Full text of DOI:10.1002/etc.5620211224

Environmental Toxicology and Chemistry, Vol. 21, No. 12, pp. 2698–2704, 2002
2002 SETAC
Printed in the USA
730-7268/02 $9.00 .00
᭧
0
ϩ
A COMPARISON OF CHELATOR-FACILITATED METAL UPTAKE BY A HALOPHYTE
AND A GLYCOPHYTE
F
IONA L. JORDAN,*† MOLLY
Department of Environmental Science, University of California, Geology 2217, Riverside, California 92521, USA
Environmental Research Laboratory, University of Arizona, 2601 North Airport Drive, Tucson, Arizona 85706, USA
Department of Soil, Water, and Environmental Science, 429 Shantz Building, University of Arizona, Tucson, Arizona 85721, USA
ROBIN-ABBOTT,‡ RAINA M. MAIER,§ and EDWARD P. G LENN‡
†
‡
§
(
Received 30 October 2001; Accepted 28 May 2002)
Abstract—Phytoextraction is the use of plants to remove contaminants, in particular metals, from soil via root uptake and trans-
location to the shoots. Efficient phytoextraction requires high-biomass plants with efficient translocating properties. Halophytes
characteristically accumulate large quantities of salts in aboveground tissue material and can have high biomass production. It has
been speculated that salt-tolerant plants may also be heavy metal tolerant and, further, may be able to accumulate metals. This
study compared growth and metal uptake by a halophyte, Atriplex nummularia, and a common glycophyte, Zea mays, in a mine-
tailing contaminated soil:mulch mixture. Two chelators, ethylenediaminetetraacetic acid (EDTA) and rhamnolipid, were used to
facilitate plant metal uptake. Despite a lower growth rate (2% growth/d) in the contaminated soil, the halophyte accumulated roughly
the same amount of metals as the glycophyte on a mass basis (30–40 mg/kg dry wt). Neither plant, however, hyperaccumulated
any of the metals tested. When treated with EDTA, specific differences in patterns of metal uptake between the two plants emerged.
The halophyte accumulated significantly more Cu (2ϫ) and Pb (1ϫ) in the shoots than the glycophyte, but root metal concentrations
were generally higher for the glycophyte, indicating that the halophyte translocated more metal from the root to the shoot than the
glycophyte. For example, Zn shoot-to-root ratios ranged from 1.4 to 2.1 for Atriplex and from 0.5 to 0.6 for Z. mays. The biodegradable
chelator rhamnolipid was not effective at enhancing shoot metal concentrations, even though radiolabeled chelator was found in
the shoot material of both plants. Our results suggest that halophytes, despite their slower growth rates, may have greater potential
to selectively phytoextract metals from contaminated soils than glycophytes.
Keywords—Halophytes
Heavy metals
Translocation
Chelator
Phytoextraction
INTRODUCTION
taxa, as having characteristically high shoot Cs concentrations.
Similarly, Williams et al. [12] found Zn concentrations in the
roots and shoots of some halophytes to reflect sedimentary
levels found in a polluted salt marsh. However, relatively little
is known about metal tolerance and accumulation in halo-
phytes.
Another requirement for efficient phytoextraction is that
the metal contaminants be in a form available for plant uptake.
This is generally not the case for weathered metal-contami-
nated soils, where the majority of the metal has become pre-
cipitated or tightly bound. Chelator applications significantly
increase plant-available metal in most metal-contaminated
soils [1–5,13]. Blaylock et al. [1] report that the addition of
One approach to phytoremediation of metal-contaminated
sites is removal of metal by phytoextraction. Phytoextraction
requires efficient transfer of metals from the roots into the
shoots to maximize metal removal by cropping of plant ma-
terials. It also requires high biomass yields. For example, the
Indian mustard, Brassica juncea, has been reported to mod-
erately accumulate a variety of metals (Pb, 3,600 mg/kg; Cu,
9
70 mg/kg; and Zn, 410 mg/kg) from contaminated soils with
reasonable biomass yields of 7.2 t/ha/month [1]. Most plants
utilized for metal accumulation are well-characterized crop
plants, including sunflower (Heliantus annus), corn (Zea
mays), pea (Pisum sitivum), and mustard (B. juncea) [2–5].
Interestingly, these plants are all classified as glycophytes or
plants lacking salt-tolerance mechanisms.
