The document examines the ability of three aquatic plants - parrot feather, creeping primrose, and water mint - to remove heavy metals from contaminated water. All three plants were able to remove iron, zinc, copper, and mercury, with average removal efficiencies of 99.8%, 76.7%, 41.62%, and 33.9% respectively. The removal rates of zinc and copper were constant, while iron and mercury removal rates depended on initial concentrations. Parrot feather showed the greatest tolerance to toxicity, followed by water mint and creeping primrose, whose growth was significantly affected. The plants accumulated metals in the order of mercury, iron, copper, then zinc.

1. Phytoaccumulation of heavy metals by aquatic plants
M. Kamala
, A.E. Ghalya,*, N. Mahmouda
, R. Côtéb
a
Biological Engineering Department, Dalhousie University, P.O. Box 1000 Halifax, Nova Scotia, Canada B3J 2X4
b
School for Resources and Environmental Studies, Dalhousie University, Halifax, Nova Scotia, Canada
Received 8 November 2002; accepted 26 March 2003
Abstract
Three aquatic plants were examined for their ability to remove heavy metals from contaminated water: parrot feather (Myriophylhum
aquaticum), creeping primrose (Ludwigina palustris), and water mint (Mentha aquatic). The plants were obtained from a Solar Aquatic System
treating municipal wastewater. All the three plants were able to remove Fe, Zn, Cu, and Hg from the contaminated water. The average removal
efficiency for the three plant species was 99.8%, 76.7%, 41.62%, and 33.9% of Hg, Fe, Cu, and Zn, respectively. The removal rates of zinc and
copper were constant (0.48 mg/l/day for Zn and 0.11 mg/l/day for Cu), whereas those of iron and mercury were dependent on the concentration
of these elements in the contaminated water and ranged from 7.00 to 0.41 mg/l/day for Fe and 0.0787 to 0.0002 mg/l/day for Hg. Parrot feather
showed greater tolerance to toxicity followed by water mint and creeping primrose. The growth of creeping primrose was significantly affected
by heavy metal toxicity. The selectivity of heavy metals for the three plant species was the same (Hg>Fe>Cu>Zn). The mass balance preformed
on the system showed that about 60.45–82.61% of the zinc and 38.96–60.75% of the copper were removed by precipitation as zinc phosphate
and copper phosphate, respectively.
D 2003 Elsevier Ltd. All rights reserved.
Keywords: Phytoremediation; Heavy metals; Parrot feather; Creeping primrose; Water mint
1. Introduction
Pollution of air, soil, and water with heavy metals is a
major environmental problem (Srivastava and Purnima,
1998). Metals cannot be easily degraded and the cleanup
usually requires their removal (Lasat, 2002). However, this
energy intensive approach can be prohibitively expensive.
Phytoremediation offers a cost-effective, nonintrusive, and
safe alternative to conventional cleanup techniques. Utilizing
the ability of certain tree, shrub, and grass species to remove,
degrade, or immobilize harmful chemicals can reduce risk
from contaminated soil, sludges, sediments, and ground
water through contaminant removal, degradation, or contain-
ment (Zavoda et al., 2001). With global heavy metal con-
tamination on the rise, plants that can process heavy metals
might provide efficient and ecologically sound approaches to
sequestration and removal (Rugh et al., 1996; Lytle et al.,
1998; Srivastava and Purnima, 1998; Lasat, 2002).
The heavy metal ions Cu2 +
, Zn2 +
, Mn2 +
, Fe2 +
, Ni2 +
, and
Co2 +
are essential micronutrients for plants, with Fe2 +
being
required in the highest concentrations (Kunze et al., 2001).
However, when present in excess, all these metals are toxic,
as are the nonessential metals Cd2 +
, Hg2 +
, and Pb2 +
. Each
plant species has different tolerance levels to the different
contaminants. Tilstone and Macnair (1997) defined heavy
metal tolerance as the ability of plants to survive concen-
trations of metals in their environment that are toxic to other
plants.
Metal uptake by plant has three patterns: (a) true exclu-
sion in which metals are restricted from entering the plant,
(b) shoot exclusion in which metals are accumulated in the
root but translocation to the shoot is restricted, and (c)
accumulation where metals are concentrated in the plant
parts. Hyperaccumulators can tolerate, uptake, and trans-
locate high levels of certain heavy metals that would be toxic
to most organisms. They are defined as plants whose leaves
may contain >100 mg/kg of Cd, >1000 mg/kg of Ni and Cu,
or >10,000 mg/kg of Zn and Mn (dry weight) when grown in
metal-rich medium (Zavoda et al., 2001). Accumulation and
particularly hyperaccumulation have attracted considerable
interest in recent years. Ebbs et al. (1997) stated that in order
0160-4120/$ - see front matter D 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0160-4120(03)00091-6
* Corresponding author. Tel.: +1-902-494-6014; fax: +1-902-423-
2423.
E-mail address: Abdel.Ghaly@Dal.Ca (A.E. Ghaly).
www.elsevier.com/locate/envint
Environment International 29 (2004) 1029–1039

2. to achieve a successful phytoremediation of soil polluted
with metals, a strategy of combining a rapid screening of
plant species possessing the ability to tolerate and accumu-
late heavy metals with agronomic practices that enhance
shoot biomass production and increase metal bioavailability
in the rhizosphere must be adapted.
