Wetlands and Aquatic Processes

Trace Element Removal from Coal Ash Leachate by a 10-Year-Old

Constructed Wetland
Z. H. Ye, S. N. Whiting, J. H. Qian, C. M. Lytle, Z.-Q. Lin, and N. Terry*
ABSTRACT

monthly average discharge limitations for Fe and Mn

are 3 and 2 mg L1, respectively, and the pH of the
water must be in the range 6 to 9 (USEPA, 2001).
Traditionally, AMD was treated by the application of
chemical bases to elevate the pH of the wastewater;
the pH increase caused Fe and Mn to precipitate from
solution due to the pH-dependent solubility of these
metals upon exposure to oxygen. Chemical treatment,
however, is costly due to the considerable requirements
of chemicals and associated labor, and furthermore, the
generation of large amounts of sludge that must be
disposed safely (Stark et al., 1995).
In the last two decades there has been a surge of
interest in the use of constructed wetlands as effective,
low-cost, low-maintenance alternatives for AMD treatment (Kleinmann, 1985; Wieder, 1989; Tarutis et al.,
1999). Constructed wetlands are able to remove trace
elements from the wastewater via a number of complex,
interactive physicochemical reactions. The primary
mechanisms responsible for the removal and retention
of Fe, Mn, and S include the formation and precipitation
of metal oxides and sulfides within the sediments (Tarutis and Unz, 1995). Indeed, 40 to 70% of total Fe removed from AMD by some wetlands was found as ferric
hydroxides from the hydrolysis of ferric iron or the
oxidation of ferrous iron (Henrot and Wieder, 1990).
Wetland ecosystems can be characterized by their
emergent macrophytes, and these plants are crucially
important in many ways for removing trace elements
from the wastewater. Although the direct uptake of
trace elements into the plant tissues appears to account
for only a small proportion of the total removal by some
wetlands (Mitsch and Wise, 1998), plants potentiate
metal retention by filtration, adsorption, and cation exchange, and through plant-induced chemical changes in
the rhizosphere (Dunbabin and Bowmer, 1992). Plants
also provide habitat and energy sources (organic carbon) to maintain and stimulate a diverse microbial population in the sediments (Skousen et al., 1994); these
microbes drive the immobilization of contaminants in
the sediments through both oxidative and reductive processes (Johnson, 1998). Precipitation of metal oxides,
following microbe-mediated oxidation, is thought to be
one of the most important removal mechanisms in wetlands (Skousen et al., 1994; Stark et al., 1996).
Wetlands are capable of removing large quantities
of trace elements from wastewater. There is, however,
considerable variation both among metals and also be-

This study investigated the ability of a 10-yr-old constructed wetland to treat metal-contaminated leachate emanating from a coal ash
pile at the Widows Creek electric utility, Alabama (USA). The two
vegetated cells, which were dominated by cattail (Typha latifolia L.)
and soft rush ( Juncus effusus L.), were very effective at removing Fe
and Cd from the wastewater, but less efficient for Zn, S, B, and Mn.
The concentrations were decreased by up to 99% for Fe, 91% for
Cd, 63% for Zn, 61% for S, 58% for Mn, and 50% for B. Higher pH
levels (6) in standing water substantially improved the removing
efficiency of the wetland for Mn only. The belowground tissues of
both cattail and soft rush had high concentrations of all elements;
only for Mn, however, did the concentration in the shoots exceed
those in the belowground tissues. The concentrations of trace elements
in fallen litter were higher than in the living shoots, but lower than in
the belowground tissues. The trace element accumulation in the plants
accounted for less than 2.5% of the annual loading of each trace
element into the wetland. The sediments were the primary sinks for
the elements removed from the wastewater. Except for Mn, the concentrations of trace elements in the upper layer (05 cm) of the sediment
profile tended to be higher than the lower layers (510 and 1015
cm). We conclude that constructed wetlands are still able to efficiently
remove metals in the long term (i.e., 10 yr after construction).