Halophytes, such as the four-wing saltbush, Atriplex num-
mularia, are plants that tolerate saline soils and accumulate
large quantities of salts in the aboveground tissue [6]. These
plants can achieve biomass yields of 20 to 30 t/ha and have
been shown to accumulate up to 40% NaCl in the biomass
1
g EDTA/kg soil resulted in a 10-fold increase in metal uptake
by B. juncea. Ethylenediaminetetraacetic acid and other syn-
thetic chelators, however, can be toxic and nonbiodegradable,
limiting their usefulness for continued soil applications [14–
1
8].
The objective of this study was to compare root and shoot
metal accumulation from an aged contaminated soil by a hal-
ophyte (A. nummularia) and a glycophyte (Z. mays) in the
presence and absence of metal chelators. Two chelators were
used, a synthetic chelator (EDTA) and a microbially produced
chelator (rhamnolipid). Rhamnolipid is produced by Pseu-
domonas aeruginosa and has been shown to chelate metals
such as Pb and Cd in both soil and solution [14–20].
[
6]. A few recent studies have provided evidence suggesting
that halophytes may be useful for phytoremediation, particu-
larly for salty soils [6–12]. For example, Thomas et al. [7]
demonstrated that the halophyte Mesembryanthemum crys-
tallinum regulates Cu uptake more effectively than Arabidop-
sis thaliana. In another study, Broadley et al. [10] identified
the family Caryophyllidea, containing a number of halophyte
MATERIALS AND METHODS
Soils
The contaminated and uncontaminated soils used in this
study were collected from an industrial site in Tucson (AZ,
*
To whom correspondence may be addressed
(
fiona@mail.ucr.edu).
2
698

Comparing metal uptake between a halophyte and a glycophyte
Environ. Toxicol. Chem. 21, 2002
2699
Table 1. Soil physical and chemical properties including total and
selected toxicity characteristic leaching procedure (TCLP) soil metal
concentrations
soil conditioner (El Toro Forest Mulch, Triple A Fertilizer,
Tucson, AZ, USA). The soil mixtures were used for all ex-
periments unless noted otherwise.
Unconta-
Contaminated minated
Soil
Mixturea
Chelators
Both EDTA and 14C-EDTA (specific activity
ϭ 0.69 ␮Ci/
Analytes
Total (mg/kg soil)
mmol) were purchased from Sigma Scientific (St. Louis, MO,
USA). Rhamnolipid, a mixture of mono- and dirhamnolipid
forms, was generously donated by Jeneil Biosurfactant (Sauk-
ville, WI, USA). The 14C-monorhamnolipid (specific activity
Lead
Zinc
2,000
2,500
21,000
150,000
170
50
60
280
1,000
1,300
10,000
75,000
—
Copper
Iron
Nitrate
Sulfate
Sodium
17,000
170
ϭ
0.435 ␮Ci/mmol) was produced and purified as described
6,500
680
890
170
—
—
by Maslin and Maier [20].
General characteristics
Plant tissue metal analysis
pH
7.9
1.4
3.2
8.5
0.9
—
—
—
—
Plant samples analyzed for metal content were oven dried
65ЊC) for 2 d, ground in a stainless-steel Wiley Mill (A.H.
Thomas, Philadelphia, PA, USA), and passed through a 20-
mesh screen. Five-hundred-milligram samples were ashed in
Salinity (dS/m)
Conductivity (mmhos/cm)
(
TCLP (mg/L extract solution)
Lead
Zinc
Copper
12.5 (
26.7 (
33.3 (
Ϯ
Ϯ
Ϯ
0.3)
7)
NDb
ND
ND
a furnace at 500
ЊC for 5 h. The ashed samples were digested
10.0 (
17.3 (
Ϯ
Ϯ
0)
4)
7)
0 (
Ϯ
0)
at 150 C in 3-M hydrochloric acid (HCl) for 3 min and brought
Њ
to 10-ml volume with ddH O. Samples were filtered through
2
a
The mixture soil is a 1:1 (mass basis) mixture of contaminated soil
and El Toro mulch. Metal values reported for the mixture soil are
estimates based on the contaminated soil. The mulch used to sup-
plement the soil had minimal TCLP levels of Pb, Zn, or Cu (data
not shown).
Whatman 42 paper (Clifton, NJ, USA) prior to analysis by
AA. All glassware was acid washed in 10% nitric acid and
rinsed thoroughly with double-distilled water prior to exper-
imentation. Standards and sample blanks subjected to the same
procedure were verified regularly.
b
ND
ϭ not determined.