The objectives of this study were (a) to assess the ability
of three aquatic plants (parrot feather, creeping primrose, and
water mint) to tolerate water contaminated with four heavy
metals (Zn, Cu, Fe, and Hg,), (b) to determine the heavy
metal selectivity for each plant, and (c) to examine the plant
ability to hydroponically treat water contaminated with
heavy metals.
2. Experimental apparatus
The experimental setup shown in Fig. 1 consists of
holding tanks, lighting systems, aeration system, and temper-
ature-monitoring system.
2.1. Holding tanks
Three boxes constructed from a 2.5-cm-thick plywood
material were used in this experiment, one box for each
plant. Each box (60 120 80 cm) was divided into two
compartments: one compartment was used as a control while
the other one received the contaminants. Each compartment
(60 60 80 cm) was filled with 55 l of water.
2.2. Lighting system
The light intensity (625 hlx/7200 cm2
) provided by the
artificial lighting system (Fig. 1b) was similar to that of
natural lighting required for aquatic plants. Each lighting unit
consisted of eight light bulbs (six 34 W cool white fluores-
cent bulbs and two Gro-lux 40 W bulbs). Each light bulb was
122 cm in length. The lighting system was placed on the top
of each box using wooden supports in such a way that it gave
a space of 60 cm clearance between the light bulbs and the
water surface in the box. This space was chosen to achieve a
good air circulation and provide the heat and light required
for plant growth.
2.3. Aeration system
Aquatic plants grown in closed systems require oxygen to
survive. Lack of oxygen causes root system to die as the
anaerobic conditions give rise to root fungal diseases. There-
fore, an aeration unit was installed in the bottom of each
compartment to provide oxygen for aquatic plants. The air
flows from the main laboratory supply to a manifold with six
outlets. Each outlet was connected to a pressure regulator
(Model 129121/510, ARO, Bryan, OH), which was con-
nected to the aerator located in each compartment. Each
aerator consisted of a main tube (26.5 cm long) with three
perforated stainless steel laterals (30 cm in length and 0.6 cm
in diameter) coming off it at right angles to the main. Tygon
tubing of 0.75 cm outside diameter was used to connect the
main air supply, manifold and aeration unit.
2.4. Temperature-monitoring system
The water and the ambient air temperature were moni-
tored during the experimental period. Six thermocouples
(TW1–TW6) were used to measure the water temperature,
one for each compartment. Three thermometers (TA1–TA3)
were used to measure the ambient air temperature above the
water surface in each wooden box. The thermocouples were
connected to a thermoelectric device (Multimite, Thermo-
electric, Brampton, Ontario) for temperature monitoring as
shown in Fig. 1.
3. Experimental procedures
3.1. Selected aquatic plants
Three aquatic plant species (Fig. 2) were used in the
present study: water mint (Mentha aquatic), parrot feather
(Myriophylhum aquaticum), and creeping primrose (Ludwi-
igina palustris). The water mint has a square stem and
opposite leaves. The stems can reach 60–80 cm high and
produce small flowers in late summer and can be grown from
root divisions or stem cuttings. The parrot feather is a
feathery aquatic plant with stem that can grow up to 30 cm
above the water surface. The short leaves grow tight whorls
and are shades of bright green. The plant reproduces by
fragments breaking from the parent plant. The creeping
primrose is a creeping aquatic weed that grows along shore-
lines in shallow water. Its leaves are arranged oppositely
along the stem. They are oval-shaped to elliptic in outline
and are approximately 1–2 cm long. Its stem is creeping and
rooting at the nodes.
These aquatic plants were obtained from Bear River
Solar Aquastick wastewater treatment facility, Halifax,
Nova Scotia, where they were grown hydroponically. The
plants were translocated to the system and kept for 2 weeks
in clean water to acclimatize before adding the contami-
nants. During the adaptation period, the plants were sup-
plied with hydroponic fertilizer ‘‘10-6-16’’ (Plant Products,
Brampton, Ontario). The manufacturer recommended 1 ml
of fertilizer per 1 l of water. The composition of fertilizer
and the calculated amounts of various nutrient elements
added to the system (g/l) are shown in Table 1.
3.2. Heavy metal contaminant preparation
Four heavy metals were investigated in this study (Fe, Cu,
Zn, Hg). These contaminants were added as ferrous ammo-
nium sulfate [Fe(NH4)2(SO4)26H2O], cupric nitrate
(Cu(NO3)23H2O), zinc nitrate (Zn(NO3)26H2O), and mer-
M. Kamal et al. / Environment International 29 (2004) 1029–1039
1030

3. Fig.
1.
Experimental
setup:
(a)
holding
tanks;
(b)
lighting
system.
M. Kamal et al. / Environment International 29 (2004) 1029–1039 1031

4. curic sulfate (HgSO4). These were purchased as reagent
grade chemicals from Fisher Scientific, Ottawa, Ontario.
The concentrations of the individual heavy metals were
chosen to be within the range of the plant tolerance shown
in Table 2 as reported by Bridwell (1978). All of the ppm
calculations were based on the individual element versus
their compound form (i.e. Hg versus mercuric sulfate). The
reagents were dissolved in distilled water to achieve the
appropriate contamination level. After the adaptation period,
the contaminants were added to the three compartments
receiving treatment in the concentrations shown in Table 2.
