediments (Craft 1996), the degradation of the riparian zones and development of
preferential flow paths (Harrison et al. 2012), the diffusion and bioturbation of
nutrients from the sediments into water column (O'Brien et al. 2012), the development
of anoxic environments and the input of polluted groundwater (Gabriele et al. 2013;
Teufl et al. 2013). Income cases, over the long term CWs can become a source of
nutrients and may require periodic restoration. Understanding the trend in nutrient
Chapter 1
4
attenuation after the restoration or indeed the optimum time for restoration is
challenging and needs further investigation.
Whilst the focus of understanding CW function is traditionally on their ability
to reduce nutrients, CWs are highly dynamic and transition between autotrophic to
heterotrophic metabolism at the ecosystem-scale, depending on available light and
organic matter. Measurements of autotrophy/heterotrophy and CW metabolism can be
complex and costly, and methodology varies between lentic and lotic systems. Low-
cost alternative approaches, or proxies, offer potential advantages to assess CW
function and their metabolism. One suggested proxy is the net variation diurnal
oxygen as delta DO (ΔDO), defined as the net effect of biogeochemical processes that
consume oxygen (e.g., respiration) and processes that deliver oxygen to the water
column (e.g., photosynthesis and reaeration or exchange with the atmosphere). If ΔDO
can be shown to be an effective proxy, under a range of hydrological conditions, this
low-cost high resolution sensor data could be easily used as an indicator of function
and nutrient dynamics in the CW.
1.2 Research aim and questions
The main goal of this thesis was to investigate stormwater nutrient attenuation
from diurnal to decadal scales, in two CWs: (a) a single stage mender SF CW and (b)
a hybrid CW with multiple alternating SF and SSF compartments. These CWs on the
SCP in WA experience a Mediterranean climate with prolonged low flow conditions
and macrophytes senescence during the dry summer, high flow conditions during the
wet winter and episodic flows after a storm event. The research also investigated DO
based metabolism as a proxy indicator of two CWs during different flow regimes.
The major research questions were:
1. How does nutrient attenuation vary over diurnal, event, seasonal and long-term
time-scales?
a) How do prolonged low flows and senescence of macrophytes during the
summer impact nutrient attenuation?
b) How does nutrient attenuation vary during episodic storm pulses?
c) How do antecedent conditions influence nutrient attenuation?
2. Can diurnal DO fluctuation be used as a proxy of wetland metabolism to gain
insight into biogeochemical processes controlling nutrient attenuation?
d) How does diurnal DO fluctuation impact nutrient attenuation?
Chapter 1
5
1.3 Research approach
Regular field sampling was undertaken from December 2012 to December
2015 to collect surface water and soil porewater samples and analysed them for
nutrients (details in Chapters 3 to 5). DO loggers, a multi-parameter water quality
sonde, oxidation-reduction potential (Eh) sensors and conductivity, temperature and
water depth (CTD) loggers were installed to receive high resolution spatial and
temporal data (details in Chapters 3 to 5). Historical data of water quality, water flow,
sediment quality, macrophytes properties, and meteorology were also collected,
synthesised and analysed. The thesis accounted nutrient attenuation in two different
types of CWs using extensive dataset which spanned over diurnal, event, seasonal and
annual time-scales. The standardised delta concentration (SDC), defined as the
standardised difference in nutrient concentrations between the inlet and outlet, was
used as an index to assess nutrient attenuation in the CWs, particularly during the low
flow conditions. SDC concept also used when there were ungauged water inputs to the
systems. Relative contribution of sediments, macrophytes, CW design, hydrological
seasonality and biogeochemical processes on overall nutrient dynamics was reported.
DO based metabolism was explored to find its effectiveness for using it as a proxy
indicator of CW function under hydrological variability.
1.4 Thesis outline
This thesis contains six chapters. A general background of stormwater nutrient
dynamics and attenuation in CWs is presented in Chapter 2. Chapters 3 to 5 are
presented in the style of manuscripts for journal publication. As a result these chapters
are self-contained and may be viewed independent of the rest of the thesis. Due to the
choice of this format, there may be some repetition in the literature review, site
description and methodology for these chapters. In Chapter 3, the research focuses on
nutrient attenuation in a surface flow CW that experiences prolonged low flows and
macrophyte senescence during the summer periods. The site was built in 2004 as a
drainage compensation basin and restored in 2010. Water quality data, water flow
data, sediment quality, macrophyte biomass, soil porewater and DO are assessed to
investigate diurnal to long-term (2004 to 2015) nutrient dynamics. In Chapter 4, the
research describes nutrient attenuation in a hybrid CW that contains alternating SF and
SSF compartments across episodic event to long term (2009 to 2015) time-scale.
Water quality data, water flow data, sediments quality, macrophytes quality and DO
Chapter 1
6
data are also assessed for this chapter. We also quantified nutrient load attenuation in
Chapter 4 over baseflow/low flow and highflow conditions. Chapter 5 shows a
comparison of nutrient attenuation in the two CWs during episodic storm events. This
chapter also deals with metabolism over fourteen storm events from 2014 to 2015 to
gain insight into metabolism dynamics due to storm perturbation. Chapter 6 presents
overall findings and conclusions, and also indicates some research gaps and
recommends future works.
Chapter 2
Constructed wetlands for urban stormwater
management: current research and future needs
Chapter 2
8
2.1 Background
One of the key functions and ecological services of urban watercourses is their
ability to attenuate nutrients from inflowing stormwater runoff. The stormwater runoff
can modify the morphological and hydrological characteristics within and around the
watercourses, and can impact groundwater flow and affect nutrient cycles of
downstream river and estuary (Walsh et al. 2012; Fletcher et al. 2013). Watercourses
nurture ecological resources and support embedded environmental and economic
services. However, excessive nutrient loads can be transported into the watercourses
by stormwater runoff, resulting from growing urbanisation, industrial activities,
shifting catchment properties and climate variability (Figure 2. 1). Excess nutrients
can trigger algal blooms, deplete dissolved oxygen (DO), hamper fish production,
fragment habitat, endanger biodiversity and lower the amenity in the water bodies
(Taylor et al. 2004; Hathaway et al. 2012). Holistic management plans are therefore
required to reduce nutrient (source control in urban catchment), mitigate nutrient
(nutrient attenuation) and subsequently protect stormwater receiving urban waterways
and support ecological services of such waterways.
The use of Integrated Stormwater Management Systems (ISWMS) and the
adaptation of Best Management Practices (BMPs) resulted in a paradigm shift in
stormwater management at both local and regional scales. In particular, some current
initiatives for stormwater management include Low Impact Design (LID) in the USA
and New Zealand, Sustainable Urban Drainage Systems (SUDS) in the UK and Water
Sensitive Urban Design (WSUD) in Australia (Fletcher et al. 2014). Under WSUD,
different aspects of the management, protection and conservation of urban water,
including stormwater, are prescribed (Deletic et al. 2013). WSUD (Box 1. 1) approach
ultimately minimizes the connection between the impervious surface and receiving
watercourses, and maximises the storage stormwater temporarily within the watershed
using retention-based or infiltration-based treatment facilities (Fletcher et al. 2013).
WSUD has promoted the use of constructed wetlands (CWs) as a retention-based
technology for attenuating nutrients from stormwater runoff (Deletic et al. 2013;
Fletcher et al. 2013; Payne et al. 2014). Soils/sediments, amended substrates, aquatic
macrophytes and microbial communities within CWs can attenuate stormwater
nutrient through hydrological, biological and biogeochemical processes (Vymazal
2013b). CWs, as engineered systems, are economically viable and environmentally
sustainable in comparison with other conventional wastewater treatment systems, with
Chapter 2
9
low construction and maintenance costs, easy operation and low energy requirements
(Puigagut et al. 2008; Kadlec & Wallace 2009). In addition, CWs can provide multi-
functional benefits and services in the urban landscape, including regulatory services
(flood control, micro climate improvement and soil erosion control), biogeochemical
services (soil formation, carbon retention and trace element storage), ecological
services (ecosystem maintenance and food web support), provisioning services (plant
biomass production, biodiversity support and stormwater harvest) and cultural
activities (recreational and aesthetic benefits, and educational programs enhancement)
(MEA 2005; De Groot 2006). As background to the subsequent research chapters
within this thesis, the present chapter provides an overview of nutrient attenuation
mechanisms within CWs, focusing on nitrogen (N) and phosphorus (P) sourced from
urban stormwater runoff. An improved understanding of nutrient attenuation is
essential to optimize the urban CW performance and support water management
practices. Finally, some knowledge gaps and future research needs are highlighted.
2.2 Urban stormwater management
2.2.1 Sources of stormwater nutrient
Urban stormwater traditionally refers to water runoff over fully or partially
impervious surfaces or points and non-point sources within urban catchments during
and/or after any precipitation event (Barbosa et al. 2012). Excess nutrient
concentrations and composition have been measured by collecting water samples
along the flow path, from the house, garden and carpark, to the ditch and outlet drain,
and finally in the downstream river, estuary or coastal zone. Stormwater quality can
be highly variable, and is a function of (a) meteorology or climate (rainfall intensity,
antecedent dry/wet period between storm events and evaporation), (b) location (site,
urban condition and land use pattern), (c) catchment characteristics (size, impervious
surface fraction, connectivity of paved surfaces and atmospheric deposition), and (d)
drainage infrastructure (separate or combined, open channels/streams or pipes, age,
cross-connections, sewer overflows, connection to surrounding groundwater or
existing septic tanks) (Gasperi et al. 2011; Zgheib et al. 2012). For instance,
sometimes intense rainfall can mobilise nutrients from the catchment and deliver them
as a higher load, the phenomenon known as the first flush (FF) effect (Deletic 1998;
Bach et al. 2010). However, subsequent rainfall dilutes nutrient concentrations. More
nutrients can be washed away from semi-impervious surfaces, while impervious
Chapter 2
10
surfaces deliver nutrients quickly to the downstream water courses. The ratio between
infrastructure coverage and open space influences the quantity of water runoff and
infiltration during and after storm events. The drainage network accelerates water
runoff through increased surface water connectivity, and reduces stormwater
infiltration into the groundwater.
Figure 2. 1 Conceptual model indicating how stormwater flows in undeveloped or natural area
and urbanized area. (a) Stormwater flows in undeveloped or natural area and (b) in urbanized
area, roof runoff is diverted into soakwells (shallow vertical infiltration wells) and road runoff
is locally infiltrated, leading to an increasing groundwater table and a reduction of
evapotranspiration. The symbols P, ET, In, Ir, SI and A represent precipitation,
evapotranspiration, infiltration, irrigation, stormwater infiltration and abstraction, respectively.
The figure is modified after Locatelli et al. (2017).
2.2.2 WSUD and WSC aspect for stormwater management
Australia has developed WSUD approach in recent decades to integrate water
supply systems, wastewater treatment processes and stormwater management into
urban planning and design, considering ecological sustainability, environmental
assimilation capacity and economic viability (Whelans et al. 1994; Lloyd 2001; Wong
et al. 2012). The concept comprises of several objectives and principles based on
Chapter 2
11
multi-disciplinary planning and practices for stormwater management (Box 2. 1). The
approach involves pro-active structural planning for urban development, that
recognises the opportunities for urban design, landscape architecture and
infrastructure interventions (Wong & Eadie 2000). WSUD consists of two parts:
‘water sensitive’ and ‘urban design’. The meaning of WSUD varies amongst
practitioners, reflecting its framework, application and coverage. Wong et al. (2012)
defined WSUD, for instance, as “a new paradigm in planning and design of urban
environments that is ‘sensitive’ to the issues of water sustainability and environmental
protection”. The agreement of the National Water Initiative by Australian
governments expresses WSUD as “the integration of urban planning with the
management, protection and conservation of the urban water cycle that ensures that
urban water management is sensitive to natural hydrological and ecological processes”
(COAG 2004). The approach also consolidates the urban water cycle, including
stormwater, groundwater and wastewater management and water supply, into urban
design in order to minimise environmental degradation, improve aesthetic and
recreational attractions (Lloyd 2001), enhance urban stream health (Mitchell et al.
2007), provide protection for waterways from being contaminated, elevate urban
discharge system (Li et al. 2009), beautify urban landscape and improve micro-
climate by enhancing evapotranspiration (Endreny 2008).
In the last couple of decades, several urban stormwater treatment technologies
and WSUD elements have been developed for achieving Water Sensitive City (WSC)
(Wong & Brown 2008; Brown et al. 2009; Wong et al. 2012). The WSC approach
aims for collaboration between multi-sectoral stakeholders in the urban planning and
designing of future cities, improving existing poorly equipped water services to
support the projected population growth and tackling economic and meteorological
uncertainties (Brown et al. 2009; Deletic et al. 2013). The approach utilizes different
elements of resilient, liveable, productive and sustainable cities with a view to
providing water security for social need and economic prosperity. Such planning shifts
the focus from water supply and wastewater disposal to more holistic systems that
integrate different water sources, enables the combination of centralized and
decentralized water operations and expands the range of functional and ecological
services provided by urban water to achieve ‘the cities of the future’ (Wong et al.
2012). WSUD elements for stormwater management can be grouped broadly as (a)
infiltration-based technologies and (b) retention-based technologies (Fletcher et al.
Chapter 2
12
2013). CWs are generally retention-based technologies that can be used to treat, re-
cycle and re-use stormwater to achieve relevant WSUD objectives.
Box 2. 1 Main objectives and principles of WSUD for stormwater management.
Major objectives of WSUD
A. Reduce impact on receiving waters
1. Minimise impacts on existing natural features, ecological processes and natural
hydrologic behaviour of catchments
2. Improve the quality of and minimise polluted stormwater discharges to the natural
environment.
3. Improve waterway health through restoring or preserving the natural hydrological regime
of catchments through treatment and reuse technologies
4. Reduce runoff and peak flows from urban developments during storm events
5. Harmonise water cycle practices across and within the institutions responsible for
waterway health, flood management, pollution prevention and protection of social
amenity
B. Promote re-use
6. Adopt a fit-for-purpose approach to the use of potential alternative sources of water
including stormwater
7. Incorporate collection, treatment and/or reuse of runoff, including roof water and other
stormwater
8. Treat stormwater to meet water quality objectives for reuse and/or discharge by capturing
sediments, pollution and nutrients through the retention and slow release of stormwater
C. Protection of “social” water services
9. Increase social amenity and aesthetic with water for the urban dwellers through multi-
purpose green space, landscaping and integrating water into the landscape to enhance
visual, social, cultural and ecological values
10. Account for the nexus between water use and wider social and resource issues
11. Promote a significant degree of water-related self-sufficiency within urban settings by
optimizing the use of water sources to minimise potable storm and waste water inflows
and outflows through the incorporation into urban design of localised water storage
Modified after Lee & Yigitcanlar (2010)
2.3 Constructed wetland and its biotic-abiotic component
In general, wetlands are highly productive and biologically diverse ecotones
between terrestrial and aquatic environments, which are characterized by shallow
water overlaying waterlogged soil and interspersed with submerged or emergent
vegetation. Wetlands are broadly two types: natural wetlands and CWs. CWs are
extensively engineered systems built for specific purposes such as providing domestic,
industrial and urban stormwater management and nutrient treatment. Overall CWs can
be categorized into different types (Table 2. 1) according to water level, saturation of
media, presence of vegetation, hydraulic and flow patterns particularly for domestic
Chapter 2
13
and industrial wastewater treatment. However these classifications can be also applied
for urban stormwater nutrient attenuation. There are two classes of CW based on
water level: (a) free water or surface flow (SF) CWs and (b) subsurface flow (SSF)
CWs (Kadlec & Wallace 2009). A combination of SF and SSF CWs generates hybrid
CW. Flows within the CWs can be vertical flow (VF) and horizontal flow (HF), and
they have specific advantages and disadvantages (Table 2. 2).
CWs contain both biotic and abiotic components. Macrophytes and microbes
are the main biotic components of CWs, while soils/sediments and filter media are the
main abiotic components. Aquatic macrophytes vary depending on water level,
geographic setting and tolerance to variable environmental conditions. Details of
macrophyte classifications in CWs are described elsewhere (Brix 1994; Vymazal
2011; Saeed & Sun 2012; Vymazal 2013a), however they can be summarised as four
categories:
(a) Emergent macrophytes: These are typically rooted on saturated or
submerged soil, containing extensive root and rhizome systems, with leaves and stems
extending out of the waterbody.
(b) Floating leaf macrophytes: These are usually rooted in submerged
sediments but leaves and flowers float on the water surface.
(c) Free-floating macrophytes: This group grows and floats freely on or under
the water surface. Root systems, if present, generally hang beneath the plant and are
not attached to the bottom sediments.
(d) Submerged macrophytes: These grow underwater, particularly in oxygen
rich areas. The photosynthetic tissues are also submerged.
Microbial organisms play important roles in the processing of nutrients in
CWs. Major microbial organisms include bacteria, algae and fungi. Bacteria can be
both autotrophic and heterotrophic. Autotrophic bacteria use carbon (C) from carbon
dioxide (CO2) while heterotrophic bacteria utilise organic carbon for the formation of
cell tissue. Phototrophic organisms use light as an energy source. Phototrophs can be
either autotroph e.g., photosynthetic bacteria or heterotrophic e.g., Sulphur (S)
reducing bacteria. In contrast chemotrophs, derive energy from chemical reactions.
Nitrosonomas and Nitrobacter are the common examples of autotrophic chemotrophs
(Reddy et al. 1984; US EPA 1993; Kadlec & Knight 1996; Paul & Clark 1996).
Sediments/soils are the main abiotic components of CWs. Stormwater nutrient usually
settles in the sediments. Different filter media are also used in the CWs to enhance
Chapter 2
14
specific nutrient attenuation. The artificial media or substrates can be made up of
natural particles i.e., sand, clay, gravel, calcite or calcium carbonate (CaCO3), peat,
zeolite and limestone; industrial by-products i.e., slag, fly ash, coal cinder and alum
sludge; or artificial products i.e., activated carbon, light weight aggregate, compost
and calcium silicate hydrate (Saeed & Sun 2012; Wu et al. 2015).
Table 2. 1 Traits to categorize constructed wetlands.
Physical
attribute
Specific
trait
Description Defined classes
for each trait
Sub-class
Hydrology
Water
position
Position of water surface
relative to soil or substrate
Surface flow –
Subsurface flow –
Flow
direction
Predominant direction of
flow through system
Horizontal –
Vertical
Down flow
Up flow
Mixed flow
Saturation
of media
Degree of saturation in
media-based systems
Free-draining –
Intermittent –
Constant –
Influent
loading
type
Position and type of
influent distribution in
media-based systems
Surface inflow –
Subsurface inflow –
Basal inflow –
Vegetation
Sessility Location of the roots:
attached in the benthic
sediments or floating
Sessile (benthic bound) –
Floating –
Growth
form
Dominant growth form of
the vegetation in relation
to the water
Emergent Herbaceous
Woody
Submerged –
Floating leaved –
Free-floating –
Modified after Fonder & Headley (2011, 2013)
2.4 Nutrient attenuation pathways
A wide range of hydrological and biogeochemical interactions between water,
macrophytes, microbes, soils, filter media and ambient air within and around the CWs
combine to support water purification and/or nutrient attenuation (Figure 2. 2). CWs
are active bioreactors, governed by material and energy fluxes that lead to
characteristic chemical reactions. Nutrient attenuation may occur through destructive
or non-destructive pathways. Nutrients can be degraded or consumed during
destructive processes resulting in maximum attenuation, while non-destructive
processes convert nutrients from one form to another.
Chapter 2
15
Table 2. 2 A comparison of surface, subsurface, vertical flow and horizontal flow CWs.
SF SSF
Ad
van
tag
es Provide the maximum attenuation if constructed
for specific nutrient
Potential scope of air-water interaction
Support macrophyte growth
Provide alternating conditions for nutrient
processing
Better integration in landscape
Nutrient attenuation in full scale
Substrates encourage nutrient attenuation
Dis
adv
anta
ges
Problems with mosquitos
Dominant aerobic or anaerobic condition can
attenuate specific nutrients and generate others
Clogging problem can arise
Shut down of water flow to alternating
compartment if flow is low
May require additional water to keep full
system flowing
VF SSF HF SSF
Ad
van
tag
es Require relatively small area compared to HF
Potential scope of oxygen supply and mixing
Potential for nitrification process
Simple hydraulics
Higher attenuation from initial stage of operation
Long flow distances
Formation of humic acids for N and P
attenuation
Potential for denitrification process
Dis
adv
anta
ges
Short flow distances
Clogging problem can occur
Limited nitrate attenuation or denitrification
Need higher technical involvement
Lower P attenuation capabilities
Require a relatively high area compared to VF
Clogging problem can occur
Sulphur transformation can affect nitrification
sensitivity
Low ammonium oxidation
Less P attenuation
Preferential flow pathways emerge
Modified after Vymazal (2007); Malaviya & Singh (2012); Saeed & Sun (2012); Vymazal (2013b); Meng et al.
(2014)
In general, ammonia (NH3), nitrate (NO3), nitrite (NO2), total Kjeldahl
nitrogen (TKN), urea, dissolved inorganic nitrogen (DIN), dissolved organic nitrogen
(DON), particular nitrogen (PN) and total nitrogen (TN) are common N-based
nutrients in urban stormwater. Furthermore, filterable reactive phosphorus (FRP),
particulate phosphorus (PP), organic phosphorus and total phosphorus (TP) are the
main P-based nutrients. However, these names can slightly differ depending on
purposes of interpretation (Chapter 3 to 5).
2.4.1 Attenuation of nitrogen
The main routes of N attenuation in the CWs are sedimentation, plant uptake,
adsorption, nitrification, ammonification, ammonia volatilization, denitrification,
dissimilatory reduction of nitrate to ammonium (DNRA), N2 fixation and organic N
burial. Moreover, recently identified processes such as partial nitrification-
denitrification (PND), anaerobic ammonia oxidation (ANAMMOX) and completely
autotrophic nitrate removal over the nitrate (CANON) are also considered as N
processing pathways.
Chapter 2
16
Figure 2. 2 Cross section of a surface flow constructed wetland with groundwater
connectivity. Nutrients can enter the system through surface flow or runoff and groundwater
inputs. Aerobic and anaerobic biogeochemical reactions transform nutrients within
soils/sediments and water. Macrophytes and microbes can also uptake nutrients during growth
phases and release them during senescence. Nutrients can be exchanged across the sediment-
water column interface.
Both dissolved and particulate nutrients in the shallow water can be affected
by the presence of microbes, whether under aerobic or anaerobic conditions. Oxygen
for aerobic microbial degradation can be supplied via atmospheric oxygen diffusion,
photosynthesis, convection/wind effects or macrophytes root transfer into the
rhizosphere. Aerobic microbial degradation is dominant in the SF CWs, while the SSF
CWs prefer anaerobic microbial cycling of nutrients.
Ammonification converts organic N to NH3 and can occur when stormwater
contains high concentrations of organic N (Savant et al. 1982). Some of the generated
NH3 can be volatilized and released as gas (Reddy et al. 1984) (Table 2. 3).
Nitrification (Table 2. 3), microbial conversion of NH3 to NO3, is an important
pathway of NH3 attenuation and the process depends on microbial activity, DO level,
temperature and incoming NH3 concentrations. Sufficient oxygen in the water column
is essential for aerobic microbial processing. Initially microbial transformation of NH3
to NO2 occurs, and then NO2 converts to NO3. Chemolithotrophic bacteria like
Nitrosomonas, Nitrosococcus and Nitrosospira are active at the first stage. Facultative
chemolithotrophic bacteria such as Nitrospira and Nitrobacter, other bacteria such as
Arthrobacter globiformis, Mycobacterium phlei and Thiosphaera, or fungi such as
Chapter 2
17
Aspergillus flavus, Penicillium or Cephalosporium play a role at the later stage
(Reddy et al. 1984; US EPA 1993; Kadlec & Knight 1996; Paul & Clark 1996).
Microbial denitrification (Table 2. 3) is a major process of N attenuation in
CWs where NO3 transforms into nitrous oxide (N2O), nitric oxide or nitrogen
monoxide (NO) and ultimately to nitrogen gas (N2). Nitrate acts as the terminal
electron acceptor for this process and organic carbon sources act as electron donor
(Vymazal 1995). Denitrifying bacteria are usually chemoheterotrophs indicating that
they derive energy through chemical reactions. Bacillus, Micrococcus and
Pseudomonas are probably the most important genera in soils, while Pseudomonas,
Aeromonas and Vibrio are common in the aquatic environment (Bachand & Horne
1999; Vymazal 2007).
DNRA (Table 2. 3) reduces NO2 and NO3 to NH3 (Dong & Sun 2007; Kadlec
& Wallace 2009). This process takes place under low oxidation-reduction potential
(Eh) conditions and usually in C rich environments, particularly in the SSF CWs
where facultative or anaerobic microbial populations are dominant (Garcia et al.
2010). PND (Table 2. 3) is the microbiological conversion of ammonia into nitrite in
oxygen and carbon-limited environments (Jianlong & Ning 2004; Jianlong & Jing
2005). This process requires 25% less oxygen and 40% less organic matter than
nitrification and denitrification, respectively (Van Dongen et al. 2001; Jianlong & Jing
2005). Moreover, lower oxygen levels inside the CW can promote the growth of
ammonium oxidizers over nitrate oxidizers (Zhang et al. 2011). ANAMMOX directly
oxidizes ammonium and nitrite into nitrogen gas. NO3 is considered an electron
acceptor and the process is influenced by the Planctomycete bacteria group under
anaerobic conditions. Different intermediate products including hydrazine and
hydroxylamine can be formed. However, the growth of ANAMMOX bacteria is
usually slow, and higher concentrations of sulphide, NH3 and NO2 in water can slow
the growth of ANAMMOX bacteria even further (Third et al. 2001; Saeed & Sun
2012). Simultaneous PND and ANAMMOX occur in CANON. CANON also depends
on the mutual co-existence of aerobic ammonium oxidizers and anaerobic
ANAMMOX bacteria under oxygen limited conditions (Third et al. 2001; Saeed &
Sun 2012).
Biomass assimilation is the process of N uptake by macrophytes,
phytoplankton or microbial biomass that can reduce dissolved N concentrations in the
CWs (Sun et al. 2005). Macrophytes are natural purifiers within CWs (Malaviya &
Chapter 2
18
Singh 2012) as they: (a) provide suitable roots-shoots-leaves media for nutrient
attachment, uptake and storage during early to mature stages of growth (Brodrick et
al. 1988), (b) provide surfaces and oxygen in the below ground biomass or
root/rhizosphere zones for the growth and attachment of microorganisms or biofilms
for nutrient processing (Cui et al. 2010; Saeed & Sun 2012), (c) provide carbon from
root exudates (due to photosynthetically fixed carbon) for denitrification (Brix 1997;
Masi 2008; Wang et al. 2012), and (d) facilitate uniform flow distribution and flow
retardation and increase nutrient contact with plant surfaces for the effective removal
of finely graded particles and associated nutrients (Wong et al. 1999). However, the
total N attenuation in macrophytes can be limited as senescent macrophytes can
release nutrients back into the CW. Therefore, macrophytes require periodic
harvesting for permanent N removal.
Adsorption of NH3 from incoming waters by cation exchange (CE) processes
can also reduce TN concentrations (Bayley et al. 2003; Cui et al. 2010; Saeed & Sun
2011). This process is dominant in SSF system, where there is increased contact
between the substrate and nutrients. N2 fixation converts gaseous nitrogen into NH3
and can increase the N stored in the CWs. Symbiotic actinomycetes (associated with
nodulated host plants) and asymbiotic (free-living) heterotrophic bacteria and blue-
green algae (cyanobacteria) can fix N2 (Stewart 1973; Johnston 1991; Paul & Clark
1996).
Organic N burial takes place in the CWs when organic N from detritus,
senescent or refractory plant material gets buried and form litter and eventually peat.
The process may occur when emergent vegetation dominates SF CWs, where the litter
layer plays an important role in nutrient attenuation (Vymazal 2007). However, due to
sediment resuspension these particulate organic nutrients can be released to the water
column prior to the formation of the more stable peat.
2.4.2 Attenuation of phosphorus
P attenuation processes can be grouped into two broad categories: (a) long
term attenuation processes (sedimentation or physical settling, accretion,
accumulation, precipitation, adsorption, litter/peat ligand exchange) and (b) short term
attenuation processes (plant establishment, biofilm growth and microbial colonization
of litter) (Bavor & Adcock 1994; Kadlec & Knight 1996; Reddy et al. 1999).
Chapter 2
19
Sedimentation or physical settling processes filter course or particulate
material and this is of particular importance for P attenuation. P species can
accumulate in the soils/sediments or filter media and thus results a higher attenuation.
Chemical precipitation converts soluble P into insoluble P. Precipitation can also
incorporate phosphate ions into inorganic minerals, for example aluminium (Al),
calcium (Ca), iron (Fe) and magnesium (Mg) minerals, forming amorphous or poorly
crystalline solids (Vymazal 2007). P-containing mineral precipitates commonly found
in CWs include apatite as Ca5(Cl,F)(PO4)3, hydroxylapatite as Ca5(OH)(PO4)3,
variscite as Al(PO4)·2H2O, strengite as Fe(PO4)·2H2O, vivianite as Fe3(PO4)2·8H2O
and wavellite as Al3(OH)3(PO4)2·5H2O (Reddy & D'Angelo 1997). P can also co-
precipitate with other minerals, such as ferric oxyhydroxide and carbonate minerals,
such as CaCO3 (Vymazal 2007). At pH > 6, calcium phosphate can also be formed
and precipitate.
P adsorption by soil substrates is a long-term mechanism for P attenuation
(Richardson 1985; Kadlec & Knight 1996). The adsorption capacity of a substrate
depends on the clay content and mineral components (Rhue & Harris 1999) as clay
materials have higher adsorptive capacity. The concentrations of amorphous Al, Fe,
Ca and total organic carbon (TOC) also influence P adsorption (Reddy et al. 1995;
Reddy et al. 1999).
CE reactions can take place where negatively charged sediment/soil particles
attach to positively charged phosphate (Richardson 1985). Phosphate displaces water
or hydroxyls from the surface of Fe and Al hydrous oxides to form monodentate and
binuclear complexes within the coordination sphere of the hydrous oxide, through
ligand exchange reactions (Faulkner & Richardson 1989). Moreover, the transfer
between dissolved and particulate phases of P takes place via a two-step process: (a)
the rapid exchanges of phosphate between the soil porewater and soil particle or
mineral surfaces (adsorption) and (b) the slow penetration of phosphate into solid
phases (absorption) (Dunne & Reddy 2005). Since P sorption capacity is finite, once
soil sorption sites become saturated by prolonged nutrient loading, the CWs may
become a source of P, instead of a sink (Kadlec & Wallace 2009).
Macrophytes can uptake dissolved inorganic P from incoming water (Brix
1994; Vymazal 2001). Most of the uptake of P is performed by plant roots, followed
by absorption through leaves and shoots (Kadlec & Knight 1996; Vymazal 2007;
Vymazal 2013a). Macrophytes uptake P at higher rates during early to mature growth
Chapter 2
20
phases. However, senescent macrophytes can release dissolved P to the CWs
(Cheesman et al. 2010). Microbiota may have high growth rates due to high P
consumption rates in the CWs. Although P uptake by microbiota and biofilms in the
CWs is rapid, the total P microbial pool may not be significant. Moreover, the
majority of dissolved P taken up by microbial biomass will be rapidly returned to the
P cycle through bacterial decomposition (Reddy et al. 1995). Bacterial communities
also contribute to the long-term P burial in the system (Gächter & Meyer 1993). In
overall, stormwater nutrient can transfer from water column to groundwater,
atmosphere, sediments, microbes or macrophytes (Table 2. 3). Selected nutrient
attenuation efficiencies measured in different CWs are summarized in Table 2. 4.
2.5 Research needs and knowledge gaps
2.5.1 Performance variability due to CW design and type
CWs provide variable N and P attenuation across the world (Table 2. 4).
Attenuation of stormwater nutrient can go up to about 95%, however, this
performance of the CWs are not same. Major factors driving this differing
performance include CW treatment area, volume and flow path and
compartmentalization, all of which influence the biogeochemical and hydrological
processing of nutrients. CWs can be selective for particular nutrients depending on
such biogeochemical and hydrological variability. The SF and SSF systems usually
exhibit good attenuation for NH3 and NO3, respectively (Saeed & Sun 2012).
However, single-staged CW, either the SF or SSF, can also release nutrients to the
water column. Hybrid CWs have potential as an emerging beneficial approach that
nutrients generated in one compartment of these systems can be attenuated in
subsequent alternating compartments. Subsequent compartments ultimately provide
optimal environments for nutrient attenuation and polish the water. Benefits of
different approaches can be maximised by careful design, but uncertainty remains as
to what aspects of CW design promote attenuation. It is critical to show how the
different types of CWs perform under the highly variable flow and biogeochemical
conditions. Furthermore, water is generally unevenly distributed in the full scale CWs
because of channelization, microtopography, pattern of macrophytes, presence of any
woody log, filtering media (Fink & Mitsch 2007) and therefore optimization of CW
performance can be challenging.
Chapter 2
21
Table 2. 3 Nitrogen and phosphorus attenuation and transformation dynamics within and around constructed wetlands.
Nutrient
Transformation route within CW
Transformation
mode
Output of nutrients in different compartments in close vicinity of CW
Surface
water
Groundwater Atmosphere Soil and
sediments
Macrophytes
and microbes
NH3 Ammonia volatilization – forms gaseous NH3 (Vymazal 2007)
Plant assimilation or uptake (Sun et al. 2005)
Microbiota assimilation (Imfeld et al. 2009)
Adsorption in soil/sediment (Cui et al. 2010)
Nitrification (autotrophic) – forms NO2, N2O and NO3
Nitrification (heterotrophic) – forms NO2, N2O and NO3 (Reddy
et al. 1984)
Cation exchange (Cui et al. 2010)
Partial nitrification denitrification – forms NO2 (Jianlong &
Ning 2004; Jianlong & Jing 2005)
ANAMMOX – forms N2
CANON – forms N2 and NO3 (Third et al. 2001)
Dilution (Kim et al. 2005; Kim et al. 2007)
Leaching to groundwater (Harper 1988)
Physicochemical
Biological
Microbiological
Physicochemicalb
Microbiological
Microbiological
Microbiological
Physicochemicalb
Microbiological
Microbiological
Physical
Physical
Discharge
or transfer
to CW’s
water as
NH3, N2O,
NO3, NOx,
DON, N2,
TKN, TN
Transport
to/from
groundwater
through
porous media
of CW as
NH3, NO2,
NO3, NOx,
DON, TKN,
TN
Emission
through
diffusion,
transportation
via/by
macrophytes,
ebullition as
NH3, N2O, N2,
NO2
Adsorption,
cation
exchange by
CW’s soil
and sediment,
or store in
soil pore
space as
NH3, NO3,
N2O, DON,
TKN, TN
Adsorption
(or storage)
by plant
surface and
microbial
assemblage
and/or
forming
biofilm as
NH3, NO3,
TKN, TN
NO2,
NO3 or
NOx
Plant assimilation or uptake
Nitrification – forms NO3 and N2O
Denitrification – forms NO2, N2O and N2 (Bachand & Horne
1999)
Dissimilatory nitrate reduction – forms NH3 (Dong & Sun
2007)
ANAMMOX – forms N2 (Van Dongen et al. 2001)
Dilution (Kim et al. 2005; Kim et al. 2007)
Leaching to groundwater (Harper 1988)
Biologicala
Microbiological
Microbiological
Microbiological
Microbiological
Physical
Physical
DON
Ammonification – forms NH3 (Savant et al. 1982)
Oxidation – forms NO2 and N2O
Burial – buried in soil and forms peat (Vymazal 2007)
Dilution (Kim et al. 2005; Kim et al. 2007)
Leaching to groundwater (Harper 1988)
Microbiological
Microbiological
Physicochemical
Physical
Physical
TN, PN Adsorption in soil or sediment (Cui et al. 2010)
Dilution (Kim et al. 2005; Kim et al. 2007)
Physicochemicalb
Physical
Chapter 2
22
Nutrient
Transformation route within CW
Transformation
mode
Output of nutrients in different compartments in close vicinity of CW
Surface
water
Groundwater Atmosphere Soil and
sediments
Macrophytes
and microbes
DOP,
FRP ,
PP.
TP
Adsorption in soil or sediment (Cui et al. 2010)
Dilution (Kim et al. 2005; Kim et al. 2007)
Leaching to groundwater (Harper 1988)
Settlement and accretion (Richardson 1985).
Precipitation – forms amorphous solid
Soil adsorption (Reddy et al. 1999)
Plant uptake or assimilation (Vymazal 2013a)
Microbiota assimilation
Microbiota assimilation/biofilm growth (Gächter & Meyer
1993)
Volatilization – forms PH3 (Bavor & Adcock 1994)
Litter/peat accretion (Bavor & Adcock 1994)
Physicochemicalb
Physical
Physical
Physical
Chemical
Physical
Biologicala
Microbiological
Microbiological
Physicochemical
Physicochemical
Discharge
or transfer
to CW’s
water as
FRP, TP
Transport to
groundwater
through via
media of CW
as FRP, TP
Emission
through
diffusion,
volatilization
of decomposed
material as
PH3
Adsorption,
settling and
peat
formation as
FRP, TP
Adsorption
by plant
surface and
microbial
assemblage
and/or
biofilm as
FRP, TP
DOC,
DIC,
TDC
Aerobic degradation – form CO2, NH3
Anaerobic degradation (Vymazal & Kröpfelová 2009; Saeed &
Sun 2012)
Sorption and sedimentation
Phytoaccumulation and phytodegradation (Imfeld et al. 2009)
Microbial degradation (Imfeld et al. 2009)
Microbiological
Microbiological
Physical
Biological
Microbiological
Discharge
or transfer
to CW’s
water as
DOC, DIC,
TDC
Transport to
groundwater
as DOC,
DIC, TDC
Emission as
CO2, CH4
Settling and
peat
formation as
DOC, DIC
Uptake or
degradation
by microbes
as CO2, CH4,
Modified after Vymazal (2007); Saeed & Sun (2012); Jahangir et al. (2016) aPermanent removal bin presence of filter media
FFM: Floating macrophytes
HSSF: Horizontal subsurface flow
VSSF: Vertical subsurface flow
CH4: Methane
PH3: Phosphine gas
Chapter 2
23
Table 2. 4 Attenuation of nutrients and physicochemical parameters from urban stormwater runoff from some site-specific investigations.