Transplant preparation
USA). The contaminated soil is a mine-tailing waste disposed
of approximately 30 years ago. The uncontaminated soil was
collected from a nearby site containing no tailings and sup-
porting plant growth. Composite samples of contaminated and
uncontaminated soil were collected and passed through a 20-
mm sieve, and a 100-g subsample of composite was sent to
an U.S. Environmental Protection Agency (U.S. EPA)–ap-
proved laboratory for total (hard and soft) metals, pH, con-
ductivity, and salinity (Table 1). These data were certified with
the appropriate quality assurance/quality control protocols
The two plants chosen for this study were Z. mays and A.
nummularia
because it is commonly used in phytoremediation studies [2–
]. Atriplex nummularia was chosen as the model halophyte
. Zea mays was chosen as the model glycophyte
5
because it is a native desert plant and has frequently been
observed growing on disturbed sites. In addition, both plant
varieties chosen were drought tolerant. More important, the
plants chosen here grew better in the contaminated soil than
other varieties/species screened directly in the contaminated
soil (no mulch amendment; data not shown).
(
CSAL Analytical Laboratory, Tucson, and Laboratory Con-
Seeds of Z. mays cv. Mayo Tuxpeno (Native Seed Search,
Tucson, AZ, USA) were sown into a commercial potting soil
sultants, Phoenix, AZ, USA). The toxicity characteristic leach-
ing procedure (TCLP) metal concentrations of both the con-
taminated and the uncontaminated soils as well as the contam-
inated soil:mulch mixture were determined and are presented
in Table 1. The TCLP metal concentrations were determined
by the following modified procedure (U.S. EPA Method 1311).
Ten-gram soil samples were incubated in a rotary shaker (200
(
El Toro Potting Soil). The seedlings were watered daily for
two weeks. Uniform seedlings were selected and transplanted
into 300-ml pots containing approximately 250 g of the con-
taminated soil:mulch mixture). For A. nummularia, uniform
established plants obtained from clippings were selected and
transplanted as described for Z. mays. In most cases, the trans-
plants were allowed to establish roots under a daily watering
regime for a week prior to receiving chelator treatment.
rpm) overnight (18
centrifuged at 6,000
Ϯ
2 h) with 200 ml of solution 1 and
g
for 10 min. The supernatant was filtered
through TCLP-certified glass filters (VWR Scientific, West-
chester, PA, USA) and analyzed for metals by flame atomic
absorption (AA) spectrophotometry (Model IL, Video 12,
Thermo Jarrell Ash, Franklin, MS, USA). One liter of extract-
ing solution 1 contained 5.7 ml glacial acetic acid and 64.3
ml of 1-N NaOH. The solution pH was adjusted with NaOH
Relative growth rates
Relative growth rates (RGRs) were determined for A. num-
mularia and Z. mays in the absence of chelator. Five plants
were sown into the contaminated soil:mulch mixture and an
uncontaminated soil:mulch mixture and watered every other
day to field capacity with tap water. The plants were harvested
after 40 d. Initial and final dry-weight measurements were
obtained for the shoot material. These samples were further
processed and analyzed for Pb content. The RGRs were cal-
culated by the compound interest formula RGR (grams new
to 4.93
trace metal concentrations. All glassware were acid washed in
0% nitric acid and rinsed thoroughly with double-distilled
Ϯ 0.05. All supplies were of reagent grade containing
1
water prior to experimentation. The TCLP standards and sam-
ple blanks were verified regularly by AA spectrophotometry.
It was determined in a preliminary plant screening (data
not shown) that in order to facilitate plant growth, organic
amendments were required in the contaminated soil. In this
study, both soils were mixed 1:1 (w/w) with forest mulch and
growth per gram existing growth per day, g/g/d)
lninitial wt)/number of days. In Table 2, the data are presented
as percentage growth per day by multiplying RGR by 100.
ϭ (lnFinal wt
Ϫ

2
700
Environ. Toxicol. Chem. 21, 2002
F.L. Jordan et al.
Table 2. Percentage growth per day and shoot Pb concentrations of
Zea mays and Atriplex nummularia transplants sown into a 1:1
mixture of contaminated (contam.) or uncontaminated (uncontam.)
Sciences (Norwalk, CT, USA). Fourteen milliliters of Hionic-
Fluor (Perkin-Elmer Life Sciences) scintillation fluid was add-
ed to the digested suspension and assayed for radioactivity in
a Packard scintillation counter (TriCarb 1600, Packard, Mer-
iden, CT, USA).
soil to mulch. Values in parentheses are standard deviation,
n ϭ 5
Percentage growth/day
g/g/d)
Shoot Pb concentrations
(mg/kg)
(
Statistical analysis
Contam.
Uncontam.
Contam.
Uncontam.