The final concentrations of the nutrients and heavy metals
added (from all sources) are illustrated in Table 3.
3.3. Experimental protocol
One box (two compartment) was used for each plant
species. Each compartment received 55 l of water and the
recommended amount of plant fertilizer. About 300 g of
each translocated plant was placed in each of the two
compartments in the holding tank assigned for that plant.
The control in each box was used to observe differences in
growth rates among each plant due to uptake of contami-
nants. The lighting system was turned on and was controlled
with a timer, which was adjusted to achieve 16 h light and 8
h dark. The pressure regulator was adjusted at 0.136 atm
during the whole experimental period to give the aeration
rate of 15 cm3
/min, which is required for the plants’
survival. The plants were left for 2 weeks to adapt to the
new environment. After the acclimatization period, the
desired amounts of heavy metals were added to the treat-
ment compartment of each holding tank. Temperature read-
ings were taken every day in order to monitor the change in
the water and the ambient air temperatures during the
experimental period. The growth of the individual plant
was also observed on a daily basis. The entire experiment
was reported.
Fig. 2. Aquatic plants used in the experiment: (a) mint plant; (b) parrot
feather; (c) creeping primrose.
Table 1
Nutrient concentrations in fertilizer and aquatic system
Element Concentration in
fertilizer (%)*
Concentration in the
aquatic system (mg/l)
Potassium 16.0 144.0
Nitrogen 10.0 108.0
Calcium 7.2 78.0
Phosphorus 6.0 28.5
Sulfur 2.9 31.5
Magnesium 2.2 24.0
Iron 0.3 3.25
Manganese 0.1042 1.12
Copper 0.0052 0.056
Zinc 0.0052 0.056
*On weight basis.
Table 2
Nutrient concentration tolerance range for most plants and the amount
added to the system
Element Tolerance range (mg/l) Amount
Minimum Maximum added
Iron 2 200 100
Zinc 2 20 25
Copper 2 5 4.5
Mercury – 2 0.5
M. Kamal et al. / Environment International 29 (2004) 1029–1039
1032

5. 3.4. Sampling and analysis
Water samples of 250 ml were collected from all
compartments at 48-h intervals for heavy metals and
nitrogen analyses. An atomic absorption spectrometer
(Varion SpectrAA, Model Number: 55B, Varion, Mulgrave,
Victoria, Australia) was used for Cu, Zn, and Fe analyses.
A Mercury Analyzer System with Bacharach-Coleman,
cold vapor atomic absorption system (Model Number:
50B, Bacharach, New Kensington, PA), was used for Hg
analysis.
Plant samples (5 g each) were collected at the start and the
end of the experiment and analyzed for the presence of heavy
metals. The plant samples were dried in a convection oven
for 24 h at 45 jC to ascertain the accumulation of each
contaminant and avoid evaporation of Hg. After drying, the
plants were ground and digested with hydrochloric–nitric–
hydrofluric–perchloric acids (30 + 10 + 10 + 5 ml/g sample)
in a closed vessel at a temperature of 100 jC, and then the
heavy metal elements were determined by the same analyt-
ical methods used for water samples.
4. Results and discussion
The ambient air and water temperature remained constant
during the whole experimental period at 24 F 1 and 17 F 1
jC, respectively. The average concentrations of the various
heavy metals in the contaminated compartments are pre-
sented in Fig. 3. The averages of the final concentrations of
the heavy metals are presented in Fig. 4. The average
concentrations of the heavy metal in the whole plants at
the end of the experiment are presented in Fig. 5. The values
are the average of two replicates. The coefficients of varia-
tion were very small (ranged from 3% to 10.5%).
Table 3
Concentrations of elements (mg/l) present in the contaminated water
Element Concentration
in water
Fertilizer
added
Reagents
added
Total
Nutrients
Potassium – 144.00 – 144.00
Nitrogen – 108.00 59.60 167.60
Calcium – 78.00 – 78.00
Sulfur – 31.50 117.16 148.66
Phosphorus – 28.50 – 28.50
Magnesium – 24.00 – 24.00
Manganese – 1.12 – 1.12
Heavy metals
Zinc 3.0 0.06 25.00 28.06
Copper 1.0 0.06 4.50 5.56
Iron 0.3 3.25 100.00 103.55
Mercury 0.001 0.00 0.50 0.501
Fig. 3. Heavy metal concentrations in the contaminated water: (a) zinc; (b) copper; (c) iron; (d) mercury.
M. Kamal et al. / Environment International 29 (2004) 1029–1039 1033

6. 4.1. Zinc
The initial concentration of zinc in the control tanks was
3.56 mg/l, which decreased by the end of the experiment to
0.46 mg/l for parrot feather, 0.4 mg/l for creeping primrose,
and 0.21 mg/l for water mint, whereas the initial concen-
tration of zinc in the contaminated tanks was 28.06 mg/l,
which was reduced to 18.4, 18.9, and 18.3 mg/l by the end of
the experiment for parrot feather, creeping primrose, and
water mint, respectively. Zinc removal in the three contami-
nated compartments is almost identical and can be described
by the following equation:
YZ ¼ 28:06 0:55747X þ 0:00467X2
ðR2
¼ 1Þ ð1Þ
where YZ is the zinc concentration in the water (ppm) and X
is the number of days from start of the experiment.