Nutrients Location CW type Attenuation efficiency References
Urban stormwater
runoff
P, SS Tampa, Florida,
USA
SF 90% TP, 94% SS Rushton et al. (1995)
Urban stormwater
runoff
P Washington, USA SF 82% TP Reinelt & Horner (1995)
Residential
stormwater runoff
P, N, SS Manassas, Virginia,
USA
SF 54.7% NH3, 25.5% TKN, 39.4% NOx, 21.7% TN,
57.9% TSS, 35.8% DOP, 45.9% TP
Carleton et al. (2000)
Nursery runoff P, N New South Wales,
Australia
SF 84% TN, 65% TP Headley et al. (2001)
Urban stormwater
runoff
P Henely Brook
Perth, Western
Australia, Australia
SF 5% P in 1st year, 10% P in 2
nd year Lund et al. (2001)
Nursery runoff N Nimes, France SF 70% NOx Merlin et al. (2002)
Residential
stormwater runoff
P, N, SS Port Jackson,
Sydney, Australia
SF 12% TP, 16% TN, 9% TKN, 9-46% TSS Birch et al. (2004)
Urban stormwater
runoff
N, P Crum Woods,
Swarthmore
College,
Pennsylvania, USA
SF 26% NOx, 59% FRP Bangs (2007)
Urban stormwater
runoff
(gully pot, liquor)
N University of
Edinburgh,
Scotland
SF 78.7% NH3 in 1st year, 95.8% NH3 in 2
nd year, 93.7%
NO3 in 1st year, 95.5% NO3 in 2
nd year
Lee & Scholz (2007)
Combined
stormwater urban
& agricultural
runoff
N, P Putrajaya wetlands,
Malaysia
Hybrid 82.1% TN, 84.3% TP Sim et al. (2008)
Highway runoff SS, N, P Island of Crete,
Greece
Hybrid 89% TSS, 49% TN, 60% TP
Terzakis et al. (2008)
Combined
stormwater urban
(14%) agricultural
(40%), forested
land (33%)
N, P Seokmoon,
Choongnam, South
Korea
SF 51.5% and 31.7% TN during growing and winter
season, respectively 50.6% and 53.0% TP during
growing and winter season, respectively
Ham et al. (2010)
Combined
wastewater runoff
and urban
stormwater runoff
N, P Danshui River
Basin, Taipei,
Taiwan
SF 78% NH3, 49% TP Ko et al. (2010)
Chapter 2
24
Nutrients Location CW type Attenuation efficiency References
Simulated
stormwater runoff
N, P Clemson, USA SF 87.9% N in 1st year and 66.9% N in 2
nd year, 80.9% P
in 1st year and 58.6% P in 2
nd year
White & Cousins (2013)
Watershed
stormwater
P Atlantic Coastal
Ridge, FL, USA
Hybrid 27% P Cohen & Brown (2007)
Urban stormwater
runoff
NO3,
NH3,
FRP
Lake Jackson, FL,
USA
SF 67 % NO3, 87%, NH3, 62% FRP Johengen & LaRock (1993)
Wastewater
treatment
TSS,
NH3,
TN, FRP
The Universitat
Politècnica de
Catalunya-
Barcelona Tech,
Spain
SF 97% TSS, 94% NH3, 46% TN and 4% FRP Ávila et al. (2013)
Tertiary treated
urban wastewaters
P, NH3,
TN
Spain Hybrid 77.0% P, 95.0% NH3, 24.4% TN Martín et al. (2013)
Conservation of
the Ganga river
TSS,
TDS,
NO3,
FRP,
NH3,
Shantikunj,
Haridwar, India
SSF 65% TSS, 78% TDS, 84% NO3, 76% FRP and 86%
NH3
Rai et al. (2013)
Ameliorating
diffused nutrient
loadings
N, P Gundowring,
Wodonga, Victoria,
Australia
SF 10 kg N yr−1
(23 g N m−2
yr−1
) and 1.24 kg P yr−1
(2.80
g P m−2
yr−1
)
Raisin et al. (1997)
TDS: Total dissolved solid
TSS: Total suspended solid
SS: suspended solid
Chapter 2
25
2.5.2 Filter media as an effective nutrient sink
Different unconventional substrates, industrial by-products, as well as natural
and artificial materials can be used as CW filter media to enhance nutrient attenuation
by CE, adsorption, precipitation and complexation (Saeed & Sun 2012; Wu et al.
2015). Frequently used filter media include organic wood mulch, rice husks, gravel,
sand, clay, laterite, peat, alum sludge, calcite, marble, vermiculite, slag, fly ash,
bentonite, dolomite, limestone, shell, zeolite, wollastonite, activated carbon, light
weight aggregates, compost and (Van de Moortel et al. 2009; Malaviya & Singh 2012;
Saeed & Sun 2012). The presence of filter media in CWs is beneficial as it performs
multiple tasks such as (a) providing active sites for nutrient adsorption (Wood &
McAtamney 1996), (b) promote diffusion of oxygen into the substrate for nutrient
processing (if a porous substrate like compost is used) (Aslam et al. 2007), (c)
generate anaerobic conditions for denitrification (if a non-porous substrate used) )
(Gray et al. 2000), (d) provide suitable conditions for ion exchange (Cui et al. 2010),
(e) provide aeration in the filter bed (Białowiec et al. 2011), (f) support biofilm for
nutrient processing (Bourgues & Hart 2007) and (g) leach C (e.g., from organic wood
mulch) for denitrification (Saeed & Sun 2011). These filter media may have limited
interaction with solar radiation and the atmosphere, therefore encouraging anoxic
conditions. The performance of filter media in the field level CWs are subject to
further study. Furthermore, any release of nutrients from filter media can make
nutrient attenuation estimation challenging. The relative contribution of filter media in
the different compartments of a CW also needs to be assessed.
Some filter media can be dried out due to lack of stormevent during the
summer in semi-arid area where rainfall during summer is near zero. Therefore, filter
media can be dried out and microbial community can be stressed. It is therefore
necessary to estimate the function of such filter media in stormwater nutrient
attenuation after prolong dry periods. Furthermore, episodic pulses of storm events
can flood filter media and nutrient dynamics in such condition need to be understood.
2.5.3 Nutrient attenuation variability during flow dynamics
CWs designed for treating wastewater receive a relatively control rate of
inflowing water (Jahangir et al. 2016), However, CWs designed for stormwater
treatment typically experience highly variable inflow depending on hydro-
meteorology. Therefore, nutrient attenuation varies widely depending on the ratio of
Chapter 2
26
CW volume to inflow rate, inflow water volume, respective HRT and ADD.
Furthermore, CWs in semi-arid areas experience prolonged summer dry periods with
low to near zero flows, yet it is unclear how antecedent dry or wet conditions impact
on performance.
CWs frequently receive water from ungauged inflows including groundwater.
It is challenging to consider multiple and ungauged sources of water and nutrient
during CW efficiency estimation. HRT is usually higher during the dry periods
compared to the wet periods. Longer HRT increases the water contact time with
aquatic macrophytes and benthos, and subsequently facilitates nutrient attenuation
(Kadlec & Knight 1996). However, it is not clear how CW response to nutrient
processing under anoxic conditions is impacted by low flows and higher HRT. It is
also not clear how an open water or SF CW respond to such abrupt flow variability.
CWs can use recirculation system to keep a system functional over the dry season.
However, such recirculation can influence water balance and nutrient dynamics that
needed to be estimated.
The intense rainfall experienced in Mediterranean climates during the wet
winter can import high nutrient loads into CWs and therefore, event mean
concentration (EMC) changes with season. However, it is challenging to estimate how
CWs respond to these changing nutrient loads after and between storms. Episodic
storm events along with groundwater seepage can flood the riparian zones of CWs.
The riparian zones are CW ecotones where intense biogeochemical processing of
nutrients takes place. However, the response of CW riparian zones under seasonal
saturation and drying dynamics over is complex and requires further study.
Water levels in Mediterranean CWs can become very low (sometimes less
than 0.3 m) and therefore the water in the system can become stagnant. Ultimately, the
change in flow and associated water level convert the system into a lotic environment.
However, the same CW receives water from episodic stormwater early in the winter
and transitions back to a lotic environment. Sometimes groundwater inputs can also
contribute to the lotic state of the CWs. The significance of periodic lentic-lotic
transitions for nutrient dynamics remains unknown. The re-saturation of previously
exposed riparian soils during transitions to lotic conditions can also impact nutrient
attenuation or release to receiving waters, however, to date this scenario has not been
investigated within the context of stormwater treating CWs.
Chapter 2
27
2.5.4 Role of macrophyte growth and decay in wetland performance
Macrophytes can accumulate nutrients in their above ground biomass (AGB)
i.e., leaves and shoots and below ground biomass (BGB) i.e., roots and rhizospheres,
and therefore attention is higher in vegetated CWs than non-vegetated CWs (Vymazal
2011; Vymazal 2013a, b). However, nutrient uptake by macrophytes depends on the
type of macrophytes present (emergent, rooted and floating), species diversity (one
species or multiple), the planted form (individual or combined), growth stage (early or
mature stage) and stormwater flow conditions in the CWs. Different macrophyte
groups or species may interact and compete for nutrient uptake, nutrient uptake can
vary when species coverage or species type changes across the CW and therefore
optimization of macrophyte establishment needs to be considered. The time required
for CW macrophyte biomass to reach equilibrium conditions and how nutrient uptake
changes over this time, is still need further investigation. It is challenging to assess all
these aspects of macrophyte dynamics because of inter-specific completion of
macrophytes for nutrients, spatial heterogeneity in macrophytes biomass and its
seasonal dynamics.
Another challenge is macrophyte senescence on nutrient uptake dynamics
(Figure 2. 3). Macrophytes can exhibit seasonal senescence that can release stored
nutrients back to the CW (Brodrick et al. 1988; Kröger et al. 2007), and therefore,
mature macrophytes need periodic harvesting (Greenway 2005; Malaviya & Singh
2012; Vymazal 2013a). Furthermore, dead macrophytes can accumulate in the open
water during low flow conditions, hindering light penetration into the water column
and changing the DO dynamics. Therefore, the impact of seasonal macrophytes
senescence on the overall nutrient dynamics needs to be better understood.
2.5.5 Does the overall ‘metabolism’ of wetlands impact their attenuation ability
Wetland metabolism can be defined as the difference in the rate of C input and
output. If a wetland accumulates C (input > output, net autotrophic), then the
metabolism is characterised as positive. When a wetland releases more C than it is
retaining (i.e., net heterotrophic) then the metabolism is considered negative (Tuttle et
al. 2008; Coletti et al. 2013). Overall aquatic metabolism is the balance between
productivity or processes that produce oxygen (such as photosynthesis) and processes
that consume oxygen (such as respiration). The total metabolic balance of an
ecosystem is also known as the net ecosystem production (NEP), which is defined as
Chapter 2
28
the difference between gross primary production (GPP) and ecosystem respiration (R)
or community respiration (CR) (Lovett et al. 2006). NEP reflects the balance between
all anabolic and catabolic processes. When NEP is positive (net autotrophic), fixation
of inorganic C through photosynthesis exceeds oxidation of organic C. Alternatively,
when NEP is negative (net heterotrophic), the re-mineralisation of organic C exceeds
the fixation of inorganic C (Lovett et al. 2006; Dodds & Cole 2007; Staehr et al.
2012b). Researchers have used DO, CO2 and isotopes as key variables to determine
metabolism in aquatic environments including lake, estuary, river and ocean (Staehr et
al. 2010; Staehr et al. 2012a). DO variability in CWs under different flow regimes,
across both short and long time-scales, is still poorly understood. Usually DO levels in
a CW are monitored during the daytime and therefore, it is important to understand
how DO changes over the diurnal cycle. It is challenging to estimate metabolism in
stormwater CWs as these systems are dynamic and complex from both hydrological
and biogeochemical perspectives. A few studies have measured DO dynamics in
CWs, monitoring GPP and R (and therefore metabolism) as an indicator of wetland
function (Cronk & Mitsch 1994a; Tuttle et al. 2008; Reeder 2011). However, in these
studies DO was monitored over a relatively short timescale, from several days to a
week, and a lake metabolism definition was used. As discussed above, during the dry
summer CWs can transition to lentic systems (under these conditions the lake
metabolism definition is appropriate). However, early in the wet winter, storm event
pulses or frequent rainfall can shift the CW to a lotic system. These lentic-lotic
transitions make it complex to estimate metabolism in the CWs in Mediterranean
climates. Also, phytoplankton are typically the main primary producers in lakes, while
dense aquatic macrophytes are typically the main primary producers in CWs. The
different seasonal dynamics of these two types of primary producers have distinct
influences on metabolism. The summer macrophyte senescence experienced in
Mediterranean climates can input significant C loadings to the CWs at a time of very
long HRT and high temperatures leading to higher rates of oxygen consumption
indicating as biological oxygen demand (BOD), chemical oxygen demand (COD) and
sediment oxygen demand (SOD). Moreover, water levels can decrease during summer
lentic conditions, due to high evaporative demand. Thus the Mediterranean climate
imposes a strong signature on the wetland metabolism because the wetland carbon
budget is influenced by both vegetation mortality (Stephens 2005) and changes in
hydrology (Reddy & DeLaune 2008; Tuttle et al. 2008; Page & Dalal 2011). Changes
Chapter 2
29
in DO concentrations, taken over appropriate time and length scales, may be a simple
but useful indicator of wetland metabolism that can be used under both lentic and lotic
conditions, however further work is needed to investigate how this indicator impacts
on nutrient attenuation under different hydrological regimes.
Figure 2. 3 Schematic representation of the main phase of above ground biomass of
macrophytes including leaf expansion, maturity and senescence. Leaf maturity begins hen leaf
expansion is over and ends with the first senescence symptoms. During the maturity phase, the
leaf faces numerous sub-lethal events leading to many chronic senescence syndromes and
recovery events. Leaf-senescence no-return syndrome is characterized by a succession of
degradation process that will lead to death. The figure is modified after Guiboileau et al.
(2010).
2.5.6 Effectiveness of CW restoration to improve nutrient attenuation
CWs attenuate nutrients effectively in the first couple of years after the
construction; however this effectiveness can decrease over time due to periodic
accumulation of nutrients and organic matter in the sediments (Craft 1996; Hancock
2002), degradation of the riparian zones and alteration of water flow path (Harrison et
al. 2012), diffusion and resuspension of nutrients from the sediments (O'Brien et al.
2012), senescence of plant material that then releases nutrients to the system (Kröger
et al. 2007), development of anoxic environments and input of polluted groundwater
(Gabriele et al. 2013; Teufl et al. 2013). In the long term, CW can be a source of
nutrients, and therefore CW restoration is a key management intervention to maintain
nutrient attenuation (Jenkins & Greenway 2007). This restoration effort may include
modification of the CW main channel (Seitzinger et al. 2002), changes in water flow
paths (Craig et al. 2008), introduction of obstacles to influence flow direction
(Groffman et al. 2005), modifications to increases the water contact with hyporheic
Chapter 2
30
zone, benthos and macrophytes (Craig et al. 2008; Hester et al. 2009; Gordon et al.
2013), provision of environments with available carbon sources (Bernhardt & Likens
2002) and the creation of denitrification hotspots (Craig et al. 2008). However, how
these restoration efforts impact on the attenuation of different nutrient species is
poorly understood. There is also little clarity on how long such restoration features
continue to improve nutrient attenuation and whether they remain effective under
highly variable hydrological and biogeochemical conditions. An assessment of the
impact of CW restoration nutrient attenuation, and how that changes with time is
needed.
Chapter 3
Nutrient attenuation in a surface flow
constructed wetland experiencing summer low
flow and macrophyte senescence
This chapter is based on the manuscript published as: T. M. Adyel, M. R. Hipsey and
C. E. Oldham (2017) Temporal dynamics of stormwater nutrient attenuation of an
urban constructed wetland experiencing summer low flows and macrophytes
senescence. Ecological Engineering, 102, 641-661.
Chapter 3
32
Abstract
The attenuation of urban stormwater nutrient is essential for maintaining
ecological health of downstream waterways. Constructed wetlands (CWs) are widely
used for stormwater management due to the economic and environmental benefits
they create by attenuating nutrients. This research assessed stormwater nutrient
attenuation in a surface flow CW located in Western Australia, by assessing data
spanning decadal, seasonal and diurnal time-scales. The study site was built in 2004
and restored in 2010, and the significance of such restoration was also assessed in this
research. The CW designed to minimize nutrient loads to downstream sensitive
waterways despite experiencing Mediterranean climates with strong hydrological
seasonality including prolonged low or no flow conditions during the dry summer
periods and nutrient-rich pulses during the wet winter periods. Furthermore, seasonal
ungauged water inputs and macrophytes senescence made the system complex to
optimize nutrient attenuation. The annual average attenuation of nitrogen (N) and
phosphorus (P) based nutrients were 2–4 times higher during the post-restoration
periods than that observed during the pre-restoration periods. The attenuation of
inorganic, organic, dissolved and total N was higher during the dry periods of the
post-restoration regimes, while P species showed higher attenuation during the wet
periods of the post-restoration regimes. Inorganic N showed greater attenuation than
attenuation of inorganic P throughout the year. Baumea articulata and Schoenoplectus
validus were the dominant macrophyte species across the site; B. articulata was
estimated to store about 3 and 2 times higher N and P pools, respectively than that
stored in S. validus. Low dissolved oxygen (DO) occurred in the site and DO showed
a decreasing trend over the decade and a distinct diurnal signature of day-time peaks
and night-time anoxia. Diurnal pattern in aquatic metabolism, defined as the daily
fluctuation of DO levels within the water column of CW, was related to solar radiation
and nutrient concentrations, and could be considered as a proxy of CW function.
Keywords
Dissolved oxygen; Macrophytes senescence; Nutrient attenuation; Stormwater;
Water Sensitive Urban Design; Wetland restoration
Chapter 3
33
3.1 Introduction
Water quality is declining in many urban areas due to non-point source
nutrients, typically delivered via stormwater (Roy et al. 2008; Wong et al. 2012),
leading to nuisance algal blooms and eutrophication in downstream waterways
(Taylor et al. 2004; Hathaway et al. 2012). Stormwater and/or associated nutrient can
also lead to habitat fragmentation, biodiversity loss and changes in the morphological
and hydrological characteristics of the watercourses (Taylor et al. 2004; Walsh et al.
2005a; Francey et al. 2010; Zgheib et al. 2012). As a result, management of
stormwater quality has become an essential strategy to improve ecological integrity of
drainage systems, and as a means to improve urban sustainability and liveability (Hatt
et al. 2007; Wong et al. 2012; Deletic et al. 2013).
A paradigm shift in stormwater management has therefore been taking place in
many cities around the world based on Integrated Stormwater Management System
(ISWMS) and through adoption of Best Management Practices (BMPs). Such
initiatives have been referred as Low Impact Development (LID) in the USA (Hinman
2005), Sustainable Urban Drainage Systems (SUDS) in the UK (CIRIA 2000) and
Water Sensitive Urban Design (WSUD) in Australia (Lloyd 2001). In general, WSUD
are based on the idea that the management, protection and conservation of the urban
water cycle must take place to ensure that the urban water environment is sensitive to
natural hydrological, economic and ecological processes (Deletic et al. 2013), and that
the ecological services of urban waterways are maintained (Wong & Brown 2009).
There are different WSUD elements are in practice for stormwater nutrient
attenuation, and constructed wetland (CW) is one of them.
CWs have found widespread adoption over the past few decades due to their
ability to treat and transform stormwater nutrient by utilizing different hydrological
and biogeochemical processes. These processes take place within the water,
soils/sediments, aquatic macrophytes and microbial assemblages of CWs (Carleton et
al. 2000; Carleton et al. 2001; Kivaisi 2001; Mitsch & Gosselink 2007; Vymazal
2007). CWs vary widely in their performance ranging from 10 to 90% for nitrogen
(N) and phosphorus (P), and in some cases they may increase nutrient levels (Rushton
et al. 1995; Saeed & Sun 2012). The performance depends on numerous factors
including the area and land use of the catchment, nutrient loading rate, hydro-
climatological variability, macrophyte biomass, hydraulic retention time (HRT),
groundwater connectivity and specific sediment/soil properties (Hamilton et al. 1993;
Chapter 3
34
Carleton et al. 2001; Gasperi et al. 2011; Zgheib et al. 2011; Zgheib et al. 2012; Tang
et al. 2013). The degree of nutrient attenuation in CWs is controlled
biogeochemically and/or hydrologically (Fink & Mitsch 2007; Huang et al. 2011;
Chang et al. 2013; Jahangir et al. 2016). N and P of stormwater experience different
transformations depending on the CW’s biogeochemical conditions. These conditions
include oxygen concentrations, availability of a carbon (C) source, macrophytes and
microorganism abundance, soil/sediment properties and degree catchment-riparian-
hyporheic exchange (Kadlec & Knight 1996). For example, a lack of oxygen
suppresses nitrification, and triggers denitrification and dissimilatory reduction of
nitrate to ammonia (DNRA) (Harrison et al. 2012). The degree to which
biogeochemical processing of nutrients can occur, however, is highly dependent on
hydrological conditions within the water and groundwater regions of the CW. Higher
HRT increases the water contact with aquatic macrophytes and benthos, and usually
facilitates nutrient attenuation (Kadlec & Knight 1996). Furthermore, riparian zone
saturation exports nutrients to the overlying water column (Ranalli & Macalady 2010).
Increased flow velocities after the storm events flush senescent and floating biomass
from the system, thereby promoting aquatic productivity through the improved light
availability and higher temperature within the water column (Tuttle et al. 2008).
However, water is unevenly distributed in the CW because of channelization,
microtopography and spatial patterns within macrophyte beds (Kadlec & Knight
1996), and hence different parts of the CW can function differently (Fink & Mitsch
2007). These variabilities complicate the optimisation of the CW design and function;
in particular the function of CWs in the Mediterranean climates is less clear. The
effects of prolonged low flows and senescence of macrophytes during the summer
periods on the CW function in the Mediterranean climates are not well understood.
Furthermore, the performance of CW also varies over time that makes it complex for
system optimisation.
In the first few years after the construction, CWs function most effectively
predominantly due to available sites for absorbing nutrients (Omari et al. 2003;
Malaviya & Singh 2012), the settling of particulate nutrients within the system (Reddy
et al. 1999; Kadlec & Wallace 2009), higher nutrient accumulation by macrophytes in
their growth phase (Mustafa et al. 2009; Vymazal 2013a), intense hyporheic exchange
(Ranalli & Macalady 2010), and high photic depth and oxygen availability within the
water column (Stein & Paul 2005). The effectiveness of CWs can decrease with time
Chapter 3
35
due to senescence and accumulation of plant material that subsequently releases
nutrients back to the system (Kröger et al. 2007), accumulation of nutrients and
organic matter in the sediments (Craft 1996; Hancock 2002), degradation of the
riparian zones and development of preferential flow paths (Harrison et al. 2012),
diffusion and bioturbation of nutrients from the sediments (O'Brien et al. 2012),
development of anoxic environments and the input of polluted groundwater (Gabriele
et al. 2013; Teufl et al. 2013). Therefore, in the long term CWs can become a source
of nutrients and may require periodic restoration. However, predicting the optimum
time for CW restoration is challenging.
CWs restoration is undertaken to re-establish the ecological functions of both
biotic and abiotic components (Fernández & Novo 2007), thereby improving nutrient
attenuation (Bukaveckas 2007; Craig et al. 2008; Gabriele et al. 2013). Physical
restoration can involve the modification of stream channels (Seitzinger et al. 2002),
changes in water flow paths (Craig et al. 2008), earthworks in the riparian zones
(Craig et al. 2008), introduction of artificial woody debris or dams to influence flow
direction (Groffman et al. 2005) and incorporation of permeable layer of engineered
materials or substrates (Saeed & Sun 2012). Restoration can also include the re-
establishment of macrophytes and routine harvesting of mature macrophytes (Brown
2000; Vymazal 2013a), modifications to increase the water contact between hyporheic
zone, benthos and macrophytes (Craig et al. 2008; Hester et al. 2009; Gordon et al.
2013), provision of environments with available C sources (Bernhardt & Likens 2002)
and the creation of denitrification hotspots (Craig et al. 2008). However, in the
Mediterranean climate, CWs experience prolonged periods of no or low surface flows
during the summer, episodic high flows after storm events and possible groundwater
input during the winter. Moreover, senescent macrophytes can release organic matter
during the period when temperatures are high and CWs dry out that ultimately leads to
anoxic conditions. Therefore, major feedbacks between hydrological and/or
biogeochemical processes need to be identified for optimal restoration and subsequent
nutrient attenuation.
In the present study, we assessed the nutrient attenuation in a surface flow CW
situated on the Swan Coastal Plain (SCP) of Western Australia (WA). The studied
CW experiences Mediterranean climates with prolonged periods of no to low flows
and macrophytes senescence, particularly during the summer periods. In dry periods
local groundwater can contribute flow within the CW. During the wet periods,
Chapter 3
36
stormwater runoff from residential, commercial and industrial sites is collected by
surface and subsurface drainage system and discharged to the CW. We explored
dynamics of nutrient attenuation over diurnal, seasonal and decadal time-scales
through synthesis of historical meteorological data, water quality and quantity data,
sediment and macrophyte quality data, and high frequency dissolved oxygen (DO)
data. Our analysis adopted a whole-system approach to investigate the influence of
restoration, seasonal prolonged low flows, episodic flows and macrophyte senescence
on the overall nutrient dynamics. The results also highlighted the variability in CW
function, and demonstrate the importance of changes in CW metabolism regarding
nutrient dynamics.
3.2 Materials and methods
3.2.1 Site background and restoration works
The study focused on the Anvil Way Compensation Basin (AWCB), situated
in the Mills Street Main Drain (MSMD) sub-catchment of WA (Figure 3. 1). MSMD
drains stormwater from a 120 ha residential and light industrial area with high density
traffic ways. The AWCB was initially built in 2004 as a drainage compensating basin
to retain small amount of stormwater and reduce the export of stormwater pollutant to
the Canning River. The basin was restored to a “living stream” in November 2010.
The restoration efforts included the removal of approximately 1400 m3 metal- and
hydrocarbon-containing sediments, the introduction of a sedimentation pond close to
the main inlet to trap particulate material and the creation of a meander through the
basin to increase water contact with the wetland macrophytes and benthos during low
flow conditions. Macrophytes were also planted at benches or riparian zones to
encourage sedimentation and nutrient removal and three low sand bars were
introduced to reduce risk of short-circuiting of the flow between the inlet and outlet.
The AWCB is now roughly triangular shaped and 0.9 ha in area.
3.2.2 Data collection
3.2.2.1 Meteorology
Solar radiation data (August 2014 to August 2015) was obtained from the
weather station of the Department of Agriculture and Food (DAFWA), South Perth,
located approximately 5 km north-west of the AWCB. Daily rainfall data (January
2011 to August 2015) was obtained from the Welshpool depot of the Department of
Chapter 3
37
Water (DoW) and the Perth Airport Station of the Bureau of Meteorology (BoM).
These stations are approximately 0.8 km north-west and 5 km north-south of the
AWCB, respectively.
3.2.2.2 Water quantity
5-min intervals water discharge and level data at the AWCB inflow and
outflow stations were obtained from DoW. A Starflow ultrasonic instrument
(UnidataTM, Australia) at the inflow and a float well sensor at the outflow were used.
Inflow and outflow data were only available since 2009 and 2012, respectively. The
water levels of four groundwater bores (G1, G2, G3 and G4) were also logged by
DoW. Bores G1to G3 and G4 began operation in April 2010 and January 2012,
respectively. Bore covers were trafficable and flush with the ground surface.
Groundwater levels were monitored on a monthly basis and depth from top of the
casing was converted to Australian Height Datum (AHD).
3.2.2.3 Water quality
Historical unfiltered and filtered water samples at fortnightly or monthly
(September 2004 to May 2015) intervals were collected from the main inlet (W1),
Mars Street Drain or MSD (W2) and main outlet (W3) by DoW. We also collected
opportunistic (November 2013 to May 2015) and pre-storm event (12 to 13 March
2015) water samples at three sites i.e., W1, W2 and W3 (Table 3. 1). Pre-storm event
samples were collected at about 2-h intervals spanning day and night regimes. Water
samples were filtered on-site using 0.45 μm membrane filter, stored on ice until the
returning to the laboratory, and then frozen until analysis. Historical water samples
were analysed for filtered ammonia (NH3), filtered total oxidised nitrogen (NOx), total
Kjeldahl nitrogen (TKN), unfiltered total nitrogen (TN), filtered reactive phosphorus
(FRP) and unfiltered total phosphorus (TP) using standard method (APHA 1998).
Historical quarterly water samples were also collected and analysed for dissolved
organic nitrogen (DON). Opportunistic and pre-storm event samples were analysed
for NH3, filtered nitrite (NO2), filtered urea nitrogen (UreaN), filtered total nitrogen
(FTN), TN, FRP, filtered total phosphorus (FTP), TP, dissolved organic carbon
(DOC) and dissolved inorganic carbon (DIC) using Lachat Quick Chem procedure
(APHA 2012). We calculated other nutrient parameters based on analysis results of
opportunistic and pre-storm event samples (Table 3. 1).
Chapter 3
38
Figure 3. 1 Map of the Anvil Way Compensation Basin showing monitoring and sampling
points. Sampling points include surface water (W1, W2 and W3), groundwater (G1, G2, G3
and G4), sediment (S1, S2, S3 and S4) and macrophyte (M1, M2, M3 and M4). Stormwater
enters to the system at the main inlet (W1) and Mars Street Drain (W2), and exits at the main
outlet (W3) into the Canning River. The AWCB is mainly dominated by B. articulata and S.
validus. P1, P2 and P3 indicate the porewater monitoring points at a transect within riparian
zone.
Canning River
DO logger
P3
P2
P1
Transect
Chapter 3
39
During the historical water sampling, in situ DO concentrations were also
recorded at three sites by DoW. During opportunistic and pre-storm sampling, we also
measured DO concentrations in the main stream water at 10-min intervals using a
miniDOT DO logger (PMA, USA) from July 2014 to May 2015. An additional DO
logger (D-opto, New Zealand) was also installed at W3 to capture DO signals at 10-
min intervals during the pre-storm event.
DoW collected and analysed groundwater samples for NH3, NOx, TKN, TN,
FRP and TP using standard method (APHA 1998). Groundwater samples were also
monitored for DO concentrations using in situ logger (Table 3. 1).
3.2.2.4 Sediment quality
DoW collected sediment samples from November 2011 to May 2015 on about
half yearly at in-stream sites i.e., at inlet (S1) and outlet (S4), and two bench sites i.e.,
S2 and S3 (Figure 3. 1) Average sediment sampling depth was about 10–15 cm.
Sediments were analysed for TKN, TP and total organic carbon (TOC). Details
sampling and analysis procedure are described elsewhere (GHD 2008).
Soil/sediment oxidation-reduction potential (Eh) and porewater nutrient
concentrations were monitored at three points (P1, P2 and P3) over a 2 m transect
extending from the water's edge to the riparian zone (Table 3. 1). P1, P2 and P3 were
approximately 60, 40 and 20 cm below the soil surface and each point was
horizontally separated by 50 cm. Three Eh sensors (Sensorex, USA) were inserted
vertically into these points and connected to a data logger (DaqPro, USA) recording at
30-min intervals. One suction cup (UTM, Germany) was also placed at each
porewater monitoring point (same depth of Eh sensors) with the other end attached to
sampling bottle of 500 mL capacity. The suction cups periodically collected water
samples from soil porewater and stored them in the sampling bottles for later analysis
(Table 3. 1).
A perforated PVC pipe was inserted up to 30 cm from soil surface in the
transect. A CTD (conductivity, temperature and depth) diver (Schlumberger,
Germany) was placed at the bottom of the pipe to measure water level/pressure. The
CTD pressure changes were compared with an on-site atmospheric baro diver (Onset
HOBO, USA). Electric conductivity (EC) and temperature were monitored at 10-min
intervals.
Chapter 3
40
Table 3. 1 Summary table of spatial and temporal dynamics of studied nutrient parameters,
data type, sampling frequency and data source for the Anvil Way Compensation Basin
assessment.
Component Unit Nutrient Data type Sampling
sites
Duration Sampling
frequency
Analytical
method
Data
source
Surface
water quality
mg/L NH3, NOx,
TKN, TN,
FRP, TP
Historical W1, W2,
W3
September
2004 to
May 2015
Fortnightly
to monthly
APHA
(1998)
DoW
DON Quarterly
NH3, NOx,
NO2, NO3,
FTN, TN,
UreaN,
DIN, PN,
DON,
NUON,
FRP, FTP,
TP, PP,
DOP, DIC,
DOC, TDC
Opportunistic W1, W2,
W3
November
2013 to
May 2015
Fortnightly
to monthly
APHA
(2012)
This
study
NH3, NOx,
FNO2, NO3,
FTN, TN,
UreaN,
DIN, PN,
DON,
NUON,
FRP, FTP,
TP, PP,
DOP, DIC,
DOC, TDC
Pre-storm
event
W1, W2,
W3
12 to 13
March
2015
2–3-h
intervals
APHA
(2012)
This
study
Groundwater
quality
mg/L NH3, NOx,
TKN, TN,
FRP, TP
Historical G1, G2,
G3, G4
April
2010 to
May 2015
Monthly APHA
(1998)
DoW
Soil
porewater
quality
mg/L NH3, NOx,
NO2, NO3,
FTN, DIN,
DON,
NUON,
FRP, FTP,
DOP, DIC,
DOC,TDC
Opportunistic P1, P2, P3 11 to 17
March
2015
2–3-h
intervals
APHA
(2012)
This
study
Sediment
quality
mg/kg TKN or TN,
TP, TOC
Historical S1, S2,
S3, S4
November
2011 to
May 2015
Variable,
mainly half
yearly
GHD
(2008)
DoW
Macrophyte
quality
mg/kg TKN or TN,
TP
Historical M1, M2,
M3, M4
October
2012 to
October
2013
Variable,
mainly half
yearly
GHD
(2008)
DoW
Analysed nutrient species: NH3: Filtered ammonia; NOx: Filtered total oxidised nitrogen; NO2: Filtered nitrite; UreaN: Filtered urea nitrogen; FTN: Filtered total nitrogen; FRP: Filtered reactive phosphorus; FTP: Filtered total phosphorus; TN: Unfiltered
total nitrogen; TP: Unfiltered total phosphorus; DIC: Dissolved inorganic carbon; DOC: Dissolved organic carbon.
Calculated nutrient species: NO3: Filtered nitrate = NOx – NO2; DIN: Filtered inorganic nitrogen = NOx + NH3; PN: Particulate
nitrogen = TN – FTN; DON: Dissolved organic nitrogen = FTN – NH3 – NOx; NUON: Filtered non-urea organic
nitrogen = DON – UreaN; PP: Particulate phosphorus = TP – FTP; DOP: Filtered organic phosphorus = FTP – FRP; TDC: Total dissolved carbon = DIC + DOC.
3.2.2.5 Macrophyte properties
A natural resource management group, the South East Regional Centre for
Urban Landcare (SERCUL), WA planted the AWCB with Schoenoplectus validus,
Chapter 3
41
Baumea articulata, B. juncea, Juncus pallidus and J. kraussii after the restoration.
Recent (June 2012) ground-truthing of species coverage found S. validus and B.
articulata as dominant species. Macrophyte samples at four points (M1, M2, M3 and
M4) were collected using 0.0625 m2 quadrat on about half yearly from October 2012
to October 2013. Macrophyte samples were analysed for TKN and TP concentrations
in the above ground biomass (AGB) such as shoots and leaves, and below ground
biomass (BGB) such as roots and rhizosphere (GHD 2008).
Floating macrophytes, weeds, algal mats, debris or a combination of all
accumulated periodically in the AWCB main stream. Images of main stream coverage
by these floating materials between 1 August 2014 and 28 June 2015 were obtained
from the NearMap (Link: http:au.nearmap.com).
3.2.3 Data analysis
3.2.3.1 Nutrient attenuation
CW efficiency is typically calculated as the difference between the influent
and effluent nutrient loads. Nutrient concentrations and water flows need to be
monitored simultaneously at the inflow and outflow of CW to calculate the load.
Quantification of nutrient attenuation within the AWCB, however, was complicated
due to presence of ungauged inputs such as MSD (W2) and possible groundwater.
Furthermore, simultaneous flow data was available at the main inflow and outflow
from October 2012, and hence the calculation of nutrient load attenuation, and how it
changed over time, was not always possible. Therefore, an alternative approach, the
standardised delta concentration, SDC (%), was adopted for assessing nutrient
attenuation over time. The overall estimation of nutrient attenuation was performed in
two ways where scenario 1 considered SDC as standardised difference in nutrient
concentrations between the inlet and outlet (Adyel et al. 2016):
𝑆𝐷𝐶 =𝐶𝑖𝑛𝑙𝑒𝑡−𝐶𝑜𝑢𝑡𝑙𝑒𝑡
𝐶𝑖𝑛𝑙𝑒𝑡 × 100 (3.1)
where 𝐶𝑖𝑛𝑙𝑒𝑡 is nutrient concentration at the inlet (mg/L) and 𝐶𝑜𝑢𝑡𝑙𝑒𝑡 is nutrient
concentration at the outlet (mg/L).
From the water balance analysis, we expected that MSD and groundwater
inputs would contribute about 20–70% water depending on the time of the year and
available rainfall (Adyel et al. 2015a; Ruibal-Conti et al. 2015). Consequently,
Chapter 3
42
scenario 2 was used to estimate averaged SDC, SDC_ave (%) where inlet nutrient
concentration was the average concentration of W1 and W2:
𝑆𝐷𝐶_𝑎𝑣𝑒 =𝐶∗
𝑖𝑛𝑙𝑒𝑡−𝐶𝑜𝑢𝑡𝑙𝑒𝑡
𝐶∗𝑖𝑛𝑙𝑒𝑡
× 100 (3.2)
where 𝐶∗𝑖𝑛𝑙𝑒𝑡 is the average nutrient concentration of the main inlet (W1) and MSD
(W2) (mg/L) and 𝐶𝑜𝑢𝑡𝑙𝑒𝑡 is the nutrient concentration at the main outlet (W3) (mg/L).