cid mine drainage (AMD) is produced during the

mining, storage, and processing of coal, and in
association with the land disposal of the byproducts of
coal combustion. Acid mine drainage is characterized
by low pH as well as iron (Fe) and manganese (Mn)
concentrations of greater than 6 and 2 mg L1, respectively (Brodie et al., 1989; Tarutis et al., 1999). Acid
mine drainage therefore constitutes a major pollution
problem (Vile and Wieder, 1993; Tarutis and Unz, 1995),
affecting more than 20 000 km of streams and rivers
in the USA (Perry and Kleinmann, 1991). The turbidity and high concentrations of trace elements of AMD
reduce both species diversity and population size of
aquatic plants, invertebrates, and fish in the streams
receiving these discharges (Koryak and Reilly, 1984).
Coal-fired power-generating facilities must achieve regulatory compliance when discharging such metal-contaminated wastewater.
Wastewater discharges in the USA must comply with
National Pollution Discharge Elimination System
(NPDES) effluent limitations. For example, the NPDES
Department of Plant and Microbial Biology, Univ. of California at
Berkeley, 111 Koshland Hall, Berkeley, CA 94720. Z.H. Ye, present
address: Institute for Natural Resource and Environmental Management, Hong Kong Baptist Univ., Hong Kong. Received 1 Aug. 2000.
*Corresponding author (nterry@nature.berkeley.edu).

Abbreviations: AMD, acid mine drainage; NPDES, National Pollution

Discharge Elimination System.

Published in J. Environ. Qual. 30:17101719 (2001).

1710

YE ET AL.: TRACE ELEMENT REMOVAL FROM COAL ASH LEACHATE

tween wetlands in the degree to which each metal is

removed (Ye et al., 2001). A further significant problem
with AMD is that it is produced for many tens or even
hundreds of years as the coal combustion by-products
continue to weather, releasing sulfates and trace elements. For wetlands to be a reliable and sustainable
treatment technology, their pollutant removal rates
must be consistently high in the long term (such as
10 yr). Some studies indicate that constructed wetlands have a finite lifespan with respect to metal retention and that they could eventually fail to remove some
elements (Weider, 1993; Horne, 2000). For example, the
capacity of wetlands to retain Fe (primarily as oxides)
might eventually be exhausted, and the acidity of the
mine water will overcome the capability of the wetland
to neutralize the water (Stark et al., 1995).
This study examined trace element removal by an
overmatured (10-yr-old) wetland constructed at the
Widows Creek Fossil Plant in Alabama. The Tennessee
Valley Authority (TVA) created this wetland as an eco-

1711

logical reclamation system treating acidic leachate from

an abandoned coal ash pile. We investigated whether
this 10-yr-old wetland could still efficiently remove trace
elements from the wastewater, and determined the fate
of the six primary contaminants (Fe, Mn, S, B, Cd, and
Zn) loaded into the wetland. Because high concentrations of trace elements have been sequestered in the
wetland sediments over a 10-yr operation, particular
attention was paid to the wetland development (vegetation) and trace element uptake by the dominant plant
species.
MATERIALS AND METHODS
Constructed Wetland
The wetland was constructed in June 1986 at the Widows
Creek Fossil Plant (Jackson County, AL) along the toe of the
coal runoff pile. Three shallow cells were excavated to create
a marsh-type surface-flow wetland with an original cattailvegetated area of 4800 m2 (Fig. 1). Cell 2 had six finger dikes

Fig. 1. Plan diagram of the Widows Creek constructed wetland in Jackson County, Alabama. Cells 1, 2, and 3 enclosed 630, 3120, and 1050
m2, respectively.

1712

J. ENVIRON. QUAL., VOL. 30, SEPTEMBEROCTOBER 2001

that extended into the cell to prevent short-circuiting of the

water flow through the wetland. In August 1987, the cattails
were almost entirely eradicated by an infestation of cattail
armyworms (Simyra henrici ). In the next year, the wetland
plants reemerged and were dominated primarily by cattail,
but with a significant number of soft rush, along with some
spike rush [Eleocharis palustris (L.) Roem. & Schult.] and
woolgrass bulrush [Scirpus cyperinus (L.) Kunth]. During this
study period (May 1996 through May 1997), Cell 1 and part
of Cell 2 (mostly within last four finger dike sections) were
vegetated predominantly with cattail and soft rush, which this
study primarily focused on. Cell 3 was not examined because
sodium hydroxide solution was continuously added to the
water flowing into this cell to ensure sufficient metal removal
by chemical precipitation.