The measured concentrations of zinc in plants obtained
from the control tanks were 291, 68, and 209 mg/kg (dry
basis) for parrot feather, creeping primrose, and water mint,
whereas the measured zinc concentration in the plants
obtained from the contaminated tanks were 549, 1243, and
1498 mg/kg (dry basis) for parrot feather, water mint, and
creeping primrose, respectively. The removal rate of zinc
from the contaminated water was approximately 0.455 mg/l/
day for all the plants.
Samecka-Cymerman and Kempers (1996) used Scapania
undulata that has an initial zinc concentration in its tissues of
37.7 mg/kg (dry basis) to remove zinc from sewage collected
from pesticide-producing factory having zinc concentration
of 248 Ag/l. The zinc concentration in the plant tissues
reached 181 mg/kg (dry basis) after 14 days.
Ebbs and Kochain (1997) tested five different spices of
Brassica for their ability to accumulate zinc in their tissues.
These included Brassica juncea (RH-30), B. juncea (acces-
sion 426308), B. juncea (accession 184290), Brassica rapa
(parkland), and Brassica napus (accession 535855) were
tested. The plants were supplied with zinc sulfate at an initial
zinc concentration of 6.5 mg/l zinc concentrations in the plant
tissues after 14 days were 750, 1375, 1400, 2100, and 2600
mg/kg (dry basis) for RH-30, accession 426308, accession
184290, Parkland, and accession 535855, respectively.
Jain et al. (1990) used water velvet (Azolla pinnata R.Br.)
and duckweed (Lemna minor L.) for removal of zinc from
polluted water. The plants were exposed to a series of
concentrations of zinc (1.0, 2.0, 4.0, and 8.0 mg/l) for 14
days, supplied as zinc nitrate [Zn(NO3)26H2O]. Zinc con-
centrations in the water velvet tissues at the end of the
experiment were 831, 1480, 2647, and 4316 mg/kg (dry
basis), whereas zinc concentrations in the duckweed tissues
at the end of the experiment were 717, 1284, 2277, and 3698
mg/kg (dry basis) for the 1.0, 2.0, 4.0, and 8.0 mg Zn/l
concentrations, respectively.
Hinchman et al. (1998) investigated the removal of zinc by
eastern gama grass grown in inert quartz sand in lysimeter
pots with continuous addition of zinc ion in the nutrient
Fig. 4. Final heavy metal concentrations in the water: (a) zinc; (b) copper; (c) iron; (d) mercury.
M. Kamal et al. / Environment International 29 (2004) 1029–1039
1034

7. solution at concentrations of 160, 600 mg Zn/kg soil, over a
period of 60 days. Leachate analyses for zinc by atomic ab-
sorption spectrophotometry indicate that plants subjected to
both levels of zinc were removing up to 70% of the zinc from
the leachate. The authors used the same experimental setup
with Hybrid Poplar with zinc doses ranging from 50 to 2000
mg Zn/kg soil over a period of 60 days. Their results showed
that up to 800 Ag Zn/g soil, Zn added in nutrient solution,
were selectively absorbed and sequestered by the plants. At
levels of zinc above 1000 mg Zn/kg soil, the zinc level in
leachate was always below 100 mg Zn/kg soil. Leaf analyses
showed 528 mg Zn/kg in mature (large) leaves, 300 mg Zn/kg
in medium size leaves, and 140 mg Zn/kg in small leaves.
The results obtained from the present study indicated that
the three plants used were superior to S. undulata used by
Samecka-Cymerman and Kempers (1996) but inferior to the
water velvet and duckweed reported by Jain et al. (1990) and
the Brassica species reported by Ebbs and Kochain (1997).
4.2. Copper
The initial copper concentration in the control tanks was
1.056, which decreased by the end of the experiment to 0.06
mg/l for parrot feather, 0.12 mg/l for creeping primrose, and
0.1 mg/l for water mint, whereas the initial copper concen-
tration in the contaminated compartments was 5.56 mg/l,
which was reduced by the end of the experiment to 3.19,
3.06, and 3.48 mg/l for parrot feather, creeping primrose,
and water mint, respectively. The three plants showed
different removal patterns of copper from the water. The
copper removal from the three contaminated tanks can be
described by the following equations:
For parrot feather
YC ¼ 5:56 0:165X þ 0:0024X2
ðR2
¼ 0:999Þ ð2Þ
For creeping primrose
YC ¼ 5:56 0:02103X þ 0:0044X2
ðR2
¼ 0:999Þ ð3Þ
For water mint
YC ¼ 5:56 0:138X þ 0:0019X2
ðR2
¼ 0:970Þ ð4Þ
where YC is the copper concentration in the water (ppm).
The measured concentrations of copper in plants
obtained from the control tanks were 18, 25, and 11 mg/
kg (dry basis) for parrot feather, creeping primrose, and
Fig. 5. Final heavy metals concentrations in the plants: (a) zinc; (b) copper; (c) iron; (d) mercury.
M. Kamal et al. / Environment International 29 (2004) 1029–1039 1035

8. water mint, whereas the measured copper concentrations in
the plants obtained from the contaminated tanks were 304,
848, and 314 mg/kg (dry basis) for parrot feather, creeping
primrose, and water mint, respectively. The average removal
rate of copper was 0.16, 0.21, and 0.14 mg/l/day for the
contaminated tanks containing parrot feather, creeping prim-
rose, and water mint, respectively.