Two-way analysis of variance (ANOVA) was conducted to
compare mean Pb, Cu, and Zn uptake by the two plants across
treatments using the SYSTAT 10 statistical software (CPSS,
Chicago, IL, USA). ANOVA was also undertaken to compare
mean relative growth rates of the two plants in the uncontam-
inated and contaminated soil as well as to compare the mean
Plants
Atriplex
Zea mays 6.0 (
ANOVAa
Plant type
2.0 (
Ϯ
Ϯ
0.1)
0.4)
5.0 (
11 (
Ϯ
Ϯ
0.7)
0.1)
31 (
41 (
Ϯ
Ϯ
3)
28)
20 (
35 (
Ϯ
Ϯ
2)
38)
F
p
ratio
value
80
0.001
1.0
chelator uptake by the two plants. Both
F ratios and p values
Ͻ
0.30
are reported in the tables at alpha 0.05.
ϭ
Soil
0.5, 7.0b
0.5, 0.02
RESULTS
F
p
ratio
value
170
0.001
b
Ͻ
Soil characteristics and plant growth
Plant and soil
F
p
Both soils used in this study were collected from the same
site. Other than the extensive heavy metal contamination found
for the contaminated soil, both soils had similar general chem-
ical characteristics (Table 1). The glycophyte (Z. mays) had a
two- to threefold higher relative growth rate than the halophyte
b
ratio
value
2.7
0.1
0.1, 0.4
0.5, 1.0
b
a
ANOVA
ϭ analysis of variance.
b
Test statistics subtracting an outlier value from the Z. mays shoot
Pb concentration data.
(
A. nummularia) in both the contaminated and the uncontam-
inated soil:mulch mixtures ( 0.001; Table 2). Growth of
p
Ͻ
both plants was inhibited in the contaminated soil:mulch mix-
ture when compared to the uncontaminated soil:mulch mixture
(p Ͻ 0.001). The halophyte showed a few minor signs of stress
when grown in the contaminated soil:mulch mixture. In con-
trast, and despite its higher relative growth rate, the glycophyte
showed profound signs of metal stress, including severe chlo-
rosis and mortality when grown in the contaminated soil:mulch
mixture.
Chelator-facilitated metal-uptake experiments
Metal uptake by both plants was measured in the presence
and absence of chelator. Both plants were treated every other
day with distilled water or a 5-mM chelator solution (either
EDTA or rhamnolipid) for a total application rate of 5 mmol
chelator/kg dry soil over a 7-d period. The plants were har-
vested 24 h after the final watering event. On harvest, plants
were cut 1 cm above the soil, and the shoots were washed
with deionized water. The roots were washed first in distilled
water to remove excess soil and then washed in deionized
water. After washing, root and shoot samples were dried for
metal analysis. All treatments were conducted in triplicate.
The experiment was duplicated for a total of six plants per
treatment.
Plant metal uptake
Metal uptake by both the halophyte (A. nummularia) and
the glycophyte (Z. mays) was measured in the presence and
absence of chelator (either 5 mmol EDTA or rhamnolipid/kg
dry soil). Plants were allowed to establish roots in the con-
taminated soil:mulch mixture one week prior to the application
of treatment solutions. In the absence of chelator, total metal
Chelator complexation of soil bound metals
(
Pb ϩ Cu ϩ Zn) accumulation by the two plants was similar,
The ability of the two chelators to complex Pb, Cu, and Zn
from the contaminated soil was determined. Triplicate soil:
mulch samples were incubated (1:4 solid–liquid) with 5 mM
ranging from 920 mg/kg for the halophyte to 950 mg/kg for
the glycophyte. Metal accumulation generally followed the
order Cu
k Zn Ͼ Pb, reflecting the fact that the metals were
NaNO (control), EDTA, or rhamnolipid (all pH
ϭ
7.5, ionic
5 mM) at 200 rpm for 20 h. Each sample was
centrifuged at 10,000 to pellet the soil. The supernatant was
3
present in the soil at 11,000, 1,250, and 1,000 mg/kg, respec-
tively (Table 1). In general, both plants accumulated 2- to 11-
fold more metal in the roots than the shoots with one exception:
The halophyte accumulated more Zn in the shoot than the root
strength
ϭ
g
filtered through a Whatman 42 filter and analyzed for Pb, Cu,
and Zn by AA.