Collection of Samples
The plants, water, and sediments were sampled in Cells 1
and 2 only. To ensure randomized sampling, three 1-m2 areas
per cell were chosen using the standard grid sampling technique (USEPA, 1989). Three whole-plant samples (root and
shoot) of the dominant species, cattail and soft rush, were
collected from each cell. The plants were collected in triplicates on a monthly basis except for November, December,
and January, when the plant shoots were senescent; samples
of plant fallen litter were collected during the winter months.
Samples were collected in triplicate on a quarterly basis from
each cell. Water samples were collected from the inlet to Cell 1
and the outlet from Cell 2. Sediment profiles (5-cm cores of
the 0- to 15-cm layers as described by Boulding, 1994) and sediment pore water (Rhizon soil moisture filter, 10 cm, 0.1-m
pore size [Ben Meadows, Canton, GA]) were collected in the
randomly assigned sampling areas. Plant density was determined from monthly counts of cattail and soft rush shoots in
five randomly selected 0.25-m2 areas within each cell. The
biomass of cattail and soft rush per unit sampling area was
determined as the product of the dry weight of each whole
shoot (an average of five shoots) and the plant density in each
sampling area. The pH and temperature were measured monthly
using a Corning Checkmate modular system (Fisher Scientific,
Pittsburgh, PA) at the center of each cell, where possible.

Trace Element Analysis

Samples of unfiltered inlet and outlet waters were digested
with a mixture of HNO3, H2O2, and HCl (USEPA, 1983). Plant
samples were dried at 70C, weighed, ground in a Wiley mill,
and acid-digested with a mixture of HNO3, H2O2, and HCl
(Zarcinas et al., 1987). The sediment profiles were cut into 5-cm
sections (05, 510, and 1015 cm), air-dried, ground, and aciddigested with HNO3 and H2O2 (Method 3050B; USEPA, 1996a).
The total concentrations of Fe, Mn, S, B, Cd, and Zn were measured in the water, plant, and sediment digests by inductively
coupled plasmaatomic emission spectroscopy (ICPAES)
(Method 6010B; USEPA, 1996b). The water-soluble trace elements in pore waters were determined directly by ICPAES.

RESULTS
Temperature and pH of the Surface Water
The pH and temperature of the surface water in
Cells 1 and 2 were not significantly different (p 0.05,
two-tailed t test), so the data presented in Fig. 2 are
averaged for measurements taken in both cells. The
initial surface water pH in Cells 1 and 2 (May 1996)

Fig. 2. Temporal changes in the pH and temperature of the surface

water in Cells 1 and 2. Error bars indicate standard error of the
mean (SE), n 6.

was low (acidic), but the pH increased to about 7 in

July (Fig. 2) due to a large amount of rainfall. In an
attempt to raise the pH of the wastewater and hence
improve the efficiency of contaminant removal, the
NaOH solution was added to Cell 1 in October, and
the pH remained high in the following months, through
March 1997. The temperature of the surface water
showed typical seasonal variation, with a summer high
of about 30C and a winter low of 5C (Fig. 2).

Reduction in Trace Element Concentrations

There was considerable variation in the concentration
of each trace element between the sampling times
(Fig. 3). The lowest concentration of each of the six
trace elements was in February 1997, which coincides
with the period of high rainfall (i.e., dilution of the
leachate). The concentrations of trace elements in the
outlet water were lower than those in the inlet in most
cases, confirming that the wetland was removing trace
elements from the wastewater. The degree to which
the concentrations of the trace elements were reduced
varied considerably between sampling times; the maximum concentration reduction rates were 99% for Fe,
91% for Cd, 63% for Zn, 61% for S, 58% for Mn, and
50% for B.

Trace Element Concentrations in the Sediments

The concentrations of Fe and S in the sediment,
70 000 and 8000 mg kg1, respectively, were the highest among the six elements studied (Fig. 4). The concen-

YE ET AL.: TRACE ELEMENT REMOVAL FROM COAL ASH LEACHATE

1713

Fig. 3. Temporal changes in trace element concentrations in the inlet water to Cell 1 and the outlet water from Cell 2. Figures above the bars
indicate the percentage reduction in trace element concentration for that month.

trations of the other elements were much lower and

were in the order of Mn B Zn Cd, which is the
same as their relative concentrations in the inlet water
(Fig. 4 compared with Fig. 3). In both Cells 1 and 2, the
concentrations of Fe, S, B, Cd, and Zn in the top (05
cm) layer tended to be higher than in the lower layers;
however, these differences were not statistically significant (analysis of variance [ANOVA], all P 0.05)
(Fig. 4). For all elements except Mn and S, concentrations in the top 15 cm of sediments of Cell 1 tended to
be higher than those in Cell 2, but again, the differences
were not statistically significant (Fig. 4; P 0.05, twotailed t test). Furthermore, the concentration of trace
elements in the sediments did not increase significantly
during the 1-yr period of this study (data not shown).
The concentrations of S, Fe, Mn, B, Zn, and Cd in the
pore water were 539, 100, 6.4, 1.24, 0.41, and 0.03 mg

L1, respectively, which are identical to their mean concentrations in the inlet water.