Although this approach was not able to provide an exact value of CW
efficiency, it was able to: (a) provide a range of performance values for different
compounds of interest, (b) highlight the importance of the impact that ungauged
sources have on the assessment, and (c) demonstrate the relative differences in
performance over time. When simultaneous flow and nutrient data are available at
main inlet and outlet, we calculated incoming and outgoing nutrient load and
subsequently load attenuation, as:
𝐿 = 𝑄𝑖 × 𝐶𝑖 (3.3)
𝐿𝐴 =∑ 𝐿𝑖−∑ 𝐿𝑜
∑ 𝐿𝑖 × 100 (3.4)
where L is the nutrient load (mg/s), Qi is the flow rate at ith timestep (m3/s), Ci is the
nutrient concentration at ith timestep (mg/L), LA is nutrient load attenuation (%), Li is
the load at the inlet (mg/s) and Lo is the load at the outlet (mg/s).
3.2.3.2 Sediment nutrient mass and porewater properties
The mass of sediment nutrient (TKN or TN, TP and TOC) was calculated for the
bench and in-stream sites assuming that each of the sampling site was representative
of a specified area of the AWCB:
𝑁𝑀𝑠 = 𝑁𝐶𝑠 × 𝐵𝑠 × 𝐴𝑠 × 𝐷𝑠 (3.5)
where NMs is the sediment nutrient mass (kg), NCs is the sediment nutrient
concentration (mg/kg), Bs is the sediment bulk density (kg/m3), As is sediment area
(m2) and Ds is sediment depth (m).
Chapter 3
43
In addition, nutrient concentrations of riparian soil/sediment porewater were
analysed to assess nutrient variability due to riparian zones saturation (Table 3. 1).
3.2.3.3 Macrophytes nutrient mass and main stream coverage
S. validus was sampled at four macrophyte sampling sites on three sampling
regimes (October 2012, May 2013 and October 2013) while B. articulata was only
sampled at M1 and M2 on October 2012 and May 2013 due to minimal coverage at
M3 and M4. S. validus was opportunistic and quickly grew into any gaps, especially
in the areas where B. articulata was sampled, while B. articulata was long lived. TKN
and TP pools in AGB and BGB at four sampling points were estimated assuming each
of the sampling site was representative of a specified area of the AWCB:
𝑁𝑀𝑏 = 𝑁𝐶𝑏 × 𝐷𝑊𝑏 × 𝐴𝑏 × 𝑄𝑏 (3.6)
where NMb is the biomass nutrient mass (kg), NCb is biomass nutrient concentration,
DWb is dry weight of the biomass (kg), Ab is the biomass area (m2) and Qb is quadrat
area as 0.0625 m2.
NearMap images of the AWCB were digitized to estimate the main stream
coverage by floating material.
3.2.3.4 Hydrological assessment
Beside the main inflow, stormwater entered to the AWCB through MSD and
probably groundwater, particularly during the storm events and wet periods. The
water input from local sources and ungauged areas into the AWCB was calculated
considering rainfall pattern from June 2012 to December 2013. The volumetric
contribution for ungauged areas (VCUA) was computed by numerical integration of
the discharge hydrograph over the specified duration of the event (Chow et al. 1999).
Selected small rainfall events of short duration (e.g. pulses) were used to highlight the
contribution of local catchment areas to the AWCB. Large rainfall events (> 25 mm)
of long duration (∼24 h) or event hydrographs presenting peak discharges above the
0.6 m3/s value, were used to highlight the potential maximum volumetric contribution
and peak discharge values from ungauged areas (Ruibal-Conti et al. 2015).
The volume of the AWCB was calculated using a Digital Elevation Model
(DEM). Height and volume ratio relationship was obtained from site construction
Chapter 3
44
plan. HRT of the AWCB was established using the volume of the system and the
outflow rate:
𝐻𝑅𝑇 =𝑉𝑎
𝑄𝑜𝑢𝑡𝑙𝑒𝑡 (3.7)
where 𝑉a is volume of the AWCB (m3) and Qoutlet is the discharge rate at the outlet
(m3/s).
3.2.3.5 DO as a function of solar radiation
The DO logger at the mid-channel water recorded DO concentrations and
saturation at 10-min intervals. Based on available data, the delta DO saturation,
ΔDOsat (%) was calculated as a proxy of CW metabolism for each day according to:
∆𝐷𝑂𝑠𝑎𝑡 = 𝐷𝑂𝑚𝑎𝑥 − 𝐷𝑂𝑚𝑖𝑛 (3.8)
where DOmax is the maximum DO saturation at the afternoon (%) and DOmin is the
minimum DO saturation at the morning (%).
At the same time, daily total exposure of solar radiation over the period
between DOmin and DOmax was computed. Average monthly ΔDOsat and solar
exposure were then considered to examine the change of DO saturation over months
and seasons.
3.3 Results
3.3.1 Long term changes in wetland properties
Nutrient concentrations at the two inlets (W1 and W2) and the outlet (W3)
were averaged over periods defined as ‘dry’ (December to May) and ‘wet’ (June to
November) under both pre- and post-restoration regimes (Table 3. 2). Mean NH3,
NOx, TN, FRP and TP concentrations at W2 were higher than that measured at W1,
irrespective of the periods or restoration regimes. DON concentrations were higher at
W1 compared to that observed at W2 during the dry periods and vice-versa during the
wet periods under both pre- and post-restoration regimes. NH3 concentrations at W3
during the wet periods of pre-restoration periods were almost double than those
measured during the wet periods of post-restoration periods. NOx concentrations at
W3 during the wet periods were higher than that observed during the dry periods
Chapter 3
45
under both pre- and post-restoration regimes. TKN and DON concentrations were
variable at W3. TN concentrations at W3 were lower during the dry periods than that
of the wet periods, while TP concentrations showed the reverse. Since 2014,
additional nutrient species were measured in the surface water (Table 3. 2). FTN, PN
and DIN concentrations at W3 during the dry periods were less than the combined
mean concentrations measured at W1 and W2. The opposite scenario was found for
NO2, FTN, UreaN, DIN, NO3 and FTP concentrations. W3 showed average higher
DIC and DOC concentrations than that measured at W1, while both species showed
the highest concentrations at W3 during the wet periods.
Annual average nutrient attenuation as SDC was assessed from 2004 to 2015
(Figure 3. 2 and Figure 3. 3). Inorganic and dissolved N fractions like NOx and NH3
showed higher attenuation than the total and organic fractions like TN and DON
(Table 3. 3). Moreover, NH3, TKN and TN attenuation increased after the restoration.
Annual mean NOx attenuation was always positive (Figure 3. 2d) with slight decrease
immediately after the restoration; however NOx showed an increasing attenuation
trend three years later (Figure 3. 2c, d). DON attenuation also increased slightly after
the restoration, however, was negative thereafter (Figure 3. 2g,h). FRP and TP showed
periodic release and attenuation during the pre-restoration, while attenuation increased
after the restoration (Figure 3. 3). Annual average FRP and TP attenuation was
negative for the first six years of monitoring. TP attenuation then increased and
reached at the maximum immediately after the restoration (Figure 3. 3b, d). However,
TP and FRP started to be released from the AWCB four years after the restoration.
Overall, the attenuation of inorganic N was greater than inorganic P. Inorganic N and
TP attenuated more effectively during the dry and wet periods, respectively (Table 3.
3). DIN showed higher attenuation than DON, while DIC and DOC released (Table 3.
3). Restoration features improved attenuation of NH3, TKN, TN, FRP and TP in the
AWCB. Using student’s t-test for independent samples, Mann-Whitney’s normalized
test and normalized Wilcoxon’s T test against the normal distribution of data, we
found that attenuation during post-restoration periods were statistically significant for
NH3 (p < 0.010), TKN (p < 0.002), TN (p < 0.001) and TP (p < 0.041).
Using student's t tests for independent samples, Mann-Whitney's normalized
test and normalized Wilcoxon's T test against the normal distribution of data, it was
found that attenuation during the post-restoration periods were statistically significant
for NH3 (p < 0.010), TKN (p < 0.002), TN (p < 0.001) and TP (p < 0.041). As the
Chapter 3
46
AWCB received water and nutrients from multiple sources, there were some
uncertainties of nutrient attenuation. SDC estimation didn’t consider the influence of
the MSD and therefore SDC_ave was applied since it combined mean concentrations
at W1 and W2 as input concentrations (Figure 3. 4 and Figure 3. 5). Nutrient
attenuation as SDC_ave increased compared to SDC for all nutrients species
irrespective of periods or restoration regimes. Moreover, SDC_ave was positive for all
species, except DIC (Table 3. 3).
Load attenuation indicated the performance of the AWCB when (a)
simultaneous flow and nutrient concentration were available at the inlet and outlet,
and (b) influence of MSD was relatively small in respect of water flow and nutrient.
However, these condition were available is some occasion since end of 2012. NH3
showed higher load attenuation among N based nutrients with the efficiency being
above 50%. This was equivalent to a mass removal between 0.1 and 11 kg/day. TKN,
TN and NOx were subsequently removed with less efficiency. Although there was
limited DON data, it suggested that this fraction was hardly removed. This was
consistent with the observation of low SDC for DON (Figure 3. 2g,h). FRP showed
higher load attenuation (up to 70%) compared to TP, however both nutrients showed
periodic release and attenuation. Furthermore, P load released during high flow
conditions caused by sudden pulse after prolonged dry or low flow conditions.
3.3.2 Nutrient accumulation or uptake in the sediment and macrophytes
The bench sites (S2 and S3) had a greater sedimentation area and therefore,
accumulated more nutrient mass than that stored in the in-stream sites (S1 and S4)
(Figure 3. 6). S4 had 3–4 times more nutrient mass than that of at S1. Nutrient mass
increased sharply at the bench sites from November 2011 to May 2013. No data was
obtained after May 2014 at the bench sites. Net assimilation of TN and TOC varied
consistently with time, while P assimilation was less clear (where t is the days since
11 November 2011):
Chapter 3
47
Table 3. 2 Spatial and temporal variability of nutrient concentrations in the Anvil Way Compensation Basin. The dry and wet periods represent
December to May and June to November, respectively.
aThe Australian and New Zealand Environment and Conservation Council (ANZECC) guideline values
bHalthy River Action plan (HRAP) target values
n/a not applicable
Nutrient
species
Dry periods Wet periods
Pre-restoration Post-restoration Pre-restoration Post-restoration
Main inlet
(W1)
(mg/L)
Mars
Street
Drain
(W2)
(mg/L)
Man
outlet
(W3)
(mg/L)
Outlet
conc.
within
target
values
(%)
Main inlet
(W1)
(mg/L)
Mars
Street
Drain
(W2)
(mg/L)
Man outlet
(W3)
(mg/L)
Outlet
conc.
within
target
values
(%)
Main inlet
(W1)
(mg/L)
Mars
Street
Drain
(W2)
(mg/L)
Man
outlet
(W3)
(mg/L)
Outlet
conc.
within
target
values
(%)
Main inlet
(W1)
(mg/L)
Mars
Street
Drain
(W2)
(mg/L)
Man outlet
(W3)
(mg/L)
Outlet
conc.
within
target
values
(%)
NH3 0.15±0.16 0.41±0.28 0.10±0.14 73.53a 0.39±0.37 0.42±0.11 0.11±0.12 42.86
a 0.15±0.16 0.42±0.24 0.15±0.21 54.28
a 0.16±0.16 0.18±0.17 0.07±0.06 60.71
a
NOx 0.16±0.18 0.16±0.18 0.05±0.06 94.12a 0.14±0.19 0.44±0.36 0.03±0.04 100
a 0.40±0.24 0.49±0.48 0.27±0.23 40
a 0.52±0.31 0.54±0.44 0.41±0.32 23.3
a
TKN 1.13±0.47 1.34±0.46 1.26±0.46 n/a 1.27±0.46 1.24±0.36 1.01±0.28 n/a 0.97±0.37 1.11±0.53 1.03±0.37 n/a 0.88±0.26 0.99±0.63 0.73±0.20 n/a
DON 0.75±0.25 0.72±0.27 0.81±0.24 n/a 0.70±0.14 0.51±0.18 0.73±0.15 n/a 0.63±0.26 0.69±0.46 0.63±0.25 n/a 0.46±0.17 0.58±0.49 0.48±0.17 n/a
TN 1.25±0.47 1.52±0.51 1.33±0.42 44.18a
23.53b
1.33±0.45 1.68±0.74 0.99±0.29 79.31a
62.07b
1.34±0.48 1.58±0.74 1.36±0.71 5143a
28.57b
1.28±0.59 1.28±0.74 1.01±0.49 46.45a
66.67a
FRP 0.07±0.06 0.17±0.19 0.08±0.05 26.47a 0.15±0.14 0.51±0.78 0.11±0.11 26.93
a 0.05±0.02 0.09±0.13 0.04±0.02 57.14
a 0.05±0.02 0.09±0.06 0.04±0.02 48.38
a
TP 0.17±0.10 0.37±0.33 0.21±0.10 0a
2.94b
0.22±0.13 0.64±0.82 0.22±0.13 17.24a
24.14b
0.14±0.05 0.20±0.19 0.18±0.20 0a
2.86b
0.12±0.05 0.17±0.15 0.10±0.04 50a
56.26b
NO2 n/a n/a n/a n/a 0.02±0.00 0.02 0.01±0.01 n/a n/a n/a n/a n/a 0.01±0.00 0.06±0.07 0.01±0.00 n/a
FTN n/a n/a n/a n/a 0.79±0.04 0.95 0.72±0.14 n/a n/a n/a n/a n/a 0.71±0.48 0.54±0.12 0.49±0.26 n/a
UreaN n/a n/a n/a n/a 0.00±0.01 0 0.00±0.00 n/a n/a n/a n/a n/a 0.04±0.03 0.02±0.01 0.02±0.02 n/a
PN n/a n/a n/a n/a 0.22±0.28 0.59 0.09±0.23 n/a n/a n/a n/a n/a 0.02±0.14 0.02±0.03 0.07±0.05 n/a
DIN n/a n/a n/a n/a 0.27±0.07 0.66 0.16±0.08 n/a n/a n/a n/a n/a 0.32±0.20 0.31±0.07 0.15±0.06 n/a
NO3 n/a n/a n/a n/a 0.17±0.05 0.15 0.05±0.07 n/a n/a n/a n/a n/a 0.28±0.21 0.23±0.04 0.12±0.07 n/a
DON n/a n/a n/a n/a 0.53±0.11 0.29 0.57±0.06 n/a n/a n/a n/a n/a 0.39±0.28 0.24±0.11 0.34±0.24 n/a
NUON n/a n/a n/a n/a 0.52±0.10 0.29 0.57±0.06 n/a n/a n/a n/a n/a 0.17±0.20 0.22±0.12 0.24±0.20 n/a
FTP n/a n/a n/a n/a 0.07±0.03 1.65 0.18±0.06 n/a n/a n/a n/a n/a 0.05±0.01 0.06±0.01 0.05±0.01 n/a
PP n/a n/a n/a n/a 0.08±0.03 0.05 0.05±0.02 n/a n/a n/a n/a n/a 0.02±0.01 0.02±0.01 0.02±0.03 n/a
DOP n/a n/a n/a n/a 0.01±0.00 0.05 0.02±0.01 n/a n/a n/a n/a n/a 0.01±0.01 0.01±0.00 0.01±0.01 n/a
DIC n/a n/a n/a n/a 10.51±2.88 2.4 15.51±3.15 n/a n/a n/a n/a n/a 10.42±8.41 5.03±2.12 10.81±8.14 n/a
DOC n/a n/a n/a n/a 10.55±1.74 7.75 12.60±0.52 n/a n/a n/a n/a n/a 6.46±3.11 4.82±0.25 6.38±2.99 n/a
TDC n/a n/a n/a n/a 21.06±1.14 10.15 28.11±2.63 n/a n/a n/a n/a n/a 16.89±10.84 9.85±1.93 17.19±10.37 n/a
Chapter 3
48
Figure 3. 2 Nutrient attenuation (as SDC) in the Anvil Way Compensation Basin over pre- and
post-restoration regimes. Bar plots (a, c, e, g and i) show attenuation over individual sampling
periods, while box plots (b, d, f, h and j) show annual average attenuation. Solid thick vertical
lines in bar and box plots represent the time (November 2010) when the Anvil Way
Compensation Basin was restored. Horizontal lines and open squares in each box plot
represent the annual median and mean SDC, respectively. Data of pre-restoration and post-
restoration periods include September 2004 to November 2010 and December 2010 to April
2015, respectively.
Chapter 3
49
Figure 3. 3 Nutrient attenuation (as SDC) in the Anvil Way Compensation Basin over pre- and
post-restoration regimes. Bar plots (a and c) show attenuation over individual sampling
periods, while box plots (b and d) show annual average attenuation. Solid thick vertical lines
in bar and box plots represent the time (November 2010) when the Anvil Way Compensation
Basin was restored. Horizontal lines and open squares in each box plot represent the annual
median and mean SDC, respectively. Data of pre-restoration and post-restoration periods
include September 2004 to November 2010 and December 2010 to April 2015, respectively.
TN = 18.64t + 62.67; R2
= 0.84 (3.9)
TOC = 240.7t + 9163; R2
= 0.92 (3.10)
TP = 1.866t + 122.6; R2
= 0.55 (3.11)
The most recent i.e., June 2012 assessment of species coverage showed that B.
articulata (jointed twig-rush) and S. validus (lake club rush) were the most widely
dispersed macrophytes across the AWCB. B. articulata and S. validus were rooted on
riparian and saturated soil, and had extensive root and rhizome systems. TKN
concentrations in the AGB and BGB of S. validus were approximately the same, while
B. articulata showed higher TKN and TP concentrations in the BGB. Nutrient uptake
efficiency of the common reed was found higher in AGB than in BGB in Malaysia
(Sim et al. 2008).
Chapter 3
50
Table 3. 3 Nutrient attenuation as SDC and SDC_ave during the dry and wet periods over pre- and post-restoration regimes of the Anvil Way
Compensation Basin. The dry and wet periods represent December to May and June to November, respectively.
Dry periods Wet periods
Pre-restoration Post-restoration Pre-restoration Post-restoration
SDC (%) SDC_ave (%) SDC (%) SDC_ave (%) SDC (%) SDC_ave (%) SDC (%) SDC_ave (%)
NH3 -24.24±247.62 67.83±69.44 60.58±41.86 71.88±38.40 -30.07±161.26 69.83±36.91 29.77±71.38 68.85±20.83
NOx 44.14±56.34 80.20±24.76 43.31±47.05 83.71±22.37 38.59±33.39 70.59±15.41 10.33±54.07 61.26±20.28
TKN -20.25±53.35 48.08±15.58 13.93±33.32 44.66±23.69 -10.89±30.71 48.08±16.43 13.57±20.68 48.25±21.75
DON -9.47±27.06 44.27±15.01 -4.24±13.61 13.65±26.59 -5.15±25.29 48.86±13.55 -9.22±39.25 12.24±21.67
TN -14.66±50.48 50.88±15.48 20.62±30.35 56.39±20.21 -3.46±42.76 55.06±11.16 19.53±18.02 57.69±15.90
FRP -29.79±79.04 56.63±39.18 1.91±79.70 28.99±80.52 1.97±58.50 59.27±22.51 16.33±28.74 61.58±29.30
TP -42.26±81.55 54.24±21.70 -16.81±65.52 49.59±35.15 -21.64±68.89 50.10±16.87 15.81±19.84 57.82±25.37
NO2 n/a n/a 45.29±38.96 55.63 n/a n/a 13.81±30.92 80.28±9.34
FTN n/a n/a 7.96±22.261 52.11 n/a n/a 20.00±21.58 63.86±4.69
UreaN n/a n/a 64.346 n/a n/a n/a 30.75±21.17 50.83±8.97
PN n/a n/a 40.623 75.24 n/a n/a 68.15±56.23 66.61±17.99
DIN n/a n/a 42.85±14.16 78.49 n/a n/a 35.63±38.26 67.47±4.17
NO3 n/a n/a 72.88±35.45 70.09 n/a n/a 29.732±51.06 63.17±0.80
DON n/a n/a -11.44±34.73 17.33 n/a n/a 7.50±15.39 59.34±3.58
NUON n/a n/a -11.77±34.26 17.33 n/a n/a -4.96±28.37 62.69±5.36
FTP n/a n/a -147.86±32.42 87.42 n/a n/a -1.50±12.93 56.53±0.44
PP n/a n/a 39.92±1.60 61.51 n/a n/a 3.75±81.09 67.79±10.83
DOP n/a n/a -18.59±43.23 65.69 n/a n/a 10.83±10.15 61.28±3.34
DIC n/a n/a -49.09±10.85 -22.13 n/a n/a -7.03±14.38 49.65±9.04
DOC n/a n/a -20.69±14.97 33.57 n/a n/a -0.24±15.39 52.66±14.32
TDC n/a n/a -33.36±5.28 13.65 n/a n/a -5.43±12.67 48.29±4.89
n/a not applicable
Chapter 3
51
Figure 3. 4 Change of attenuation (as SDC_ave) of N species by the Anvil Way Compensation
Basin over the pre- and post-restoration regimes. Bar plots show attenuation over individual
sampling periods, while box plots show annual average attenuation. 100% SDC_ave indicates
compete attenuation while negative SDC_ave indicates nutrient release. Solid thick vertical
lines in bar and box plots represent the time when the Anvil Way Compensation Basin was
restored (November 2010). Horizontal lines and open squares in each box plot represent the
annual median and mean SDC_ave, respectively. Data of the pre-restoration periods include
September 2004 to November 2010, whilst the post-restoration periods include December
2010 to April 2015.
Chapter 3
52
Figure 3. 5 Change of attenuation (as SDC_ave) of P species by the Anvil Way Compensation
Basin over the pre- and post-restoration regimes. 100% SDC_ave indicates compete
attenuation while negative SDC_ave indicates nutrient release. Bar plots show attenuation
over individual sampling periods, while box plots show annual average attenuation. Solid
thick vertical lines in bar and box plots represent the time when the CW was restored
(November 2010). Horizontal lines and open squares in each box plot represent the annual
median and mean SDC_ave, respectively. Data of the pre-restoration periods include
September 2004 to November 2010, whilst the post-restoration periods include December
2010 to April 2015.
However, the significantly higher BGB of S. validus and B. articulata in the
AWCB confirmed TKN and TP mass per unit area were also higher in the BGB. The
macrophytes showed maximum nutrient storage in the ABG and BGB at the age of
two and half years. The storage of nutrient mass started to decline in the following few
months (Figure 3. 7). Importantly, the macrophytes experienced seasonal senescence,
and seasonal biomass decline in the two year old B. articulata at M1 was three times
higher than that observed at M2 (Table 3. 4).
Overall S. validus senescence was 2–3 times higher than senescence of B.
articulata and after 2.5 years of plantation, new stem and leaves of S. validus started
to grow, percentage of mature plants increased and percentage of dead plants
decreased.
Chapter 3
53
Figure 3. 6 Nutrient pools of (a) TN, (b) TP and (c) TOC at the in-stream (S1 and S4) and
bench (S2 and S3) sediments of the Anvil Way Compensation Basin from November 2011 to
May 2015. Data are not available since May 2014 at the bench sites.
3.3.3 Hydrological variability on nutrients dynamics
Although there was very low flow during the dry summer periods in WA,
seasonal variability in hydrology was expected to drive variability in attenuation. The
maximum rainfall around the AWCB occurred during the winter between July and
September, followed by a sharp reduction over spring and summer between
September and January (Figure 3. 8a). A seasonal pattern of flow rate and discharge
was observed at the outflow, with high flow rate and discharge during the wet months
(Figure 3. 8b). Total outflow discharge was sometimes higher than the total input,
particularly during the wet periods. The water balance indicated approximately 20–
70% of the water came from unaccounted sources, such as MSD and possible
groundwater inputs (Ruibal-Conti et al. 2015) depending on the rainfall characteristics
(amount and duration of rainfall). VCUA between June and December in 2012 and
2013 was estimated for selected rainfall events. VCUA was generally small (13–27%)
for mid to large size rainfall events (> 25 mm) and large (40–80%) for small and
Chapter 3
54
frequent rainfall events, particularly between the late wet and early dry periods
(Figure 3. 9). VCUA and runoff generating areas varied over the season in accordance
with the antecedent wetness conditions (Ruibal-Conti et al. 2015).
Figure 3. 7 Nutrient pools in AGB and BGB of the Anvil Way Compensation Basin from
October 2012 to October 2013. (a) TKN in AGB, (b) TKN in BGB, (e) TP in AGB, and (f) TP
in BGB of S. validus; (c) TKN in AGB, (d) TKN in BGB, (g) TP in AGB, and (h) TP in BGB
of B. articulata. B. articulata are collected only in October 2012 and May 2013 at site M1 and
M2. AGB and BGB denote above ground biomass and below ground biomass, respectively.
Chapter 3
55
Table 3. 4 Spatial and temporal changes of physical properties of above ground biomass of aquatic macrophytes per quadrate in the Anvil Way
Compensation Basin.
Sampling
date
Site
name
Species No of Stems No of
Inflorescences
Stem height (m) Leaf stage (%)
Min. Max. Average New Mature Senescent
10/10/2012 M1 BA 77.33±27.12 3.00±1.54 0.267±0.17 2.47±0.26 1.5±0.21 2.33±1.33 78.00±4.16 19.67±4.67
M2 BA 57.67±8.97 3.67±2.18 0.14±0.05 2.19±0.04 1.38±0.12 2.33±0.33 91.00± 2.00 6.67±1.67
M1 SV 47.67±10.39 6.00±1.53 0.08±0.04 1.59±0.09 0.82±0.09 4.33±1.86 44.00±23.43 51.67±21.6
M2 SV 71.67±9.28 10.00±2.08 0.17±2.08 2.17±0.05 1.47±0.06 1.33±0.33 79.33±5.84 19.33±5.67
M3 SV 57.00±11.14 6.67±2.73 0.04±0.01 1.47±0.04 0.89±0.05 7.33±3.93 51.00±4.58 41.67±8.33
M4 SV 54.33±3.48 9.00±1.00 0.07±0.01 1.78±0.22 1.68±0.38 3.00±1.00 52.67±6.22 44.33±5.67
13/05/2013 M1 BA 78.33±6.98 3.00±1.53 0.23±0.14 2.31±0.15 1.56±0.14 5.00±2.88 65.00±10.41 30.00±10.00
M2 BA 114.00±19.69 16.67±6.23 0.21±0.09 2.45±0.08 1.52±0.10 2.00±0.58 58.00±5.19 40.00±5.77
M1 SV 85.00±4.16 11.00±2.52 0.09±0.04 1.73±0.24 1.19±0.19 4.00±3.00 44.67±5.17 51.33±4.67
M2 SV 108.33±24.04 16.67±7.67 0.05±0.01 1.87±0.33 1.27±0.35 7.30±2.67 39.33±0.67 53.33±3.33
M3 SV 90.33±13.29 22.67±1.67 0.07±0.05 1.75±0.01 1.26±0.06 5.33±2.66 48.00±3.00 46.67±1.67
M4 SV 66.00±20.52 20.00±4.36 0.05±0.03 1.85±0.12 1.31±0.15 11.00±4.93 52.33±3.71 36.67±6.01
15/10/2013 M1 SV 84.67±7.68 21.33±4.84 0.13±0.08 1.65±0.08 0.94±0.03 3.00±2.00 86.00±3.00 11.00±1.00
M2 SV 80.33±14.62 15.00±3.78 0.04±0.01 1.58±0.14 0.77±0.08 4.67±0.88 84.33±5.78 11.00±4.93
M3 SV 115.00±21.50 21.00±8.08 0.08±0.01 1.73±0.09 0.79±0.22 3.33±0.88 53.33±9.53 43.33±8.82
M4 SV 123.67±38.58 16.67±3.99 0.09±0.03 2.14±0.06 1.01±0.21 1.17±13.20 52.17±13.20 46.67±13.64
BA: B. articulata
SV: S. validus
Chapter 3
56
Figure 3. 8 (a) Monthly total rainfall recorded at Perth Airport Station, located 5 km north-
south of the Anvil Way Compensation Basin, from 2012 to 2015, and (b) monthly total
discharge and monthly average flow rate at main outflow station of the Anvil Way
Compensation Basin from 2012 to 2014.
Figure 3. 9 Selected rainfall amount and rainfall duration at Welshpool Weather Station,
located 0.8 km north-west of the Anvil Way Compensation Basin, along with flow discharge
in 2012 and 2013. (a) Rainfall amount and duration in 2012, (b) inflow and outflow water
volume, and VCUA in 2012, (c) rainfall amount and duration in 2013, and (d) inflow and
outflow water volume and VCUA in 2013. VCUA denotes volumetric contribution from
ungauged areas.
Chapter 3
57
The possible contribution of this ungauged water may bring with additional
nutrient inputs from MSD and/or groundwater. Nutrient concentrations at MSD are
given in Table 3. 5, while groundwater nutrients since April 2010 are shown in Table
3. 5. NH3 concentrations were higher during the wet periods across all the four bores
while NOx, TKN and TN concentrations were higher during the wet periods, except at
G1. FRP and TP concentrations were 3–10 times higher at G2 than that measured at
other bores and concentrations were relatively close during at the dry and wet periods
(Table 3. 5). NH3 concentrations at W3 were higher than that measured at
groundwater bores during the wet periods and vice-versa during the dry periods. NOx
concentrations were higher at W3 than that observed at G2 but lower than at G4. TN
concentrations during the wet periods in bores, except G2, were higher than
concentrations at W3. FRP and TP concentrations at W3 were lower than
concentration at G2, however higher than other bores. As the AWCB received water
and nutrients from multiple sources, there were some uncertainties of nutrient
attenuation. SDC estimation didn't consider the influence of W2 and therefore
SDC_ave was applied since it combined mean concentrations at W1 and W2 as input
concentrations (Figure 3. 4 and 3. 5). SDC_ave increased compared to SDC for all
nutrients irrespective of periods or restoration regimes. Moreover, SDC_ave was
positive for all species, except DIC. The relationship between SDC and HRT was not
clear as nutrients showed both attenuation and release at the same HRT. For example,
during the wet periods, we observed maximum attenuation and release of NOx and
TN when HRT was 0.3–0.5 day. We observed limited correlation between SDC and
HRT when outflow water levels were low due to limited rainfall and associated flows.
Under such conditions, up to 70% attenuation and 10% release of NOx were observed
at 1–2 days HRT.
Table 3. 5 Spatial and temporal variation of groundwater nutrient concentrations at four bores
in the Anvil Way Compensation Basin from 2010 to April 2015. The dry and wet periods
represent December to May and June to November, respectively.
Nutrient
(mg/L)
Bore
G1 G2 G3 G4
Dry Wet Dry Wet Dry Wet Dry Wet
NH3 0.19±0.06 0.26±0.05 0.18±0.02 0.22±0.04 0.25±0.05 0.33±0.12 0.39±0.08 0.48±0.14
NOx 2.49±2.10 2.26±1.57 0.013±0.00 0.244±0.69 0.097±0.12 1.44±2.95 0.11±0.17 3.14±4.24
TKN 0.79±0.25 0.78±0.31 0.48±0.15 0.55±0.12 0.50±0.09 0.86±0.63 0.90±0.25 0.94±0.26
TN 3.29±2.06 3.26±1.67 0.49±0.15 0.78±0.75 0.59±0.18 2.25±3.30 1.00±0.31 4.06±4.08
FRP 0.06±0.01 0.05±0.01 0.22±0.07 0.22±0.07 0.02±0.00 0.02±0.00 0.07±0.03 0.08±0.03
TP 0.07±0.02 0.07±0.02 0.26±0.07 0.23±0.08 0.03±0.00 0.03±0.01 0.08±0.03 0.09±0.03 Bores G1, G2 and G3 were installed in April 2010 and G4 was installed in January 2012
Chapter 3
58
Hydrological variability including rainfall, groundwater input or seepage can
influence the properties of riparian zones of CW. Eh of riparian soil and nutrient
concentrations in soil porewater influenced by soil saturation after rainfall or seasonal
groundwater seepage in the AWCB. P1, the deepest point of the riparian transect, was
saturated throughout the sampling periods and showed highly reduced Eh (Figure 3.
10). P3 situated at the top of the transect (in the sand bar), and was saturated
periodically and displayed a relatively oxidised state. NH3 concentrations increased at
the three sites after soil saturation, and interestingly P3 showed the highest NH3
concentrations. Average NH3 concentrations in porewater were higher than that found
in groundwater bores and outlet. NOx concentrations at P1 and P2 changed slightly
after soil saturation while P3 showed the highest NOx concentrations. FRP
concentrations were also maximum at P3, where P1 and P2 experienced with FRP
increase slightly. Soil porewater samples were rich with DOC and DIC and porewater
DIC concentrations were lower at P1 and P2 compared to that at P3 after the first
rainfall event on 13 March 2015. DIC and DOC concentrations increased at all sites
following the second round soil saturation (Figure 3. 10). Average porewater DIC and
DOC concentrations were higher than that measured at the surface water inlet and
outlet.
3.3.4 Diurnal changes in wetland function
DO is considered as an important parameter for nutrient dynamics within
CWs. However, we had only spot DO readings at W1, W2 and W3 during the day
time on fortnightly or monthly intervals from 2004 to 2015. DO concentrations
showed a clear seasonal trend, with the maxima and minima during the wet and dry
periods, respectively. A long term decrease in the average DO levels was observed at
W1 and W3. DO concentrations dropped to around zero, and anoxic water has been
recently observed at W1 and W3 for several months over the dry periods. DO maxima
also decreased during the wet periods to around 7 mg/L in 2014 (Figure 3. 11a).
Groundwater DO concentrations were consistently low, ranging from 0 to 2 mg/L,
with few exception and no obvious seasonal trend (Figure 3. 11b). High resolution DO
data of the AWCB from 2014 to 2015 indicated a strong diurnal DO signal where
nigh-time anoxia (< 1 mg/L) was observed for most of the dry months (Figure 3. 11c)
and the maximum DO concentrations were recorded during the sunny days. DO levels
of the AWCB responded to the exposure of solar radiation (Figure 3. 12a, b). Higher
Chapter 3
59
DO concentrations and saturation fluctuation were found during the dry months.
Furthermore, ΔDO saturation over time, particularly during the dry periods, was
correlated with solar radiation exposure (R2 = 0.82) (Figure 3. 12a). Almost flat or
zero DO concentration and saturation were also observed during the sunny days.
There might be several causes behind the occurrence of low DO, and one of
the assumptions was formation of weeds, algal mats, floating macrophytes, debris or
combination of all on water surface and edge of the main stream. Although controlled
floating macrophyes or bed attenuated nutrient in other studies (Chua et al. 2012;
Chang et al. 2013), floating material in the AWCB formed a thick layer in the main
stream. Such floating material of the AWCB provided shade to the stream water that
suppress light penetration and photosynthesis, and consequently reduced DO levels.
We observed different scenarios of the coverage of the main stream by floating
material using the digitized NearMap images from 1 August 2014. These scenarios
included: (a) 1 August 2014: water was flowing and little scattered material started to
grow on the main stream, (b) 27 October 2014: slight accumulation of floating
material on the edge of the main stream, (c) 7 December 2014: accumulation of
floating material progressed towards the edge and part of the main stream, (d) 7
February 2015: occurrence of dense floating material over the main inlet,
sedimentation pond and main stream, (e) 7 March 2015: dense floating material over
main inlet, sedimentation pond and main stream, (f) 24 April 2015: clearance of all
debris and floating material from main stream, and (g) 28 June 2015: main stream
without any floating material. Attempts were taken to compare the relative coverage
or shade of the main channel with DO saturation fluctuation. The maximum stream
coverage (about 70%) and occurrences of flat DO were observed in February and
March 2015. The minimum ΔDO saturation was recorded at the time of maximum
shade while ΔDO saturation increased in April 2015 after the manual removal of
floating material (Figure 3. 12c).
The diurnal pattern of nutrient concentrations of the surface water was
monitored along with DO levels and solar exposure for 2 days (12 and 13 March
2015) before a storm event (Figure 3. 13). The 12 March was sunny and had hypoxic
water (DO∼ 0.2–0.7 mg/L) at the main stream, while DO concentrations were almost
zero at night. The following day was cloudy (Figure 3. 13a) and DO concentrations
remained unchanged with only a slight increase of DO levels for a few hours after a
rainfall pulse (Figure 3. 13b).
Chapter 3
60
Figure 3. 10 Spatial and temporal trend of (a) water level and porewater oxidation-reduction
potential within a riparian zone transect, and (b–e) nutrients from 11 to 17 March 2015. Water
level and porewater oxidation-reduction potential data are collected at 10 and 30-min
intervals, respectively. P1, P2 and P3 sites are at saturated zone, transitional zone between
saturation and unsaturation zone, and unsaturated sand bar, respectively.
Chapter 3
61
Figure 3. 11 Changes of delta DO saturation in response of solar exposure and stream
coverage with floating macrophytes in the Anvil Way Compensation Basin. (a) Change of
delta DO saturation (difference between the highest DO saturation during the afternoon and
the lowest DO saturation during the morning) as a function of solar exposure (sum of solar
radiation exposed over time to achieve delta DO saturation) during the wet and dry periods.
The wet and dry periods include June to November and December to May, respectively. (b)
Standard deviation of delta DO saturation as a function of solar exposure at the wet and dry
periods. (c) Coverage of main stream by senescent macrophytes along with floating material
and respective mean delta DO saturation over time. NearMap provides main stream coverage
images.
Nutrient species (Figure 3. 13c-f) followed a diurnal trend with DO
concentrations and sunlight. NH3 concentrations at W1 and W2 decreased as
progressed towards the day-time and relatively higher DO levels, and vice-versa at
night. DON also followed diurnal trend with the maximum concentrations at W3 was
measured at the mid-day. FRP at W1 and W3 showed a diurnal pattern with increased
concentrations from the afternoon to next day early morning due to lack of DO.
However, as sunlight appeared and DO concentrations increased, FRP concentrations
decreased at W1 and W3. DIC concentrations were lower at W3 compared to that
found at W1, while DOC released from the system, particularly from evening to
morning.
Chapter 3
62
Figure 3. 12 (a) Changes of delta DO saturation in response of solar exposure and stream
coverage with floating macrophytes in the Anvil Way Compensation Basin. (a) Change of
delta DO saturation (difference between the highest DO saturation during the afternoon and
the lowest DO saturation during the morning) as a function of solar exposure (sum of solar
radiation exposed over time to achieve delta DO saturation) during the wet and dry periods.
The wet and dry periods include June to November and December to May, respectively. (b)
Standard deviation of delta DO saturation as a function of solar exposure at the wet and dry
periods. (c) Coverage of main stream by senescent macrophytes along with floating material
and respective mean delta DO saturation over time. NearMap provides main stream coverage
images.
