Thesis Constructed Wetland
Thesis Constructed Wetland
Thesis Constructed Wetland
3
). In Europe, for example, increases in nitrate levels in rivers over the
past 50 years correlated with increased use of N fertilizer (Howarth et al., 1996).
In the US, an estimated 7% of US drinking waterwells have been shut down be-
cause of agricultural-related nitrate contamination, and 44,000 infants are at risk
for nitrate toxicity from contaminated drinking water (Horne, 2001). Enhanced
algal productivity stimulated by nitrate pollution from agricultural activities is a
primary cause of the growing number of hypoxic zones in coastal waters around
the world (Daigle, 2003), including the dead zone in the Gulf of Mexico (Weir,
2005).
1
Concerns related to N pollution have stimulated interests in the use of
constructed treatment wetlands (CTWs) to treat nitrate pollution from non-
point sources (Braskerud, 2002; Kovacic et al., 2006; Mitsch et al., 2005; Reilly
et al., 2000) as well as domestic, industrial and livestock wastewaters, and land-
ll leachate (Moshiri, 1993). Application of CTWs to water quality management
attracted a greater interest due to their inherent characteristics such as low cost
and simple operation, and has been implemented in many places including in de-
veloping countries (Haberl, 1999), thus, resulting in a globally increased portion
of land covered by wetlands (IWA, 2000).
1.2 Constructed Treatment Wetlands
Constructed treatment wetlands are engineered natural treatment systems
primarily composed of aquatic plants (hydrophytes), hydric soil and water to keep
the soils waterlogged for extended period and are commonly designed as a ow-
through system having a hydrologic feature of either a free water surface sys-
tem (FWS) or subsurface ow system (SFS)(Greenway, 2004). While they are
commonly designed for water quality improvement, they also advance enhanced
aestheics, landscape, and storm mitigation. However, their treatment eciency
appears to be critically dependent on weather conditions. Variation in the amount
of rainfall and evapotranspiration leads to variable detention time and inconsistent
treatment rates. At times when inuent ow volume exceeds storage volume of
the wetland, portion of the inow exits the wetland untreated, thus leading to dif-
culties to achieve treatment goals consistently. Flood induced loss of carbon from
wetland has also been reported to cause substantial decrease in nitrate removal
(Reilly et al., 2000).
2
1.3 Nitrate Removal in Constructed Treatment
Wetlands
Nitrate removal in CTWs is assumed to be accompolished via dissimilatory
denitrication by heterotrophic bacteria utilizing nitrate as a terminal electron
acceptor to produce energy in anaerobic environment. The complete denitrica-
tion process involves sequential reductive steps starting with reduction of NO
3
to
nitrite (NO
2
). Nitrite can then be further reduced to ammonium or nitric oxide
(NO), which in turn gets reduced to nitrous oxide (N
2
O), and nally to dinitrogen
(N
2
) gas according to Eqn. (1.1) (Madigan and Martinkor, 2006). While nitrate
can also undergo assimilatory reduction by plants, fungi and bacteria (Guerrero
et al., 1981), dissimilatry denitrication is considered to be a major pathway for
nitrate removal in CTWs (Bachand and Horne, 2000b,c; Lund et al., 2000). The
inability of denitriers to synthesize denitrifying enzymes in the presence of oxygen
(Madigan and Martinkor, 2006) makes the predominantly reduced environment of
CTWs an ideal ecosystems for nitrate reduction (Kadlec and Knight, 1996). Deni-
trication in CTWs occurs at the reduced layer of episediment (Fleming-Singer and
Horne, 2002). Nitrate concentration (Poe et al., 2003), quality of organic matter
at the sediment/water interface (Russell et al., 1994), activity of the periphytons
(Sirivedhin and Gray, 2006), type of hydrophytes (Bachand and Horne, 2000c;
Bastviken et al., 2005), quantity and quality of plant biomass (Wen et al., 2010)
and temperature (Bachand and Horne, 2000c; Wood et al., 1999) were reported
as important factors in cotrolling denitrication in CTWs.
NO
3
NO
2
NO N
2
O N
2
(1.1)
3
1.4 Thesis Objectives
Emergent wetland plants are important component of CTWs, and these
plants aect N processing in many ways including: (1) partially controlling ow
patterns through the wetland, (2) blocking the wetland water surface from wind
and sunlight, thereby lowering phytoplankton growth, rates of reaeration, and the
potential for warming of water, (3) providing submerged surfaces for microbial
attachment, and (4) supplying the degradable organic carbon (C) that drive het-
erotrophic bacterial activity (Greenway, 2007; Kadlec, 2008; Thullen et al., 2005).
The importance of hydrophytes to nitrate removal in CTWs has been repeatedly
shown through experiments comparing removal rates of planted versus unplanted
CTWs (Kyambadde et al., 2005; Lin et al., 2007; Tanner, 2001). Reduction in
biomass due to increased harvesting of wetland plants have also been reported
to decrease nitrate removal from wetland (Martin et al., 2003). Because wetland
plants vary by the quality and quantity of C they supply to wetlands, as well as
by the size of surface area they provide for microbial attachment, dierent plant
species can dierently aect removal rates of pollutants (Corstanje et al., 2006;
Horne and Fleming-Singer, 2005). The few studies that have evaluated nitrate
removal as a function of wetland plant type, commonly comparing Typha spp.