(
Figs. 1 and 2) regardless of treatment. Table 3 shows detailed
test statistics on the mean metal uptake data presented for the
shoots (Fig. 1) and the roots (Fig. 2). With respect to Pb uptake
Chelator uptake by plants
An experiment was designed to quantify chelator uptake
by the halophyte and the glycophyte; 1 mmol C-labeled rham-
nolipid or EDTA was added in 50 ml water to each established
into the shoots, both plant type (
were significant ( 0.005), but the interaction term (plant
type treatment) was not. Atriplex nummularia accumulated
slightly more Pb in the shoot material than Z. mays but only
when treated with EDTA ( 0.05). A different pattern
emerged with respect to Pb uptake into the roots. Both treat-
ment effects and the interaction term were significant, but no
difference was observed in Pb uptake between the two plant
types. With respect to Cu uptake into the roots, plant type,
treatment, and the interaction term were all significant. In this
case, Z. mays accumulated significantly more Cu than A. num-
p Ͻ 0.05) and treatment effects
14
p Ͻ
ϫ
transplant (specific activity of 0.1
␮Ci/plant). Triplicate sets
of plants were incubated in sealed, transparent, Plexiglas cham-
bers and harvested after 24 h. Plants were separated into root
and shoot material, allowed to dry at room temperature for 2
d, and then macerated into fine particles. Samples (0.25 g)
were digested with 2 ml of Na-hypochlorite at 60ЊC according
to the instructions provided by Packard Biosciences (http://
ww2.packardinst.com/packard/ecom/), now Perkin-Elmer Life
p Ͻ

Comparing metal uptake between a halophyte and a glycophyte
Environ. Toxicol. Chem. 21, 2002
2701
Fig. 1. Shoot metal (Pb, Cu, and Zn) concentrations per kilogram
plant biomass (dry wt) receiving a total of 5 mM ethylenediamine-
tetraacetic acid (EDTA), 5 mM rhamnolipid, or no chelator treatment.
Both plant types received (in total) 1 mmol of chelator/kg dry soil
over a one-week period prior to harvesting. Transplants were allowed
to establish new roots in the contaminated soil:mulch mixture for one
week prior to a one-week chelator application period. Error bars rep-
Fig. 2. Root metal (Pb, Cu, and Zn) concentrations per kilogram plant
biomass (dry wt) receiving a total of 5 mM ethylenediaminetetraacetic
acid (EDTA), 5mM rhamnolipid, or no chelator treatment. Both plants
received (in total) 1 mmol of chelator/kg dry soil over a one-week
period prior to harvesting. Transplants were allowed to establish new
roots in the contaminated soil:mulch mixture for one week prior to a
one-week chelator application period. Error bars represent standard
resent standard error (
n
ϭ
6).
error (n ϭ 6).
mularia across all treatment types (p Ͻ 0.005). With respect
Zn in the roots than A. nummularia over all treatments. Only
plant type and treatment were significant with respect to Zn
uptake into the shoots. In contrast to root Zn uptake, A. num-
mularia accumulated significantly more Zn in the shoots than
Z. mays regardless of treatment.
to Cu uptake into the shoots, only treatment effects were sig-
nificant. Lastly, with respect to Zn metal uptake into the root,
all test statistics were significant. Zea mays accumulated more
We compared the ability of EDTA to complex and solubilize
metals from the soil with its ability to enhance metal uptake
by plants. The EDTA increased metal solubilization by 62-
fold (Pb) to 138-fold (Cu) (data not shown) over soils extracted
Table 3. Two-way analysis of variance (␣ ϭ 0.05) test statistics for
mean Atriplex nummularia and Zea mays shoot and root metal
concentrations presented in Figures 1 and 2
Pb
Cu
Zn
with NaNO . The EDTA significantly increased the amount of
3
Pb, Cu, and Zn accumulated by both plants in both the shoot
and the roots, with total metal accumulation ranging from
Shoot metal concentrations
Plant type
1
,500 mg/kg for the halophyte to 2,000 mg/kg for the gly-
F
p
ratio
value
5
0.03
1
0.3
35
cophyte. This reflects an approximate doubling of metal uptake
by both plants in the presence of EDTA. Ethylenediaminetet-
raacetic acid particularly facilitated the translocation of metals
from the roots to shoots in all cases. This is reflected in the
significant increase in shoot-to-root metal ratios (Table 4). For
example, shoot-to-root ratios increased twofold over those not
receiving treatment for the halophyte and one-half-fold for the
glycophyte. Considering how effective EDTA was at com-
plexing metals from the soil, only a small fraction of the
EDTA-complexed metal was actually taken up by the plants.