3.4. Discussion
3.4.1 Restoration improves attenuation of specific nutrient species
Nutrient attenuation after CW restoration depends on the effectiveness of the
biogeochemical and hydrological controls included or modified during the restoration.
The AWCB restoration aimed to improve nutrient attenuation and transformation by
inclusion of (a) a sedimentation pond close to the main inlet, (b) a meandering low
flow path, (c) densely vegetated stream banks and riparian zones, (d) multiple sand
bars, (e) a low flow diversion from MSD to the sedimentation pond, and (f) an
adjustable outlet weir. The possible consequences of these restoration initiatives on
nutrient dynamics in the AWCB are summarized in Table 3. 6.
Chapter 3
63
Figure 3. 13 Diurnal pattern of meteorology and water quality in the Anvil Way
Compensation Basin from 12 to 13 March 2015. (a) Rainfall, solar radiation and outflow rate,
(b) main stream water temperature, and DO concentrations at main stream and outlet (W3),
and (c–f) nutrient concentrations at inlet (W1) and outlet (W3). Rainfall, solar radiation, water
temperature and DO data are recorded at 10-min intervals, while outflow rate is recorded at 5-
min intervals.
Chapter 3
64
Table 3. 6 Conceptual consequences of the Anvil Way Compensation Basin restoration initiatives on nutrient attenuation and transformation
pathways.
Nutrient attenuation pathway enhancement
Restoration initiatives
Enhance
water
travel
time and
contact
with
system
Increase
surface
area-to-
volume
ratio
Increase
favourable
conditions for
denitrification
Provide
suitable
conditions
for
nitrification
Enhance
groundwater
-surface
water
interaction
Increase
interaction
with
macrophytes
Promote
opportunities
to interact
with soil or
sediment
Enhance
particle
settling
Promote
condition
for
nutrient
sorption
Removal of sludgy sediments √ √
Creation of sedimentation pond
close to main inlet √ √ √
Introduction of multiple sand bars √
Low flow diversion from MSD √
Creation of meandering stream or
channel √ √ √ √ √ √
Modification of flow path
through the basin √ √ √ √ √ √ √ √ √
Densely vegetated shallow
benches or riparian zone √ √ √ √ √ √
Modification of riparian soil zone √ √ √ √ √
Modification of outlet structure √
Chapter 3
65
Studies of other CWs have found that restoration encouraged nutrient
processing by increasing contact and interaction between stream water, vegetated
riparian zones and biogeochemical active sites (Verhoeven et al. 2006).. Additionally,
nutrient uptake and transformation within the CWs usually increased due to higher
surface transient storage (Baker et al. 2012). Before restoration, our study site was a
compensating basin where stormwater passed through the system via the outlet
quickly (approximately 1–2 h HRT). HRT of the AWCB increased after the
restoration due to inclusion of the meandering flow path, sand bars and dense
macrophytes. Although the relationship of HRT and SDC was not strong enough,
overall attenuation of the nutrients increased after the restoration (Figure 3. 2 and
Figure 3. 3). Other factors therefore contributed in overall attenuation of the system,
for example, modification of riparian zones and plantation of macrophytes
significantly contributed nutrient uptake (Figure 3. 6 and Figure 3. 7).
Interestingly, we found NOx attenuation in the AWCB decreased for the first
three years after the restoration (Figure 3. 2b, d). An available C source and low
oxygen levels are important for NOx attenuation or denitrification (Saeed & Sun
2012). Removal of organic-rich sediments during the AWCB restoration reduced the
C availability and suitable conditions for the denitrifying bacteria to process NOx.
Oxygen levels in the water column of the restored AWCB increased which can trigger
NH3 attenuation and NOx generation. Moreover, NOx attenuation is higher in fine
grained sediments than coarse sediments and therefore, replacement of sludgy
sediments with coarse grained soils/sediments within the AWCB during restoration
can also reduce NOx attenuation. Since restoration, organic matter from the inflowing
water and decay of plant material has slowly accumulated in the AWCB and oxygen
levels have declined; hence the NOx attenuation has subsequently increased (Figure 3.
2b, d).
We found that although FRP and TP attenuation increased immediately after
the restoration, P release began again two years later (Figure 3. 3). In CWs, sorption is
one of the leading mechanisms for P attenuation and essentially follows first order
reactions with removal of P to the new soils being proportional to the concentration of
P in the surface water (Kadlec & Knight 1996). During the restoration, earth works
incorporated new soil into the AWCB that have initially improved P attenuation,
however by two years had become a P source. P sorption for any soil sites is finite,
and once soil sorption sites become saturated due to nutrient loading, the system
Chapter 3
66
becomes a P source rather than a sink (Bridgham et al. 2001). Moreover,
accumulation of P from incoming water and senescent plant material and their
subsequent release or resuspension can convert the system as P source rather sink.
3.4.2 Hydrological and biogeochemical controls on nutrient attenuation
Hydrological variability is reported to influence physical and chemical
processing of nutrients within the CW (Mitsch & Gosselink 2000). Water is unevenly
distributed in the AWCB due to presence of microtopography (including benches and
sand bars), meander flow path, multiple water inputs, possible groundwater seepage
and dense macrophytes. The AWCB experienced prolonged low flows over the
summer and episodic water pulses with possible groundwater inputs during the winter
(Figure 3. 9). These entire variables shaped nutrient attenuation within the AWCB. In
general, the AWCB showed a reduced capacity to attenuate N during high flows of the
wet periods, compared to that of the dry periods. The wet time groundwater
concentrations of NH3 (all bores), NOx (all bores), TKN (bore G1, G3 and G4), TN
(bore G1, G3 and G4), FRP (bore G2) and TP (bore G2) were sometimes higher than
nutrient concentrations observed at W3. Therefore, anoxic nutrient-rich groundwater
can contribute additional nutrient concentrations to the AWCB. Furthermore, MSD
delivered stormwater and passed quickly at the outlet providing high concentrations at
the outlet indicating overall nutrient enrichment.
Nutrient attenuation depends on available ambient conditions that prevail
within CW. N attenuation in the AWCB was temperature dependent, for example,
NH3 and NOx were attenuated about 2 and 4 times higher, respectively during the dry
periods compared to that of the wet periods (Table 3. 3). During post-restoration,
overall NOx removal was 7 times higher during the dry periods than that measured
during the wet periods. NOx attenuation possibly increases during the dry periods
because warmer water increased microbial activities to break down organic carbon,
subsequently reduces DO levels and provides available C for denitrification (Rousseau
et al. 2008). Sediments of the AWCB were rich with organic matter (Figure 3. 6) and
DOC concentrations were higher at the main outlet than that observed at the main inlet
(Table 3. 3) supports the conditions for NOx attenuation. Organic carbon for NOx
attenuation in the AWCB can be delivered from inflow water, decaying of senescent
macrophytes and sediment resuspension. DON, NUON and DOP released from the
AWCB during the dry periods while DOP attenuation occurred during the wet
Chapter 3
67
periods. FRP and TP were attenuated more during the wet periods than that measured
during the dry periods, and dilution was considered as important reason for
attenuation. Dissolved N and P species showed higher attenuation than the particulate
species. Diurnal nutrient pattern (Figure 3. 13) showed that NH3 attenuation decreased
during the day-time likely due to available solar radiation and high oxygen levels.
After hydrological variability, sediments and macrophytes are important
contributor of nutrient attenuation in the AWCB. Soils/sediments and macrophytes in
riparian and in-stream areas of CWs act as buffering filters for nutrients, where
diverse ecological, hydrological and biological processes take place (Hayashi &
Rosenberry 2002; Ranalli & Macalady 2010). In the AWCB, bench sites contained 2–
3 times higher nutrient mass than that at in-stream sites (Figure 3. 6). However, bench
sites are subject to saturation by inflow water during moderate to high flows
(particularly during the wet periods) and senescence of macrophytes, and limited
wash-off of stored nutrients during the dry periods. Inflow nutrients and senescent
plant material can settled down in the bench sites, while nutrient at in-stream sites
were subject to continuous wash-off. However, sediment nutrient pools in the AWCB
also periodically decreased indicating some release of accumulated nutrients (Figure
3. 6). P precipitates with different metallic cations, like aluminium (Al), calcium (Ca),
iron (Fe) and manganese (Mn) that form amorphous or poorly crystalline solids and
co-precipitate with other minerals, such as ferric oxyhydroxide and calcite (Vymazal
2007). Al and Fe levels in water and sediments decreased in the line with P
concentrations in the AWCB after the restoration (Adyel et al. 2015a; Ruibal-Conti et
al. 2015) suggesting that P was precipitated with these metals. The top sediment layer
of the AWCB was rich with organic matter (Figure 3. 6) and therefore, NOx
processing might be triggered in the sediments. However, organic-rich bench and in-
stream sediment sites can release NH3 and NOx from their porewater to the water
column. Gabreiele et al. (2013) found that low DO concentrations in the CW
sediments indicated as NH3 supply due to microbial mineralization of organic matter
and DNRA. Soils/sediments release P through (a) diffusion due to P concentration
gradient between soils/sediments and overlying water column and (b) advection due to
microbial activities, bioturbation and vertical movement of porewater (Reddy et al.
1999). When riparian zones of the AWCB saturated after prolonged dry periods,
nutrients release occurred (Figure 3. 13). Soil saturation after dry periods usually
causes osmotic shock and microbial cell lysis to mobilize and release P (Turner &
Chapter 3
68
Haygarth 2001; Dupas et al. 2015). Furthermore, soil saturation and associated anoxic
conditions create reductive dissolution of Fe (hydr)oxides that also contribute to P
release from soil to the porewater (Dupas et al. 2015). FRP peak was observed after
riparian zone saturation, while the maximum P was found at the top most soil layer
(P3) of the AWCB riparian transect. Similar results were also observed in the AWCB
during a big storm event in March 2015 that flooded riparian zones. Therefore,
solubilisation-mobilization-transport can be the mechanism of P to be released after
saturation of the soils within the AWCB.
Seasonal groundwater level fluctuation and seepage can modify Eh of riparian
soils to release already adsorbed nutrients (Hoffmann et al. 2009). A decrease in Eh at
P2 and P3 of the AWCB riparian transect would be related to increase organic matter
and subsequent oxygen consumption. DO levels in the flooded riparian zones of the
AWCB during storm event were near zero. G2 is situated close to the main outlet of
the AWCB and contained very high FRP and TP concentrations (Table 3. 4). The
AWCB also acts C source over last couple of years. DOC entering at W1 of the
AWCB can be degraded as it travelled towards W3, however, replaced with new DOC
produced within the system and ultimately net export of the species occurred.
Groundwater seepage can contribute some nutrients export to the AWCB; however,
the magnitude of this potential export can't be projected without continuous
monitoring of groundwater quality.
After sediments, aquatic macrophytes are also responsible for nutrient
dynamics with in a CW. During the AWCB restoration, aquatic macrophytes were
planted across the AWCB and these macrophytes demonstrated nutrient uptake and
storage during early to mature growth stage (Figure 3. 7). Macrophytes of the AWCB
formed a horizontal ecotone between riparian zone and open water, and vertical
ecotone between groundwater-connected sediments and overlying water. Macrophytes
in a CW can (a) provide suitable root-shoot-leaf zones for nutrient transformation
including attachment, uptake and storage during early and mature stage, (b) provide
surfaces for the growth of microbial communities and biofilm, and (c) release oxygen
to support nitrification and supply C from root exudates (due to photosynthetically
fixed C) for denitrification (Brix 1994; Brix 1997; Cui et al. 2010; Saeed & Sun
2012). The nutrient pools within the AWCB were higher in BGB i.e., the roots and
rhizospheres than that measured in AGB i.e., shoots and leaves. The ratio of N and P
pools in AGB and BGB in the AWCB was 1:4.6 and 1:6.8, respectively.
Chapter 3
69
Macrophytes species coverage and biomass change can happen significantly
not only during the macrophyte establishment period but also on a seasonal basis. The
lack of ground-truthed of macrophytes species coverage data during the biomass
sampling prevents meaningful estimation of nutrient pools across the AWCB. We,
therefore, tested two different scenarios to estimate relative nutrient pools in
macrophytes: if the AWCB was 100% dominated by (a) S. validus or (b) B. articulata.
If the AWCB was 100% dominant by S. validus, then the total vegetation nutrient
pools would be 665 kg TKN and 141 kg TP. Similarly, if the AWCB was 100%
covered by B. articulata, then the total vegetation nutrient pools would be 1605 kg
TKN and 256 kg TP. The only ground-truthing (in 2012) indicated that the AWCB
was indeed dominated by S. validus, thus the TKN value assuming 100% S. validus
should be considered a conservative estimate of the vegetation N pools. These TKN
pools are similar in magnitude to the sediments/soils TKN pools in 2011, although
almost an order of magnitude lower than sediments/soils TKN pool in 2013. Nutrient
storage per unit area was significantly greater for B. articulata than S. validus,
suggesting that promoting the establishment and growth of B. articulata could
improve nutrient retention across the AWCB. A distinct seasonal pattern of
macrophytes senescence was observed in the AWCB (Table 3. 4). The seasonal
dynamics of S. validus also suggested that nutrients would be transferred to other
pools within the AWCB during the summer, including possible release to the
soils/sediments and CW water through leaching that can increase nutrient
concentrations at the outlet. However, litter or decay matter on the top of the sediment
is likely to have limited oxygen diffusion capacity to the lower sediment layer that
eventually create anoxic conditions and turn the environment suitable for
denitrification (Bastviken et al. 2005). All these results indicated some inter- and
intra-specific completion of macrophytes to store stormwater nutrient.
Overall differences in TN and TP attenuation can be attributed to several
reasons: (a) the AWCB inflow had proportionally more organic N than organic P, (b)
uptake of N species was 4–7 times higher than P species in macrophytes, (c) sediment
N pools were about 100 times greater than sediment P pools, (d) periodic anoxic and
C rich conditions favour NOx attenuation, while oxic conditions are favourable for
NH3 and FRP attenuation, and (e) internal release of P species was dominant than N
release. Nutrient species can be attenuated and transported into different media such
as surface water, soils/sediments, groundwater, macrophytes, microbes and overlying
Chapter 3
70
air column through different processes in the CW as summarized in Chapter 2 (Table
2. 3).
3.4.3 DO as a proxy indicator of wetland metabolism
Aquatic metabolism can be defined as the balance of gross productivity and
respiration within the water column of CWs (Tuttle et al. 2008) and we used delta DO
as a proxy indicator of CW metabolism (Adyel et al. 2016). DO levels within the CWs
indicate the balance between biogeochemical processes that consume oxygen such as
respiration, and processes that deliver oxygen to the water column such as
photosynthesis and reaeration or exchange from the atmosphere. The AWCB was
characterized by low surface water DO levels, particularly during the dry periods and
also displayed a distinct diel DO signal with almost zero DO concentration at night.
Different studies found possible reasons for low DO levels or DO depletion in the
CWs as: (a) oxygen demand as indicated by biochemical oxygen demand (BOD) and
chemical oxygen demand (COD) (Carpenter & Lodge 1986), (b) DO consumption
during breakdown of senescent macrophytes, (c) DO consumption by sediments as
indicated by sediment oxygen demand (SOD) (Chimney et al. 2006), (d) floating
macrophytes and debris via shading that inhibits photosynthesis and DO exchange
between atmosphere and water (Caraco et al. 2006; Nahlik & Mitsch 2006), (e)
reduction of wind action in standing macrophytes that limit atmospheric reaeration of
the water column (Kirk 1994), and (f) possible seasonal mixing of surface water with
hypoxic groundwater.
We also monitored how metabolism changed with meteorological conditions.
The dry periods showed 2–3 times higher ΔDO (R2 = 0.82) compared to that observed
during the wet periods in the AWCB. Anoxic conditions were frequently observed
during the dry periods with almost no ΔDO when the main water channel was covered
with floating material. The floating material usually provides shade to the water
column, inhibits light penetration to water and inhibits oxygen exchange between
atmosphere and water. Therefore, low DO levels of the AWCB water column can
limit photosynthesis and increase oxygen demanding substances. Anoxic conditions
below dense macrophytes and near sediments have reported in other aquatic system
(Carpenter & Lodge 1986). DO levels in a CW beneath floating macrophytes were
consistently below 4.0 mg/L throughout the day (Reeder 2011). Low DO
concentrations were also reported under floating mats of Lemna and Pithophora
Chapter 3
71
(Lewis & Bender 1961) and thicker mats of Lemna and Wolffia (Morris & Barker
1977). Although floating macrophytes generate oxygen, much of which mainly
emitted to the atmosphere with very limited diffusion to the water column (Morris &
Barker 1977). When senescent floating macrophytes were manually removed from the
AWCB, DO levels increased due to light availability and higher temperature, both
promoted photosynthesis. Removal of dense plant material was previously found to be
effective for improving DO levels (Kaenel et al. 2000). Higher DO levels in the water
column were also reported in area with less macrophytes and high water turbulence
(Bunch et al. 2010). Diurnal cycles in the AWCB indicated that DO levels also
increased after a rain event pulse. This has been observed in other CWs that episodic
pulses and associated increased flow rates flush senescence cells and toxic pigments
out of the CW, trigger nutrients uptake, increase reproduction rates of aquatic
producers and promote photosynthesis and respiration (Stevenson 1996; Tuttle et al.
2008). Low DO levels were also observed in the AWCB for several days, particularly
during the dry periods, even when there was limited plant cover in the main stream.
The cause of these low DO levels can be the intense SOD and consumption of DO to
break down senescent plant material. The outlet of the AWCB displayed higher DOC
concentrations than at the inlet, particularly during the dry periods and these
additional DOC were likely generated within the system through sediment flux and
decay of senescent macrophytes. High SOD and decomposition of DOC were also
reported as a source of low DO levels in other CW (Chimney et al. 2006).
3.5 Conclusions
Stormwater nutrient attenuation in a surface flow CW was assessed in this
study over diurnal to decadal time-scales. The CW was restored after five years of
operation, and consequences of such restoration were also explored in this research.
The main findings of the research included:
1) The CW showed variable nutrient attenuation indicating the influence of
hydrological and biogeochemical processes. The analysis therefore can be considered
as an important empirical evidence on CW function and how this change at temporal
scale. Analysis results can also serve as a benchmark of CW performance in diverse
hydrological regimes like low or near zero flow to pulse flow conditions. This
research indicated that restoration can improve attenuation of specific nutrients over
time, and CW can switch between nutrient source and sink.
Chapter 3
72
2) A distinct diurnal DO signature was observed in the CW with anoxia at
night, indicating how CW shifted from oxic to anoxic conditions in a 24-h cycle. The
research also indicated how nutrients responded over diurnal cycle i.e., to solar
radiation and DO concentrations. The findings highlighted the importance of sampling
time targeting particular nutrient for the assessment of CW function at diel scale.
3) The study also investigated the importance of sediments and aquatic
macrophytes as sink of incoming nutrients. The sediments were a significantly greater
sink of nutrients than that observed in two dominant macrophytes i.e., B. articulata
and S. validus. First macrophyte species stored about 3 and 2 times higher TKN and
TP pools, respectively than that stored in second species, indicating inter-specific
competition of nutrient among macrophytes at different growth phases. Seasonal
macrophytes senescence indicated stored nutrient can be released to the system or
transmitted to other pools and converted the system as nutrient source. Moreover,
senescent macrophytes along with other debris covered the main stream and provided
shad to the water that reduced DO concentrations.
4) The research explored wetland metabolism based on diurnal change of high
frequency DO data, as a proxy of CW function. The study untimely showed that
metabolism was linked with solar radiation exposure, and able to correlate with degree
of nutrient attenuation. This investigation, therefore, created the opportunity for using
in situ sensors for indicating real-time function of stormwater treating CW. The
intense diel metabolism can lead to large variability in nutrient attenuation and should
be considered in assessing wetland performance rather than solely relying on long-
term monthly snapshots. Mature and senescent macrophytes need periodic harvest to
minimize leaching of nutrients.
5) Initiatives need to take to improve DO levels within the CW and this could
be achieved through removal of floating debris, re-circulation of effluents during the
dry periods and supply artificial aeration pumping air to the system.
Chapter 4
Stormwater nutrient attenuation in a multi-
compartment constructed wetland
This chapter is based on the manuscript published as: T. M. Adyel, C. E. Oldham and
M. R. Hipsey (2016) Stormwater nutrient attenuation in a constructed wetland with
alternating surface and subsurface flow pathways: event to annual dynamics. Water
Research, 107, 66-82.
Chapter 4
74
Abstract
Among different Water Sensitive Urban Design (WSUD) options, constructed
wetlands (CWs) are widely used to protect and support downstream urban waterways
from stormwater nutrients. This analysis assessed the nutrient attenuation ability of a
novel CW in Western Australia that combined multiple alternating surface flow (SF)
and laterite-based subsurface flow (SSF) compartments within a parkland context to
improve the urban landscape and amenity. The CW was designed to maximise
nutrient reduction despite experiencing a large range of hydrologic conditions, from
low transit time nutrient-rich pulses during the wet periods to prolonged low to zero
flow conditions during the dry periods. The CW design was further complicated by
the possibility of ungauged water inputs after wet antecedent conditions, seasonal
macrophyte senescence and a recirculation system to maintain flow during the dry
periods. From analysis of data over a range of time scales, we determined that overall
the CW attenuated up to 62% total nitrogen (TN) and 99% total phosphorus (TP)
loads during dry weather conditions, and 54–76% TN and 27–68% TP during episodic
flow pulses. N species attenuation was dominant in the SF compartments, while P
species were attenuated mostly within the SSF compartments. Nutrient accumulation
in the sediments, and above and below ground biomass of the macrophytes were
found to increase during the early stages of operation, suggesting the system reached
equilibrium within four years. Further, by comparing trends in nutrient attenuation
within the context of diel changes in high frequency oxygen data from different
compartments, it was demonstrated that changes in dissolved oxygen were related to
changes in nutrient concentration across the CW, although interpretation of this was
complicated by changing hydro-climatological conditions. The implementation of this
CW concept in a highly seasonal Mediterranean climate demonstrates that urban
liveability and environmental health can both be improved through careful design.
Keywords
Dissolved oxygen; Ecosystem service; Laterite; Multi-compartment; Wetland
metabolism
Chapter 4
75
4.1 Introduction
Stormwater runoff is an important source of nutrients in urban environments,
mobilising and redistributing both organic and inorganic forms from point and non-
point sources within the catchment (Hatt et al. 2009; Gasperi et al. 2011; Burns et al.
2012; Zgheib et al. 2012). Due to the rapid expansion of impervious areas in cities
around the globe, the magnitude of nutrient loads associated with untreated
stormwater are increasing, and subsequently creating significant challenges for
downstream receiving water environments (Walsh et al. 2005b; Brown et al. 2009;
Burns et al. 2012; Fletcher et al. 2013). These challenges include issues associated
with water quality impairment due to eutrophication such as the growth of algal
blooms, the generation of anoxia or hypoxia, and, in extreme cases, the large-scale
mortality of fish and aquatic biodiversity, all of which reduce the recreational amenity
of waterways within the urban landscape (Taylor et al. 2004; Miller & Boulton 2005;
Kaushal et al. 2008; Cuffney et al. 2010; Hathaway et al. 2012). There are, therefore,
increasing efforts to monitor and actively manage stormwater quality to safeguard and
restore urban waterways.
A number of approaches are currently using to reduce stormwater nutrient
loads from urban areas. These approaches have collectively been referred as Water
Sensitive Urban Design (WSUD) in Australia, Sustainable Urban Drainage Systems
(SUDS) in the UK and Low Impact Design (LID) in the USA and New Zealand
(Fletcher et al. 2014). A commonly used practice within such approaches is the
implementation of a constructed wetland (CW) for stormwater retention and nutrient
attenuation prior to discharge into a sensitive downstream watercourse (Hatt et al.
2006; Burns et al. 2012; Fletcher et al. 2013). CWs are engineered green bioreactors
(Mitsch & Jorgensen 2003), that attenuate nutrient loads via a range of
biogeochemical and hydrological processes occur within the water, soils/sediments,
aquatic macrophytes and associated microbial communities (Fink & Mitsch 2007;
Huang et al. 2011; Chang et al. 2013; Wu et al. 2014; Vymazal & Kröpfelová 2015).
Biogeochemical processes depend on the availability of oxygen and carbon sources,
abundance of macrophytes and microorganisms, soil/sediment properties, formation of
biofilms and the presence of artificial filter media (Huang et al. 2011; Malaviya &
Singh 2012; Saeed & Sun 2012). Hydrological processes, such as dilution and riparian
and/or hyporheic exchanges, can control the extent of biogeochemical transformations
within the CWs (Fink & Mitsch 2007). In particular, higher hydraulic retention time
Chapter 4
76
(HRT) increases the contact of water with different elements of the CW, which is
thought to promote nutrient attenuation.
Individual CW can be classified according to the predominant direction of
flow, the position of the water surface relative to soil or substrate, the presence or
absence of macrophytes, and the degree of saturation of the underlying soil (Fonder &
Headley 2013). Open water or surface flow (SF) and subsurface flow (SSF) CWs are
the two most common types for treating urban stormwater. In isolation, a SF or SSF
CW would usually attenuate certain nutrient species more than other species. For
example, CWs with continuous flow and ample oxygen exhibit efficient ammonia
(NH3) attenuation by promoting nitrification, whereas CWs with low oxygen
concentrations show greater nitrate (NO3) attenuation by promoting denitrification.
CWs with both SF and SSF systems (i.e., a ‘multi-compartment’ wetland) can
combine the advantages of single stage SF and SSF CWs and provide better overall
nutrient attenuation (Ávila et al. 2014; Ayaz et al. 2015; Vymazal & Kröpfelová
2015). Since the hydrological and biogeochemical conditions differ within
compartments of such CWs, some parts of the system can experience an aerobic
environment, while other parts are anaerobic, and some may be net autotrophic whilst
others are net heterotrophic, leading to different attenuation efficiencies for nitrogen
(N), phosphorus (P) and inorganic and organic species of each. Whilst multi-
compartment systems are promising, the optimal design of such systems for nutrient
attenuation remains uncertain. Furthermore, it is even less clear how to optimise
design of a multi-compartment CW system that experiences a Mediterranean climate,
where the flow rates become very low during the dry summer and some compartments
may dry out. In this case, additional water supply and/or recirculation may be required
to keep the CW active and maintain amenity of the system. Episodic water pulses
during the wet winter periods make nutrient attenuation more difficult due to low
HRT that limit the extent of biogeochemical reactions. Therefore, large flow
variability throughout the year makes the optimization for nutrient attenuation within
multi-compartment systems challenging. Although some studies have investigated the
long-term performance of single stage CW (Gu et al. 2006; Mustafa et al. 2009), to
the best of our knowledge there are no published data on the long-term performance of
a multi-compartment CW in an urban landscape for stormwater nutrient attenuation.
In order to optimise the design of multi-compartment CW, a better
understanding of their long-term nutrient attenuation dynamics and responses to flow
Chapter 4
77
variability is therefore necessary. In the present study, we assessed the performance of
a multi-compartment CW on the Swan Coastal Plain (SCP) in Western Australia
(WA), implemented to support the restoration of the nearby Canning River. The CW
was designed to include three SF and two laterite-based SSF compartments, and a
summer recirculation scheme. We explored the dynamics of nutrient attenuation over
event, seasonal and annual scales over a 7-year periods (2009–2015) through a
synthesis of historical meteorological data, water quality and quantity data, high-
frequency oxygen data, and sediment and macrophyte quality data. We aimed to
explore the following questions: (a) What is the comparative contribution of different
compartments of the study site to nutrient attenuation? (b) How do flow conditions
(low flow/base flow, high flow) influence the nutrient attenuation of the system? (c)
What is the particular contribution of sediments and macrophytes to nutrient
reduction? (d) How do dissolved oxygen (DO) concentrations vary between the
compartments and can daily oxygen fluctuations be used as a proxy of wetland
process dominance? The research findings highlight the relative difference between
the SF and SSF compartments, and the large seasonal variability in performance based
on changes in oxygen dynamics and retention time. The results were used to estimate
the overall nutrient reduction that occurred since wetland construction and aspects of
the design that enhance performance were discussed. Finally, we also demonstrated
that the multi-compartment design offers additional eco-recreational amenity,
providing other ecological services important for biodiversity protection and
increasing urban liveability.
4.2 Materials and methods
4.2.1 Site background
The study site, the Wharf Street Constructed Wetland (WSCW), is situated in
Canning, WA (Figure 4. 1). The climate of the site is Mediterranean with most of the
rain occurs from June to September (long term annual average 770 mm) and dry hot
summers with temperatures reaching an average maximum of 32 °C in January and
February. The catchment is characterised by a shallow groundwater table (often 2–3 m
below ground) and highly transmissive Bassendean sands. During the summer dry
months, groundwater top-up, recirculated water and drained irrigation water are the
principal water sources to the CW. During the wet season, stormwater runoff from
Chapter 4
78
residential, commercial and industrial areas is captured by surface and subsurface
drainage systems and discharged into the CW.
The WSCW has a total area of approximately 1ha and receives stormwater
from approximately 129ha of urban area drained by the Wharf Street Main Drain
(WSMD), before discharging into the Canning River. The WSCW started operation in
2009, while laterite-based SSF compartments were only incorporated in October 2012
and the system has been hydraulically connected since then. The SSF compartments
were initially built (in 2009) with recycled concrete materials (RCM), however, the
RCM were heavily abraded during installation and degraded further in situ providing
elevated alkalinity in the SSF components; several attempts were made to reduce the
alkalinity without success. In June 2012, approximately 1500 m³ of RCM were
removed (all of the SSF1 and roughly half of the SSF2) and replaced by 25–40 mm
screened and washed laterite aggregates. After incorporation of the laterite-based SSF
compartments, a continuous flow through the WSCW was finally established. The
WSCW is lined by in situ clays that restrict groundwater exchange with the system.
For our current research, we divided the WSCW into four different
compartments (Figure 4. 1): (a) compartment 1: SF1 with inlet (W1) and outlet (W2),
(b) compartment 2: laterite-based SSF1 with inlet (W3) and outlet (W4), (c)
compartment 3: laterite-based SSF2 with inlet (W5) and outlet (W6), and (d)
compartment 4: SF2 with inlet (W7) and outlet (W8). The first two open water bodies
of the WSCW are connected through a narrow vegetated channel and considered as
the SF1 compartment. Some portion of water can be by-passed the SSF1 compartment
during the high flow conditions of the wet periods. The minimum water level in the
WSCW was maintained during the dry summer by a pump-back system that
recirculated water from W8 to W1 (Figure 4. 1). It was estimated that the recirculation
system operated for 157.5 h per week (at night-time) at a flow rate of 4500 L/hr
during the dry periods and delivered about 300 m3 water during 2011 and about
2245 m3 from October 2013. Local groundwater was also used intermittently to top-up
the WSCW, keep the system active and maintain the aesthetic landscape during the
hot summer when natural water flows were nearly zero. Additional ungauged water,
especially during the stormwater pulses, entered the system through the ornamental
lake (OL) and Bebington Court Drain (BCD). The BCD catchment contains mainly
residential area with semi to high traffic way, and enters approximately 50 m upstream
of the main outlet of the WSCW.
Chapter 4
79
Figure 4. 1 (a) Map of the Wharf Street Constructed Wetland indicating sampling locations of surface water, flow, sediment and macrophyte and DO.
The ornamental lake (L) and Bebington Court Drain (W7) deliver ungauged water during episodic storm events. Flow rates were measured at inflow
(1) station from 2010 to 2013 and inflow since 2013. (b) Cross-section of the flow path through alternating SF and SSF compartments. Additional
pipes allow SSF1 to be by-passed under high flow conditions and half of SSF2 is covered by grass. Elevation from inlet to outlet is about 0.3 m.
a)
b)
W1 W8
Chapter 4
80
4.2.2 Data collection
4.2.2.1 Meteorology
Solar radiation data was obtained from the Department of Agriculture and
Food (DAFWA) weather station located approximately 3.8 km north-west of the
WSCW. Daily rainfall data was obtained from the Department of Water (DoW) and
the Bureau of Meteorology (BoM) stations that were approximately 0.3 km north-west
and 7 km north-south of the WSCW, respectively.
4.2.2.2 Water quantity
5-min intervals water discharge and level data were collected at the WSCW
inflow and outflow stations. The main inflow was measured via a structure comprising
a rock riffle (primary flow control) with a set of three by-pass pipes (under the riffle)
and a hydrostatic water level probe to provide water stage information. The outflow
was measured at the concrete V-notch weir using a water level transducer in a float
well system. A detailed water balance of the WSCW was previously provided by
(Adyel et al. 2015b).
4.2.2.3 Water quality
Fortnightly or monthly unfiltered and filtered water samples (July 2009 to August
2015) were collected from the inlet and outlet of the various compartments as
indicated above. Additional opportunistic water samples were also collected from
2013 to 2015. Water samples were filtered on-site using 0.45 μm membrane filters,
stored on ice until returning to the laboratory, and then frozen until analysis (Table
4.1). Historical water samples were analysed for filtered ammonia (NH3), filtered total
oxidised nitrogen (NOx), total Kjeldahl nitrogen (TKN), dissolved organic nitrogen
(DON), unfiltered total nitrogen (TN), filtered reactive phosphorus (FRP) and
unfiltered total phosphorus (TP). Opportunistic samples were analysed for NH3,
filtered nitrite (NO2), filtered urea nitrogen (UreaN), filtered total nitrogen (FTN), TN,
FRP, filtered total phosphorus (FTP), TP, dissolved organic carbon (DOC) and
dissolved inorganic carbon (DIC) using Lachat Quick Chem procedure (APHA 2012).
We calculated other nutrient parameters i.e., NOx, filtered nitrate (NO3), particulate
nitrogen (PN), dissolved inorganic nitrogen (DIN), DON, filtered non-urea organic
nitrogen (NUON), particulate phosphorus (PP) and dissolved organic phosphorus
(DOP) and total dissolved carbon (TDC) based on the analysis results of opportunistic
Chapter 4
81
samples (Table 4. 1). During the historical water sampling, in situ DO concentrations
were recorded. We also measured 10-min intervals DO concentrations and saturation
at the middle of SF1 and SSF1 compartments from July 2014 to July 2015 using a
miniDOT DO logger (PMA, USA).
Table 4. 1 Summary table of spatial and temporal dynamics of studied parameters, data type,
sampling frequency and data source for the Wharf Street Constructed Wetland assessment.
Component Parameter Data type Sites Duration Sampling
frequency
Analytical
method
Data
source
Surface
water quality
NH3, NOx,
TKN,
DON, TN,
FRP, TP
Historical W1, W2,
W3, W4,
W5, W6,
W7, W8
July 2009
to May
2015
Baseflow
sampling
(fortnightly
to monthly)
APHA
(1998)
DoW
TN, TP W1, W8 Event
sampling (2
to 3-hr
intervals)
APHA
(1998)
DoW
NH3, NOx,
NO2, NO3,
FTN, TN,
UreaN,
DIN, PN,
DON,
NUON,
FRP, FTP,
TP, PP,
DOP, DIC,
DOC, TDC
Opportunistic W1, W2,
W3, W4,
W5, W6,
W7, W8
November
2013 to
May 2015
Baseflow
sampling
(fortnightly
to monthly)
APHA
(2012)
This
study
DO Historical W1, W2,
W3, W4,
W5, W6,
W7, W8
July 2009
to May
2015
Baseflow
sampling
(fortnightly
to monthly)
Logger DoW
Continuous Within
SF1 and
SSF1
July 2014
to July
2015
10-min
intervals
miniDOT
DO logger
This
study
Sediment
quality
TKN or
TN, TP,
TOC
Historical S1, S2,
S3
June 2010
to May
2014
Mainly half
yearly
GHD
(2008)
DoW
Macrophyte
quality
TN, TP Historical M1, M2,
M3
November
2010 to
May 2012
Mainly half
yearly
GHD
(2008)
DoW
Analysed nutrient species: NH3: Filtered ammonia; NOx: Filtered total oxidised nitrogen; NO2: Filtered nitrite; UreaN: Filtered
urea nitrogen; FTN: Filtered total nitrogen; FRP: Filtered reactive phosphorus; FTP: Filtered total phosphorus; TN: Unfiltered
total nitrogen; TP: Unfiltered total phosphorus; DIC: Dissolved inorganic carbon; DOC: Dissolved organic carbon.
Calculated nutrient species: NO3: Filtered nitrate = NOx - NO2; DIN: Filtered inorganic nitrogen = NOx + NH3; PN: Particulate
nitrogen = TN – FTN; DON: Dissolved organic nitrogen = FTN – NH3 – NOx; NUON: Filtered non-urea organic
nitrogen = DON – UreaN; PP = Particulate phosphorus = TP – FTP; DOP: Filtered organic phosphorus = FTP – FRP; TDC: Total dissolved carbon = DIC + DOC.
Chapter 4
82
4.2.2.4. Sediment and macrophyte properties
Sediments were sampled at three locations such as S1, S2 and S3 (Figure 4. 1),
on average every six months from June 2010 to May 2014. Total number of samples
in each point was 17. S1, S2 and S3 were located at the first open water pond of SF1
compartment, the end of SF1 compartment and the open pond after SSF1
compartment, respectively. Average sediment sampling depth was around 10–15 cm.
The sediment sampling methodology is described in detailed by (GHD 2008).
Sediments were analysed for TKN or TN, TP and total organic carbon (TOC).
The WSCW was planted with aquatic macrophytes, including Baumea
articulata, B. rubiginosa, B. preissii, Carex appressa, Juncus kraussii, Schoenoplectus
validus and Typha domingensis by the South East Regional Centre for Urban
Landcare (SERCUL). Macrophyte samples (n = 51) were collected using quadrats
(0.065 m2) at three locations (M1, M2 and M3) (Figure 4. 1), at four to seven month-
intervals from November 2010 to May 2012. Macrophyte samples were analysed for
TKN or TN and TP in the above ground biomass (AGB) such as shoots and leaves and
below ground biomass (BGB) such as roots and rhizospheres (GHD 2008).
4.2.3 Data analysis
4.2.3.1 Hydrological assessment and its contribution on nutrient attenuation
A water balance was estimated for the WSCW using 5-min intervals flow data
from the main inflow and outflow. When the total incoming water volume (Vinfow) was
higher than the outgoing water volume (Voutfow), the water balance was considered
negative and the WSCW gained water (Voutflow < Vinfow). The water balance was
considered positive when volume of outgoing water was higher than volume of
incoming water (Voutflow > Vinfow); under these conditions the WSCW lost water.