Plant Productivity and Trace

Element Accumulation
The shoot counts in each sampling quadrat showed
that soft rush had a much higher density of shoots than
cattail over the entire study period (Fig. 5). However,
because the shoot morphology of the two species is very
different, the larger stature of the cattail resulted in a
greater standing biomass than soft rush (Fig. 5). In the
winter, the senescence of the standing plants resulted
in the accumulation of a significant amount of fallen
litter. During the months of November, December, and
January, the amount of fallen litter in the sampling areas
averaged 260 g m2 (Fig. 5).

1714

J. ENVIRON. QUAL., VOL. 30, SEPTEMBEROCTOBER 2001

Fig. 4. Mean trace element concentrations in the sediments of Cell 1 and Cell 2 measured at three depths in the sediment profile: 0 to 5 cm, 5
to 10 cm, and 10 to 15 cm. Means are calculated from measurements taken over the entire study (standard error of the mean [SE], n 5).

Notably, the aboveground tissues of soft rush had

higher concentrations of all six elements compared with
cattail (Fig. 6). For the belowground tissues of cattail
and soft rush, the concentration of all trace elements
except B tended to be higher in soft rush than in cattail
(Fig. 7). Because the concentrations of the six elements
in the aboveground and belowground tissues of both
cattail and soft rush were similar in Cells 1 and 2 (data
not shown for each cell), the data presented in Fig. 6
and 7 are averages calculated from the samples of each
species collected in both cells. The concentrations of Fe
and Mn in the shoots of soft rush were approximately
fourfold greater than the concentrations in the shoots of
cattail. There were no seasonal patterns of trace element
accumulation in the aboveground tissues of either species (Fig. 6). As for the shoots, there was marked variation between the sampling times in the concentration
of trace elements in the belowground tissues, which did
not follow any apparent seasonal patterns. For compari-

son, the trace element uptake by the plants is summarized as averages calculated from samples collected over
the entire study period (Table 1). The concentrations
of all elements (except Mn) in the belowground tissues
were typically greater than the concentrations in the
aboveground tissues and, in some cases, more than 60
times greater. The trace element concentrations in the
fallen litter were comparable with the belowground tissues (Table 1). Moreover, the concentrations of the six
trace elements in both the aboveground and belowground tissues followed the same order as their concentrations in both the sediments and the inlet water,
that is, Fe S Mn B Zn Cd (Fig. 4 and Table 1).

Trace Element Loading and

Retention by the Plants
The total mass of each element accumulated in the
standing biomass and fallen litter of soft rush and cattail

YE ET AL.: TRACE ELEMENT REMOVAL FROM COAL ASH LEACHATE

Fig. 5. Changes in plant density and shoot biomass for cattail and
soft rush. Values are averages of samples collected from both cells.

(Table 2) was calculated from the average biomass of

cattail shoots (509 g m2 ), soft rush shoots (242 g m2 ),
and fallen litter (259 g m2 ), and the mean concentration
of trace elements in these tissues (Table 1). The total
mass of trace elements accumulated within each of these
plant components was small compared with the trace
element loadings. The maximum was for Fe accumulated in the fallen litter (13.4 g m2 ), and the minimum
was for Cd accumulation by both cattail and soft rush
(0.0001 g m2 ).
The estimated loading rates of the six trace elements
in the inlet water to the wetland were calculated as the
product of the average element concentrations in the
inlet water and average flow rate of water into Cell 1
for each quarter on a unit area basis. Because we are
interested in the role of plants in trace element retention, the mass of trace elements accumulated by these
components is presented as a percentage of the annual
loading in the wetland (Table 2). It can be seen that
the direct uptake of trace elements by the plants is
very small when compared with the loading. Again, Fe
accumulation by the fallen litter was the highest. Even
so, this only represented 1.38% of the loading. For the
other five elements, the accumulation within the
aboveground tissues or fallen litter of the plants was
generally less than 1% of the mass of each element
flowing into the wetland during the year.