In contrast to EDTA, rhamnolipid treatment had no effect on
metal accumulation in comparison to those plants receiving
water alone (Figs. 1 and 2). Rhamnolipid was also much less
effective at complexing and removing metal from the contam-
inated soil during the soil washing experiment, removing only
Ͻ
0.001
Treatment
F
p
ratio
value
21
0.001
24
0.001
3
0.03
Ͻ
Ͻ
Plant type and treatment
F
p
ratio
value
2
0.2
2
0.1
1
0.3
Root metal concentrations
Plant type
F
p
ratio
value
2
0.2
15
0.001
4
0.05
Ͻ
Treatment
F
p
ratio
value
7
37
0.001
17
0.001
Ͻ
Ͻ
0.001
Plant type and treatment
4
- to 33-fold more metal than the NaNO control (data not
3
F
p
ratio
value
4
0.02
12
0.001
4
0.02
shown).
The halophyte and the glycophyte exhibited specific dif-
Ͻ

2
702
Environ. Toxicol. Chem. 21, 2002
F.L. Jordan et al.
Table 4. Mean shoot-to-root metal ratios and analysis of variance (
␣
the addition of EDTA significantly increased this difference
between the two plants to 2.4-fold, 3.8-fold, and 4.1-fold for
Pb, Cu, and Zn, respectively. These data indicate that the hal-
ophyte translocates all three metals, but in particular Zn, more
effectively to shoots than the glycophyte, especially in the
presence of EDTA.
ϭ
0.05) test statistics found for the different chelator treatments for
Atriplex nummularia and Zea mays. Standard deviation reported (Ϯ)
for
n ϭ 18
Pb
Shoot-to-root metal ratio
Cu
Zn
Treatment/plant type
EDTAa
Plant chelator uptake
Chelator-facilitated plant metal uptake is believed to be due,
in part, to increased metal solubility. It may also be due to
preferential root-to-shoot translocation of organo-bound heavy
metal complexes, that is, Pb-EDTA [4,5]. Since both chelators
employed in this study increased the soluble fraction of the
soil-bound metals (data not shown), we hypothesized that
rhamnolipid’s failure to enhance metal uptake was due to either
exclusion from the root or failure to pass the metal from the
root to the shoot. To test this hypothesis, we exposed both
plants to soil solutions containing 14C-EDTA or C-rhamno-
lipid. Surprisingly, both plants translocated considerably less
EDTA (only 0.01–0.09% of the total applied to each plant)
Atriplex
1.00
Ϯ
Ϯ
0.29
0.10
0.73
0.19
Ϯ
Ϯ
0.37
0.05
2.14
0.52
Ϯ
Ϯ
0.99
0.14
Zea mays 0.41
Rhamnolipid
Atriplex
0.38
Ϯ
Ϯ
0.14
0.09
0.50
0.19
Ϯ
Ϯ
0.39
0.12
1.75
0.60
Ϯ
Ϯ
0.56
0.19
Zea mays 0.23
Water
Atriplex
Zea mays 0.17
ANOVA
Plant type
0.37
Ϯ
Ϯ
0.09
0.05
0.13
0.09
Ϯ
Ϯ
0.05
0.02
1.35
0.48
Ϯ
Ϯ
0.89
0.17
14
F
p
ratio
value
90
Ͻ
40
Ͻ
110
Ͻ
0.001
0.001
0.001
than rhamnolipid (Table 5;
p ϭ 0.003). A small fraction of
Treatment
the added rhamnolipid (0.06–0.24%) was found in the plant
tissue with the amount found in the roots exceeding that found
in the shoots by two- to fourfold. Mineralization of both che-
lators was found to be negligible over the 24-h incubation
period (data not shown), signifying that the radiolabel found
in the plant was representative of the intact chelator and not
a by-product of rhizosphere metabolism.
F
p
ratio
value
70
60
5
Ͻ
0.001
Ͻ
0.001
0.007
Plant type and treatment
F
p
ratio
value
20
0.001
40
4
0.01
Ͻ
Ͻ
0.001
a
EDTA
ϭ ethylenediaminetetraacetic acid.