Hydrograph studies were performed considering the inflow and outflow rate
under different rainfall patterns and antecedent conditions (dry and wet). We
partitioned the flows into two regimes: (a) high flow, when the flow rate was >
0.005 m3/s and (b) base flow, when the flow rate was < 0.005 m
3/s. For base flows, the
yearly mean nutrient concentrations at the inlet and outlet were derived from the
historical and opportunistic data in order to define typical dry weather concentration
(DWC). Event mean concentration (EMC) as mg/L was estimated based on the
detailed event analyses where intra-event data were used to estimate load reductions:
Chapter 4
83
𝐸𝑀𝐶 = ∑ 𝑉𝑖 ×𝑛𝑖 𝐶𝑖/𝑉𝐸 (4.1)
where Vi is the volume proportional to the flow rate at ith timestep (m3), Ci is the
nutrient concentration at ith timestep (mg/L), n is the total number of samples and
VEVE is the total runoff volume across an event (m3).
Since the effectiveness of event nutrient attenuation depends on the size of the
event, the relationship between HRT and attenuation was factored into and applied to
each event that occurred. Load reduction was calculated on the occasions when
simultaneous nutrient concentration and flow data at the main inlet and outlet were
available. In this case, the total volume of water that passed through the inlet and
outlet, total nutrient load, and EMC of nutrient at W1 and W8 were calculated as:
𝐿 = 𝑄𝑖 × 𝐶𝑖 (4.2)
𝐿𝐴 =∑ 𝐿𝑖−∑ 𝐿𝑜
∑ 𝐿𝑖 × 100 (4.3)
where L is the load (mg/s), Qi is the flow rate at ith timestep (m3/s), Ci is the nutrient
concentration at ith timestep (mg/L), LA is the load attenuation (%), Li is the load at
the inlet (mg/s) and Lo is the load at the outlet (mg/s).
4.3.2.2 Nutrient attenuation within compartments
Load estimation and reduction through each compartment was not possible due
to lack of flow data in each compartment. Therefore, nutrient attenuation was
estimated by computing the standardised delta concentration (SDC) as %, defined for
each compartment as:
𝑆𝐷𝐶 =𝐶𝑖𝑛𝑙𝑒𝑡−𝐶𝑜𝑢𝑡𝑙𝑒𝑡
𝐶𝑖𝑛𝑙𝑒𝑡 × 100 (4.4)
where Cinlet is the nutrient concentration at the inlet (mg/L) and Coutlet is the nutrient
concentration at the outlet (mg/L) of the compartment. Details of the SDC approach
were provided elsewhere (Adyel et al. 2015a; Adyel et al. 2015b). Moreover, multiple
sources W6 and W7 delivered ungauged water and nutrients to the SF2 compartment.
Therefore, SDC_ave, a modified version of SDC, was applied that considered average
Chapter 4
84
nutrient concentration of W6 and W7 as the inflow concentration for the SF2
compartment.
4.3.2.3 Nutrient pool estimation in sediments and macrophytes
The sediment TKN or TN and TP pools were estimated for three sampling
locations assuming that each sampling location was representative of a specified area
as:
𝑁𝑀𝑠 = 𝑁𝐶𝑠 × 𝐵𝑠 × 𝐴𝑠 × 𝐷𝑠 (4.5)
where NMs is the sediment nutrient mass (kg), NCs is the sediment nutrient
concentration (mg/kg), Bs is the sediment bulk density (kg/m3), As is the sediment area
(m2) and Ds is the sediment depth (m).
TKN or TN and TP pools in AGB and BGB were estimated, assuming that each of the
sampling site was representative of a specified area of the WSCW, as:
𝑁𝑀𝑏 = 𝑁𝐶𝑏 × 𝐷𝑊𝑏 × 𝐴𝑏 × 𝑄𝑏 (4.6)
where NMb is the biomass nutrient pool (kg), NCb is the biomass nutrient
concentration (mg/kg), DWb is dry weight of the biomass (kg), Ab is the biomass area
(m2) and Qb is the quadrat area (0.0625 m
2).
4.3.2.4 Oxygen variability
Nutrient attenuation was correlated to changes in DO concentrations or delta
DO (ΔDO) as mg/L or DO saturation (ΔDOsat) as %, defined as:
ΔDO = DOmax - DOmin (4.7)
where DOmax is the maximum DO concentration (mg/L) or saturation (%) during the
afternoon and DOmin is the minimum DO concentration (mg/L) or saturation (%)
during the morning.
Total daily exposure of solar radiation over the period achieving DOmin and
DOmax was computed from available meteorology data. The average monthly ΔDOsat
Chapter 4
85
and solar exposure were then calculated to investigate the change in DO saturation
over months and seasons.
4.3 Results
4.3.1 Long term dynamics of nutrient quality
Nutrient concentrations changed along the multiple compartments, with higher
concentrations at the main inlet and BCD (W7) (Figure 4. 2). There was a sharp
reduction in the average concentrations of NH3, FRP and TP at the outlet of the SF1
compartment, while NOx and TN concentrations slightly reduced and DON
concentrations increased. The SSF1 compartments received nutrients at lower
concentrations as they had been treated in the SF1 compartment, and further
processing within the SSF1 decreased nutrient concentrations (Figure 4. 2). However,
in the SSF2 compartment, the average outlet (W6) concentrations of NH3, FRP and
TP were slightly higher compared to that at W4. In addition to W6, W7 delivered
nutrients to the SF2 compartments and therefore influenced the nutrient dynamics of
the compartment and overall system. At the main outlet (W8), TN concentrations were
within the Australian and New Zealand Environment and Conservation Council
(ANZECC) guideline (1.2 mg/L) and Healthy River Action Plan (HRAP) (1.0 mg/L)
target for 89 and 81% of the sampling events, respectively. TP concentrations at W8
were within the ANZECC guideline (0.065 mg/L) and HRAP target (0.1 mg/L) for 95
and 86%, respectively of the sampling events.
Nutrient concentrations at the inlets and outlets of the different compartments
were averaged over the periods defined as ‘dry’ (December to May) and ‘wet’ (June to
November) (Table 4. 2). NH3, TKN and DON concentrations at W8 compared to at
W1 were higher during the dry periods, while NOx and TN concentrations were
higher during the wet periods. FRP and TP showed relatively similar concentrations at
W1 and W8 irrespective of the periods. Since 2013, additional nutrient species were
measured in the surface water samples (Table 4. 2). The wet periods showed marginal
higher FTN, DIN, NUON and PP and concentrations at W8 than that measured during
the dry periods, while the dry periods showed two to three times higher concentrations
than the wet periods concentrations particularly for FTP and DOP. NO3 concentrations
at W8 during the wet periods were 20 times higher than that measured during the dry
periods. DOC concentrations were three to four times higher during the wet periods
than the dry periods (Table 4. 2).
Chapter 4
86
Figure 4. 2 Trends of (a) NH3, (b) NOx, (c) TKN, (d) DON, (e) TN, (f) FRP and (g) TP
nutrient concentrations along the Wharf Street Constructed Wetland from 2009 to 2015.
Horizontal lines and open squares in each box plot represent the annual median and mean
concentration, respectively. W1, W2, W4, W6, W7 and W8 represent the inlet of the
SF1/main inlet, outlet of the SF1, outlet of the SSF, outlet of the SSF2, Bebington Court Drain
and outlet of the SF2/main outlet, respectively. Horizontal solid and dotted lines of each box
represent the ANZECC guideline and HRAP target, respectively.
Chapter 4
87
Table 4. 2 Spatial and temporal variability of nutrient concentration (mg/L) in the WSCW. The dry and wet period represents six months of December
to May and June to November, respectively.
SF1 SSF1 SSF2 SF2
Inlet (W1) Outlet (W2) Inlet (W3) Outlet (W4) Inlet (W5) Outlet (W6) Inlet (W7) Outlet (W8)
Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet
2009-2015
NH3 0.13±0.02 0.07±0.01 0.30±0.19 0.03±0.00 0.03±0.01 0.06±0.02 0.04±0.02 0.01±0.00 0.01±0.00 0.02±0.00 0.04±0.01 0.02±0.00 0.08±0.02 0.03±0.00 0.03±0.01 0.02±0.00
NOx 0.08±0.02 0.33±0.05 0.09±0.04 0.25±0.05 0.04±0.01 0.20±0.06 0.06±0.01 0.18±0.04 0.04±0.02 0.22±0.08 0.11±0.04 0.21±0.06 1.08±0.14 1.91±0.28 0.07±0.02 0.30±0.06
TKN 0.78±0.09 0.94±0.12 0.94±0.18 0.64±0.03 0.75±0.05 0.61±0.04 0.48±0.14 0.32±0.03 0.79±0.13 0.60±0.09 0.41±0.08 0.29±0.04 0.48±0.05 0.51±0.05 0.66±0.04 0.59±0.03
DON 0.31±0.07 0.52±0.04 0.56±0.07 0.47±0.04 0.42±0.07 0.47±0.08 0.36±0.06 0.31±0.05 0.39±0.04 0.52±0.11 0.30±0.05 0.28±0.05 0.30±0.05 0.36±0.06 0.46±0.04 0.45±0.04
TN 0.82±0.08 1.37±0.12 0.84±0.09 0.83±0.05 0.76±0.05 0.81±0.07 0.49±0.10 0.56±0.05 0.88±0.14 0.89±0.12 0.48±0.08 0.59±0.08 1.65±0.19 2.26±0.26 0.67±0.04 0.90±0.08
FRP 0.05±0.01 0.03±0.00 0.02±0.01 0.02±0.00 0.02±0.01 0.02±0.00 0.01±0.01 0.01±0.00 0.03±0.02 0.02±0.00 0.01±0.00 0.01±0.00 0.04±0.01 0.02±0.00 0.02±0.00 0.02±0.00
TP 0.19±0.04 0.15±0.03 0.07±0.01 0.05±0.00 0.08±0.01 0.04±0.00 0.03±0.02 0.01±0.00 0.08±0.02 0.04±0.01 0.02±0.01 0.02±0.00 0.15±0.08 0.04±0.00 0.04±0.01 0.04±0.01
2013-2015
NO2 0.01±0.00 0.01±0.00 0.01±0.00 0.01±0.00 0.00 0.01±0.00 0.01 0.00±0.00 n/a 0.73±0.07 n/a 0.01±0.00 n/a n/a 0.01±0.00 0.01±0.00
FTN 0.55±0.05 0.95±0.10 0.45±0.06 0.69±0.04 0.38 0.77±0.06 0.48 0.66±0.09 n/a 0.02±0.00 n/a 0.55±0.09 n/a n/a 0.56±0.10 0.72±0.07
UreaN n/a 0.02±0.01 0.00 0.02±0.00 0.05±0.03 n/a 0.02±0.01 n/a 0.25±0.43 n/a 0.01±0.00 n/a n/a n/a 0.03±0.01
PN 0.42±0.19 0.58±0.42 0.35±0.16 0.19±0.06 0.40 0.03±0.04 0.09 0.07±0.11 n/a 0.22±0.09 n/a 0.09±0.07 n/a n/a 0.18±0.07 0.01±0.03
DIN 0.14±0.10 0.39±0.11 0.07±0.06 0.22±0.07 0.01 0.19±0.06 0.14 0.27±0.09 n/a 0.20±0.08 n/a 0.22±0.05 n/a n/a 0.11±0.08 0.25±0.08
NO3 0.08±0.06 0.33±0.10 0.05±0.04 0.20±0.07 0.01 0.16±0.06 0.12 0.26±0.09 n/a 0.51±0.08 n/a 0.20±0.05 n/a n/a 0.08±0.08 0.23±0.07
NUON 0.40±0.11 0.55±0.05 0.38±0.11 0.46±0.05 0.36 0.53±0.07 0.34 0.38±0.03 n/a 0.02±0.01 n/a 0.32±0.13 n/a n/a 0.46 0.49±0.04
FTP 0.05±0.01 0.05±0.01 0.05±0.02 0.03±0.01 0.02 0.02±0.00 0.00 0.01±0.00 n/a 0.04±0.02 n/a 0.01±0.00 n/a n/a 0.06±0.04 0.03±0.01
PP 0.14±0.02 0.19±0.10 0.09±0.04 0.04±0.01 0.09 0.02±0.00 0.01 0.00±0.00 n/a 0.01±0.00 n/a 0.01±0.01 n/a n/a 0.01±0.04 0.02±0.00
DOP 0.01±0.00 0.01±0.00 0.01±0.00 0.01±0.00 0.01 0.00±0.00 0.00 0.00±0.00 n/a 8.92±1.96 n/a 0.00±0.00 n/a n/a 0.04±0.03 0.01±0.00
DIC 31.08±7.81 16.38±4.05 26.01±7.70 13.23±3.52 27.81 13.10±3.10 28.50 15.85±4.87 n/a 45.73±6.93 n/a 6.39±3.47 n/a n/a 26.34±5.09 12.17±2.89
DOC 8.61±1.64 32.68±8.43 9.44±2.51 30.31±7.61 9.64 35.70±8.95 8.95 34.96±8.27 n/a 54.65±5.45 n/a 44.97±6.39 n/a n/a 8.63±2.65 27.70±7.04
TDC 39.68±9.14 49.06±6.08 35.45±10.15 43.54±5.87 37.45 48.80±6.12 37.45 50.81±3.91 n/a 0.73±0.07 n/a 51.35±5.46 n/a n/a 34.97±7.52 39.88±6.08 aANZECC guideline values bHRAP target values
n/a not applicable
Chapter 4
88
4.3.2 Nutrient attenuation across compartments
The overall nutrient attenuation as SDC in the WSCW was estimated by
considering nutrient concentrations at W1 and W8. The annual average attenuation of
NH3 (62.7%; p < 0.0001), FRP (57.6%; p < 0.0001) and TP (65.8%; p < 0.0001) were
always positive, while TKN, NOx, DON and TN showed periodic attenuation and
release (Figure 4. 3). The average attenuation of NOx and TN was around 15%, while
DON showed negative attenuation (−58.23%). The SF1 compartment showed mostly
NOx attenuation with a few NOx releases, while the SSF1 showed NOx releases at the
beginning of operation followed by alternating release and attenuation (Figure 4. 4).
The SF1 showed alternating annual attenuation and release of TN, while annual
average TN attenuation at the SSF1 was always positive.
Based on the mean SDC, the different compartments of the WSCW showed
variable attenuation and release of nutrients (Figure 4. 5, Table 4. 3 and Table 4. 4).
Total attenuation of DIN during the dry periods was almost 1.5 times higher than that
during the wet periods, while DON and DOP were released from the system during
the dry periods. DOC attenuation during the dry periods was three times higher than
that measured during the wet periods, while DIC released from the WSCW during the
wet periods. DOC also released from the SF1 compartment.
4.3.3 Impact of hydrological variability on nutrient attenuation
The water balance and nutrient dynamics of the WSCW were influenced by
hydrological variability including rainfall, flow pattern, water volume input, water
recirculation and antecedent conditions. A negative water balance (Voutflow < Vinfow)
was observed during rainfall events after antecedent dry conditions (Figure 4. 6). Only
50% of the total inflow volume reached the outflow, with the remaining water either
stored in the different compartments or lost from the system via evapotranspiration. A
positive water balance (Voutflow > Vinfow) was observed when recirculation pumping
coincided with rainfall after wet antecedent periods (Figure 4. 6). The volumetric
contribution from ungauged areas (VCUA) to the WSCW decreased over time with
average values of 56% (2011), 39% (2012) and 27% (2013). This may have been
impacted by the decrease in mean annual rainfall and the increase in dry antecedent
conditions.
Chapter 4
89
Figure 4. 3 Nutrient attenuation in the Wharf Street Constructed Wetland from 2009 to 2015.
Bar plots (a, c, e, g, i, k and m) show attenuation over individual sampling periods, where box
plots (b, d, f, h, j, l and n) show annual average attenuation. 100% SDC indicates compete
attenuation while negative SDC indicates nutrient release. Horizontal lines and open squares
in each box plot represent the annual median and mean SDC, respectively.
Chapter 4
90
Figure 4. 4 Nutrient attenuation in the SF1 (a, b, e, f, i and j) and SSF1 (c, d, g, h, k and l)
compartment of the Wharf Street Constructed Wetland from 2009 to 2015. Bar plots (a, c, e,
g, i and k) show attenuation over individual sampling periods, where box plots (b, d, f, h, j and
l) show annual average attenuation. 100% SDC indicates compete attenuation while negative
SDC indicates nutrient release. Horizontal lines and open squares in each box plot represent
the annual median and mean SDC, respectively.
Based on the simultaneous flow and nutrient monitoring data captured over six
storm events, we assessed the EMCs and loads of TN and TP at the inlet and outlet. A
typical flow pattern during a storm event (17–20 November 2009) and subsequent
nutrient concentrations are presented in Figure 4. 7. The event was accompanied with
30 mm of rainfall and about 65% of the incoming water reached the outlet. The
WSCW attenuated 35 and 62% of the incoming TN and TP loads, respectively.
Chapter 4
91
Figure 4. 5 Nutrient attenuation within different compartments of the Wharf Street Constructed Wetland: (a) SF1, (b) SSF1, (c) SSF2, (d) SF2 and (e)
overall system. 100% SDC indicates compete attenuation while negative SDC indicates nutrient release. Horizontal lines and open squares in each
box plot represent the median and mean SDC, respectively.
Chapter 4
92
Table 4. 3 Nutrient attenuation (%) during the dry and wet period in the Wharf Street Constructed Wetland. The dry and wet period represents six
months of December to May and June to November, respectively.
SF1 SSF1 SSF2 SF2
Dry Wet Dry Wet Dry Wet Dry Wet
NH3 39.53±46.64 57.45±6.80 -101.02±86.78 69.17±5.78 -102.62±60.99 14.47±9.61 40.87±24.95 12.25±13.08
NOx 21.63±27.08 49.23±5.96 -273.82±136.12 -238.58±89.08 -532.12±168.95 -271.48±98.65 91.17±2.93 82.21±5.84
TKN -48.67±23.68 14.54±4.44 38.30±13.92 47.02±3.77 47.10±6.71 49.09±4.50 -62.79±16.74 -38.05±11.68
DON -181.85±118.97 0.24±3.10 42.57±5.06 42.63±4.88 23.55±12.39 44.85±6.60 -68.71±27.01 -45.86±17.45
TN -6.77±13.53 27.65±5.07 37.23±10.01 30.35±4.03 39.44±6.56 31.20±7.22 46.14±6.55 43.18±9.08
FRP 54.47±7.84 34.07±12.20 -19.39±69.55 57.48±5.72 57.38±21.70 3.91±41.27 50.71±9.14 44.47±6.77
TP 54.38±6.44 54.78±4.51 63.59±16.84 81.97±1.45 74.32±1.64 56.65±8.14 19.83±11.65 0.58±12.50
Table 4. 4 Nutrient attenuation (%) during the dry and wet period in the Wharf Street Constructed Wetland. The dry and wet period represents six
months of December to May and June to November, respectively.
SF1 SSF1 SSF2 Overall
Dry Wet Dry Wet Wet Dry Wet
NO2 38.58±12.81 15.12±8.04 -26.35 42.97±6.22 15.97±21.30 34.41±15.41 21.35±9.74
FTN 16.79±9.39 24.83±4.82 n/a 20.34±6.43 25.18±9.78 -1.67±12.66 22.57±5.70
UreaN n/a 21.24±9.54 78.54 54.29±13.81 26.62±17.66 n/a -46.09±60.43
PN -41.79±109.29 80.79±13.49 -838.84 252.22±92.77 83.81±28.79 40.30±34.65 139.87±36.43
DIN 67.02±11.00 37.96±7.73 -1529.95 -47.02±34.97 -50.68±54.11 32.69±13.07 19.80±17.31
NO3 67.30±15.07 37.58±9.84 7.02 -96.54±50.74 -62.93±63.96 44.46±29.62 -9.53±27.48
DON 7.36±6.54 16.93±3.14 7.02 24.76±5.33 42.75±22.99 -4.66±13.96 16.82±4.37
NUON 7.45±6.52 17.32±3.47 82.51 22.61±5.43 43.65±23.61 -11.40 17.11±5.98
FTP 14.62±13.49 38.26±7.62 90.55 74.69±3.00 56.65±15.53 -6.50±42.23 49.11±7.87
PP 41.04±24.60 79.91±20.09 92.77 89.60±4.80 78.48±15.52 93.90±23.36 87.84±12.77
DOP -71.69±98.35 307.24±271.61 -2.47 128.67±7.44 106.82±24.99 -1008.98±1039.86 586.02±539.0
DIC 18.49±10.04 18.64±9.38 7.14 -21.38±12.68 34.64±28.84 12.99±5.69 -15.68±49.16
DOC -5.74±10.96 6.98±4.19 0.00 1.18±7.35 -4.26±18.11 5.66±15.78 15.25±2.74
TDC 13.49±9.74 11.66±4.44 n/a -8.05±5.33 2.80±14.81 11.44±5.73 19.39±5.43
n/a not applicable
Chapter 4
93
Figure 4. 6 Hydrograph at the inlet and outlet of the Wharf Street Constructed Wetland during
different hydro-climatological conditions: (a) a small rainfall under the wet antecedent
condition, (b) a small rainfall under the dry antecedent condition, (c) a large rainfall under the
dry antecedent condition and (d) frequent rainfall events under the wet antecedent condition.
TN and TP attenuation during the storm events were linked to HRT; the longer
the HRT, the greater nutrient attenuation. We calculated nutrient loads during base
flows or DWC (Table 4. 5). The WSCW attenuated up to 62% TN and 99% TP loads
during the DWC. The WSCW also attenuated about 46% TN and 48% TP loads
Chapter 4
94
during the high flows. Throughout the operation the system attenuated about 45 and
65% incoming TN and TP loads, respectively from being discharged to the
downstream Canning River.
There were uncertainties in nutrient attenuation estimation when the WSCW
received ungauged water from BCD and/or OL overflow. Under these conditions,
SDC_ave was used, and the average and median of SDC_ave were lower than that of
SDC for all nutrients in the SF2 compartment.
Figure 4. 7 Nutrient concentration and flow patterns at the main inlet (W1) and main outlet
(W8) of the Wharf Street Constructed Wetland during a storm event (17–21 November 2009).
(a) Flow and TN concentration at W1, (b) flow and TP concentration at W1, (c) flow and TP
concentration at W8 and (d) flow and TP concentration at W8. Dotted lines in each box
represent the respective EMC of the nutrients. Tables summarise total water discharge, total
nutrient load and EMC at W1 and W8 during the storm event.
Chapter 4
95
Table 4. 5 Estimation of nutrient load attenuation in the Wharf Street Constructed Wetland
during the base and high flow conditions.
TN TP
Year Flow
type
Inlet
load
(kg)
Outlet
load
(kg)
Δ Load
(kg)
Load
removal
(%)
Inlet
load
(kg)
Outlet
load (kg)
Δ Load
(kg)
Load
removal
(%)
2010 Base 162.09 83.27 78.82 49 14.492 5.11 9.37 65
High 991.90 242.95 748.94 76 99.71 65.79 33.91 34
2011 Base 18.89 14.02 4.86 26 1.8102 0.49 1.32 73
High 290.17 129.31 160.85 55 29.17 20.60 8.56 29
2012 Base 4.696 1.80 2.89 62 0.939 0.59 0.34 36
High 166.07 67.66 98.41 59 16.69 10.76 5.92 36
2013 Base 25.71 24.95 0.75 3 3.84 0.04 3.79 99
High 324.30 150.57 173.73 54 32.60 23.79 8.81 27
2014 Base 52.25 52.86 -0.60 -1 3.96 3.583 0.37 10
High 490.21 100.66 389.55 79 83.10 26.73 56.36 68
4.3.4 Nutrients storage within sediments and macrophytes
Sediments of the WSCW were potential long-term sink of nutrients. TN or
TKN and TP concentrations were highest at S1 and increased with time. Sediments
were rich with TOC, with concentrations higher at S1 than at S2 between June 2010
and May 2012, while a reverse scenario was observed between October 2012 and
October 2013. The maximum and minimum sediment nutrient pools were observed at
S2 and S3, respectively (Figure 4. 8). There was a sharp increase in nutrient pools at
all sites from October 2013 to May 2014. TN and TOC pools were 8–15 and 100–
8000 times higher, respectively than TP pools. TN and TP pools increased during the
early stage of operation (June 2010 to April 2011), had reduced by October 2011 and
increased thereafter. The net accumulation of sediment nutrients with time, where t is
days since 29 June 2010 or the first sampling date:
TN=0.0027t2–3.2448t+2397; R
2= 0.39 (4.8)
TP=0.0005t2–0.8548t+478.59; R
2= 0.56 (4.9)
TOC=0.0314t2 –40.446t+24649; R
2= 0.55 (4.10)
Aquatic macrophytes such as B. articulata, B. rubiginosa, B. preissii and
S. validus accumulated nutrients (TKN and TP) in AGB and BGB (Figure 4. 9).
B. articulata and S. validus contained higher TKN and TP nutrient pools than other
species, while BGB accumulated more nutrients than that measured in AGB.
Chapter 4
96
However, macrophytes showed seasonal senescence (Table 4. 6) that can contribute
internal nutrient release.
Figure 4. 8 Nutrient pools of (a) TN, (b) TP and (c) TOC from June 2010 to May 2014 at three
sediment sampling sites of the Wharf Street Constructed Wetland. Error bars indicate standard
error.
Figure 4. 9 Dry biomass and nutrient pools in AGB and BGB of B. articulata from November
2010 to May 2012 at three sampling sites of the Wharf Street Constructed Wetland. (a) Dry
biomass in AGB, (b) dry biomass in BGB, (c) TN pools in AGB, (d) TN pools in BGB, (e) TP
pools in AGB and (f) TP pools in BGB. Error bars indicate standard error. AGB and BGB
indicate above ground biomass and below ground biomass, respectively.
Chapter 4
97
Table 4. 6 Spatial and temporal variability of physical properties of above ground biomass of
B. articulata per quadrate in the Wharf Street Constructed Wetland.
Time Site No of
Inflores-
cences
Leaf Stage (%) Stem height (m) No of
Stem Mature New Senescent Mean Max. Min.
22/11/2010 M1 3.00±
1.73
79.00±
8.08
5.33±
2.33
15.67±
6.17
1.38±
0.19
2.31±
0.12
0.15±
0.14
108.67±
26.82
22/11/2010 M2 10.00±
0.88
99.00±
11.55
1.00±
0.33
40.00±
11.26
1.66±
0.20
2.48±
0.23
0.64±
0.14
88.00±
2.60
22/11/2010 M3 9.00±
5.57
80.00±
5.13
4.67±
2.73
15.33±
2.40
1.65±
0.16
2.86±
0.15
0.35±
0.18
88.33±
11.20
02/06/2011 M1 6.67±
3.38
65.00±
7.64
6.67±
1.67
28.33±
6.01
1.48±
0.13
2.51±
0.23
0.11±
0.03
90.33±
15.17
02/06/2011 M3 0.33±
0.33
64.33±
19.80
3.00±
1.00
32.67±
20.22
1.92±
0.08
2.38±
0.23
0.17±
0.05
50.67±
3.33
06/10/2011 M1 12.33±
4.41
86.67±
1.20
3.33±
1.20
10.00±
0.00
1.72±
0.08
2.47±
0.19
0.25±
0.16
109.67±
4.84
06/10/2011 M3 6.67±
2.33
91.67±
1.67
4.00±
1.00
4.33±
0.67
1.72±
0.08
2.76±
0.22
0.13±
0.05
93.00±
9.81
03/05/2012 M1 10.67±
2.40
68.00±
9.17
3.67±
1.33
28.33±
8.82
1.69±
0.18
2.59±
0.09
0.20±
0.06
101.00±
7.77
03/05/2012 M3 3.33±
1.20
64.33±
17.70
2.33±
1.33
33.33±
18.33
1.79±
0.34
2.81±
0.35
0.15±
0.08
90.33±
16.15
4.3.5 DO variability
DO levels were higher at the outlet than that at the inlet of the SF
compartments during 65% of sampling times (p < 0.0001). However, DO
concentrations decreased at the outlet of the SSF compartments compared to that
measured at the inlet (Figure 4. 10). Near zero DO levels were observed at the outlet
of the SSF compartments, specifically during the dry periods. A diurnal DO signal
was observed in the SF1 compartment, particularly during the dry periods, with the
maxima and minima DO levels during day and night-time, respectively. DO levels
within the SSF1 increased during the wet periods, after rainfall and flow pulses.
Night-time anoxia (DO < 1 mg/L) was observed during most of the dry months in the
SSF compartments. The trend between DO concentrations and nutrient concentrations
(Figure 4. 11) and ΔDO and SDC (Figure 4. 10) at different compartments was
investigated. NH3 concentrations at the outlet of the SF compartments reduced when
DO concentrations and ΔDO. However, NH3 processing usually generates NOx in
CW, and higher DO levels suppress NOx attenuation. In the presence of ample DO,
FRP attenuated while higher FRP concentrations prevailed at lower DO
concentrations. NOx attenuation in the SF compartments was dominant (R2 = 0.52;
p < 0.0001) at very low ΔDO, probably due to denitrification. NH3 attenuation
Chapter 4
98
increased (R2 = 0.51; p < 0.0001) and NOx attenuation suppressed in the SF
compartments during the wet periods when higher ΔDO was experienced. FRP
attenuation also responded to ΔDO in the SF1 compartment, however, not as intensely
as NH3. The SSF compartments showed variable nutrient attenuation in response to
ΔDO. Although DO levels reduced in the SSF compartments, attenuation of NH3 and
FRP, and a release of NOx were observed. Moreover, the SSF compartments received
nutrients at low concentrations and therefore, a slight increase/release in
concentrations at the outlet would show large SDC (Figure 4. 10).
4.4 Discussion
4.4.1 Variability in performance within the system and over time
Variability in hydrology, CW design and effective surface area of the CW
influenced nutrient attenuation in the WSCW over time. The WSCW experiences
prolonged low flows and low water levels during the dry summer, and therefore the
recirculation system is needed to maintain a minimum flow rate. Water can be stored
in the open water bodies or SF compartments and the SSF compartments during the
dryer periods and therefore a negative water balance was frequently observed; this
extended the HRT and led to higher nutrient load attenuation (up to 62% TN and 99%
TP).
Incoming water during the storm events contributed to the outflow with a
positive water balance and water discharged for up to 2 days (Figure 4. 6); this was
likely sourced from ungauged inputs such as BCD, direct precipitation and/or
overflow of the OL. The ungauged inputs delivered 27–56% of incoming water
depending on rainfall pattern, season and antecedent dry or wet conditions. The first
few hours of the storm events delivered nutrient in higher concentrations (Figure 4. 7)
due to the first flush effect. Some portion of water travelled from the outlet of SF1 to
the inlet of SFF2 by-passing SSF1 compartment during the high flow conditions that
can reduce overall attenuation capacity. However, during such high flow conditions,
dilution and particle settling are probably important processes of nutrient attenuation.
The presence of macrophytes and tree logs in the SF compartments and laterite in the
SSF compartments facilitate the settling of particulate nutrients by providing
roughness.
Chapter 4
99
Figure 4. 10 Spatial and temporal trend of DO concentrations in water samples of (a) SF1, (b)
SSF1, (c) SF2 and (d) SSF2 compartment of the Wharf Street Constructed Wetland from 2009
to 2015. Diurnal DO signature at the mid zone of the (e) SF1 and (f) SSF1 compartment
during rainy (6–9 September 2014) and sunny (1–4 March 2015) conditions. Relationship of
ΔDO and SDC NH3 at the (g) SF and (h) SSF compartments, SDC NOx at the (i) SF and (j)
SSF compartments, and SDC FRP at the (k) SF and (l) SSF compartments.
Chapter 4
100
Figure 4. 11 Relationship between DO and nutrient concentrations in different compartments
of the Wharf Street Constructed Wetland. DO and NH3 concentrations at the (a) SF1, (b)
SSF1, (c) SSF2, and (d) SF2; DO and NOx concentrations at the (e) SF1, (f) SSF1, (g) SSF2,
and (h) SF2; DO and FRP concentrations at the (i) SF1, (j) SSF1, (k) SSF2 and (l) SF2.
The variable redox conditions across the different compartments, and the
variable exposure to solar radiation also impacted nutrient attenuation. The SF and
SSF compartments show higher attenuation of N and P based nutrients, respectively.
The SF compartments are open to air and light those provide the aerobic environment
essential for nitrification and ensure the stability of P absorption. The SF
compartments also provide large area of sediment surfaces and macrophytes that
contribute significantly to the attenuation of incoming nutrients. The SSF
compartments provide increased surface area and anaerobic environments suitable for
the growth of biofilms. The SSF compartments in other CWs have been shown to
support nutrient attenuation by promoting plant root growth and increasing upward,
Chapter 4
101
downward and sideways mixing of incoming water (Headley et al. 2005; Konnerup et
al. 2009). Nutrient attenuation also depends on inlet concentrations in the different
compartments (Figure 4. 12 and Figure 4. 13).
Figure 4. 12 Relationship between DO and nutrient concentrations in different compartments
of the Wharf Street Constructed Wetland. DO and NH3 concentrations at the (a) SF1, (b)
SSF1, (c) SSF2, and (d) SF2; DO and NOx concentrations at the (e) SF1, (f) SSF1, (g) SSF2,
and (h) SF2; DO and FRP concentrations at the (i) SF1, (j) SSF1, (k) SSF2 and (l) SF2.
Interestingly, DON and NOx periodically released from the SF and SSF
compartments, respectively. The decay of macrophyes and organic matter, and
sediment resuspension can generate DON in the SF systems (Kröger et al. 2007).
Negatively charged ions have a low affinity to laterite, and this may have impacted
NOx attenuation in the SSF compartments. Furthermore, NH3 attenuation generates
NOx that would then increase overall NOx concentrations at the outlet. However, both
DON and NOx were attenuated in the subsequent alternating compartments,
Chapter 4
102
highlighting the benefits of multi-compartment CW. The WSCW attenuated up to
99% of TP loads, and the sediments, macrophytes and laterite were important sinks of
P. In particular, laterite has previously been shown to adsorb P via ligand exchange
reaction, where phosphate displaces water or hydroxyls from the surface of hydrous
oxides of iron (Fe) and aluminium (Al) (Wood & McAtamney 1996). Cerezo et al.
(2001) found that the addition of Fe improved P retention significantly from 55 to
66% in a CW under semi-arid conditions. We similarly observed correlations between
improved P retention and Al or Fe attenuation in the WSCW, with co-precipitation
likely facilitating attenuation. The SSF compartments were also planted with Carex
that are expected to uptake P.
The alternating SF and SSF compartments utilised in the WSCW exhibited
improved nutrient attenuation than previously found in a single stage CW within 4 km
of the current field site (Ruibal-Conti et al. 2015; Adyel et al. 2016). Furthermore, the
WSCW exhibited higher nutrient attenuation than that found in other multi-chamber
CWs for stormwater nutrient attenuation (Malaviya & Singh 2012). This is
particularly impressive given that most of the previously investigated multi-
compartment CWs were laboratory scale or relatively small with controlled flow rate
and inflow loading, while the WSCW is a full-scale system setting in an urban
catchment, and exposed to prolonged low flows during the dry summer and episodic
pulses during the wet winter.
Different reactions and processes take place simultaneously in multi-
compartment CWs to control nutrient attenuation. Although direct nutrient attenuation
processes were not measured in the WSCW, four distinct patterns (Figure 4. 14) of
NH3 and NOx dynamics across the different compartments were observed. These
patterns included (a) pattern 1: attenuation of both species, (b) pattern 2: NH3 release
and NOx attenuation, (c) pattern 3: NH3 attenuation and NOx release, and (d) pattern
4: release of both species. These patterns can be explained in relation to prevailing
conditions in each compartment, which provides some insight into likely dominant
processes. Pattern 1 was dominant in both SF1 and SF2 compartments probably due to
plant uptake, biomass assimilation and adsorption onto soils, sediments and substrates.
Furthermore, NH3 can also be attenuated via nitrification, partial nitrification and
denitrification (PND) and anaerobic ammonium oxidation (ANAMMOX). The end
product of NH3 attenuation can be further attenuated via denitrification, dissimilatory
nitrate reduction (DNRA) and PND. Connolly et al. (2004) suggested that NH3
Chapter 4
103
removal in a vertical flow SSF system occurred firstly by adsorption onto the media,
followed by nitrification and then denitrification. These steps are likely to be similar
in the SSF compartments of the WSCW. Pattern 2 occurred occasionally when
nitrification was dominant and denitrification was suppressed. However, NOx release
and NH3 attenuation (pattern 3) was observed more frequently throughout the year in
the SSF compartments. N can also be released (pattern 4) due to desorption, and the
decay of macrophytes and organic matter. While we can suggest the controlling
processes based on ambient conditions, further investigations of the processes
occurring in the different compartments are needed.
Figure 4. 13 Relationship between inlet nutrient concentrations and nutrient attenuation in
different compartments of the Wharf Street Constructed Wetland. NH3 concentrations and
NH3 attenuation at the (a) SF1, (b) SSF1 (c), SSF2, (d) SF2; NOx concentrations and NOx
attenuation at the (e) SF1, (f) SSF1 (g), SSF2, (h) SF2, TN concentrations and TN attenuation
at the (i) SF1, (j) SSF1 (k), SSF2 (l) SF2. Positive SDC and negative SDC indicate nutrient
attenuation and nutrient release, respectively.
Chapter 4
104
Figure 4. 14 (a) Conceptual relationships between different attenuation pathways of NH3 and
NOX. Measured relationships between NH3 and NOx attenuation in the (b) SF and (c) SSF
compartments of the Wharf Street Constructed Wetland.
Over time, the sediments of the WSCW accumulated one-third of the incoming
nutrient loads. The larger the sedimentation area, the larger the sediment nutrient
pools. For example, S2 was situated in the largest sedimentation area and contained 3–
4 times higher N pools and 2–3 times higher P pools than that found in S1 and S3
(Figure 4. 8). Al and Fe levels increased in the WSCW sediments along with P pools
(Ruibal-Conti et al. 2015; Adyel et al. 2016) supported our previous suggestion that P
co-precipitated with Al/Fe minerals. Other possible routes of nutrient accumulation in
the sediments of CWs include sedimentation, precipitation, chemical reaction,
Chapter 4
105
adsorption, peat formation and burial (Reddy et al. 1999; Malaviya & Singh 2012).
Occasionally, the sediments nutrient pools of the WSCW decreased over time,
indicating that the sediments acted as a source, possibly due to sediment resuspension,
and nutrient releases driven by the periodic drying and wetting of the riparian zones.