Qian et al. (1999) used 12 different plants [fuzzy water
clover (Marsilea drummondii), iris-leaved rush (Juncus
xiphioides E. Mey.), mare tail (Hippuris vulgaris L.),
monkeyflower (Mimulus guttatus Fisch.), parrot feather
(Myriophyllum brasiliense Camb.), sedge (Cyperus pseudo-
vegetus), smart weed (Polygonum hydropiperoides L.),
smooth cordgrass (Spartina altemiflora Loisel), striped rush
(Baumia rubiginosa), umbrella plant (Cyperus altemifolius
L.), water lettuce (Pistia stratiotes L.), water zinnia (Wedelia
trilobata Hitchc.)] grown hydroponically in greenhouse for
treating copper-contaminated water. The plants were sup-
plied with 1 mg Cu/l as copper sulfate for 10 days and the
entire hydroponic system was maintained under controlled
conditions with a 16-h daylight and a constant temperature
of 25 F 2 jC. The chemical analyses showed that the copper
concentrations in the shoot of the 12 plants were approx-
imately 95, 75, 65, 60, 50, 40, 40, 35, 25, 25, 15, and 15 mg/
kg (dry basis), whereas the concentrations of copper in the
roots were approximately 1150, 650, 600, 550, 450, 450,
400, 350, 310, 310, 300, and 200 mg/kg (dry basis) for
fuzzy water clover, iris-leaved rush, mare tail, monkey-
flower, parrot feather, sedge, smart weed, smooth cordgrass,
striped rush, umbrella plant, water lettuce, and water zinnia,
respectively.
Zhu et al. (1999) used water hyacinth (Eichhornia cras-
sipes) grown hydroponically for treating copper-contami-
nated water with initial concentration of 10 mg/l for 14 days
supplied as copper sulfate. The plants were maintained under
controlled conditions with a 16-h daylight and a temperature
of 25–28 jC. The copper concentrations were 130 mg/kg
(dry basis) in the plants’ shoots and 2800 mg/kg (dry basis)
in the roots.
Zayed et al. (1998) used duckweed (L. minor L.) for the
treatment of copper-contaminated water with different con-
centrations (0.1, 0.2, 0.5, 1.0, 2.0, 5.0, and 10 mg/l) supplied
as copper sulfate for 8 days. The copper concentrations in the
plants were approximately 100, 150, 250, 500, 375, 1000,
and 3400 mg/kg (dry basis) for the 0.1, 0.2, 0.5, 1.0, 2.0, 5.0,
and 10 mg Cu/l concentrations, respectively.
Samecka-Cymerman and Kempers (1996) used S. undu-
lata that has an initial copper concentration in its tissues of
2.33 mg/kg (dry weight) to remove copper from sewage
collected from pesticide-producing factory having copper
concentration of 31 Ag/l. The copper concentration in the
plant tissues reached 14.9 mg/kg (dry weight) after 14 days.
It appears from the results obtained from the present
study that the three plants used were superior accumulators
of copper to the plants used by Qian et al. (1999) and
Samecka-Cymerman and Kempers (1996). However, the
water hyacinth reported by Zhu et al. (1999) showed better
performance.
4.3. Iron
The initial iron concentration in the control tanks was
3.55, which decreased by the end of the experiment to 0.60
mg/l for parrot feather, 1.46 mg/l for creeping primrose, and
1.18 mg/l for water mint, whereas the initial iron concen-
tration in the contaminated tanks was 103.55 mg/l, which
was reduced by the end of the experiment to 27.9, 37.80, and
7.30 mg/l for parrot feather, creeping primrose, and water
mint, respectively. Although the three plants showed differ-
ent removal rates, the removal rate for each plant decreased
with time in all contaminated tanks. The removal of iron
from the contaminated water can be described by the
following equations:
For parrot feather
YI ¼ 103:55 7:3529X þ 0:1788X2
ðR2
¼ 0:999Þ ð5Þ
For creeping primrose
YI ¼ 103:55 6:0713X þ 0:1401X2
ðR2
¼ 1Þ ð6Þ
For water mint
YI ¼ 103:55 10:525X þ 0:2851X2
ðR2
¼ 0:999Þ ð7Þ
where YI is the iron concentration in the water (ppm).
The measured concentrations of iron in plants obtained
from the control tanks were 914, 1480, and 1840 mg/kg (dry
basis) for parrot feather, creeping primrose, and water mint,
whereas the measured iron concentration in the plants
obtained from the contaminated compartments were
38,800, 46,300, and 32,100 mg/kg (dry basis) for parrot
feather, creeping primrose, and water mint, respectively. The
removal rate of iron from the contaminated water ranged
from 7.00 to 0.41 mg/l/day for the parrot feather tank, from
5.79 to 0.62 mg/l/day for the creeping primrose tank, and
from 9.96 to 0.25 mg/l/day for the water mint tank, depend-
ing on the concentration.
Jain et al. (1988) used duckweed (L. minor L.) for the
treatment of iron-contaminated water with different concen-
trations (2, 6, and 10 mg/l) supplied as iron nitrate for 24 h.