DISCUSSION
ferences in metal translocation patterns. In fact, the halophyte
had significantly higher shoot-to-root metal accumulation ra-
tios than the corn across all treatments (Table 4). Specifically,
the halophyte shoot-to-root ratios were 2.2-fold, 1.4-fold, and
Metals, when present in excess in the soil, interfere with
proper enzymatic functions and can inhibit overall plant
growth. Establishment of plants for phytoextraction in heavily
contaminated soils may be quite difficult. The mine-tailings
soil employed in this study is considered highly metalliferrous,
2
.8-fold greater than the glycophyte shoot-to-root ratios for
Pb, Cu, and Zn, respectively. As was mentioned previously,
Table 5. Millimole chelator taken up per kilogram dry plant biomass, corresponding ratio of metal to chelator in Atriplex nummularia and Zea
mays, and analysis of variance (ANOVA) test statistics (␣ ϭ 0.05). Values in parentheses are the standard deviation for triplicate samples
mmol Chelator/kg plant
EDTAb
mg Metal/mg chelatora
EDTA Rhamnolipid
Rhamnolipid
Plant material
Shoots
Atriplex
Zea mays
0.10 (
0.04 (
Ϯ
Ϯ
0.01)
0.01)
0.80 (
0.20 (
Ϯ
Ϯ
0.40)
0.02)
0.80 (
1.4
Ϯ
Ϯ
0.10)
0.25)
0.06 (
0.10 (
Ϯ
Ϯ
0.03)
0.01)
(
Roots
Atriplex
Zea mays
0.10 (
0.20 (
Ϯ
Ϯ
0.03)
0.10)
1.10 (
0.70 (
Ϯ
Ϯ
0.40)
0.30)
2.8 (
3.2 (
Ϯ
Ϯ
1.0)
2.4)
0.20 (
0.30 (
Ϯ
Ϯ
0.02)
0.20)
Shoots
Roots
Shoots
Roots
ANOVA
Plant type
F
p
ratio
value
15
0.003
0.7
0.4
30
Ͻ0.001
0.3
0.6
Treatment
F
p
ratio
value
20
0.002
10
0.002
230
0.001
30
Ͻ
Ͻ
0.001
Plant type and treatment
F
p
ratio
value
8
0.02
2
0.2
20
0.002
0.06
0.8
a
Total mg metal (Pb
ϩ
Zn
ϩ Cu)/mg chelator recovered from corresponding plant material (dry wt).
b
EDTA
ϭ ethylenediaminetetraacetic acid.

Comparing metal uptake between a halophyte and a glycophyte
Environ. Toxicol. Chem. 21, 2002
2703
consisting of 15% Fe, 2% Cu, 0.2% Pb, and 0.25% Zn, among
other metals. Very few plants are actually found growing di-
rectly in the mine tailings on site. It is presumed that the high
heavy metal loading found for this soil is solely responsible
for the lack of natural plant establishment in the contaminated
soil both in the greenhouse and on site. The contaminated soil
is slightly more acidic and saline than the uncontaminated soil,
but when mixed with mulch and soil conditioner, the contam-
inated soil supported adequate plant growth.
Effective phytoextraction requires that plants be both metal
tolerant and accumulate metal in the aboveground tissue. It
has been suggested that salt-tolerant plants may be better
adapted to coping with environmental stresses, including
heavy metals [7,21], than salt-sensitive (glycophytic) crop
shoots than the glycophyte, especially Zn. Likewise, Williams
et al. [23] report that a number of halophytes (Atriplex in-
cluded) growing in a heavy metal–polluted salt marsh in the
United Kingdom exhibited similar shoot-to-root ratios. The
selective translocation of Zn (among a variety of other metals
found in our soil) by Atriplex would suggest that salt tolerance
confers an advantage in regulating ion uptake over salt sen-
sitivity. Although we compared only one representative hal-
ophyte to one representative glycophyte in this study, a similar
study comparing the response to Cu stress between Messem-
bryathemum crystallinum (another halophyte) and Arabidop-
sis thaliana (a glycophyte) demonstrated enhanced Cu uptake
and higher Cu tolerance by M. crystallinum [7]. Taking into
consideration our data and those reported elsewhere [6–11], it
can be suggested that halophytes may be superior plants for
long-term establishment in co-metal-contaminated soils.
It could also be argued that since Z. mays is a monocot
and A. nummularia a dicot, the comparison in metal uptake
between these two plants is somehow complicated by the an-
atomical and physiological differences between these plants.
Nevertheless, our results are consistent with current salt-tol-
erance research [6,21,24]. In general, halophytes such as Atri-
plex are more restrictive in regulating salt/mineral intake via
the roots than glycophytes. Furthermore, some halophytes
characteristically translocate and store salts/minerals in vac-
uoles found in the leaf tissue, providing for greater osmotic
control/water movement [6,24]. Therefore, they tend to trans-
locate ions from roots to shoots more effectively than gly-
cophytes. Thus, the basic differences between halophytes and
glycophytes observed with respect to Na uptake and transport
appear to apply to heavy metals as well, though the differences
we observed were not nearly as large as those observed for
Na.