Detailed investigation of the triggers for nutrient release from sediments is needed for
optimization of the system.
After the sediments, aquatic macrophytes in CWs are also important
facilitators of nutrient uptake and assimilation via suitable root-shoot-leaf zones.
Macrophytes of the WSCW accumulated about one-fifth of the incoming nutrient
loads, indicating higher storage than that previously found in temperate climates
(Vymazal 2013b). Macrophytes accumulated more nutrients during the early to
growth stage, particularly for the first two to three years of operation. BGB showed 2
to 3 times higher nutrient pools than that stored in AGB. AGB also demonstrated
seasonal senescence contributing to internal loading of nutrients. B. articulata and
S. validus showed the greatest capacity to assimilate nutrients and therefore the
establishment and growth of these two species could improve nutrient retention across
the system. However, the presence of unsampled macrophytes and the absence of up-
to-date species mapping made macrophyte nutrient pools estimation challenging.
Periodic species coverage count is therefore essential for better estimation of
macrophyte nutrient pools, while the harvest of mature macrophytes prior to summer
senescence is required to ensure the permanent removal of plant-stored nutrients.
Moreover, modelling approach can be incorporated to determine effectiveness of the
system over times.
Some urban water models are capable of predicting the performance of single
stage CWs; however prediction of the performance of multi-compartment CW is
challenging due to variable flow conditions. Typical computational approach such as
MUSIC - Model for Urban Stormwater Improvement Conceptualisation (MDT 2005)
conceptualises the performance of a single stage CW as a 1st order decay reaction
rate, without considering complexities such as different inlet concentrations, ungauged
water inputs, flow patterns, sediment or macrophyte uptake within the context of a
multi-compartment CW. This model underestimated nutrient attenuation during storm
events and overestimated during low flows in the WSCW using default parameters.
Alternative approaches therefore need to be considered for better prediction of
Chapter 4
106
nutrient attenuation in multi-compartment systems and ultimately the optimization of
their design.
4.4.2 ΔDO as an indicator of wetland metabolism
DO fluctuations (ΔDO) can be used as an indicator of wetland metabolism,
representing the balance between oxygen-consuming heterotrophic processes and
oxygen-delivering autotrophic processes. The strong seasonal drivers in hydrology
and productivity of the CW influence large changes in oxygen metabolism. ΔDO in
the SF1 compartment responded positively to solar radiation with statistical
significance (R2 = 0.76; p < 0.0001) during the dry periods (Figure 4. 15).
Biological, chemical and sediment oxygen demand remain intense in the
warmer and lentic water (Tuttle et al. 2008); all of these contribute to lower the DO
level. Interestingly, we also found periods of low DO and ΔDO during the summer
periods even when the exposure to solar radiation was high (Figure 4. 15a). Likely
causes of this scenario were the thick algal mats, scum, the aquatic fern (Azolla) and
other debris that accumulated as a floating mass in the SF1 compartment during the
summer when water levels dropped to less than 0.3 m. Such floating materials shade
the open water, and inhibit both photosynthesis and atmospheric DO exchange
(Nahlik & Mitsch 2006; Adyel et al. 2017a). At the same time, oxygen consumption
promoted by the respiration of the living floating material and its summer senescence,
settling and decay can increase both water column and sediment oxygen demand
(Chimney et al. 2006). Like many wetlands in Mediterranean environments, the
WSCW alternated between a lentic and lotic phase over the year with subsequent
impact on DO dynamics. DO saturation increased in the lotic phase of the WSCW
compared to lentic phase, most likely due to re-aeration and the flushing or settling of
oxygen demanding substances. Positive ΔDO as a proxy of wetland metabolism was
linked with nutrient attenuation, particularly in the SF compartments. Increased DO
levels in the SF compartments due to reaeration promoted attenuation of NH3 and
FRP. DO levels reduce in the SSF system more likely due to DO consumption by
biofilms, heterotrophic processes and limited interactions with solar radiation and air
(Rajabzadeh et al. 2015).
A ΔDO ∼1–6 mg/L was correlated with up to 90% NH3 and 80% FRP
attenuation in the SF1 compartment indicating autotrophic process. NH3 and FRP
attenuation also occurred in the SSF compartment under negative ΔDO, therefore
Chapter 4
107
other process including adsorption were probably dominant (Reddy et al. 1999;
Connolly et al. 2004); this needs further investigation. The interpretation of CW
metabolism under different hydro-climatological conditions is complicated and
therefore, requires further exploration.
Figure 4. 15 (a) Monthly ΔDO saturation as a function of solar exposure during the wet and
dry periods in the Wharf Street Constructed Wetland. R2 value is 0.76 during the dry periods.
The wet and dry periods consist of June to November and December to May, respectively.
Horizontal and vertical error bar represent standard error. (b) Standard deviation of ΔDO
saturation during same solar exposure.
4.4.3 Benefits of multi-functional system
In addition to nutrient attenuation, the WSCW can provide non-market
ecosystem services. The site integrates residential areas, a nature park and
downstream sensitive waterways, providing a passive eco-recreation and educational
asset consolidated by safety, amenity and accessibility (Figure 4. 16). The site
enhances the local habitat by returning endemic species and habitat types for flora and
fauna (Synnix 2007). This site experienced mosquito problems (Russell 1999) before
establishment and the CW design ensured a water velocity sufficient to discourage
mosquito larvae. Vegetated wetlands also contain natural predators that control
mosquito population (Greenway et al. 2003).
Chapter 4
108
Figure 4. 16 Some multi-functional aspects of the Wharf Street Constructed Wetland: (a)
cycling and walking path, (b) the SF1 compartment as a habitat of waterfowl, (c–d) amenity
facilities as (c) barbeque and (d) fishing platform.
The WSCW also has economic significance, and economic value can be
placed on its ecological services as described for other CWs (Moore & Hunt 2012). A
CW designed to treat high nutrient inflows was estimated to be worth USD 3.3 million
over 20 years (Yang et al. 2008). A careful quantification of the value of the
ecological services provided by the WSCW is required to compare it with other man-
made ecosystems. Interestingly, a study in a nearby catchment of WSMD found that
house values increased with decreasing distance to a wetland (Tapsuwan et al. 2009).
The WSCW is located close to a residential area and we would expect it has influence
on the market value of nearby properties. The WSCW multi-compartment landscape
and green infrastructure design concept can be implemented in other areas to enhance
urban livability and ecological sustainability.
a
)
b
)
c
)
d
)
Chapter 4
109
4.5 Conclusions
The monitoring, management and treatment of urban stormwater nutrients are
essential to safeguard and restore downstream waterways. In this study, we assessed
nutrient attenuation in the hybrid WSCW. Major outcomes of this research included:
1) The research demonstrated the variability in nutrient attenuation from the
event-scale with high frequency data to the annual-scale, and therefore, analysis
provided important empirical evidence justifying how the system changed over time.
Results can serve as an important benchmark for future CW design in the area
experiencing hydrologically diverse conditions i.e., near zero flow to pulse flow
conditions.
2) The study used a novel wetland design and demonstrated the benefit of
alternating SF and laterite-based SSF compartments on the overall nutrient dynamics
in a Mediterranean climate, highlighting that each compartment has its own
characteristic hydrological and biogeochemical variability. The SF and laterite-based
SSF compartments of the WSCW showed higher attenuation of N and P, respectively.
3) The research indicated relative contribution of sediments and aquatic
macrophytes for capturing incoming nutrient loads. The sediments were a larger pool
of nutrients than the dominant macrophytes such as B. articulata and S. validus.
However, seasonal macrophyte senescence and sediments resuspension appeared to
deliver nutrients back to the system.
4) The research explored “wetland metabolism” based on a diurnal changes in
high-frequency oxygen data, as a proxy for CW function and performance. This was
the first example in CW to show diurnal changes in oxygen (or metabolism) were
linked to weather conditions and able to be correlated with changes in nutrient
concentrations and importantly, the degree of nutrient attenuation. This opened up the
possibility for in situ sensors to be used to provide a near-real time measure of
stormwater wetland function.
5) The implementation of this CW concept in a highly seasonal Mediterranean
climate demonstrated that urban liveability and waterways protection can be improved
through careful design. Future CW developments could be justified by performing a
non-market cost-benefit analysis.
Chapter 4
110
[This page has left blank intentionally]
Chapter 5
Storm event-scale nutrient attenuation and
metabolism in constructed wetlands
This chapter is based on manuscript published as: T. M. Adyel, M. R. Hipsey and C.
E. Oldham (2017) Storm event-scale nutrient attenuation in constructed wetlands
experiencing a Mediterranean climate: A comparison of surface flow and hybrid
surface-subsurface flow system. Science of the Total Environment, 598, 1001-1014.
Chapter 5
112
Abstract
Among different Water Sensitive Urban Design options, constructed wetlands
(CWs) are used to protect and restore downstream water quality by attenuating
nutrients generated by stormwater runoff. This research compared the nutrient
attenuation ability during a diverse population of storm events of two CWs: (a) a
hybrid CW with multiple alternating surface flow (SF) and laterite-based subsurface
flow (SSF) compartments, and (b) a single stage SF CW. Within-storm variability,
nutrient concentrations were assessed at 2 to 3-h intervals at both the main inlet and
outlet of each CW. Dissolved oxygen concentrations of the surface waters were also
monitored at 10-min intervals using high frequency in situ sensors. Nutrient loads into
the CWs were observed to be higher when a high rainfall event occurred, particularly
after longer antecedent dry conditions. Longer hydraulic retention times promoted
higher attenuation at both sites. However, the relative extent of nutrient attenuation
differed between the CW types; the mean total nitrogen (TN) attenuation in the hybrid
and SF CW was 45 and 48%, respectively. The hybrid CW attenuated 67% total
phosphorus (TP) loads on average, while the SF CW acted as a net TP source.
Periodic storm events transitioned the lentic CW into a lotic CW and caused riparian
zone saturation; it was therefore hypothesized that such saturation of organic matter
rich-riparian zones led to release of TP in the system. The hybrid CW attenuated the
released TP in the downstream laterite-based SSF compartments. Diel oxygen
metabolism calculated before and after the storm events was found to be strongly
correlated with water temperature, solar exposure and antecedent dry condition during
the pre-storm conditions. Furthermore, the SF CW showed a significant relationship
between overall nutrient load attenuation and the change in oxygen metabolism during
the storm perturbation, suggesting oxygen variation could be a useful proxy indicator
of CW function.
Keywords
Dissolved oxygen; EMC; metabolism, nutrient load; storm event
Chapter 5
113
5.1 Introduction
Increasing catchment imperviousness associated with industrial and residential
growth in cities with conventional drainage systems has led to increased volumes and
peak flows from storm events. This stormwater runoff carries a range of pollutants and
nutrients, integrating both point and non-point sources within the urban catchment
(Filoso et al. 2015; Ma et al. 2017b, a). Depending on the pattern of the storm event,
the antecedent dry and wet conditions can influence the periodic build up as well as
the extent of the subsequent wash-off and export of nutrients. Nutrients ultimately
make their way into downstream waterways and can lead to water quality impairment,
such as excessive algal growth or eutrophication, dissolved oxygen (DO) depletion,
fish kills, amenity reduction and disruption of ecological services (Taylor et al. 2004;
Burns et al. 2012; Chen et al. 2012). Furthermore, these storm events can cause flash
floods and river bank erosion, and thereby alter the morphology and hydrograph
characteristics of waterways (Walsh et al. 2012). It is, therefore, increasingly common
practice to retain stormwater within catchments for a period before it is released to
downstream water bodies as part of Water Sensitive Urban Design (WSUD) (Lloyd
2001). One of the frequently implemented WSUD elements for water quality
improvement is the use of constructed wetlands (CWs) to attenuate episodic
stormwater flows and nutrient concentrations (Roy et al. 2008; Fletcher et al. 2013;
Fletcher et al. 2014).
The total load of nutrients exported during episodic storm events depends on
the hydro-meteorological conditions including the amount, duration and frequency of
rainfall, antecedent conditions (dry or wet), and the extent of runoff generation
Episodic water pulses make nutrient attenuation within CWs challenging due to the
low hydraulic retention time (HRT), since this limits the extent to which
biogeochemical reactions can take place. As outlined in previous studies, CWs in
Mediterranean climate experience prolonged periods of low flows, and the
performance of these systems during base flow conditions is highly variable (Adyel et
al. 2016; Adyel et al. 2017a). Furthermore, it is not clear how effectively these CWs
attenuate nutrients during episodic storm events or pulses when loads are the highest.
During low flows, CWs can become stagnant and transition into a lentic stage,
however, increased flow rates and water levels during storm pulses shift them into a
lotic stage. Understanding the significance of this lentic-lotic transition on the overall
nutrient dynamics of a CW is challenging and has not been investigated in detail. We
Chapter 5
114
have previously found that diurnal change in CW metabolism (the balance of
processes that generate or consume oxygen) was correlated with solar radiation during
base flow conditions (Adyel et al. 2016; Adyel et al. 2017a). However, shifts in
metabolism during storm events have not previously been monitored in CWs. In turn,
understanding the possible linkages between CW metabolism and nutrient attenuation
would be beneficial for CW design and optimization of its function.
In this study, nutrient attenuation during a range of storm events was compared
in two CWs located 4 km apart on the Swan Coastal Plain (SCP) of Western Australia.
These CWs were (a) a hybrid CW with alternating surface flow (SF) and laterite-
based subsurface flow (SSF) compartments, and (b) a single stage SF CW with a
meandering flow path. This research sought to determine the effectiveness of such
CWs for event-scale nutrient attenuation, under a range of antecedent and hydro-
biogeochemical conditions. Oxygen metabolism was explored as a proxy of CW
function and potential indicator of nutrient attenuation. It is envisioned that through
better understanding of the variability of CW function during storm events, this study
will help support the design of improved urban water systems for the protection of the
downstream water quality.
5.2 Materials and methods
5.2.1 Study sites
The first site was the hybrid or multi-compartment Wharf Street Constructed
Wetland (WSCW) that receives stormwater from a 129 ha urban watershed, before
discharging into the Canning River. Approximately 1 ha of the WSCW (Figure 5. 1)
consists of three vegetated and open water SF compartments and two laterite-based
SSF compartments. The WSCW was built in 2009, while the SSF compartments were
incorporated in October 2012. Details of the WSCW are provided in Adyel et al.
(2016).
The second study site was the Anvil Way Compensation Basin (AWCB), an
approximately 0.9 ha meandering single stage SF CW (Figure 5. 2), that receives
stormwater from a neighbouring 360 ha residential, heavy traffic way and light-
industrial catchment. The site was initially built in 2004 as a drainage compensating
basin to retain small amounts of stormwater and reduce the export of stormwater
nutrients to the Canning River. The AWCB was restored to a “living stream” in
Chapter 5
115
November 2010. Details of the AWCB are described in Adyel et al. (2017a). A
comparison of main features of two CWs is given in Table 5. 1.
5.2.2 Data and field campaign
The study aimed to compare the response of the CWs to several storm events
over the period 2009 and 2015. The assessment of events combined a range of
hydrographic and water quality data. In total 14 events were assessed for changes in
wetland metabolism and 9 events were chosen for nutrient attenuation analysis based
on logistical constraints and data availability. Further procedures of sampling, analysis
and data synthesis are given in following sections.
5.2.2.1 Meteorological and water quantity data
Meteorological data were obtained from the Department of Agriculture and
Food (DAFWA) weather station located approximately 3.8 km north-west of the
WSCW and 5 km north-west of the AWCB. Daily rainfall data was obtained from the
Department of Water (DoW) and the Bureau of Meteorology (BoM). The DoW
station was approximately 0.3 km north-west of the WSCW and 0.8 km north-west of
the AWCB; while the BoM station was approximately 7 km north-east of the WSCW
and 5 km north-east of the AWCB. 5-min intervals water discharge and level data at
the inflow and outflow stations of the WSCW and AWCB were obtained from DoW.
In brief, the main inflow of the WSCW was measured via a structure comprising a
rock riffle (primary flow control) with a set of three by-pass pipes (under the riffle)
and a hydrostatic water level probe to provide water stage information. The outflow of
the WSCW was measured at the concrete V-notch weir using a water level transducer
in a float well system (Adyel et al. 2015b). A Starflow ultrasonic instrument
(UnidataTM, Australia) at the inflow and a float well sensor at the outflow of the
AWCB were used to measure discharge (Ruibal-Conti et al. 2015)
Chapter 5
116
Figure 5. 1 (a) Map of the Wharf Street Constructed Wetland indicating different sampling points. Water samples were collected at the main inlet
(W1) and main outlet (W8) during storm event flows and W1 to W8 during base flows. The ornamental lake (L) and Bebington Court Drain (W7)
deliver ungauged water during the episodic storm events. Flow rate was measured at inflow(1) station from 2010 to 2013 and inflow since 2013. (b)
Cross-section of flow path through the alternating SF and SSF compartments of the Wharf Street Constructed Wetland. Additional pipes allow the
SSF1 to be by-passed under high flow conditions and half of the SSF2 is covered by grass.
a)
b)
W1 W8
Chapter 5
117
Figure 5. 2 (a) Map of the Anvil Way Compensation Basin indicating different sampling
points. Water samples were collected at the main inlet (W1) and main outlet (W2) during
storm event flows, and W1, Mars Street Drain or MSD (W2) during base flows. W2 and
possibly groundwater deliver ungauged water during the episodic storm events and wet
periods. S1-S4 and M1-M4 indicate sediments and macrophytes sampling points, respectively.
(b) Cross section of the transect within the riparian zone indicating instrumental set-up and
sampling points. P1, P2 and P3 indicate the porewater monitoring points.
P3
P2
P1
P3
P2
P1
DO loggerDO logger
a)
Chapter 5
118
Table 5. 1 Comparative description of the Wharf Street Constructed Wetland and the Anvil
Way Compensation Basin.
WSCW AWCB
CW type Multi- compartment or hybrid
(multiple alternating SF and SSF
compartments)
SF
Year of establishment 2009, laterite incorporation in 2012 2004, restoration in 2009
CW Area 1ha 0.9ha
CW shape Multiple rectangular and oval Triangular
Catchment area 129 ha 180 to 360 haa
Specific filter media Laterite in SSF compartments Absence
Groundwater connectivity Absence, but groundwater top-up
performed during the dry summer periods
Possible groundwater
connection, particularly
during the wet winter
Recirculation system Present, particularly during the summer
periods
Absence
Inflow Main inlet, BCD and OL overflow Main inlet, MSD and
possibly groundwater
Outflow Main outflow Main outflow
Water storage Different SF and SSF compartments Meandering channel
Soil type Branded red soil Sandy
Dominant macrophytes Baumea articulata, B. rubiginosa, B.
preissii, Carex appressa, Juncus kraussii,
Schoenoplectus validus and Typha
domingensis
S. validus, B. articulata, B.
juncea, Juncus pallidus and
J. kraussii
BCD: Bebington Court Drain OL: The ornamental lake
MSD: Mars Street Drain aDepending on whether the upstream compensation basin network is hydraulically connected to AWCB
5.2.2 Nutrient data
Nutrient concentrations of water samples were measured at 2 to 3-h intervals
throughout the hydrograph for each storm event at the inlet and outlet of the CWs.
Unfiltered water samples were collected and analysed for unfiltered total nitrogen
(TN) and unfiltered total phosphorus (TP) concentrations (APHA 1998, 2012). The
AWCB water samples were additionally collected and filtered on-site using 0.45 μm
membrane filter, stored on ice until returning to the laboratory, and then frozen until
analysis. Water samples of the AWCB were analysed for filtered ammonia (NH3),
filtered total oxidised nitrogen (NOx), filtered nitrite (NO2), filtered urea nitrogen
(UreaN), filtered total nitrogen (FTN), filtered reactive phosphorus (FRP), filtered
total phosphorus (FTP), dissolved organic carbon (DOC) and dissolved inorganic
carbon (DIC) using Lachat Quick Chem procedure (APHA 2012). Dissolved
inorganic nitrogen (DIN), dissolved organic nitrogen (DON), dissolved non-urea
organic nitrogen (NUON), dissolved organic phosphorus (DOP) and total dissolved
carbon (TDC) within the water samples of the AWCB were calculated based on the
Chapter 5
119
direct measurements (Adyel et al. 2017a). Detailed sampling duration, number of
samples and hydro-meteorological characteristics including rainfall amount,
antecedent dry days (ADD) and inflow volume of the different events are provided in
Table 5. 2. . Events of the AWCB and WSCW were denoted as AE1 to AE3 and WE1
to WE6, respectively.
5.2.2.3 Dissolved oxygen data and physicochemical data
In situ DO concentrations and saturation were recorded at 10-min intervals
using a miniDOT DO logger (PMA, USA) at the top 20 cm of water column of CWs
(first SF compartment of the WSCW and meandering channel of the AWCB). These
points represented the area subject to influence by periodic solar exposure and flow
variation. DO dynamics during the pre- and post-storm events were considered during
14 occasions from August 2014 to July 2015. An additional DO logger (D-opto, New
Zealand) was also installed at the main outlet (W3) of the AWCB during first event
(AE1), and at the main inlet (W1) during second event (AE2) and third event (AE3) to
measure DO levels at 10-min intervals. Additional DO data were also compared with
DO data that obtained from meandering water column of the AWCB. Moreover,
physicochemical parameters of the meandering water column of the AWCB were also
monitored in every minute during AE1 and AE2 using an AAQ water quality sonde
(Advance Technology, Japan).
Table 5. 2 Event campaigns and associated hydro-meteorology for the assessment of the
Wharf Street Constructed Wetland and the Anvil Way Compensation Basin.
Event Date Event
sampling
duration
(hrs)
No of
samples
at inlet &
outlet
Rainfall
(mm)
ADD
(day)
Inflow
volume
(m3)
Outflow
volume
(m3)
Site 1: The WSCW
Event 1 (WE1) 17-20 November 2009 70.00 15 & 15 31.20 3 17056 11700
Event 2 (WE2) 22-24 March 2010 50.00 9 & 8 40.80 23 6408 6599
Event 3 (WE3) 20-21 May 2011 27.00 8 & 8 30.30 1 1928 2518
Event 4 (WE4) 1-3 February 2012 48.00 7 & 8 20.50 4 1329 444
Event 5 (WE5) 7-10 May 2014 73.00 13 &13 58.00 0 9722 9144
Event 6 (WE6) 17-20 June 2014 70.00 12 &12 59.20 5 18361 14970
Site 2: The AWCB
Event 1 (AE1) 11-19 March 2015 28.00 12 &12 2.60 38 1239 336
Event 2 (AE2) 6-12 April 2015 47.00 12 &12 21.20 0 17210 19699
Event 3 (AE3) 15-18 May 2015 95.00 12 & 11 65.20 11 86605 111495
Chapter 5
120
5.2.2.4 Soil porewater monitoring at the AWCB
Periodic wetting and drying can change the oxidation reduction potential (Eh)
of soils and water, and Eh would then influence nutrient processing within the riparian
zone and porewater. Eh and porewater nutrient concentrations of the riparian soils of
the AWCB were therefore additionally monitored at three points (P1, P2 and P3)
along a transect extending from the water's edge through to the riparian zone (Figure
5. 2b). P1, P2 and P3 were approximately 60, 40 and 20 cm below the soil surface and
each point was horizontally separated by 50 cm. Three Eh sensors (Sensorex, USA)
were inserted vertically into these points and connected to a data logger (DaqPro,
USA), recording at 30-min intervals. One suction cup (UTM, Germany) was also
placed at each point (same depth of Eh sensors) with the other end attached to
sampling bottles (500 mL capacity). The suction cups periodically collected water
samples from the soil porewater and stored them for later analysis. A perforated PVC
pipe was inserted up to 40 cm below the soil surface in the transect. A CTD
(conductivity, temperature and depth) diver (Schlumberger, Germany) was placed at
the bottom of the pipe. The CTD pressure changes were compared with an on-site
atmospheric Baro diver (Onset HOBO, USA). Electrical conductivity (EC),
temperature and water pressure were monitored at 10-min intervals.
5.2.3 Water balance
Water balance was estimated during nine events (Table 5. 2), using gauged
flows at the main inflow and outflow of the two CWs. When total incoming water
volume (Vinflow) was higher than outgoing water volume (Voutflow), the scenario was
considered a negative water balance (Voutflow < Vinflow). A positive water balance
(Voutflow > Vinflow) was observed when the volume of outgoing water was higher than
the volume of incoming water. These are described in Adyel et al. (2016) for the
WSCW and in Adyel et al. (2017a) for the AWCB.
5.2.4 Nutrient attenuation estimation during events
Nutrient loads passing through the inlet and outlet of the two sites were
calculated to estimate nutrient load attenuation:
𝐿 = 𝑄𝑖 × 𝐶𝑖 (5.1)
Chapter 5
121
𝐿𝐴 =∑ 𝐿𝑖−∑ 𝐿𝑜
∑ 𝐿𝑖 × 100 (5.2)
where L is the load (mg/s), Qi is the flow rate at ith timestep (m3/s), Ci is the nutrient
concentration at ith timestep (mg/L), 𝐿𝐴 is load attenuation (%), 𝐿𝑖 is the load at the
inlet (mg/s) and 𝐿𝑜 is the load at the outlet (mg/s). The total number of sample
collected per event is shown in Table 5. 2.
The event mean concentration (EMC) as mg/L of nutrients at the inlet and
main outlet of two sites were also estimated to indicate volume and time weighted
nutrient concentrations during events as:
𝐸𝑀𝐶 =∑ 𝑉𝑖
𝑛𝑖=1 ×𝐶𝑖
𝑉𝐸 (5.3)
where Vi is the volume proportional to the flow rate at ith timestep (m3), Ci is the
nutrient concentration at ith timestep (mg/L), n is the total number of events
considered and 𝑉𝐸 is the total runoff volume across an event (m3).
The site mean concentration (SMC) as mg/L was defined as the mean EMC
value recorded at the inlet and outlet of the site during all studied events, and was
calculated as:
𝑆𝑀𝐶 = ∑ 𝐸𝑀𝐶𝐸𝑛𝑖=1 × 𝑉𝐸/𝑉𝐴𝐸 (5.4)
where EMCE is the EMC for a particular event (mg/L), VE is the total runoff volume
across an event (m3), n is the total number of events considered for a specific site and
𝑉𝐴𝐸 is the total volume discharged during all the events (m3)
The event flux (EF) as kg/ha was defined as the total delivery of nutrients from
respective catchments during an event, and was calculated as:
𝐸𝐹 =𝐸𝑀𝐶 ×𝑉𝐸
𝐴 (5.5)
where A is the active surface of the catchment that effectively contributed to runoff
(ha), EMC is the event mean concentration (mg/L) and 𝑉𝐸 is the total runoff volume
across an event (m3).
Chapter 5
122
5.2.5 Metabolism estimation
Metabolism as ∆DO (%) is defined as the daily fluctuation of maximum and
minimum DO saturation in the CW, as:
𝛥𝐷𝑂 = 𝐷𝑂𝑚𝑎𝑥 − 𝐷𝑂𝑚𝑖𝑛 (5.6)
where DOmax is the maximum DO saturation during the afternoon (%) and DOmin is the
minimum DO saturation during the morning (%).
Metabolism was further analysed over four consecutive days before and after
storm perturbation. Average metabolism over four consecutive days before the storm
perturbation, ∆DO̅̅ ̅̅ ̅̅pre (%), was estimated as:
∆𝐷𝑂̅̅ ̅̅ ̅̅𝑝𝑟𝑒 =
∑ (𝐷𝑂𝑚𝑎𝑥𝑖𝑛𝑖=1 − 𝐷𝑂𝑚𝑖𝑛𝑖
)
𝑛 (5.7)
where 𝐷𝑂𝑚𝑎𝑥𝑖 is the average maximum DO saturation per day during the afternoon
before the storm perturbation over four consecutive days (%), 𝐷𝑂𝑚𝑖𝑛𝑖 is the average
minimum DO saturation per day during the morning before the storm perturbation
over four consecutive days (%) and n is number of studied days i.e., 4 days.
Similarly, average metabolism over four consecutive days after the storm
perturbation, ∆DO̅̅ ̅̅ ̅̅post (%), was estimated as:
∆𝐷𝑂̅̅ ̅̅ ̅̅𝑝𝑜𝑠𝑡 =
∑ (𝐷𝑂𝑚𝑎𝑥𝑖𝑛𝑖=1 − 𝐷𝑂𝑚𝑖𝑛𝑖
)
𝑛 (5.8)
where 𝐷𝑂𝑚𝑎𝑥𝑖 is the average of maximum DO saturation per day during the afternoon
after the storm perturbation over four consecutive days (%), 𝐷𝑂𝑚𝑖𝑛𝑖 is the average of
minimum DO saturation per day during the morning after the storm perturbation over
four consecutive days (%) and n is number of studied days i.e., 4 days.
The overall average change in metabolism due to a storm perturbation,
∆(∆DO̅̅ ̅̅ ̅̅ ) (%), was estimated as:
∆(∆𝐷𝑂̅̅ ̅̅ ̅̅ ) = ∆𝐷𝑂̅̅ ̅̅ ̅̅𝑝𝑜𝑠𝑡 − ∆𝐷𝑂̅̅ ̅̅ ̅̅
𝑝𝑟𝑒 (5.9)
Chapter 5
123
where ∆𝐷𝑂̅̅ ̅̅ ̅̅𝑝𝑜𝑠𝑡 is average metabolism over four consecutive days after the storm
perturbation (%) and ∆𝐷𝑂̅̅ ̅̅ ̅̅𝑝𝑟𝑒 is average metabolism over four consecutive days
before the storm perturbation (%).
5.3 Results
5.3.1 Water balance during events
The negative water balance was observed during WE1, WE4, WE5, WE6 and
AE1, while the positive water balance was observed during WE2, WE3, AE2 and
AE3. For the former, stormwater flooded the CWs, and was stored within the SF and
SSF compartments of the WSCW and the meandering channel of the AWCB.
Additional water other than main inflow came from ungauged sources during the
positive water balance. These ungauged sources included BCD and OL in the WSCW,
and the MSD and possible groundwater in the AWCB. Event hydrographs showed
ongoing storm flows at the outlet for 2–24 h after the cessation of the storm event
indicating ungauged water inputs.
5.3.2 Water quality and nutrient attenuation during events
Physicochemical parameters in the main channel of the AWCB varied over
events. AE1 delivered more turbid water than occurred during AE2. EC reduced by
two-thirds in AE2 compared to that of AE1. AE2 flushed out one-fourth of the
chlorophyll-a that previously existed in the AWCB. pH was slightly reduced within
the AWCB during AE2 (6.8–7) compared to AE1 (7–7.5). No such data were
available in the WSCW.
Nutrient concentrations at the inlet and outlet of the studied CWs varied across
the events. The average outlet TN concentration was higher in the WSCW than that
measured in the AWCB (Figure 5. 3a). A reverse scenario was observed in case of TP
(Figure 5. 3b). The outlet concentrations of TN in both CWs were within the
prescribed Australian and New Zealand Environmental Conservation Council
(ANZECC) guideline and Healthy River Action Plan (HRAP) target. However, TP
concentrations at the outlet of the AWCB increased and exceeded the ANZECC
guideline and HRAP target. This additional TP was likely generated within the system
or from ungauged sources, and had limited time to be attenuated in the AWCB. Water
samples from the AWCB were also analysed for other nutrient parameters. NH3,
Chapter 5
124
DON, FRP, DOP, DIC and DOC concentrations decreased within the AWCB while
NOx and DIN concentrations increased (Figure 5. 4).
Figure 5. 3 (a) TN and (b) TP concentrations at the inlet and outlet of the Wharf Street
Constructed Wetland and Anvil Way Compensation Basin during different storm events.
Solid circles and solid horizontal lines within individual box indicate the mean and median
value of nutrient at respective storm events. Dashed vertical lines separate two CWs. Bottom
and upper parts of the shaded horizontal panels of TN indicate HRAP target and ANZECC
guideline, respectively and reverse for TP.
Nutrient EMCs were calculated to explore how nutrient concentrations varied
over events in two CWs. EMCs of dissolved and particulate nutrients varied across
spatial and temporal scales. The WSCW received stormwater containing higher TN
and TP EMCs than that received by the AWCB (Table 5. 3). Between the two CWs,
TP EMCs at the outlet were higher in the AWCB. AE1 delivered stormwater at the
inlet of the AWCB with higher EMC of NH3, FTN, DON, FRP, FTP, DOP, DOC,
DIC and TDC. The inlet EMC for NOx, DIN, NO3 and PP was higher during AE3
(Table 5. 4). The average EMC of NH3, DIN, DON, DOP, DOC and DIC decreased
Chapter 5
125
at the outlet compared to that measured at the inlet; however, the reverse scenario was
observed for NOx and FRP (Table S1). SMCs of incoming and outgoing nutrients also
varied across the CWs. SMC of TN and TP was higher at the WSCW than that of the
AWCB (Table 5. 3).
Figure 5. 4 Nutrient concentrations at the inlet and outlet of the Anvil Way Compensation
Basin during three storm events.
Incoming nutrient loads also showed variability across the two different
catchments. The MSMD catchment delivered a higher nutrient load than that of the
WSMD (Table 5. 3). For example, the AWCB received up to 78.21 kg TN and
7.74 kg TP, while the WSCW received up to 19.14 kg TN and 2.12 kg TP during the
events. Moreover, the MSMD catchment delivered slightly higher nutrient mass per
unit area than that of the WSMD. The AWCB also received higher average inorganic
loads than organic loads of N, P and carbon (C) (Table 5. 4). Storm event nutrient
attenuation varied between the WSCW and AWCB depending on flow conditions.
Typical hydrographs and pollutographs over the storm events are shown in Figure 5.
5. On average, the WSCW attenuated 45% TN and 67% TP loads. Although the
AWCB attenuated 48% TN loads, the CW also released 23% TN loads during AE3.
The AWCB attenuated all nutrient species during AE1. Loads of NH3, NOX and DIN
in the AWCB were attenuated as about 24, 35 and 31%, respectively during AE1,
while NH3, DON, FRP, DOP, DOC and DIC loads generated during AE2 and AE3.
Chapter 5
126
Overall, the AWCB attenuated inorganic N while released organic N, both inorganic
and organic P and C during the events.
Table 5. 3 Variation of EMC, SMC, EF, load and load attenuation of TN and TP during storm
events in the Wharf Street Constructed Wetland and the Anvil Way Compensation Basin.
5.3.3 Changes of soil porewater properties due to riparian saturation
Saturation can influence the characteristics of riparian zones during
hydrological variability including rainfall and groundwater seepage. Saturation due to
water level change modified Eh, EC and temperature of the riparian zone and
ultimately the nutrient dynamics within soil porewater of the AWCB. Temperatures
within the riparian soil decreased due to saturation and ranged from 15 to 21 °C. EC
sharply reduced after saturation occurred. When the saturated zone started to dry
again, EC of that area increased (Figure 5. 6a). Eh of the riparian soil varied with
saturation and monitoring depth. Eh slowly decreased following the soil saturation and
showed a distinct pattern (Figure 5. 6b). The deepest point of the transect exhibited
reduced conditions (Eh ~ − 340 to − 310 mV), while Eh was slightly higher (ranging
from − 310 to − 295 mV) at P2. P1, the point 20 cm below the soil surface, showed
high variability in Eh, ranging from reducing to oxidizing conditions (Eh ~ − 200 to
Parameter
WSCW AWCB
Storm events
WE1 WE2 WE3 WE4 WE5 WE6 AE1 AE2 AE3
TN EMCinflow (mg/L) 0.83 2.0 1.28 0.90 1.07 1.04 1.02 0.38 0.90
TN EMCoutflow (mg/L) 0.78 1.38 0.72 0.88 0.64 0.72 0.21 0.45 1.11
TN SMCinflow (mg/L) 1.09 0.82
TN SMCoutflow (mg/L) 0.82 1.00
TP EMCinflow (mg/L) 0.11 0.33 0.23 0.20 0.18 0.09 0.21 0.09 0.09
TP EMCoutflow (mg/L) 0.06 0.14 0.06 0.05 0.06 0.05 0.05 0.15 0.16
TP SMCinflow (mg/L) 0.15 0.09
TP SMCoutflow (mg/L) 0.07 0.16
TN loadinflow (kg) 14.09 12.84 2.46 1.20 10.36 19.14 1.26 6.58 78.21
TN loadoutflow (kg) 9.13 9.10 1.28 0.39 5.84 10.82 0.26 0.73 96.43
TN load reduction (kg) 4.96 3.74 1.18 0.81 4.52 8.32 1.00 5.85 -18.22
TN load reduction (%) 35.20 29.13 47.97 67.50 43.63 43.47 79.37 88.91 -23.30
TP loadinflow (kg) 1.96 2.12 0.45 0.27 1.72 1.61 0.26 1.59 7.74
TP loadoutflow (kg) 0.75 0.95 0.14 0.02 0.59 0.73 0.07 2.51 14.19
TP load reduction (kg) 1.21 1.17 0.31 0.25 1.13 0.88 0.19 -0.92 -6.45
TP load reduction (%) 61.73 55.19 68.89 92.59 65.70 54.66 73.08 -57.86 -83.33
TN EF (kg/ha) 0.11 0.10 0.02 0.01 0.08 0.15 < 0.01 0.02 0.26
TP EF (kg/ha) 0.01 0.02 0.00 0.00 0.01 0.01 < 0.01 0.01 0.03
TN SEF (kg/ha) 0.08 0.14
TP SEF (kg/ha) < 0.01 0.02
Chapter 5
127
140 mV). A reduced Eh at P2 and P1 was likely related to increased organic matter
content of the AWCB.
Table 5. 4 EMC and load variability at the inlet and outlet for nutrient species in the Anvil
Way Compensation Basin during different storm events.