The iron concentrations in the plants were approximately
259.5, 328.4, and 463.2 mg/kg (dry basis) for the 2, 6, and 10
mg Fe/l concentrations, respectively.
Jain et al. (1989) used duckweed (L. minor L.) and water
velvet (A. pinnata R.Br.) for the treatment of iron-contami-
nated water with different concentrations of iron (1.0, 2.0,
4.0, and 8.0 mg/l) supplied as [NH4Fe(SO4)12H2O] for 14
days. The iron concentrations in the duckweed were approx-
imately 1221, 2308, 4268, and 6826 mg/kg (dry basis),
whereas iron concentrations in the water velvet were approx-
imately 1363, 2638, 5135, and 9676 mg/kg (dry basis) for
1.0, 2.0, 4.0, and 8.0 mg Fe/l, respectively.
The results obtained from the present study indicated that
the three plants used were superior to the duckweed (L.
M. Kamal et al. / Environment International 29 (2004) 1029–1039
1036

9. minor L.) used by Jain et al. (1988, 1989) and water velvet
(A. pinnata R.Br.) used by Jain et al. (1989). It appears also
from the results presented in the literature that plant accu-
mulation of iron depends on the type of salt added and the
initial concentration.
4.4. Mercury
The initial concentration of mercury in the control tanks
was 1.00 Ag/l, which decreased by the end of the experiment
to 0.05 Ag/l for parrot feather, 0.04 Ag/l for creeping
primrose, and 0.6 Ag/l for water mint, whereas the initial
concentration of mercury in the contaminated tanks was
501.00 Ag/l, which was reduced by the end of the experiment
to 0.15, 1.3, and 0.02 Ag/l for parrot feather, creeping
primrose, and water mint, respectively. Mercury removal
from the three contaminated tanks (by the three plants) is
almost identical and can be described by the following
equation:
YM ¼ 501 88:211X þ 5:113X2
0:0976X3
ðR2
¼ 0:99Þ ð8Þ
where YM is mercury concentration in the water (ppm).
The measured concentrations of mercury in plants
obtained from the control tanks were 1.01, 0.83, and
0.92 mg/kg (dry basis) for parrot feather, creeping prim-
rose, and water mint, whereas the measured mercury
concentration in the plants obtained from the contaminated
tank were 208, 335, and 179 mg/kg (dry basis) for parrot
feather, water mint, and creeping primrose, respectively.
The removal rate of mercury from the contaminated water
started at 0.07807 and decreased with time reaching
0.00002 mg/l/day by the end of the experiment for all
contaminated tanks.
Samecka-Cymerman and Kempers (1996) used S. undu-
lata that has an initial mercury concentration in its tissues of
0.05 mg/kg (dry weight) to remove mercury from sewage
collected from pesticide-producing factory having a mer-
cury concentration of 5 Ag/l. The mercury concentration in
the plant tissues reached 2.4 mg/kg (dry weight) after 14
days.
Qian et al. (1999) used 12 different plants [fuzzy water
clover (M. drummondii), iris-leaved rush (J. xiphioides E.
Mey.), mare tail (H. vulgaris L.), monkeyflower (M. guttatus
Fisch.), parrot feather (M. brasiliense Camb.), sedge (C.
pseudovegetus), smart weed (P. hydropiperoides L.), smooth
cordgrass (S. altemiflora Loisel), striped rush (B. rubigi-
nosa), umbrella plant (C. altemifolius L.), water lettuce (P.
stratiotes L.), water zinnia (W. trilobata Hitchc.)] grown
hydroponically in greenhouse for treating mercury-contami-
nated water. Plants were supplied with 1 mg Hg/l for 10 days
as mercuric chloride, the entire hydroponic was maintained
under controlled conditions with a 16-h daylight and a day
and night temperature of 25 F 2 jC. Mercury concentrations
in the shoot of the 12 plants were approximately 90, 60, 40,
35, 30, 27, 25, 20, 15, 10, 5, and 5 mg/kg (dry basis) for
fuzzy water clover, iris-leaved rush, mare tail, monkey-
flower, parrot feather, sedge, smart weed, smooth cordgrass,
striped rush, umbrella plant, water lettuce, and water zinnia,
respectively.
The results of this study showed that the three plants used
were superior to the S. undulata used by Samecka-Cymer-
man and Kempers (1996) and all plants used by QiAn et al.
(1999).
4.5. Removal efficiency and element selectivity
The removal efficiency of each plant for the different
heavy metal ions is shown in Table 4. The removal
efficiency of for zinc was 34.42%, 32.63%, and 34.77%,
for copper was 42.58%, 44.92%, and 30.89%, for iron was
73.06%, 63.68%, and 92.92%, and for mercury was
99.97%, 99.74%, and 99.99% for parrot feather, creeping
primrose, and water mint, respectively. Therefore, the ion
selectivity for the three cultivars was HgFeCuZn.
4.6. Mass balance
A mass balance was performed on the system in order to
determine the elements’ removal pathways. The results are
shown in Table 5.