DISCUSSION
Trace Element Removal by Wetlands
and Role of Vegetation
It has been suggested that constructed wetlands have
a finite and short life span with respect to metal reten-

1715

tion (Wieder, 1993; Horne, 2000). However, this 1-yr

study indicates that the 10-yr-old constructed wetland
reduced the concentrations of the primary contaminants
in the wastewater, and was still able to almost completely remove Fe and Cd from the wastewater (Fig. 3).
At the Widows Creek wetland the concentrations of
Fe in the outlet were between 0.16 and 22 mg L1,
irrespective of the high variability in Fe concentrations
in the inlet water (9.5 to 234 mg L1 ). The concentrations of Fe in the outlet water were below the NPDES
discharge limits in the months of November 1996 and
May 1997. Moreover, Cd was below the NPDES discharge limit in May 1996, February 1997, and May 1997,
and Zn met the NPDES limit (0.09 mg L1 ) in all months
except May 1997. Our results are consistent with the
findings of other studies, which demonstrated that wetlands are more effective at removing Fe than other
elements, such as Mn and S, from AMD (e.g., Wieder,
1989; Mitsch and Wise, 1998; Ye et al., 2001).
The fate of the six primary contaminants within the
wetland was determined with particular reference to the
roles of the plants in trace element removal. The shoot
biomass of cattail and soft rush indicated that they were
tolerant to the high concentrations of trace elements in
the water and sediments within the wetland. The standing biomass of cattail shoots was 509 g m2, which is
comparable with the 447 and 502 g m2 given by Fennessy and Mitsch (1989) and Mitsch and Wise (1998),
respectively, for cattails grown in wetlands treating
AMD. The aboveground biomass of soft rush was also
in the range reported for that species by Vymazal (1995).
When averaged over the study period (Table 1), however, the mean concentrations of trace elements in the
plants were within the ranges reported by other authors
for these species growing in metal-contaminated wetlands (Taylor and Crowder, 1983; Fernandes and Henriques, 1990). Notably, both species accumulated more
Mn in their shoots than their belowground tissues, a
finding also reported by Taylor and Crowder (1983) and
Fernandes and Henriques (1990). Our results indicate
that the majority of the Fe, S, B, and Cd taken up or
immobilized by the plants were localized within the
belowground tissues. This localization of trace elements
within the root system, and limited transfer to the
shoots, may be a corollary mechanism of tolerance toward toxic trace elements by these wetland plants (Taylor and Crowder, 1983; Ye et al., 1997).
The senescence of the plants in the winter results in
a considerable mass of fallen litter in the wetland cells.
The concentrations of Fe, S, B, and Cd, but especially
Fe and Cd, in or bound onto this fallen litter were higher
than those in the aboveground tissues of cattail and soft
rush (Table 1). In fact, the concentrations of Fe and Cd
in the fallen litter were 161 and 58 times higher than
those in the shoots of cattail, respectively. Our field observations documented a significant quantity of brown precipitate on the fallen litter, most probably composed of
hydrated iron oxides. This high concentration of Cd in
the fallen litter may also be a result of the high binding
affinities between Cd and organic matter (Ross, 1994).

1716

J. ENVIRON. QUAL., VOL. 30, SEPTEMBEROCTOBER 2001

Fig. 6. Trace element concentrations in the aboveground tissues of cattail (black circles) and soft rush (unfilled circles). Monthly means are
calculated from plant samples collected from both cells, error bars indicate standard error of the mean (SE).

This indicates that the plant detritus in the wetland

might play a significant role in the removal of elements
such as Fe and Cd by adsorptionabsorption or by providing sites for their precipitation from the water.
Trace element uptake and accumulation in the plant
shoots does not contribute significantly to the overall
removal of trace elements by the wetland. Means masses
of 2.2 g of S, 1 g of Mn and Fe, and 0.05 g of B, Zn,
and Cd were accumulated in aboveground tissues of the

plants (cattail and soft rush) per square meter (Table 2).
With respect to the total loadings of trace elements
in the inflow, the aboveground plant tissues therefore
retained less than 1.9% of the annual loading (Table 2).
This highlights the minor role of metal uptake by the
plant shoots in the efficiency of metal removal from
wastewater by wetlands. Similarly, Mitsch and Wise
(1998) reported the bioaccumulation of Fe in plant
shoots as only accounting for 0.07% of the annual load-

Table 1. Mean concentrations of the six trace elements in the living and senescent tissue of cattail and soft rush (mean SE, n 12,
except fallen litter, where n 3).
Trace element concentrations
Element

Cattail shoots

Soft rush shoots

Cattail roots

Soft rush roots

Fallen litter

kg1

Fe
Mn
S
B
Zn
Cd

320
1 097
2 495
46
27
0.24

31
114
260
4
7
0.08

1 217
312
3 853
41
54
0.5

176
38
143
5
8
0.2

mg
41 318
200
6 026
113
42
10

5 248
15
699
14
4
1

68 469
117
6 885
248
55
27

8 574
13
567
42
4
5

51 626
502
4 128
181
34
14

6 811
102
502
16
5
2

1717

YE ET AL.: TRACE ELEMENT REMOVAL FROM COAL ASH LEACHATE

Fig. 7. Trace element concentrations in the belowground tissues of cattail (black circles) and soft rush (unfilled circles). Monthly means are
calculated from plant samples collected from both cells, error bars indicate standard error of the mean (SE).