Parameter Events Parameter Events
AE1 AE2 AE3 AE1 AE2 AE3
NH
3
EMCinflow (mg/L) 0.44 0.03 0.22
FR
P
EMCinflow (mg/L) 0.13 0.05 0.05
EMCoutflow (mg/L) 0.04 0.03 0.23 EMCoutflow (mg/L) 0.04 0.10 0.11
Loadinflow (kg) 0.55 0.45 18.98 Loadinflow (kg) 0.16 0.94 4.52
Loadoutflow (kg) 0.05 0.52 19.49 Loadoutflow (kg) 0.05 1.71 9.68
Load reduction (kg) 0.50 -0.07 -0.51 Load reduction (kg) 0.11 -0.77 -5.16
Load reduction (%) 90.91 -15.56 -2.69 Load reduction (%) 68.75 -81.91 -114.16
NO
x
EMCinflow (mg/L) 0.01 0.09 0.32
FT
P
EMCinflow (mg/L) 0.20 0.07 0.06
EMCoutflow (mg/L) 0.00 0.07 0.36 EMCoutflow (mg/L) 0.06 0.13 0.12
Loadinflow (kg) 0.01 1.55 27.38 Loadinflow (kg) 0.25 1.16 5.13
Loadoutflow (kg) 0.00 1.25 31.26 Loadoutflow (kg) 0.07 2.16 10.67
Load reduction (kg) 0.01 0.30 -3.88 Load reduction (kg) 0.18 -1.00 -5.54
Load reduction (%) 100.00 19.35 -14.17 Load reduction (%) 72.00 -86.21 -107.99
FT
N
EMCinflow (mg/L) 0.98 0.33 0.88 P
P
EMCinflow (mg/L) 0.02 0.02 0.03
EMCoutflow (mg/L) 0.21 0.41 1.01 EMCoutflow (mg/L) 0.01 0.02 0.04
Loadinflow (kg) 1.22 5.61 76.05 Loadinflow (kg) 0.02 0.43 2.61
Loadoutflow (kg) 0.26 7.00 87.89 Loadoutflow (kg) 0.01 0.36 3.53
Load reduction (kg) 0.96 -1.39 -11.84 Load reduction (kg) 0.01 0.07 -0.92
Load reduction (%) 78.69 -24.78 -15.57 Load reduction (%) 50.00 16.28 -35.25
PN
EMCinflow (mg/L) 0.03 0.06 0.06
DO
P
EMCinflow (mg/L) 0.07 0.01 0.01
EMCoutflow (mg/L) 0.01 0.04 0.14 EMCoutflow (mg/L) 0.02 0.03 0.01
Loadinflow (kg) 0.04 0.96 5.52 Loadinflow (kg) 0.09 0.22 0.61
Loadoutflow (kg) 0.01 0.73 11.71 Loadoutflow (kg) 0.02 0.45 0.98
Load reduction (kg) 0.03 0.23 -6.19 Load reduction (kg) 0.07 -0.23 -0.37
Load reduction (%) 75.00 23.96 -112.14 Load reduction (%) 77.78 -104.55 -60.66
DIN
EMCinflow (mg/L) 0.45 0.12 0.54
DO
C
EMCinflow (mg/L) 11.40 4.56 3.93
EMCoutflow (mg/L) 0.04 0.10 0.59 EMCoutflow (mg/L) 3.39 6.34 6.08
Loadinflow (kg) 0.56 2.00 46.35 Loadinflow (kg) 14.14 78.53 340.79
Loadoutflow (kg) 0.05 1.77 50.75 Loadoutflow (kg) 4.20 109.06 526.80
Load reduction (kg) 0.51 0.23 -4.40 Load reduction (kg) 9.94 -30.53 -186.01
Load reduction (%) 91.07 11.50 -9.49 Load reduction (%) 70.30 -38.88 -54.58
NO
3
EMCinflow (mg/L) 0.00 0.08 0.30
DIC
EMCinflow (mg/L) 27.33 5.74 2.98
EMCoutflow (mg/L) 0.00 0.06 0.33 EMCoutflow (mg/L) 7.91 8.51 3.29
Loadinflow (kg) 0.00 1.43 25.85 Loadinflow (kg) 33.89 98.83 258.27
Loadoutflow (kg) 0.00 1.10 28.95 Loadoutflow (kg) 9.81 146.51 285.27
Load reduction (kg) 0.00 0.33 -3.10 Load reduction (kg) 24.08 -47.68 -27.00
Load reduction (%) 0.00 23.08 -11.99 Load reduction (%) 71.05 -48.24 -10.45
DO
N
EMCinflow (mg/L) 0.53 0.21 0.34
TD
C
EMCinflow (mg/L) 38.75 10.31 6.92
EMCoutflow (mg/L) 0.16 0.30 0.43 EMCoutflow (mg/L) 11.30 14.85 9.37
Loadinflow (kg) 0.66 3.61 29.70 Loadinflow (kg) 48.04 177.37 599.45
Loadoutflow (kg) 0.20 5.23 37.14 Loadoutflow (kg) 14.01 255.57 811.87
Load reduction (kg) 0.46 -1.62 -7.44 Load reduction (kg) 34.03 -78.20 -212.42
Load reduction (%) 69.70 -44.88 -25.05 Load reduction (%) 70.84 -44.09 -35.44
Chapter 5
128
Chapter 5
129
Figure 5. 5 Stormwater hydrographs, pollutographs and nutrients dynamics in the Wharf
Street Constructed Wetland from 7 to 11 May 2014: (a) flow rate and TN concentrations at the
inlet, (b) flow rate and TP concentrations at the inlet, (c) flow rate and TN concentrations at
the outlet and (d) flow rate and TP concentrations at the outlet. Stormwater hydrographs,
pollutographs and nutrients dynamics in the Anvil Way Compensation Basin from 12 to 14
March 2015: (e) flow rate and TN concentrations at the inlet, (f) flow rate and TP
concentrations at the inlet, (g) flow rate and TN concentrations at the outlet and (h) flow rate
and TP concentrations at the outlet. Horizontal dashed lines in each box indicate EMC of
nutrients. Tables indicate the total water volume passed through the inlet and outlet, inlet and
outlet load, inlet and outlet EMC, load attenuation and volume of stormwater treated during
the event.
Figure 5. 6 Temporal variability of soil porewater properties in the transect of the Anvil Way
Compensation Basin along with rainfall. (a) Changes in EC and temperature and (b) Eh along
with water level variations during and after rainfall.
Soil porewater samples were analysed to explore the variability in nutrient
concentrations during soil saturation (Figure 5. 7). Nutrient concentrations varied with
sampling depth. Interestingly, P3 showed higher concentrations of NH3, NOx, DIN,
DON, FRP, FTP and DOP compared to other two points. Average DOC and DIC
concentrations were maximum at P1 during AE1, and concentrations slightly
decreased in the following events. Soil porewater nutrient concentrations were
compared with surface water nutrient concentrations. NH3, DON and DIN
Chapter 5
130
concentrations were 2–3 times higher in the porewater than that observed in the inlet.
NOx, FRP and DOP concentrations were lower in the porewater compared to that of
inlet water, while other parameters i.e., NH3, DIN, DON, DIC and DOC showed
higher porewater concentrations compared to that measured at the inlet.
Figure 5. 7 Nutrient concentrations in porewater samples of the Anvil Way Compensation
Basin during events AE1 and AE2.
5.3.4 DO dynamics during events
High resolution DO data was monitored at 10-min intervals from July 2014 to
June 2015 in the SF1 compartment of the WSCW and the central channel of the
AWCB. DO showed a distinct diurnal signal, with night-time anoxia and day-time
peaks. Fourteen storm events were further assessed to investigate DO dynamics within
the CWs during pre- and post-storm events. Meteorological dry periods (November to
May) showed DO concentrations < 2 mg/L or saturation levels about 5% during the
pre-event in the AWCB at the day-time. However, the pre-event during the wet season
(June to October) exhibited DO concentrations above 6 mg/L. The WSCW showed
slightly higher DO concentrations and saturation levels than that measured in the
AWCB in the pre- and post-event during both dry and wet seasons. The maximum DO
levels were observed between 12:00 to 15:00 h on a diurnal scale during the pre-event.
Chapter 5
131
Metabolism was estimated over four consecutive days during the pre-and post-event
periods to indicate the function of the CWs at the event scale. Metabolism due to
storm perturbation increased after some storm events in both CWs (Figure 5. 8),
however also showed no change in other storm events; it is unclear the reason for the
difference and this needs further exploration. The post-event metabolism was always
higher in the WSCW compared to that of the AWCB. Metabolism under the pre-event
in the AWCB was correlated with increased solar exposure (R2 = 0.67), water
temperature (R2 = 0.47) and ADD (R
2 = 0.55). After the events, the observed
correlations disappeared as changes in CW volume, flow rate and HRT re-set the
AWCB (Figure 5. 9). A similar pattern was observed in the SF1 compartment of the
WSCW.
The SF1 compartment of the WSCW showed a total of 5 and 6 events with
higher metabolism during the pre- and post-event, respectively, while the AWCB
showed 5 and 3 events with higher metabolism during the pre- and post-event,
respectively. Therefore, positive metabolism was less dominant in the AWCB
compared to the WSCW. However, the SF compartments of the WSCW exhibited
almost double the metabolism even during the pre- and post-event, compared to that
measured in the meandering channel of the AWCB. High ΔDO was observed in the
CWs during the pre-event after prolonged ADD e.g., January 2015 in both sites. DO
concentrations in the water increased after that event, however the water remained
100% saturated for several days and minimal DO fluctuations occurred throughout the
post-event periods.
5.4 Discussion
5.4.1 Hydro-climate and catchment character determine nutrient export
Climatological conditions including rainfall and the number of ADD, and
catchment characteristics including area and land use determine nutrient export to the
CWs. The first few hours of the incoming stormwater usually wash away any loosely
bound soil particles from the catchment and cause higher nutrient loads and turbidity.
Nutrient exports were also higher, when the event was preceded by either long ADD
or when the rainfall intensity was high (Table 5. 2). Storm events with lower rainfall
amount and shorter ADD can deliver limited nutrient loads compared to loads that
were delivered in previous events, and therefore variation in intra-storm nutrient EMC
and loads were observed in the studied CWs (Figure 5. 10). Among three events
Chapter 5
132
recorded at the AWCB, AE2 consisted lower concentrations of TN and TP. AE2
experienced with single ADD and therefore, nutrients likely to be washed-off from the
catchment prior to the event. ADD also influenced nutrient accumulation in other
catchments, and large rain events were found to have higher erosive capacity (Inamdar
et al. 2015; Gunaratne et al. 2016).
Figure 5. 8 Temporal variability of metabolism in the (a) SF1 compartment of the Wharf
Street Constructed Wetland and (b) the Anvil Way Compensation Basin during four
consecutive days of pre-storm events (∆𝐃𝐎̅̅ ̅̅ ̅̅𝐩𝐫𝐞) and post-storm events (∆DO̅̅ ̅̅ ̅̅
post). Solid
circles and lines in individual boxes indicate mean and median values, respectively. Error bars
of the box plots indicate standard error. Green and yellow panels indicate that significant
positive and negative metabolism, respectively took place during the pre-event conditions
compared to post-event conditions.
Chapter 5
133
Figure 5. 9 Principal component analysis of different parameters (average over four
consecutive days) during the (a) pre-event and (b) post-event in the Anvil Way Compensation
Basin. Dis: Discharge and SR: Solar radiation exposure.
Hydrological variability also influences nutrient dynamics. Stormwater flows
have the capacity to flush turbid water from the CW (Tuttle et al. 2008) or settle
within the CW, and therefore less turbid water was observed in the CWs at 2 to 3 h
after the cessation of the storm events, when the flow rate was about 0.02 m3/s at the
outflow. The CWs received ungauged water making up to 70% of storm flows; the
ungauged contribution to the WSCW was less than that of AWCB. More importantly,
the AWCB received anoxic groundwater that can lower DO levels in the CW, and
impact the overall nutrient load estimation. The ungauged water also impacted the
overall performance of the system under different flow conditions. For example, when
the contribution of ungauged water was considered, the estimated performance of the
WSCW and AWCB was about 40% higher than the performance when ungauged
ΔDO SR Tem
ADD
Dis
CW volume
HRT
-1
-0.5
0
0.5
1
-1 -0.5 0 0.5 1F
2 (
19.5
6 %
)
F1 (57.60 %)
axes (F1 and F2: 77.16 %) a)
ΔDO
SR
Tem
Rainfall
Dis
CW volume
HRT
-1
-0.5
0
0.5
1
-1 -0.5 0 0.5 1
F2 (
26.6
4 %
)
F1 (48.80 %)
axes (F1 and F2: 75.44 %) b)
Chapter 5
134
water wasn't considered during base flows (Adyel et al. 2016; Adyel et al. 2017a).
Moreover, stormwater during the events saturates or floods the riparian zones, and
shifts the CWs from a lentic to lotic stage. Such saturation after prolonged dry periods
released nutrients to the system (Table 5. 3).
Catchment properties including soil types, land use pattern, residential or
industrial setting and imperviousness influence nutrient exports to the CWs during
storm events. The area of the MSMD is larger than that of the WSMD. The MSMD
catchment is a combination of residential and semi-industrial area with high traffic
land use, while the WSMD is predominantly a residential area with high traffic.
Therefore, it is expected the MSMD generate higher runoff than that of the WSMD
(Table 5. 2). However, the WSMD delivered higher nutrient loads and showed a
higher SMC, possibly arising from residential gardening lots or activities within the
nearby residential areas. However, the WSMD delivered higher nutrient loads and
showed a higher SMC, possibly arising from residential gardening lots or activities
within the nearby residential areas. Further study is necessary to indicate the exact
contribution of each garden lot or significance of lot size on nutrient delivery after a
storm event to support targeted management.
Figure 5. 10 Intra-storm variability flow and EMC of TN and TP in the (a) Wharf Street
Constructed Wetland and (b) Anvil Way Compensation Basin.
Chapter 5
135
5.4.2 CWs design and hydrological variability control nutrient attenuation
CW design, effective treatment areas, flow patterns, conditions of
soils/sediments, and the presence of filter media and macrophytes can all be important
drivers of nutrient attenuation. The total available effective area of nutrient attenuation
was higher in the WSCW compare to the AWCB. The WSCW was constructed and
retrofitted more recently than that of the AWCB, and the hybrid nature of the WSCW
provided conditions more suitable for nutrient attenuation. A retrofitted stream-
constructed wetland showed reduced suspended solids and N export to downstream
waters (Filoso & Palmer 2011; Filoso et al. 2015). Moreover, the WSCW contained
alternating SF and SSF compartments that provided aerobic and anaerobic conditions,
respectively for nutrient processing. Similar conditions have previously been observed
in other CWs (Headley et al. 2005; Konnerup et al. 2009; Headley & Tanner 2012) to
support nutrient attenuation. In the WSCW, laterite-based SSF compartments
provided additional roughness for the vertical and horizontal mixing of water and
increased HRT to promote settling of particulate nutrients. Furthermore, laterite is
known to adsorb FRP via ligand exchange reactions, where phosphate displaces water
or hydroxyls from the surface of iron (Fe) and aluminium (Al) hydrous oxides (Wood
& McAtamney 1996). Therefore, the WSCW showed higher P attenuation.
Hydrological variability shaped nutrient dynamics in the CWs during the
events. Decreased flow rate in the CWs usually increased water travel time, and this
increased water travel time correlated with increased attenuation of TN (R2 = 0.45)
and TP (R2 = 0.83) in the WSCW, and attenuation of TN (R
2 = 0.51) and TP
(R2 = 0.96) in the AWCB (Figure 5. 10a) during the events. High flow and low HRT
during events saturated the riparian zones of both CWs and converted the system from
lentic to lotic phase. Subsequently, the riparian zones released nutrients, and therefore,
the CW acted as a net producer of nutrients. Although the WSCW released some
nutrients during riparian zone saturation/flooding, the released nutrients were
attenuated in the subsequent SSF compartments. As there is no alternating
compartment in the AWCB, the released nutrients had very limited opportunity to be
attenuated before downstream release (Table 5. 3). Moreover, high flows result in
changing inundation areas with new areas becoming incorporated into the CW. Hence,
sediment nutrient fluxes or suspension can take place during events that deliver some
nutrients back into the system. In addition, stormwater from MSD travelled very
quickly within the AWCB, and get minimal chance to be attenuated. Therefore, outlet
Chapter 5
136
concentrations of the AWCB indicated nutrient release from the system, which is
similar to other studies that also have shown that flooding of CWs can release
nutrients (Dupas et al. 2015). Careful investigation of the relationship between flow
and nutrient attenuation is further needed to determine how travel time and reaction or
exposure time interact to control the overall attenuation.
Event flow and saturation may trigger rapid changes of Eh within the riparian
zone and thus initiate rapid nutrient attenuation or release to the water column.
Seasonal groundwater level fluctuations and seepage have previously been shown to
modify Eh in riparian soils and to release already adsorbed nutrients (Hoffmann et al.
2009). Nutrient dynamics in the inflowing or outgoing water were compared within
the riparian porewater of the AWCB. NOx, FRP and DOP concentrations were lower
in the porewater compared to the inflowing stormwater. Biogeochemical processing
can reduce NOx concentrations in the riparian soil via denitrification. Bioavailable
porewater C can be utilized for NOx processing, and therefore NOx concentrations in
the porewater samples were less relative to the inflowing water. However, P releases
were observed from the same soils during large events (AE2 and AE3).
Soils/sediments usually release P in CWs through (a) diffusion due to P concentration
gradient between soils/sediments and the overlying water column and (b) advection
due to microbial activities, bioturbation and vertical movement of porewater (Reddy et
al. 1999). Soil saturation after prolonged dry periods can also cause osmotic shock
and microbial cell lysis that mobilizes and releases P in the CWs (Turner & Haygarth
2001; Dupas et al. 2015). Soil saturation and associated anoxic conditions can also
generate reductive dissolution of Fe(hydr)oxides that release P to the porewater
(Dupas et al. 2015). Ultimately, solubilisation-mobilisation-transport is the likely
route of P processing or release after the saturation of riparian soils in the AWCB.
DOC and DIC concentrations were 3–4 times higher in the porewater than that
at the inlet of the AWCB. Accumulation of organic matter from the decay of plant
biomass, whether from locally planted macrophytes or from material deposited during
seasonal flood events, is a possible source of organic nutrients in the riparian transect.
Sediment re-suspension and transport during events can also deliver the particulate C
accumulated across the riparian zone back to the main basin of the AWCB. DOC
entering the AWCB can be attenuated as it travelled through the wetland, however it
can also be replaced by new DOC produced within the system, ultimately resulting in
a net export. However, the SSF compartments of the WSCW contained no or limited
Chapter 5
137
dead macrophytes and sediments, and provided roughness of incoming stormwater
ultimately showed higher C attenuation.
Macrophytes, both living and senescent, are important component of CWs.
Both CWs contained a variety of macrophytes (Table 5. 1)
where B. articulata and S. validus were the most common species. Macrophytes are
natural purifiers of CWs that increase hydraulic roughness, promote uniform flow,
enhance sedimentation of particles, provide surface area for small-particle adhesion,
and protect sediments from erosion during the events. Nutrient uptake or release by
macrophytes during the storm events was not measured directly in this study and
needs further investigation.
Figure 5. 11 (a) Nutrient loads attenuation as a function of travel time of water flow in the
Wharf Street Constructed Wetland and Anvil Way Compensation Basin. (b) Relationship
between nutrient load attenuation and changes in metabolism due to storm perturbation in the
AWCB.
Chapter 5
138
5.4.3 Diurnal DO and metabolism shape nutrient dynamics
Three events in the AWCB were assessed to provide insights into how DO
influence biogeochemical processing of nutrients in diurnal scale. DO concentrations
were usually low during base flow or lentic stage. However, DO concentrations
increased during high flow or episodic pulses that shift CW to lotic stage. Frequent
rainfall also increased DO concentrations due to aeration, and flushing of anoxic water
and other oxygen-consuming substances (Tuttle et al. 2008; Reeder 2011).
Interestingly, nutrients at the outlet exhibited a diurnal trend responding to DO
concentrations and available sunlight. For example, NH3 was attenuated when ample
solar radiation was available, and therefore, nitrification was likely significant.
NOX attenuation was dominant at night and under low DO levels. Therefore, at the
event scale when the system became autotrophic, NH3 attenuation was dominant,
while heterotrophic state supported NOx attenuation via denitrification. FRP
attenuation was also dominant during day-time probably through autotrophic process
under high DO concentrations and solar radiation. The process declined at night and
additional FRP was released from the system most likely from sediments experiencing
low oxygen. The CW was net autotrophic when diurnal DO concentrations increased
from morning to afternoon, indicating that average daily photosynthesis exceeded
respiration.
The use of wetland metabolism as a proxy indicator of CW function was also
considered. Metabolism estimation in stormwater treating CWs is complex as these
systems are dynamic from both hydrological and biogeochemical perspectives. DO
based metabolism was investigated in CWs (Cronk & Mitsch 1994b; Tuttle et al.
2008; Reeder 2011) over short timescales i.e., several days to a week using the lake
metabolism concept (Staehr et al. 2010). Metabolism of a lentic CW during the dry
summer in a Mediterranean climate can be defined using the lake metabolism concept.
However, storm events or frequent rainfall can shift the CW from a lentic to a lotic
system where the lake metabolism concept is not applicable. Therefore, metabolism
was estimated over four consecutive days before and after storm perturbation during
lentic condition. Water temperature, solar exposure, ADD, water volume and water
level influenced DO and daily DO level fluctuation, and ultimately the wetland
metabolism during the events in the AWCB. We attempted to link metabolism due to
the storm perturbation with nutrient attenuation in the AWCB. Positive metabolism
due to perturbation occurred when the post-event metabolism was higher than the pre-
Chapter 5
139
event metabolism, and negative metabolism occurred when the pre-event metabolism
was higher than the post-event metabolism. It was assumed that nutrients can be
attenuated or released in the CWs due to autotrophic or heterotrophic processes. These
processes use DO and provide a metabolic signature. The positive metabolism showed
a linear link with attenuation of nutrients including NH3 (R2 = 0.84), NOx (R
2 = 0.9),
TN (R2 = 0.76), FRP (R
2 = 0.98), TP (R
2 = 0.75), DIC (R
2 = 0.67) and DOC
(R2 = 0.96) (Figure 5. 11b). Autotrophic microbial processes likely consumed or
converted nutrients to simpler compounds and positive metabolism took place.
Therefore, data supported the concept that microbial processing of nutrients or CW
function could be assessed using a proxy indicator of metabolism. The six storm
events in the WSCW could not be used for this analysis due to the absence of diurnal
DO data. Future research should investigate metabolism in different compartments of
hybrid CW systems during storm events. The time taken by CWs to respond to a
storm event and to return to its pre-storm conditions was assessed across fourteen
events (Figure 5. 8). The SF of the AWCB could maintain its pre-event metabolism
i.e., it was more resistant to the storm-perturbation. The SF AWCB developed low DO
levels compared to the WSCW, due to presence of sludgy sediment, senescent
macrophytes and anoxic groundwater inputs. All of these can contribute to low
metabolism and make the CW more resistant to DO fluctuations during events. Due to
macrophyte senescence and organic rich nutrients in the inflows of the CWs, oxygen
consuming process can be intensified throughout the year and therefore induced
higher daily DO fluctuations. In another study, disturbances to diurnal DO changes
were associated with storm pulses, and the recovery times for stream metabolism were
found as two categories: < 5 days (30% of the disturbances) or > 15 days (70% of the
disturbances) (O'Connor et al. 2012). Large flow pulses can cause full-bed
mobilization that disrupts stream metabolism by destroying periphyton habitats with
long recovery times. Estimation of metabolism resistance and recovery, and their
drivers in stormwater treatment CWs need to be further investigated under different
rainfall and flow patterns to better predict CW performance.
5.5 Conclusions
This work investigated stormwater nutrient attenuation during episodic storm
events in two urban CWs: a hybrid CW with alternating SF and laterite-based SSF
compartments and a meandering SF CW. Major outcome of this research included:
Chapter 5
140
1) The research indicated that rainfall amount, ADD and catchment
characteristics influenced nutrient wash-off from the catchment after the storm events
and its subsequent delivery to the CWs. The research provided important empirical
evidence showing how the CWs changed over time between anoxic-lentic and oxic-
lotic conditions.
2) The research demonstrated the variability of nutrient attenuation across two
CWs during the storm events. The multi-compartment or hybrid WSCW was more
effective than that of single stage AWCB due to having alternating compartments, a
recirculation system, juvenile macrophytes and the absence of anoxic groundwater as
in the former CW. The study can therefore provide a benchmark for future CW
design.
3) The research indicated both hydrological and biogeochemical processes
were important for nutrient attenuation and transformation not only during events, but
also during the days before and after the storm events. Storm events after prolonged
dry conditions caused riparian zone saturation and subsequently released of P and C
based nutrients. However, laterite-based SSF compartments of the WSCW acted as an
effective P sink.
4) The study showed that CWs can transition between oxic and anoxic
conditions within a diurnal cycle. The study also used high frequency DO data to
explore CW metabolism due to storm perturbation and to indicate its effectiveness as
a proxy indicator of CW function. Positive metabolism was linked with nutrient
attenuation, particularly in the AWCB during the storm events. Water temperature,
solar exposure and ADD influenced such metabolism before any storm event. This
metabolism approach should be explored in other CWs for the prediction of nutrient
attenuation, using easily available and inexpensive DO sensors that can provide real-
time data. This study also indicated that AWCB showed resistance to the metabolism
recovery after the storm perturbations.
Chapter 6
Conclusions
Chapter 6
142
This research investigated stormwater nutrient attenuation in two constructed
wetlands (CWs) on the Swan Coastal Plain in Western Australia at diurnal to decadal
time-scale. Studied CWs were varied in type, flow pattern, surface area and catchment
characteristics. The first site was the Anvil Way Compensation Basin (AWCB), a
meandering surface flow (SF) system that was built in 2004 and retrofitted in 2010.
The second site was the Wharf Street Constructed Wetland (WSCW), a hybrid CW
with multiple alternating SF and laterite-based subsurface flow (SSF) compartments.
Although the WSCW was initially constructed in 2010, a retrofication effort was
conducted in 2012 when laterite-based SSF compartments were incorporated. Both
CWs experience a Mediterranean climate including the prolonged dry summer period
with low or sometimes near zero flows and nutrient-rich episodic storm pulses during
the wet winter. These CWs were further complicated by the summer macrophyte
senescence. This thesis explored the impact of different hydro-climatological and
subsequent biogeochemical variabilities on the nutrient attenuation in two CWs.
This research is amongst the first to investigate stormwater nutrient attenuation
at the same sites across a range of scales, from diurnal to long term (up to 12 years)
time-scale. It is also a rare example of an investigation of CW performance under a
Mediterranean climate with variable i.e., near zero and pulse flow conditions (Chapter
3 to 5). The study also assesses the effectiveness of CW restoration or retrofication on
the nutrient attenuation (Chapter 3) and demonstrates the impact of alternating SF and
SSF compartments on the nutrient attenuation in a hybrid CW (Chapter 4 and 5). The
research used a novel index for estimating nutrient attenuation when ungauged water
sources contribute significantly to the CWs (Chapter 3 to 4). The research also used
wetland metabolism by exploring dissolved oxygen (DO) dynamics as a proxy for the
CW functions. This metabolism index was linked to the nutrient attenuation during the
base flows (Chapter 4) and the episodic storm pulses (Chapter 5).
Antecedent conditions shaped the nutrient delivery from catchment to the CW.
High nutrient load export took place during the event accompanied with higher
antecedent dry days and rainfall amount. The relative extent of nutrient attenuation
varied across spatial and temporal scales. The multi-compartment or hybrid WSCW
showed a better nutrient attenuation per unit area than that of the single stage SF
AWCW by providing alternating multiple compartments, laterite substrate, negligible
groundwater connectivity, water recirculation option, and extending hydraulic
retention time (HRT). Nutrient attenuation was impacted by variation in design, the
Chapter 6
143
effective treatment area, the presence of macrophytes and their age, the available filter
media and its roughness, the availability of alternating aerobic or anaerobic
conditions, and the exposure of wetland compartments to light and air. CW performed
very well (up to 95% efficiency) during first couple of years of operation, however the
perforce reduced in long term and system acted a nutrient source. Nitrogen-based
species attenuation was dominant in the SF compartments of the hybrid CW, while
phosphorus-based species attenuated initially in the SF compartments and then further
polished in the laterite-based SSF compartments (Adyel et al. 2017a).
The CWs experienced strong seasonal hydrology. During the summer
condition, groundwater probably the main source of water in the AWCB, and
recirculation and groundwater top up at the WSCW. Hydrological variabilities i.e.,
flow and HRT shape nutrient attenuation more intensely in hybrid CW than that of SF
CW. Hybrid CW increased nutrient attenuation by extending HRT both during the
base and event flow. Although the SF CW was sensitive to HRT during the event
flows, there was high variability in attenuation during the low flow and HRT
conditions. At diurnal scale CW respond to available solar radiation and DO levels.
For example, ammonia (NH3) was attenuated when ample solar radiation and DO
were available. Filtered total oxidized nitrogen (NOX) attenuation was dominant,
particularly at night and under low DO levels. Therefore, at the event scale when the
system became autotrophic, NH3 attenuation was dominant, while heterotrophic state
supported NOx attenuation. Therefore, care should take during sampling collection
plan to assess CW performance in respect of particular nutrient species. Nutrient
attenuation also varied across the events depending on antecedent condition, total
rainfall, inflowing water and catchment area. Therefore intra storm variability of
nutrient loads and subsequent load attenuation was observed in CW.
The CW restoration had a significant influence on the nutrient attenuation.
Retrofitted six-year old SF CW showed increased the nutrient attenuation, except for
NOx. Replacement recycled concrete material with laterite increased nutrient
attenuation in the hybrid system through providing active sites for nutrient adsorption
and extending water travel time. Sediments and macrophytes stored about one-third
and one-fifth of the incoming nutrient loads, respectively. Nutrient accumulation was
higher during early to mature growth stages, particularly for the first two years of CW
operation, and below ground biomass i.e. roots and rhizospheres stored more nutrients
than above ground biomass i.e., shoots and leaves. However, summer macrophytes
Chapter 6
144
senescence and sediment flux increased internal loading of “new” material contribute
lower the attenuation capacity of the system. Saturation of riparian zone after
prolonged dry periods led to NOx attention and P release in SF system; however,
released P attenuated in subsequent alternating SSF compartment of the hybrid
system. The lowest NOx concentrations at the deepest point of the AWCB transect
indicating that denitrification was dominant in the porewater. Furthermore senescent
macrophytes and floating debris shaded the water column, prevented light penetration
and resulted in lower daily DO levels.
DO in the CWs showed intense diurnal trends, particularly in the dry summer,
with anoxia at night and maxima during the day-time. As macrophytes shading
reduced DO levels, it subsequently reduced daily metabolism of the CWs. The CW
was net autotrophic when diurnal DO concentrations increased from morning to
afternoon, indicating that average daily photosynthesis exceeded respiration. Diel
oxygen metabolism was correlated with water temperature, solar exposure and
antecedent dry days during the pre-storm conditions. Significant relationship between
overall nutrient attenuation and change metabolism during the storm perturbation in
the SF system suggested this metabolism could be a useful proxy indicator of CW
function (Adyel et al. 2017b).
Spatial heterogeneity of riparian soils in CWs can affect the water distribution
as well as the nutrient attenuation or release under Mediterranean climates. Further
investigation of nutrient dynamics in the riparian - hyporheic zone needs to be
considered. Moreover, the contribution of groundwater to the nutrient inputs during
the events requires more studies in the SF CW. Dimensionless numbers such as the
Damköhler number can be applied to investigate the relative timescales of reaction
and travel/exposure of nutrients in the CWs under different flow regimes. Further
investigation using hydrodynamic-biogeochemical models can parameterize the
nutrient attenuation rates. A careful design of CW needs to be considered, not only for
nutrient attenuation but also to integrate it within residential areas and downstream
sensitive waterways, providing a passive eco-recreation and educational asset
consolidated by safety, amenity and accessibility.
References
145
References
Adyel T.M., Hipsey M.R. & Oldham C.E. (2017a). Temporal dynamics of stormwater
nutrient attenuation of an urban constructed wetland experiencing summer low flows
and macrophyte senescence. Ecological Engineering, 102, 641-661.
Adyel T.M., Oldham C.E. & Hipsey M.R. (2017b). Storm event-scale nutrient attenuation in
constructed wetlands experienceing a mediterranean climate: A comparision of a
surface flow and hybrid surface-subsurface flow system. Science of the Total
Environment, 598, 1001-1014.
Adyel T.M., Oldham C.E. & Hipsey M.R. (2016). Stormwater nutrient attenuation in a
constructed wetland with alternating surface and subsurface flow pathways: Event to
annual dynamics. Water Research, 107, 66-82.
Adyel T.M., Conti A.L.R., Ocampo C., Coletti J.Z., Hipsey M.R. & Oldham C. (2015a).
Ecohydrology and stormwater nutrient attenuation performance of constructed
wetland in Western Australia. 36th IAHR World Congress, The Hague, the
Netherlands, pp. 1-11.
Adyel T.M., Ocampo C., Tareque H., Oldham C.E. & Hipsey M.R. (2015b). Performance
assessment of Wharf St Constructed Wetland 2009-2014, Cooperative Research
Centre for Water Sensitive Cities, Australia, pp. 1-74.
APHA (1998). Standard Methods for the Examination of Water and Wastewater, twentith ed.
American Public Health Association, American Water Works Association, Water
Environment Fedeartaion, Washington, DC, USA.
APHA (2012). Standard Methods for the Examination of Water and Wastewater, twenty-
second ed. American Public Health Association, American Water Works Association,
Water Environment Fedeartaion, Washington, DC, USA.
Aslam M.M., Malik M., Baig M.A., Qazi I.A. & Iqbal J. (2007). Treatment performances of
compost-based and gravel-based vertical flow wetlands operated identically for
refinery wastewater treatment in Pakistan. Ecological Engineering, 30, 34-42.
Ávila C., Garfí M. & García J. (2013). Three-stage hybrid constructed wetland system for
wastewater treatment and reuse in warm climate regions. Ecological Engineering, 61,
Part A, 43-49.
Ávila C., Matamoros V., Reyes-Contreras C., Piña B., Casado M., Mita L., Rivetti C., Barata
C., García J. & Bayona J.M. (2014). Attenuation of emerging organic contaminants in
a hybrid constructed wetland system under different hydraulic loading rates and their
associated toxicological effects in wastewater. Science of the Total Environment, 470–
471, 1272-1280.
Ayaz S.Ç., Aktaş Ö., Akça L. & Fındık N. (2015). Effluent quality and reuse potential of
domestic wastewater treated in a pilot-scale hybrid constructed wetland system.
Journal of Environmental Management, 156, 115-120.
Bach P.M., McCarthy D.T. & Deletic A. (2010). Redefining the stormwater first flush
phenomenon. Water Research, 44, 2487-2498.
Bachand P.A.M. & Horne A.J. (1999). Denitrification in constructed free-water surface
wetlands: I. Very high nitrate removal rates in a macrocosm study. Ecological
Engineering, 14, 9-15.
Baker D.W., Bledsoe B.P. & Price J.M. (2012). Stream nitrate uptake and transient storage
over a gradient of geomorphic complexity, north‐central Colorado, USA.
Hydrological Processes, 26, 3241-3252.
Bangs C. (2007). Urban stormwater treatment by experimental constructed wetlands. Senior
Design Project, Department of Engineering, PA, USA.
References
146
Barbosa A.E., Fernandes J.N. & David L.M. (2012). Key issues for sustainable urban
stormwater management. Water Research, 46, 6787-6798.
Bastviken S.K., Eriksson P., Premrov A. & Tonderski K. (2005). Potential denitrification in
wetland sediments with different plant species detritus. Ecological Engineering, 25,
183-190.
Bavor H.J. & Adcock P.W. (1994). Relating phosphorus distribution within wetlands to their
long term performance. 2nd Annual Conference, Soil and Water Mangement for
Urban Development, Creative Stormwater Management, Sydney, Australia
Bayley M.L., Davison L. & Headley T.R. (2003). Nitrogen removal from domestic effluent
using subsurface flow constructed wetlands: influence of depth, hydraulic residence
time and pre-nitrification. Water Science and Technology, 48, 175-182.
Bernhardt E.S. & Likens G.E. (2002). Dissolved organic carbon enrichment alters nitrogen
dynamics in a forest stream. Ecology, 83, 1689-1700.
Białowiec A., Janczukowicz W. & Randerson P.F. (2011). Nitrogen removal from wastewater
in vertical flow constructed wetlands containing LWA/gravel layers and reed
vegetation. Ecological Engineering, 37, 897-902.
Birch G.F., Matthai C., Fazeli M.S. & Suh J.Y. (2004). Efficiency of a constructed wetland in
removing contaminants from stormwater. Wetlands, 24, 459-466.
Bourgues S. & Hart B.T.H. (2007). Nitrogen removal capacity of wetlands: sediment versus
epiphytic biofilms. Water Science and Technology, 55, 175-182.
Bridgham S.D., Johnston C.A., Schubauer-Berigan J.P. & Weishampel P. (2001). Phosphorus
sorption dynamics in soils and coupling with surface and pore water in riverine
wetlands. Soil Science Society of America Journal, 65, 577-588.
Brix H. (1994). Functions of macrophytes in constructed wetlands. Water Science and
Technology, 29, 71-78.
Brix H. (1997). Do macrophytes play a role in constructed treatment wetlands? Water Science
and Technology, 35, 11-17.
Brodrick S.J., Cullen P. & Maher W. (1988). Denitrification in a natural wetland receiving
secondary treated effluent. Water Research, 22, 431-439.
Brown K. (2000). Urban stream restoration practices: an initial assessment. The Center for
Watershed Protection. Elliot City, MD, USA.
Brown R.R., Keath N. & Wong T.H.F. (2009b). Urban water management in cities: historical,
current and future regimes. Water Science and Technology, 59, 847-855.
Bukaveckas P.A. (2007). Effects of channel restoration on water velocity, transient storage,
and nutrient uptake in a channelized stream. Environmental Science & Technology,
41, 1570-1576.
Bunch A.J., Allen M.S. & Gwinn D.C. (2010). Spatial and temporal hypoxia dynamics in
dense emergent macrophytes in a Florida lake. Wetlands, 30, 429-435.
Burns M.J., Fletcher T.D., Walsh C.J., Ladson A.R. & Hatt B.E. (2012). Hydrologic
shortcomings of conventional urban stormwater management and opportunities for
reform. Landscape and Urban Planning, 105, 230-240.
Caraco N., Cole J., Findlay S. & Wigand C. (2006). Vascular plants as engineers of oxygen in
aquatic systems. BioScience, 56, 219-225.
Carleton J.N., Grizzard T.J., Godrej A.N. & Post H.E. (2001). Factors affecting the
performance of stormwater treatment wetlands. Water Research, 35, 1552-1562.
Carleton J.N., Grizzard T.J., Godrej A.N., Post H.E., Lampe L. & Kenel P.P. (2000).
Performance of a constructed wetlands in treating urban stormwater runoff. Water
Environment Research, 295-304.