The amounts of zinc removed from water were much
higher than that utilized by plants in both the control and
treatment tanks. Zinc may have been removed from water
through a chemical pathway that involved the formation
and precipitation of Zn3(PO4)2. Phosphorus added to the
water with the plant nutrient in the form of P2O5 may
have reacted with zinc nitrate [Zn(NO3)2] to form insolu-
ble zinc phosphate [Zn3(PO4)2]. The following equation
Table 4
Heavy metal removal efficiencies
Element Initial Parrot feather Creeping primrose Water Mint
concentration
Concentration (mg/l) Efficiency Concentration (mg/l) Efficiency Concentration (mg/l) Efficiency
(mg/l)
Final Removal
(%)
Final Removal
(%)
Final Removal
(%)
Zn 28.056 18.4 9.656 34.42 18.9 9.156 32.63 18.3 9.756 34.77
Cu 5.556 3.19 2.366 42.58 3.06 2.496 44.92 3.48 1.716 30.89
Fe 103.55 27.9 75.65 73.06 37.8 65.75 63.68 7.3 96.25 92.92
Hg 0.501 0.00015 0.50085 99.97 0.0013 0.4997 99.74 0.00004 0.50096 99.99
M. Kamal et al. / Environment International 29 (2004) 1029–1039 1037

10. describes the chemical reaction that may have taken place
in water.
3½ZnðNO3Þ2 6H2O þ P2O5 þ 0:5O2
! Zn3ðPO4Þ2 þ 18H2O þ 6NO
3 ð9Þ
On the average, about 17–40% of zinc removed from
water was utilized by plants depending on the plant type. The
other portion (60–83%) may have precipitated as zinc
phosphate [Zn3(PO4)2]. The results also showed that plant
uptake of zinc was dependent on the initial concentration in
water.
The amounts of copper removed from water were also
higher than that utilized by the plants in both control and
contaminated tanks. Copper may have been removed from
water through the formation and precipitation of Cu3(PO4)2.
Phosphorus added to the water as plant nutrient in the form
of P2O5 may have reacted with copper nitrate [Cu(NO3)2] to
form copper phosphate [Cu3(PO4)2]. The following equation
describes the chemical reaction that may have taken place in
the water.
3½CuðNO3Þ2 3H2O þ P2O5 þ 1:5O2
! Cu3ðPO4Þ2 3H2O þ 6H2O þ 6NO
3 ð10Þ
On the average, about 39–60% of copper removed from
water was utilized by plants depending on the plant type.
The remaining portion (40–61%) was precipitated as copper
phosphate [Cu3(PO4)2]. The plant uptake of copper and/or
precipitation of copper phosphate depend on the initial
concentration of copper in water.
The amounts of iron removed from water in the con-
taminated tanks by the plants ranged from 94.36% to
99.35%, whereas that removed from the water in the control
tanks ranged from 94.42% to 99.32% depending on the type
of plant. The amounts of mercury removed from water in the
contaminated tanks by the plants ranged from 90.05% to
93.36%, whereas that removed from the water in the control
Table 5
Mass balance
Element Treatment Path Parrot feather Creeping primrose Water mint
(mg) (%) (mg) (%) (mg) (%)
Zinc
Control
Total 130 100.00 132 100.00 142 100.00
Plant uptake 35 26.92 6 4.55 23 16.19
Precipitation 95 73.08 126 95.45 119 83.81
Contaminated
Total 483 100.00 458 100.00 488 100.00
Plant uptake 84 17.39 134 29.26 193 39.55
Precipitation 399 82.61 324 70.74 295 60.45
Copper
Control
Total 49.5 100.00 46.8 100.00 47.80 100.00
Plant uptake 2.0 4.04 2.8 5.98 0.87 1.82
Precipitation 47.8 95.96 44.0 94.02 46.93 98.18
Contaminated
Total 118.5 100.00 125.0 100.00 104.00 100.00
Plant uptake 46.5 39.24 76.3 61.04 52.90 50.86
Precipitation 72.0 60.75 48.7 38.96 56.20 49.13
Iron
Control
Total 147.5 100.00 116.5 100.00 118.5 100.00
Plant uptake 146.5 99.32 110.0 94.42 112.6 95.02
Precipitation 1.0 0.68 6.5 5.58 5.9 4.98
Contaminated
Total 3812.3 100.00 3287.5 100.00 4975.5 100.00
Plant uptake 3787.5 99.35 3102.0 94.36 4812.5 96.72
Precipitation 24.8 0.65 185.5 5.64 163.0 3.28
Mercury
Control
Total 0.048 100.00 0.048 100.00 0.073 100.00
Plant uptake 0.045 93.75 0.047 97.92 0.020 72.60
Precipitation 0.003 6.25 0.001 2.08 0.053 27.40
Contaminated
Total 26.82 100.00 27.09 100.00 27.75 100.00
Plant uptake 25.04 93.36 24.99 92.25 24.99 90.05
Precipitation 1.78 6.64 2.10 7.75 2.76 9.95
M. Kamal et al. / Environment International 29 (2004) 1029–1039
1038

11. tanks ranged from 72.60% to 97.92% depending on the type
of plant. Parrot feather showed the highest removal effi-
ciency for iron and mercury.
4.7. Plant tolerance
Generally, all the experimental plants showed a slight
reduction in the plant growth, branching, leaf size, and root
system. Parrot feather shows promising results in terms of
tolerance to the heavy metals concentration in the aqueous
solution. A dark green color of the parrot feather leaves
grown in the contaminated compartment was apparent. The
creeping primrose showed the least tolerance to toxicity
since the plants begin to show abnormal darkening of leaves,
stems, and roots. By the end of the experiment, almost all the
creeping plants were dead. Therefore, the experimental
plants could be ordered as follows in terms of heavy metals
tolerance: parrot featherwater mintcreeping primrose.