ing of Fe into a cattail-dominated wetland. Similarly,

the fallen litter retained a maximum of 1.4% of the
annual loading of each of the six elements despite containing higher concentrations of Fe, Cd, Mn, and B than
the living shoots (Table 2).
The concentrations of the trace elements in the sediments were very high (Fig. 4), and were within the
ranges reported by previous investigators in the other
cattail-dominated wetlands (Taylor and Crowder, 1983).
The wetland sediment was a major pool of the trace

elements removed by the wetland, which agrees with

the findings of many other studies (e.g., Fernandes and
Henriques, 1990). Our data indicate that there was direct deposition of the trace elements to the sediments;
there was a trend for higher concentrations of the trace
elements except Mn in the top layers (05 cm) of Cell 1,
and to a lesser extent in Cell 2 (Fig. 4), than in the lower
layers (510 or 1015 cm). A similar localized deposition
was found at the Springdale constructed wetland in
Pennsylvania, also removing trace elements from coal

Table 2. Annual trace element loading into the wetland, the total mass of trace elements accumulated in the plant material, and the
contribution of trace element accumulation in the plants to removal by the wetland. Total treatment area of the wetland 3600 m2,
average flow rate in inlet 70 L min1. The data for quarterly flow rate in the inlet were provided by the Widows Creek Fossil Plant
in Alabama, based on averages of frequent measurements they recorded.
Fe
Annual loading into the wetland, g
Accumulation in plants, g m2
Cattail shoots
Soft rush shoots
Fallen litter
Percent of the annual loading
Cattail and soft rush shoots
Fallen litter

m 2

yr1

967

Mn
80

S
4875

B
24

Zn

Cd

1.41

0.49

0.16
0.29
13.4

0.56
0.076
0.13

1.27
0.93
1.07

0.014
0.013
0.01

0.02
0.01
0.05

0.0001
0.0001
0.004

0.05
1.38

0.80
0.16

0.05
0.02

1.90
0.63

0.14
0.20

0.05
0.76

1718

J. ENVIRON. QUAL., VOL. 30, SEPTEMBEROCTOBER 2001

combustion by-product leachate (Ye et al., 2001). Although the sediment contains a large volume of pore
water, the low element concentrations in this fraction
do not contribute significantly to the overall element
budget in the wetland.

Factors Influencing the Effectiveness

of Wetland Systems to Treat
Acid Mine Drainage
The concentration of Fe in the inlet water was quite
variable with time, but the Fe concentrations in the
outlet water remained fairly constant (Fig. 3). It appears,
therefore, that the efficiency of Fe removal by the first
two cells of the wetland was unaffected by the rate of
Fe input. This may, in part, be due to the low Fe loading
rate in this wetland (2.6 g m2 d1 ). For example,
Brodie (1993) indicated that Fe removal by wetlands is
generally efficient for loading rates up to 13 g m2 d1.
Conversely, in February of our study, the dilution effect
of the heavy rain resulted in low concentrations of all
trace elements in the inlet water. The heavy rainfall was
also associated with a low concentration reduction (%)
of Fe, Mn, S, and Cd in this month (Fig. 3). This suggests
that the addition of rainfall water to the wetland decreased residence time of wastewater in wetlands, along
with low-standing biomass, and therefore, decreased the
ability of wetlands to remove trace elements from the
wastewater (Stark et al., 1994).
Abiotic factors also affect the ability of wetlands in
trace element removal from wastewater. The process of
abiotic iron oxidation or precipitation is primarily pH
dependent, increasing 100 times for every unit increase
of pH above 3.5 (Hedin, 1989). Furthermore, Mn requires a higher pH than Fe to be precipitated from
solution; Skousen et al. (1994) indicated that oxidized
iron precipitates as ferric hydroxide as the pH increases
above 3.5, while Mn hydroxides require a pH of at least
7 to precipitate. The results presented here show that
pH in the AMD is an important factor in Mn removal
by the constructed wetland (Fig. 2). Before the addition
of caustic soda into Cell 1, the maximum reduction for
Mn was 26%, with a mean pH of 3.9. When NaOH was
drip-fed into Cell 1, the pH in the outlet rose to greater
than 6 and the average reduction of Mn increased to
58% (Fig. 3). In contrast, the reduction in the concentration of the other elements did not significantly change
with caustic soda addition. Similarly, our previous study
of the Springdale constructed wetland (Ye et al., 2001)
showed that where the average pH in water was 7.1,
the average reduction in the concentration of Mn after
passing through the wetland was 91%. Therefore, when
managing constructed wetlands treating AMD, the pH
of the surface water may require chemical manipulation
(e.g., addition of chemical bases) to ensure that Mn can
be removed efficiently to comply with the NPDES Mn
discharge limit.