References
147
Carpenter S.R. & Lodge D.M. (1986). Effects of submersed macrophytes on ecosystem
processes. Aquatic Botany, 26, 341-370.
Cerezo R.G., Suárez M.L. & Vidal-Abarca M.R. (2001). The performance of a multi-stage
system of constructed wetlands for urban wastewater treatment in a semiarid region of
SE Spain. Ecological Engineering, 16, 501-517.
Chang N.-B., Xuan Z., Marimon Z., Islam K. & Wanielista M.P. (2013). Exploring
hydrobiogeochemical processes of floating treatment wetlands in a subtropical
stormwater wet detention pond. Ecological Engineering, 54, 66-76.
Cheesman A.W., Turner B.L., Inglett P.W. & Reddy K.R. (2010). Phosphorus transformations
during decomposition of wetland macrophytes. Environmental Science & Technology,
44, 9265-9271.
Chen N., Wu J. & Hong H. (2012). Effect of storm events on riverine nitrogen dynamics in a
subtropical watershed, southeastern China. Science of the Total Environment, 431,
357-365.
Chimney M.J., Wenkert L. & Pietro K.C. (2006). Patterns of vertical stratification in a
subtropical constructed wetland in south Florida (USA). Ecological Engineering, 27,
322-330.
Chow V.T., Maidment D.R. & Mays L.W. (1999). Applied Hydrology. McGraw-Hill
Education.
Chua L.H.C., Tan S.B.K., Sim C.H. & Goyal M.K. (2012). Treatment of baseflow from an
urban catchment by a floating wetland system. Ecological Engineering, 49, 170-180.
CIRIA (2000). Sustainable Urban Drainage Systems - Design Manual for Scotland and
Northern Ireland. In: Report No. C521. CIRIA, Dundee, Scotland.
COAG (2004). Intergovernmental Agreement on a National Water Initiative. Commonwealth
of Australia and the Governments of New South Wales, Victoria, Queensland, South
Australia, the Australian Capital Territory and the Northern Territory, Australia.
Cohen M.J. & Brown M.T. (2007). A model examining hierarchical wetland networks for
watershed stormwater management. Ecological Modelling, 201, 179-193.
Coletti J.Z., Hinz C., Vogwill R. & Hipsey M.R. (2013). Hydrological controls on carbon
metabolism in wetlands. Ecological Modelling, 249, 3-18.
Connolly R., Zhao Y., Sun G. & Allen S. (2004). Removal of ammoniacal-nitrogen from an
artificial landfill leachate in downflow reed beds. Process Biochemistry, 39, 1971-
1976.
Craft C. (1996). Dynamics of nitrogen and phosphorus retention during wetland ecosystem
succession. Wetlands Ecology and Management, 4, 177-187.
Craig L.S., Palmer M.A., Richardson D.C., Filoso S., Bernhardt E.S., Bledsoe B.P., Doyle
M.W., Groffman P.M., Hassett B.A. & Kaushal S.S. (2008). Stream restoration
strategies for reducing river nitrogen loads. Frontiers in Ecology and the
Environment, 6, 529-538.
Cronk J.K. & Mitsch W.J. (1994a). Aquatic metabolism in four newly constructed freshwater
wetlands with different hydrologic inputs. Ecological Engineering, 3, 449-468.
Cronk J.K. & Mitsch W.J. (1994b). Periphyton productivity on artificial and natural surfaces
in constructed freshwater wetlands under different hydrologic regimes. Aquatic
Botany, 48, 325-341.
Cuffney T.F., Brightbill R.A., May J.T. & Waite I.R. (2010). Responses of benthic
macroinvertebrates to environmental changes associated with urbanization in nine
metropolitan areas. Ecological Applications, 20, 1384-1401.
References
148
Cui L., Ouyang Y., Lou Q., Yang F., Chen Y., Zhu W. & Luo S. (2010). Removal of nutrients
from wastewater with Canna indica L. under different vertical-flow constructed
wetland conditions. Ecological Engineering, 36, 1083-1088.
De Groot R. (2006). Function-analysis and valuation as a tool to assess land use conflicts in
planning for sustainable, multi-functional landscapes. Landscape and Urban
Planning, 75, 175-186.
Deletic A. (1998). The first flush load of urban surface runoff. Water Research, 32, 2462-
2470.
Deletic A., Ghadouani A., Keller J. & Wong T. (2013). Revolutionising Urban Water
Management. Water Journal, 7, 62-70.
Dodds W.K. & Cole J.J. (2007). Expanding the concept of trophic state in aquatic ecosystems:
it’s not just the autotrophs. Aquatic Sciences-Research Across Boundaries, 69, 427-
439.
Dong Z. & Sun T. (2007). A potential new process for improving nitrogen removal in
constructed wetlands—promoting coexistence of partial-nitrification and
ANAMMOX. Ecological Engineering, 31, 69-78.
Dunne E.J. & Reddy K.R. (2005). Phosphorus biogeochemistry of wetlands in agricultural
watersheds. Nutrient management in agricultural watersheds: a wetland solution.
Wageningen, The Netherlands, Wageningen Academic Publishers, pp. 105-119.
Dupas R., Gruau G., Gu S., Humbert G., Jaffrézic A. & Gascuel-Odoux C. (2015).
Groundwater control of biogeochemical processes causing phosphorus release from
riparian wetlands. Water Research, 84, 307-314.
Endreny T. (2008). Naturalizing urban watershed hydrology to mitigate urban heat-island
effects. Hydrological Processes, 22, 461-463.
Faulkner S.P. & Richardson C.J. (1989). Physical and chemical characteristics of freshwater
wetland soils. Constructed Wetlands for Wastewater Treatment, 41-72.
Fernández J.B.G. & Novo F.G. (2007). High-intensity versus low-intensity restoration
alternatives of a tidal marsh in Guadalquivir estuary, SW Spain. Ecological
Engineering, 30, 112-121.
Filoso S. & Palmer M.A. (2011). Assessing stream restoration effectiveness at reducing
nitrogen export to downstream waters. Ecological applications, 21, 1989-2006.
Filoso S., Smith S.M.C., Williams M.R. & Palmer M.A. (2015). The efficacy of constructed
stream–wetland complexes at reducing the flux of suspended solids to Chesapeake
Bay. Environmental Science & Technology, 49, 8986-8994.
Fink D.F. & Mitsch W.J. (2007). Hydrology and nutrient biogeochemistry in a created river
diversion oxbow wetland. Ecological Engineering, 30, 93-102.
Fletcher T.D., Andrieu H. & Hamel P. (2013). Understanding, management and modelling of
urban hydrology and its consequences for receiving waters: A state of the art.
Advances in Water Resources, 51, 261-279.
Fletcher T.D., Shuster W., Hunt W.F., Ashley R., Butler D., Arthur S., Trowsdale S., Barraud
S., Semadeni-Davies A. & Bertrand-Krajewski J.-L. (2014). SUDS, LID, BMPs,
WSUD and more–The evolution and application of terminology surrounding urban
drainage. Urban Water Journal, 1-18.
Fonder N. & Headley T. (2011). Systematic classification, nomenclature and reporting for
constructed treatment wetlands. Water and Nutrient Management in Natural and
Constructed Wetlands. Springer, pp. 191-219.
Fonder N. & Headley T. (2013). The taxonomy of treatment wetlands: A proposed
classification and nomenclature system. Ecological Engineering, 51, 203-211.
References
149
Francey M., Fletcher T.D., Deletic A. & Duncan H. (2010). New insights into the quality of
urban storm water in south eastern Australia. Journal of Environmental Engineering,
136, 381-390.
Gabriele W., Welti N. & Hein T. (2013). Limitations of stream restoration for nitrogen
retention in agricultural headwater streams. Ecological Engineering, 60, 224-234.
Gächter R. & Meyer J.S. (1993). The role of microorganisms in mobilization and fixation of
phosphorus in sediments. Proceedings of the Third International Workshop on
Phosphorus in Sediments. Springer, pp. 103-121.
Garcia J., Rousseau D.P.L., Morato J., Lesage E.L.S., Matamoros V. & Bayona J.M. (2010).
Contaminant removal processes in subsurface-flow constructed wetlands: a review.
Critical Reviews in Environmental Science and Technology, 40, 561-661.
Gasperi J., Zgheib S., Cladière M., Rocher V., Moilleron R. & Chebbo G. (2011). Priority
pollutants in urban stormwater: Part 2–case of combined sewers. Water Research, 46,
6693–6703.
GHD (2008). Swan River Trust Report for Constructed Wetland Monitoring Plan General
Monitoring Methods and Procedures. GHD, pp. 1-49.
Gordon R.P., Lautz L.K. & Daniluk T.L. (2013). Spatial patterns of hyporheic exchange and
biogeochemical cycling around cross‐vane restoration structures: Implications for
stream restoration design. Water Resources Research, 49, 2040-2055.
Gray S., Kinross J., Read P. & Marland A. (2000). The nutrient assimilative capacity of maerl
as a substrate in constructed wetland systems for waste treatment. Water Research,
34, 2183-2190.
Greenway M. (2005). The role of constructed wetlands in secondary effluent treatment and
water reuse in subtropical and and Australia. Ecological Engineering, 25, 501-509.
Greenway M., Dale P. & Chapman H. (2003). An assessment of mosquito breeding and
control in 4 surface flow wetlands in tropical-sub-tropical Australia. Water Science &
Technology, 48, 249-256.
Groffman P.M., Dorsey A.M. & Mayer P.M. (2005). N processing within geomorphic
structures in urban streams. Journal of the North American Benthological Society, 24,
613-625.
Gu B., Chimney M.J., Newman J. & Nungesser M.K. (2006). Limnological characteristics of
a subtropical constructed wetland in south Florida (USA). Ecological Engineering,
27, 345-360.
Guiboileau A., Sormani R., Meyer C. & Masclaux-Daubresse C. (2010). Senescence and
death of plant organs: nutrient recycling and developmental regulation. Comptes
rendus biologies, 333, 382-391.
Gunaratne G.L., Vogwill R.I.J. & Hipsey M.R. (2016). Effect of seasonal flushing on nutrient
export characteristics of an urbanizing, remote, ungauged coastal catchment.
Hydrological Sciences Journal, 1-18.
Ham J., Yoon C.G., Kim H-J. & Kim H-C. (2010). Modeling the effects of constructed
wetland on nonpoint source pollution control and reservior water quality
improvemnet. Journal of Environmnetal Sciences, 22, 834-839.
Hamilton H., Nix P.G. & Sobolewski A. (1993). An overview of constructed wetlands as
alternatives to conventional waste treatment systems. Water Pollution Research
Journal of Canada, 28, 529-548.
Hancock P.J. (2002). Human impacts on the stream–groundwater exchange zone.
Environmental Management, 29, 763-781.
References
150
Harper H.H. (1988). Effects of stormwater management systems on groundwater quality.
FDEP Project# WM190. Florida Department of Environmental Regulation,
Tallahassee, FL, USA.
Harrison M.D., Groffman P.M., Mayer P.M. & Kaushal S.S. (2012). Microbial biomass and
activity in geomorphic features in forested and urban restored and degraded streams.
Ecological Engineering, 38, 1-10.
Hathaway J.M., Moore T.L., Burkholder J.M. & Hunt W.F. (2012). Temporal analysis of
stormwater control measure effluent based on windows of harmful algal bloom
(HAB) sensitivity: Are annual nutrient EMCs appropriate during HAB-sensitive
seasons? Ecological Engineering, 49, 41-47.
Hatt B.E., Deletic A. & Fletcher T.D. (2006). Integrated treatment and recycling of
stormwater: a review of Australian practice. Journal of Environmental Management,
79, 102-113.
Hatt B.E., Fletcher T.D. & Deletic A. (2007). Treatment performance of gravel filter media:
Implications for design and application of stormwater infiltration systems. Water
Research, 41, 2513-2524.
Hatt B.E., Fletcher T.D. & Deletic A. (2009). Hydrologic and pollutant removal performance
of stormwater biofiltration systems at the field scale. Journal of Hydrology, 365, 310-
321.
Hayashi M. & Rosenberry D.O. (2002). Effects of ground water exchange on the hydrology
and ecology of surface water. Groundwater, 40, 309-316.
Headley T.R., Herity E. & Davison L. (2005). Treatment at different depths and vertical
mixing within a 1-m deep horizontal subsurface-flow wetland. Ecological
Engineering, 25, 567-582.
Headley T.R., Huett D.O. & Davison L. (2001). The removal of nutrients from plant nursery
irrigation runoff in subsurface horizontal-flow wetlands. Water Science and
Technology, 44, 77-84.
Headley T.R. & Tanner C.C. (2012). Constructed Wetlands With Floating Emergent
Macrophytes: An Innovative Stormwater Treatment Technology. Critical Reviews in
Environmental Science and Technology, 42, 2261-2310.
Hester E.T., Doyle M.W. & Poole G.C. (2009). The influence of in‐stream structures on
summer water temperatures via induced hyporheic exchange. Limnology and
Oceanography, 54, 355-367.
Hinman C. (2005). Low Impact Development: Technical Guidance Manual for Puget Sound.
Washington State University, USA.
Hoffmann C.C., Kjaergaard C., Uusi-Kämppä J., Hansen H.C.B. & Kronvang B. (2009).
Phosphorus retention in riparian buffers: review of their efficiency. Journal of
Environmental Quality, 38, 1942-1955.
Huang J.C., Mitsch W.J. & Johnson D.L. (2011). Estimating biogeochemical and biotic
interactions between a stream channel and a created riparian wetland: A medium-
scale physical model. Ecological Engineering, 37, 1035-1049.
Imfeld G., Braeckevelt M., Kuschk P. & Richnow H.H. (2009). Monitoring and assessing
processes of organic chemicals removal in constructed wetlands. Chemosphere, 74,
349-362.
Inamdar S., Dhillon G., Singh S., Parr T. & Qin Z. (2015). Particulate nitrogen exports in
stream runoff exceed dissolved nitrogen forms during large tropical storms in a
temperate, headwater, forested watershed. Journal of Geophysical Research:
Biogeosciences, 120, 1548-1566.
References
151
Jahangir M.M.R., Richards K.G., Healy M.G., Gill L., Müller C., Johnston P. & Fenton O.
(2016). Carbon and nitrogen dynamics and greenhouse gas emissions in constructed
wetlands treating wastewater: a review. Hydrology and Earth System Sciences 20,
109-123.
Jianlong W. & Jing K. (2005). The characteristics of anaerobic ammonium oxidation
(ANAMMOX) by granular sludge from an EGSB reactor. Process Biochemistry, 40,
1973-1978.
Jianlong W. & Ning Y. (2004). Partial nitrification under limited dissolved oxygen conditions.
Process Biochemistry, 39, 1223-1229.
Johengen T.H. & LaRock P.A. (1993). Quantifying nutrient removal processes within a
constructed wetland designed to treat urban stormwater runoff. Ecological
Engineering, 2, 347-366.
Johnston C.A. (1991). Sediment and nutrient retention by freshwater wetlands: effects on
surface water quality. Critical Reviews in Environmental Science and Technology, 21,
491-565.
Kadlec R.H. (2000). The inadequacy of first-order treatment wetland models. Ecological
Engineering, 15, 105-119.
Kadlec R.H. & Knight R.L. (1996). Treatment Wetlnds. CRC Press, Boca Raton, FL, USA.
Kadlec R.H. & Wallace S.D. (2009). Treatment Wetlands. CRC Press, Boca Raton, FL, USA.
Kaenel B.R., Buehrer H. & Uehlinger U. (2000). Effects of aquatic plant management on
stream metabolism and oxygen balance in streams. Freshwater Biology, 45, 85-95.
Kaushal S.S., Groffman P.M., Band L.E., Shields C.A., Morgan R.P., Palmer M.A., Belt K.T.,
Swan C.M., Findlay S.E.G. & Fisher G.T. (2008). Interaction between urbanization
and climate variability amplifies watershed nitrate export in Maryland. Environmental
Science & Technology, 42, 5872-5878.
Kim L.-H., Kayhanian M., Zoh K.-D. & Stenstrom M.K. (2005). Modeling of highway
stormwater runoff. Science of the Total Environment, 348, 1-18.
Kim L.-H., Ko S.-O., Jeong S. & Yoon J. (2007). Characteristics of washed-off pollutants and
dynamic EMCs in parking lots and bridges during a storm. Science of the Total
Environment, 376, 178-184.
Kirk J.T.O. (1994). Light and photosynthesis in aquatic ecosystems. Cambridge university
Press.
Kivaisi A.K. (2001). The potential for constructed wetlands for wastewater treatment and
reuse in developing countries: a review. Ecological Engineering, 16, 545-560.
Ko C.-H., Chang F.-C., Lee T.-M., Chen P.-Y., Chen H.-H., Hsieh H.-L. & Guan C.-Y.
(2010). Impact of flood damage on pollutant removal efficiencies of a subtropical
urban constructed wetland. Science of the Total Environment, 408, 4328-4333.
Konnerup D., Koottatep T. & Brix H. (2009). Treatment of domestic wastewater in tropical,
subsurface flow constructed wetlands planted with Canna and Heliconia. Ecological
Engineering, 35, 248-257.
Kröger R., Holland M., Moore M. & Cooper C. (2007). Plant senescence: a mechanism for
nutrient release in temperate agricultural wetlands. Environmental Pollution, 146,
114-119.
Lee B.-H. & Scholz M. (2007). What is the role of Phragmites australis in experimental
constructed wetland filters treating urban runoff? Ecological Engineering, 29, 87-95.
Lee S. & Yigitcanlar T. (2010). Sustainable urban stormwater management: water sensitive
urban design perceptions, drivers and barriers. Rethinking Sustainable Development:
Urban Management, Engineering, and Design, IGI Global, Engineering Science
Reference, Hershey Pa, pp. 1-10.
References
152
Lewis W.M. & Bender M. (1961). Effect of a cover of duckweeds and the alga Pithophora
upon the dissolved oxygen and free carbon dioxide of small ponds. Ecology, 42, 602-
603.
Li Y.S., Robin P., Qiu J.P., Morand P., Luth, Landrain B. & Cluzeau D. (2009). Optimal
Transfer between Extensive Wastewater Treatment Systems Combining
Vermifiltration, Macrophyte Lagoons and Constructed Wetlands. Proceedings of the
International Symposium on Environmental Science and Technology, China, pp.
2464-2469.
Lloyd S.D. (2001). Water Sensitive Urban Design in The Australian Context. Synthesis of a
conference held 30-31 August 2000, Melbourne, Australia. Cooperative Research
Centre for Catchment Hydrology Melbourne, Australia, pp. 1-26.
Locatelli L., Mark O., Mikkelsen P.S., Arnbjerg-Nielsen K., Deletic A., Roldin M. & Binning
P.J. (2017). Hydrologic impact of urbanization with extensive stormwater infiltration.
Journal of Hydrology, 544, 524-537.
Lovett G.M., Cole J.J. & Pace M.L. (2006). Is net ecosystem production equal to ecosystem
carbon accumulation? Ecosystems, 9, 152-155.
Lund M.A., Lavery P.S. & Froend R.F. (2001). Removing filterable reactive phosphorus from
highly coloured stormwater using constructed wetlands. Water Science & Technology,
44, 85-92.
Ma Y., Liu A., Egodawatta P., McGree J. & Goonetilleke A. (2017a). Assessment and
management of human health risk from toxic metals and polycyclic aromatic
hydrocarbons in urban stormwater arising from anthropogenic activities and traffic
congestion. Science of the Total Environment, 579, 202-211.
Ma Y., Liu A., Egodawatta P., McGree J. & Goonetilleke A. (2017b). Quantitative assessment
of human health risk posed by polycyclic aromatic hydrocarbons in urban road dust.
Science of the Total Environment, 575, 895-904.
Malaviya P. & Singh A. (2012). Constructed Wetlands for Management of Urban Stormwater
Runoff. Critical Reviews in Environmental Science and Technology, 42, 2153-2214.
Managi S., Goonetilleke A. & Wilson C. (2016). Water resources: Embed stormwater use in
city planning. Nature, 532, 37-37.
Martín M., Gargallo S., Hernández-Crespo C. & Oliver N. (2013). Phosphorus and nitrogen
removal from tertiary treated urban wastewaters by a vertical flow constructed
wetland. Ecological Engineering, 61, 34-42.
Masi F. (2008). Enhanced denitrification by a hybrid HF-FWS constructed wetland in a large-
scale wastewater treatment plant. In Wastewater Treatment, Plant Dynamics and
Management in Constructed and Natural Wetlands. Springer, pp. 267-275.
MDT (2005). MUSIC User Manual. CRC for Catchment Hydrology, Australia.
MEA (2005). Millennium Ecosystem Assessment, Ecosystems and Human Well-being:
Wetlands and Water Synthesis. World Resource Institute Washington DC, pp. 1-80.
Meng P., Pei H., Hu W., Shao Y. & Li Z. (2014). How to increase microbial degradation in
constructed wetlands: influencing factors and improvement measures. Bioresource
Technology, 157, 316-326.
Merlin G., Sedilot C., Herbert N. & Grasselly D. ( 2002). Treatment of tomato greenhouses
drainage solutions by experimental constructed wetlands. 8th International
Conference on Wetland Systems for Water Pollution Control, pp. 526–533.
Miller W. & Boulton A.J. (2005). Managing and rehabilitating ecosystem processes in
regional urban streams in Australia. Hydrobiologia, 552, 121-133.
References
153
Mitchell V.G., Deletic A., Fletcher T.D., Hatt B.E. & McCarthy D.T. (2007). Achieving
multiple benefits from stormwater harvesting. Water Science and Technology, 55,
135-144.
Mitsch W.J. & Gosselink J.G. (2000). The value of wetlands: importance of scale and
landscape setting. Ecological Economics, 35, 25-33.
Mitsch W.J. & Gosselink J.G. (2007). Wetlands. John Wiley & Sons, New York, USA.
Mitsch W.J. & Jorgensen S.E. (2003). Ecological engineering: A field whose time has come.
Ecological Engineering, 20, 363-377.
Moore T.L.C. & Hunt W.F. (2012). Ecosystem service provision by stormwater wetlands and
ponds–A means for evaluation? Water Research, 46, 6811-6823.
Morris P. & Barker W. (1977). Oxygen transport rates through mats of Lemna minor and
Wolffia sp. and oxygen tension within and below the mat. Canadian Journal of
Botany, 55, 1926-1932.
Mustafa A., Scholz M., Harrington R. & Carroll P. (2009). Long-term performance of a
representative integrated constructed wetland treating farmyard runoff. Ecological
Engineering, 35, 779-790.
Nahlik A.M. & Mitsch W.J. (2006). Tropical treatment wetlands dominated by free-floating
macrophytes for water quality improvement in Costa Rica. Ecological Engineering,
28, 246-257.
O'Brien J.M., Hamilton S.K., Kinsman-Costello L.E., Lennon J.T. & Ostrom N.E. (2012).
Nitrogen transformations in a through‐flow wetland revealed using whole‐ecosystem
pulsed 15
N additions. Limnology and Oceanography, 57, 221-234.
O'Connor B.L., Harvey J.W. & McPhillips L.E. (2012). Thresholds of flow‐induced bed
disturbances and their effects on stream metabolism in an agricultural river. Water
Resources Research, 48.
Omari K., Revitt M., Shutes B. & Garelick H. (2003). Hydrocarbon removal in an
experimental gravel bed constructed wetland. Water Science & Technology, 48, 275-
281.
Page K. & Dalal R. (2011). Contribution of natural and drained wetland systems to carbon
stocks, CO2, N2O, and CH4 fluxes: an Australian perspective. Soil Research, 49, 377-
388.
Paul E.A. & Clark F.E. (1996). Ammonification and nitrification. Soil Microbiology and
Biochemistry, Academic Press, San Diego, pp. 182-183.
Payne E.G.I., Fletcher T.D., Russell D.G., Grace M.R., Cavagnaro T.R., Evrard V., Deletic
A., Hatt B.E. & Cook P.L.M. (2014). Temporary Storage or Permanent Removal? The
Division of Nitrogen between Biotic Assimilation and Denitrification in Stormwater
Biofiltration Systems. Plos One, 9(3): e90890.
Puigagut J., Caselles-Osorio A., Vaello N. & García J. (2008). Fractionation, biodegradability
and particle-size distribution of organic matter in horizontal subsurface-flow
constructed wetlands. Wastewater Treatment, Plant Dynamics and Management in
Constructed and Natural Wetlands, pp. 289-297.
Rai U.N., Tripathi R.D., Singh N.K., Upadhyay A.K., Dwivedi S., Shukla M.K., Mallick S.,
Singh S.N. & Nautiyal C.S. (2013). Constructed wetland as an ecotechnological tool
for pollution treatment for conservation of Ganga river. Bioresource Technology, 148,
535-541.
Raisin G.W., Mitchell D.S. & Croome R.L. (1997). The effectiveness of a small constructed
wetland in ameliorating diffuse nutrient loadsing from an australian rural catchment.
Ecological Engineering, 9, 19-35.
References
154
Rajabzadeh A.R., Legge R.L. & Weber K.P. (2015). Multiphysics modelling of flow
dynamics, biofilm development and wastewater treatment in a subsurface vertical
flow constructed wetland mesocosm. Ecological Engineering, 74, 107-116.
Ranalli A.J. & Macalady D.L. (2010). The importance of the riparian zone and in-stream
processes in nitrate attenuation in undisturbed and agricultural watersheds–A review
of the scientific literature. Journal of Hydrology, 389, 406-415.
Reddy K.R. & D'Angelo E.M. (1997). Biogeochemical indicators to evaluate pollutant
removal efficiency in constructed wetlands. Water Science and Technology, 35, 1-10.
Reddy K.R. & DeLaune R.D. (2008). Biogeochemistry of Wetlands: Science and
Applications, CRC Press.
Reddy K.R., Diaz O.A., Scinto L.J. & Agami M. (1995). Phosphorus dynamics in selected
wetlands and streams of the Lake Okeechobee Basin. Ecological Engineering, 5, 183-
207.
Reddy K.R., Kadlec R.H., Flaig E. & Gale P.M. (1999). Phosphorus retention in streams and
wetlands: a review. Critical Reviews in Environmental Science and Technology, 29,
83-146.
Reddy K.R., Patrick W.H. & Broadbent F.E. (1984). Nitrogen transformations and loss in
flooded soils and sediments. Critical Reviews in Environmental Science and
Technology, 13, 273-309.
Reeder B.C. (2011). Assessing constructed wetland functional success using diel changes in
dissolved oxygen, pH, and temperature in submerged, emergent, and open-water
habitats in the Beaver Creek Wetlands Complex, Kentucky (USA). Ecological
Engineering, 37, 1772-1778.
Reinelt L.E. & Horner R.R. (1995). Pollutant removal from stormwater runoff by palustrine
wetlands based on comprehensive budgets. Ecological Engineering, 4, 77-97.
Rhue R.D. & Harris W.G. (1999). Phosphorus sorption/desorption reactions in soils and
sediments. Phosphorus Biogeochemistry in Subtropical Ecosystems. CRC Press, pp.
187-207.
Richardson C.J. (1985). Mechanisms controlling phosphorus retention capacity in freshwater
wetlands. Science, 228, 1424-1426.
Rousseau D., Lesage E., Story A., Vanrolleghem P.A. & De Pauw N. (2008). Constructed
wetlands for water reclamation. Desalination, 218, 181-189.
Roy A.H., Wenger S.J., Fletcher T.D., Walsh C.J., Ladson A.R., Shuster W.D., Thurston
H.W. & Brown R.R. (2008). Impediments and solutions to sustainable, watershed-
scale urban stormwater management: lessons from Australia and the United States.
Environmental Management, 42, 344-359.
Ruibal-Conti A.L., Ocampo C., Adyel T.M., Hipsey M.R. & Oldham C.E. (2015).
Performance assessment of the Anvil Way Compensation Basin living stream: 2004 -
2013. Cooperative Research Centre for Water Sensitive Cities, Australia, pp. 1-83.
Rushton B., Miller C. & Hull C. (1995). Residence time as a pollutant removal mechanism in
stormwater detention ponds. 4th Biennial Stormwater Research Conference
Southwest Florida Water Management District, Brooksville, FL, USA, pp. 210–221.
Russell R.C. (1999). Constructed wetlands and mosquitoes: health hazards and management
options—an Australian perspective. Ecological Engineering, 12, 107-124.
Saeed T. & Sun G. (2012). A review on nitrogen and organics removal mechanisms in
subsurface flow constructed wetlands: Dependency on environmental parameters,
operating conditions and supporting media. Journal of Environmental Management,
112, 429-448.
References
155
Saeed T. & Sun G.Z. (2011). Kinetic modelling of nitrogen and organics removal in vertical
and horizontal flow wetlands. Water Research, 45, 3137-3152.
Savant N.K., De Datta S.K. & Brady N.C. (1982). Nitrogen Transformations in Wetland Rice
Soils. Advances in Agronomy. Academic Press, pp. 241-302.
Seitzinger S.P., Styles R.V., Boyer E.W., Alexander R.B., Billen G., Howarth R.W., Mayer B.
& Van Breemen N. (2002). Nitrogen retention in rivers: model development and
application to watersheds in the northeastern USA. Biogeochemistry, 57, 199-237.
Sim C.H., Yusoff M.K., Shutes B., Ho S.C. & Mansor M. (2008). Nutrient removal in a pilot
and full scale constructed wetland, Putrajaya city, Malaysia. Journal of
Environmental Management, 88, 307-317.
Staehr P.A., Bade D., Van de Bogert M.C., Koch G.R., Williamson C., Hanson P., Cole J.J. &
Kratz T. (2010). Lake metabolism and the diel oxygen technique: State of the science.
Limnology and Oceanography: Methods, 8, 628-644.
Staehr P.A., Testa J.M., Kemp W.M., Cole J.J., Sand-Jensen K. & Smith S.V. (2012a). The
metabolism of aquatic ecosystems: history, applications, and future challenges.
Aquatic Sciences, 74, 15-29.
Staehr P.A., Testa J.M., Kemp W.M., Cole J.J., Sand-Jensen K. & Smith S.V. (2012b). The
metabolism of aquatic ecosystems: history, applications, and future challenges.
Aquatic Sciences, 74, 15-29.
Stein O.R. & Paul B. (2005). Temperature, plants, and oxygen: How does season affect
constructed wetland performance? Journal of Environmental Science and Health, 40,
1331-1342.
Stephens S.L. (2005). Forest fire causes and extent on United States Forest Service lands.
International Journal of Wildland Fire, 14, 213-222.
Stevenson R.J. (1996). An Introduction to Algal Ecology in Freashwater Benthic Habitats.
Algal Ecology: Freshwater Benthic Ecosystem, Academic Press, New York, pp. 3-30.
Stewart W.D.P. (1973). Nitrogen fixation by photosynthetic microorganisms. Annual Reviews
in Microbiology, 27, 283-316.
Sun G., Zhao Y. & Allen S. (2005). Enhanced removal of organic matter and ammoniacal-
nitrogen in a column experiment of tidal flow constructed wetland system. Journal of
Biotechnology, 115, 189-197.
Synnix (2007). Wharf Street Constructed Wetlands Concept design. Synnix Environment PI,
pp. 1-70.
Tang J.Y., Aryal R., Deletic A., Gernjak W., Glenn E., McCarthy D. & Escher B.I. (2013).
Toxicity characterization of urban stormwater with bioanalytical tools. Water
Research, 47, 5594-5606.
Tapsuwan S., Ingram G., Burton M. & Brennan D. (2009). Capitalized amenity value of urban
wetlands: a hedonic property price approach to urban wetlands in Perth, Western
Australia. Australian Journal of Agricultural and Resource Economics, 53, 527-545.
Taylor S.L., Roberts S.C., Walsh C.J. & Hatt B.E. (2004). Catchment urbanisation and
increased benthic algal biomass in streams: linking mechanisms to management.
Freshwater Biology, 49, 835-851.
Terzakis S., Fountoulakis M.S., Georgaki I., Albantakis D., Sabathianakis I., Karathanasis
A.D., Kalogerakis N. & Manios T. (2008). Constructed wetlands treating highway
runoff in the central Mediterranean region. Chemosphere, 72, 141-149.
Teufl B., Weigelhofer G., Fuchsberger J. & Hein T. (2013). Effects of hydromorphology and
riparian vegetation on the sediment quality of agricultural low-order streams:
consequences for stream restoration. Environmental Science and Pollution Research,
20, 1781-1793.
References
156
Third K.A., Sliekers A.O., Kuenen J.G. & Jetten M.S.M. (2001). The CANON system
(completely autotrophic nitrogen-removal over nitrite) under ammonium limitation:
interaction and competition between three groups of bacteria. Systematic and Applied
Microbiology, 24, 588-596.
Turner B.L. & Haygarth P.M. (2001). Biogeochemistry: phosphorus solubilization in rewetted
soils. Nature, 411, 258-258.
Tuttle C.L., Zhang L. & Mitsch W.J. (2008). Aquatic metabolism as an indicator of the
ecological effects of hydrologic pulsing in flow-through wetlands. Ecological
Indicators, 8, 795-806.
US EPA (1993). Nitrogen Control, Office of Research and Development. EPA/625/R-93/010,
Washington, DC, USA.
Van de Moortel A.M.K., Rousseau D.P.L., Tack F.M.G. & De Pauw N. (2009). A
comparative study of surface and subsurface flow constructed wetlands for treatment
of combined sewer overflows: a greenhouse experiment. Ecological Engineering, 35,
175-183.
Van Dongen L., Jetten M.S., Van Loosdrecht M.C.M. & Van Loosdrecht M.C.M. (2001). The
Combined Sharon/Anammox Process: A sustainable method for N-removal from
sludge water, STOWA Report 2000-25, IWA Publishing, pp. 1-64.
Verhoeven J.T.A., Arheimer B., Yin C. & Hefting M.M. (2006). Regional and global
concerns over wetlands and water quality. Trends in Ecology & Evolution, 21, 96-
103.
Vymazal J. (2001). Constructed wetlands for wastewater treatment in the Czech Republic.
Water Science and Technology, 44, 369-374.
Vymazal J. (2007). Removal of nutrients in various types of constructed wetlands. Science of
the Total Environment, 380, 48-65.
Vymazal J. (2011). Plants used in constructed wetlands with horizontal subsurface flow: a
review. Hydrobiologia, 674, 133-156.
Vymazal J. (2013a). Emergent plants used in free water surface constructed wetlands: A
review. Ecological Engineering, 61, 582-592.
Vymazal J. (2013b). The use of hybrid constructed wetlands for wastewater treatment with
special attention to nitrogen removal: a review of a recent development. Water
Research, 47, 4795-4811.
Vymazal J. & Kröpfelová 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. & Kröpfelová L. (2015). Multistage hybrid constructed wetland for enhanced
removal of nitrogen. Ecological Engineering, 84, 202-208.
Walsh C.J., Fletcher T.D. & Burns M.J. (2012). Urban stormwater runoff: a new class of
environmental flow problem. Plos One, 7, e45814.
Walsh C.J., Fletcher T.D. & Ladson A.R. (2005a). Stream restoration in urban catchments
through redesigning stormwater systems: looking to the catchment to save the stream.
Journal of the North American Benthological Society, 24, 690-705.
Walsh C.J., Roy A.H., Feminella J.W., Cottingham P.D., Groffman P.M. & Morgan R.P.
(2005b). The urban stream syndrome: current knowledge and the search for a cure.
Journal of the North American Benthological Society, 24, 706-723.
Wang R., Baldy V., Périssol C. & Korboulewsky N. (2012). Influence of plants on microbial
activity in a vertical-downflow wetland system treating waste activated sludge with
high organic matter concentrations. Journal of Environmental Management, 95, S158-
S164.
References
157
Whelans C., Manshell H.G. & Thompson P. (1994). Planning and management guidelines for
water sensetive urban (residential) design. Department of Planning and Urban
Developement of Western Australia.
White S.A. & Cousins M.M (2013). Floating treatmnet wetland aided remediation of nitrogen
and phosporus from simulated stormwater runoff. Ecological Economics, 61, 207-
215.
Wong T. & Brown R. (2008). Transitioning to water sensitive cities: ensuring resilience
through a new hydro-social contract. Proceedings of the 11th International
Conference on Urban Drainage, Edinburgh, Edinburgh, UK.
Wong T.H.F., Allen R., Beringer J., Brown R.R., Deletic A., Fletcher T.D., Gangadharan L.,
Gernjak W., Jakob C., O’Loan T., Reeder M., Tapper N. & Walsh C. (2012).
Blueprint 2012–Stormwater Management in a Water Sensitive City. Centre for Water
Sensitive Cities, Melbourne, Australia, pp. 1-60.
Wong T.H.F., Breen P.F., Somes N.L.G. & Lloyd S.D. (1999). Managing Urban Stormwater
Using Constructed Wetlands. Australia pp. 1-40.
Wong T.H.F. & Brown R.R. (2009). The water sensitive city: principles for practice. Water
Science and Technology, 60, 673-682.
Wong T.H.F. & Eadie M.L. (2000). Water sensitive urban design—A paradigm shift in urban
design. Proceedings of the Xth World Water Congress, pp. 12-16.
Wood R.B. & McAtamney C.F. (1996). Constructed wetlands for waste water treatment: the
use of laterite in the bed medium in phosphorus and heavy metal removal.
Hydrobiologia, 340, 323-331.
Wu H., Zhang J., Ngo H.H., Guo W., Hu Z., Liang S., Fan J. & Liu H. (2015). A review on
the sustainability of constructed wetlands for wastewater treatment: Design and
operation. Bioresource Technology, 175, 594-601.
Wu S., Kuschk P., Brix H., Vymazal J. & 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.
Yang W., Chang J., Xu B., Peng C. & Ge Y. (2008). Ecosystem service value assessment for
constructed wetlands: A case study in Hangzhou, China. Ecological Economics, 68,
116-125.
Zgheib S., Moilleron R. & Chebbo G. (2012). Priority pollutants in urban stormwater: Part 1–
Case of separate storm sewers. Water Research, 46, 6683-6692.
Zgheib S., Moilleron R., Saad M. & Chebbo G. (2011). Partition of pollution between
dissolved and particulate phases: What about emerging substances in urban
stormwater catchments? Water Research, 45, 913-925.
Zhang L., Xia X., Zhao Y., Xi B., Yan Y., Guo X., Xiong Y. & Zhan J. (2011). The
ammonium nitrogen oxidation process in horizontal subsurface flow constructed
wetlands. Ecological Engineering, 37, 1614-1619.
References
158
[This page has left blank intentionally]