5. Conclusion
The three aquatic plants investigated (parrot feather, water
mint, and creeping primrose) have the ability to remove
heavy metals (Zn, Cu, Fe, and Hg) from contaminated water.
The removal efficiency was 99.8%, 76.7%, 41.62%, and
33.9% for Hg, Fe, Cu, and Zn, respectively. The removal
rates for Zn and Cu were constant (0.48 mg/l/day for Zn and
0.11 mg/l/day for Cu), whereas those for Fe and Hg were
dependent on their concentrations in the contaminated water
and ranged from 7 to 0.41 mg/l/day for Fe and 0. 0787 to
0.0002 mg/l/day for Hg. Creeping primrose has the least
tolerance to heavy metal toxicity whereas parrot feather has
the greatest tolerance. The selectivity of heavy metals by all
plants was FeHgCuZn. The three plant species could be
considered hyperaccumulators for Fe ions. There appears to
be a chemical pathway for removal of zinc and copper.
About 60.45–82.61% of the Zn and 49.13–60.75% of the
Cu were removed by precipitation as zinc phosphate and
copper phosphate, respectively.
Acknowledgements
The authors wish to thank Mr. Jack Vissers for his
valuable technical assistance and Mr. John Pyke for his help
with the chemical analyses. Special thanks go to Mr. Mark
Vanzummer for donating the plants used in this study. The
project was funded by Environment Canada and EJLP
foundation.
References
Bridwell R. Hydroponic Gardening. Santa Barbara, California: Wood-
bridge Press; 1978.
Ebbs SD, Kochain LV. Toxicity of zinc and copper to Brassica species:
implication for phytoremediation. J Environ Qual 1997;26:776–81.
Ebbs SD, Lasat MM, Brady DJ, Cornish J, Gordon R, Kochain LV. Phy-
toextraction of cadmium and zinc from a contaminated soil. J Environ
Qual 1997;26(5):1424–30.
Hinchman RR, Nergi MC, Gatliff EG. Phytoremediation: using green plants
to clean up contaminated soil, ground water and wastewater. 1998. http://
www.treemediation.com/technical/phytoremediation_1998.pdf.
Jain SK, Gujral GS, Jha NK, Vasudevan P. Heavy metals uptake by Pleuro-
tus sajor-caju from metal-enriched Duckweed substrate. Biol Wastes
1988;24:275–82.
Jain SK, Vasudevan P, Jha NK. Removal of some heavy metals from
polluted water by aquatic plants: studies on duckweed and water velvet.
Biol Wastes 1989;28:115–26.
Jain SK, Vasudevan P, Jha NK. Azolla pinnata R.Br. and Lemna minor L. for
removal of lead and zinc from polluted water. Water Res 1990;24(2):
177–83.
Kunze R, Frommer WB, Flügge UI. Metabolic engineering in plants: the
role of membrane transport. Metab Eng 2001;4:57–66.
Lasat MM. Phytoextraction of toxic metals: a review of biological mecha-
nisms. J Environ Qual 2002;31(1):109–20.
Lytle CM, Lytle FW, Yang N, Qian JH, Hansen D, Zayed A, Terry N.
Reduction of Cr (VI) to Cr (III) by wetland plants: potential for in-situ
heavy metal detoxification. Environ Sci Technol 1998;32:3087–93.
Qian JH, Zayed A, Zhu ML, Yu M, Terry N. Phytoaccumulation of trace
elements by wetland plants: III. Uptake and accumulation of ten trace
elements by twelve plant species. J Environ Qual 1999;28(5):1448–56.
Rugh CL, Wilde HD, Stack NM, Thompson DM, Summers AO, Meagher
RB. Mercuric ion reduction and resistance in transgenic Arabidopsis
thaliana plants expressing a modified bacterial merAgene. Proc Natl
Acad Sci U S A 1996;93(8):3182–8187.
Samecka-Cymerman RB, Kempers AJ. Bioaccumulation of heavy metals
by aquatic macrophytes around Wroclaw, Poland. Ecotoxicol Environ
Saf 1996;35:242–7.
Srivastava AK, Purnima X. Phytoremediation for heavy metals—a land
plant based sustainable strategy for environmental decontamination.
Proc Natl Acad Sci, India, Sect B Biol Sci 1998;68(3–5):199–215.
Tilstone GH, Macnair MR. The consequence of selection for copper toler-
ance on the uptake and accumulation of copper in Mimulus guttatus.
Ann Bot 1997;80:747–51.
Zavoda J, Cutright T, Szpak J, Fallon E. Uptake, selectivity, and inhibition
of hydroponics treatment of contaminants. J Environ Eng 2001;127(6):
502–8.
Zayed AS, Gwthaman S, Terry N. Phytoaccumulation of trace elements by
wetland plants: I. Duckweed. J Environ Qual 1998;27(3):715–21.
Zhu YL, Zayed AM, Qian J-H, de Souza M, Terry N. Phytoaccumulation of
trace elements by wetland plants: II. Water hyacinth. J Environ Qual
1999;28(1):339–44.
M. Kamal et al. / Environment International 29 (2004) 1029–1039 1039