CONCLUSIONS
This study demonstrated that constructed wetlands
sustain high efficiency of trace element removal from

wastewater in the long term, such as 10 yr after their

construction. The concentrations of Fe and Cd in the
inlet water were substantially reduced (up to 99%) in
some months during the study year. Other trace element
concentrations were reduced, but to a lesser extent (less
than 63%). Most notably, the direct uptake of the trace
elements by cattail and soft rush plants in the wetland
accounted for less than 2.5% of the annual element
loading rates. It appears that the sediments were the
primary sink for the trace elements removed. In this
study, the effects of plantmicrobial interactions on the
localized removal of trace elements were not examined,
but were almost certainly a crucial factor in trace element removal by constructed wetlands. We are currently
elucidating plantmicrobe interactions in the rhizosphere and determining their role in the mechanisms of
trace element sequestration.
ACKNOWLEDGMENTS
This study was supported by a grant from the Electric Power
Research Institute (EPRI). The authors also thank Widows
Creek Fossil Plant in Alabama for helping with this project.

REFERENCES
Boulding, J.R. 1994. Description and sampling of contaminated soils.
2nd ed. Lewis Publ., Boca Raton, FL.
Brodie, G.A. 1993. Stage, aerobic constructed wetlands to treat acid
drainage: Case history of Fabius Impoundment 1 and overview of
the Tennessee Valley Authoritys program. p. 157165. In G.A.
Moshiri (ed.) Constructed wetlands for water quality improvement.
Lewis Publ., Boca Raton, FL.
Brodie, G.A., D.A. Hammer, and D.A. Tomljanovich. 1989. Treatment of acid drainage from coal facilities with man-made wetlands.
p. 903921. In D.A. Hammer (ed.) Constructed wetland for wastewater treatment: Municipal, industrial and agricultural. Proc. from
the 1st Int. Conf. on Constructed Wetlands for Wastewater Treatment, Chattanooga, TN. 1317 June 1988. Lewis Publ., Chelsea, MI.
Dunbabin, J.S., and K.H. Bowmer. 1992. Potential use of constructed
wetlands for treatment of industrial wastewaters containing metals.
Sci. Total Environ. 111:151168.
Fennessy, M.S., and W.J. Mitsch. 1989. Treating coal mine drainage
with an artificial wetland. Res. J. Water Pollut. Control Fed. 61:
16911701.
Fernandes, J.C., and F.S. Henriques. 1990. Metal levels in soils and
cattail (Typha latifolia L.) plants in a pyrites mine area at Lousal,
Portugal. Int. J. Environ. Stud. 36:205210.
Hedin, R.S. 1989. Treatment of coal mine drainage with constructed
wetland. p. 349362. In S.K. Majumdar et al. (ed.) Wetlands ecology
and conservation: Emphasis in Pennsylvania. The Pennsylvania
Acad. of Sci., Easton, PA.
Henrot, J., and R.K. Wieder. 1990. Processes of iron and manganese
retention in laboratory peat microcosms subjected to acid mine
drainage. J. Environ. Qual. 19:312320.
Horne, A.J. 2000. Phytoremediation by constructed wetlands. p. 1339.
In N. Terry and G. Banuelos (ed.) Phytoremediation of contaminated soil and water. Lewis Publ., New York.
Johnson, D.B. 1998. Biodiversity and ecology of acidophilic microorganisms. FEMS Microbiol. Ecol. 27:307317.
Kleinmann, R.L.P. 1985. Treatment of acid mine water by wetlands.
p. 1735. In Control of acid mine drainage. Bureau of Mines Inf.
Circ. 9027. U.S. Dep. of the Interior, Washington, DC.
Koryak, M., and R.J. Reilly. 1984. Vascular riffle flora of Appalachian
streams: The ecology and effects of acid mine drainage on Justicia
americana (L.) Vahl. Proc. Pennsylvania Acad. Sci. 58:5563.
Mitsch, W.J., and K.M. Wise. 1998. Water quality, fate of metals, and
predictive model validation of a constructed wetland treating acid
mine drainage. Water Res. 32:18881900.