(cattail) with Scirpus spp. (bulrush), have shown either little eect (McIntyre and
Riha, 1991; Zhu and Sikora, 1995), or greater removal by cattail (Bachand and
Horne, 2000a; Hume et al., 2002).
The objective of this research was to test the working hypothesis that nitrate
removal in CTWs is greater using cattail versus bulrush. These two plant species
were selected because they are commonly used in CTWs, and they are two species
that are commonly compared in the literature regarding nitrate removal. Water
in the small, batch-mode mesocosms with plants and associated sediments were
spiked with nitrate, and nitrate was monitored over time to estimate relative rates
4
of nitrate removal. Experimental mesocosms were used because the approach
captures some of the ecological complexity of full-scale treatment wetlands while
allowing for the replication needed for statistical comparison between treatments
(Kangas and Adey, 1996). In contrast to other mesocosms studies that used sand
or gravel (Iamchaturapatr et al., 2007; Zhu and Sikora, 1995), sediments collected
from cattail and bulrush stands from a regional CTW were used. An additional
unique aspect of this study was the intensive monitoring of dissolved oxygen (DO)
in the water column of the mesocosms. Given that cattail litter typically has more
labile C and N relative to bulrush (Hume et al., 2002), higher microbial activity,
higher nitrate removal, and lower DO were expected in the cattail mesocosms.
5
Chapter 2
Materials and Methods
2.1 Experimental Design
Triplicate mesocosms were constructed consisting of two plant treatments
and associated mineral-dominated sediment: cattail (Typha spp.) and bulrush
(Scirpus spp.). Plants, sediment and water from cattail and bulrush stands were
collected during the summer of 2006 from a mature CTW in Moscow, Idaho,
used to polish secondary euent. In the laboratory, sediment and plants were
transferred into glass aquariums measuring 50.8 cm in length, 25.4 cm in width,
and 30.5 cm in height. Plant density in each mesocosm was approximately 50
plants/m
2
. Once plants took root and sediments stabilized to a thickness of around
15 cm, mesocosms were gently ooded with 13 L of wetland water to a depth of
10 cm. The mesocosms, operated as batch systems, were spiked with nitrate to
obtain a nal concentration of 19 mg N/L. Minor evaporative losses of around
400 mL/d were compensated for by adding deionized water to the aquariums once
every day. Room temperature was maintained at 18.5
C and plants were exposed
to 8 h/d of indoor plant lighting as well as natural light from nearby windows.
Duplicate control aquaria with no plants and sediments were set up and operated
in the same manner as experimental ones.
6
2.2 Sampling and Analysis
Sediments collected from the stands where wetland plants were collected
were analyzed (in triplicate) for a range of parameters including pH, water content
(drying at 105
C for 24 h), loss on ignition (combustion at 550
C), and total C and
N content (dry combustion at 1350
C; LECO CNS 2000 Analyzer, LECO Corp.
MI). DO was measured in all six mesocosms with Hach standard luminescent DO
(LDO) IntelliCAL probes attached to HQ40d digital meter/data loggers (Hach
Company, Loveland, CO). DO probes were deployed at a water depth of 5 cm
and water column DO was automatically measured and logged every 30 min over
a 2-month duration. Approximately 1140 DO data points were collected in each
mesocosm. Three weeks after the start of DO monitoring, mesocosms were spiked
with nitrate and monitored for nitrate and ammonia for 14 days. Water samples
were collected at decreasing frequency as the experiment progressed, ve to four
times per day initially and once per day towards the end of the incubation. For each
sampling event, duplicate 2 mL water samples were collected from each mesocosm
at a depth of 5 cm using a syringe. Samples were ltered using a 0.45 pore size
syringe tip-lter and analyzed for nitrate and ammonia by ow injection analysis
on a Lachat 8500 QuikChem auto analyzer (Lachat Instruments, Milwaukee,WI).
Ammonia rarely exceeded 0.1 mg N/L, suggesting that N cycling in the wetland
mesocosms was dominated by nitrate. Thus, the remainder of this paper focuses
on nitrate dynamics. Various physico-chemical properties of the sediments used
for the experiment are provided in Table (2.1).
Table 2.1: Physical and chemical properties of wetland sediments
Sediment pH Conductivity Loss on Total Total Texture
mS/cm Ignition% Nitrogen% Carbon%
Cattail 7.19 0.78 5 0.15 2.1 Silt Loam
Bulrush 7.21 0.79 7 0.17 2.5 Silt Loam
7
2.3 Data Analysis
2.3.1 Data Modeling
Assuming a rst-order rate for nitrate removal under batch conditions at
constant volume and uniform nitrate concentration, the equation for concentration
over time is
C(t) = C
o
e
k
v
t
(2.1)
where C(t) (mg N/L) is concentration of nitrate at time t (d), Co (mg N/L) is
initial nitrate concentration, and k
v
(d