YE ET AL.: TRACE ELEMENT REMOVAL FROM COAL ASH LEACHATE

Perry, A., and R.L.P. Kleinmann. 1991. The use of constructed wetlands in the treatment of acid mine drainage. Nat. Resour. Forum
15(3):178184.
Ross, S.M. 1994. Retention, transportation and mobility of toxic metals
in soil. p. 63152. In S.M. Ross (ed.) Toxic metals in soilplant
systems. John Wiley & Sons, Chichester, UK.
Skousen, J., A. Sexstone, K. Garbutt, and J. Sencindiver. 1994. Acid
mine drainage treatment with wetlands and anoxic limestone drains.
p. 263281. In D.M. Kent (ed.) Applied wetland science and technology. Lewis Publ., Boca Raton, FL.
Stark, L.R., R.P. Brooks, F.M. Williams, S.E. Stevens, Jr., and L.K.
Davis. 1994. Water quality during storm events from two constructed wetlands receiving mine drainage. Water Resour. Bull.
30:639650.
Stark, L.R., F.M. Williams, W.R. Wenerick, P.J. Wuest, and C.A.
Urban. 1995. The effects of carbon supplementation and plant
species on iron retention in mesocosm treatment wetlands. Wetlands 15:5867.
Stark, L.R., F.M. Williams, W.R. Wenerick, P.J. Wuest, and C. Urban.
1996. The effects of substrate type, surface water depth, and flow
rate on manganese retention in mesocosm wetlands. J. Environ.
Qual. 25:97106.
Tarutis, W.J., Jr., L.R. Stark, and F.M. Williams. 1999. Sizing and
performance estimation of coal mine drainage wetlands. Ecol.
Eng. 12:353372.
Tarutis, W.J., Jr., and R.F. Unz. 1995. Iron and manganese release
in coal mine drainage wetland microcosms. Water Sci. Technol.
32:187192.
Taylor, G.J., and A.A. Crowder. 1983. Uptake and accumulation of
copper, nickel and iron by Typha latifolia grown in solution culture.
Can. J. Bot. 61:18251830.
USEPA. 1983. Methods for chemical analysis of water and waste.
USEPA 600/4-79-020. USEPA, Washington, DC.

1719

USEPA. 1989. Soil sampling quality assurance users guide. USEPA600/

8-89/046. USEPA, Washington, DC.
USEPA. 1996a. Methods for chemical analysis of water and waste.
Method 3050B. USEPA, Washington, DC.
USEPA. 1996b. Inductively coupled plasmaatomic emission spectroscopy. Method 6010 B. USEPA, Washington, DC.
USEPA. 2001. National Pollutant Discharge Elimination System
(NPDES) permitting program [Online]. Available at http://www.
epa.gov/owm/npdes.htm (verified 22 May 2001). USEPA Office of
Wastewater Manage., Washington, DC.
Vile, M.A., and R.K. Wieder. 1993. Alkalinity generation by Fe(III)
reduction versus sulfate reduction in wetlands constructed for acid
mine drainage treatment. Water Air Soil Pollut. 69:425441.
Vymazal, J. 1995. Algae and element cycling in wetland. Lewis Publ./
CRC Press, Boca Raton, FL.
Wieder, R.K. 1989. A survey of constructed wetlands for acid coal
mine drainage treatment in the eastern United States. Wetlands
9:299315.
Wieder, R.K. 1993. Ion input/output budgets for five wetlands constructed for acid coal mine drainage treatment. Water Air Soil
Pollut. 71:231270.
Ye, Z.H., A.J.M. Baker, M.H. Wong, and A.J. Willis. 1997. Zinc, lead
and cadmium tolerance, uptake and accumulation in populations
of Typha latifolia L. New Phytol. 136:469480.
Ye, Z.H., S.N. Whiting, Z.-Q. Lin, C.M. Lytle, J.H. Qian, and N.
Terry. 2001. Removal and distribution of Fe, Mn, Co and Ni within
a Pennsylvania constructed wetland treating coal combustion byproduct leachate. J. Environ. Qual. (in press).
Zarcinas, B.A., B. Cartwright, and L.R. Spouncer. 1987. Nitric acid
digestion and multi-element analysis of plant material by inductively coupled plasma spectrometry. Commun. Soil Sci. Plant
Anal. 18:131146.