plants (Z. mays, B. juncea, P. sitivum, and so on) commonly
chosen for phytoextraction research. In this study, growth of
both the halophyte and the glycophyte was stunted in the con-
taminated soil:mulch mixture compared to the uncontaminated
soil:mulch mixture. When comparing the two plant types, the
glycophyte had a faster growth rate than the halophyte re-
gardless of the soil:mulch mixture. Despite their slower growth
rate, halophytes such as A. nummularia are known to achieve
biomass yields as high as 30 t/ha even under salt stress con-
ditions [6]. Further, unlike Z. mays, which is an annual plant,
A. nummularia is a perennial plant that continues to accu-
mulate biomass at an exponential rate throughout the growing
season (E.P. Glenn, personal communication). It should be
noted that the slower growth rate of the halophyte did not
correspond to a decrease in metal uptake. In fact, both plants
accumulated similar levels of metals (on a mass basis), but
the symptoms of metal toxicity (stunted growth and chlorosis)
demonstrated by the halophyte were far less severe than those
demonstrated by the glycophyte when sown in the contami-
nated soil:mulch mixture.
Results from this study confirm that EDTA can solubilize
metals from the soil and facilitate metal uptake by plants [1–
5,13]. One major drawback to the use of synthetic chelators
such as EDTA is that they are highly recalcitrant to microbial
degradation and may persist in the soil for decades [16,17].
In fact, concerns have been raised regarding the use of syn-
thetic chelators in the field [16–18]. In this study, we compared
the ability of EDTA and a biodegradable chelator, rhamnolipid,
to facilitate plant metal uptake. Rhamnolipid, like EDTA, com-
plexed soil-bound metals. However, rhamnolipid was not ef-
fective at enhancing metal uptake by either the halophyte or
the glycophyte, even though it was found in the shoot material
of both plants. One reason for this may be that, compared to
EDTA, rhamnolipid has a significantly lower binding affinity
for metals. For example, the stability constant for EDTA-Pb
is 17.88 compared to 8.58 for rhamnolipid-Pb, which repre-
sents a nine-order-of-magnitude difference in the relative
strength of the two complexes [19]. A second possibility is
that the metal-bound rhamnolipid complex is excluded from
the root entirely, whereas the free rhamnolipid molecule is not.
Continued research is required to identify environmentally
compatible metal chelators that may be used to aid in phy-
toextraction scenarios.
Both plants accumulated metals in the order Cu
Ͼ Zn Ͼ
Pb, reflecting the metal concentrations found in the contami-
nated soil and/or the bioavailability of the metal. Both Cu and
Zn are essential for maintaining plant metabolic processes, but
in excess concentrations both metals are phytotoxic. Lead, on
the other hand, is not an essential metal. In addition, Cu and
Zn are considerably more bioavailable than Pb in soils. Both
Zn and Cu tend to form ionic instead of covalent chemical
bonds, making them less stable and prone to resolubilization
[
22]. Conversely, Pb forms strong covalent complexes, such
as inorganic oxides, carbonates, or sulfides, that are much more
stable. Even though Pb (2,000 mg/kg dry soil) and Zn (2,500
mg/kg dry) are at similar concentrations in the soil, less Pb
(Ͻ0.01% of the total) was accumulated in the aboveground
plant material than Zn (0.1%). This was especially true for the
halophyte. Williams et al. [12] observed similar metal loading
of essential (Cu, Zn, Mn, and Fe) rather than nonessential
metals (Pb and Cd) for a group of salt marsh plants in the
United Kingdom. In our study, as in theirs, neither the halo-
phyte nor the glycophyte hyperaccumulated any one metal (Cu,
Pb, or Zn) from the contaminated soil:mulch mixture. In fact,
both plants accumulated significantly less than 1% of their
biomass in total (Pb
ϩ
Cu
ϩ
Zn) metal.
To summarize, our results provide further evidence that salt
tolerance may play a role in conferring metal tolerance in
halophytes. The fact that Atriplex spp. and other halophytes
selectively translocate Zn from contaminated soils and that an
increase in stress response factors was observed for M. crys-
tallinum when exposed to increasing Cu concentrations sug-
gests that halophytes may have an advantage over the gly-
Despite the fact that the halophyte did not hyperaccumulate
metals while growing in the highly metalliferrous soil (as it
ϩ
Ϫ
does Na and Cl in salty soils), its metal accumulation pattern
differed significantly from the glycophyte. These differences
became more pronounced when EDTA was applied to the soil.
The halophyte translocated more metal from the roots to the

2
704
Environ. Toxicol. Chem. 21, 2002
F.L. Jordan et al.
1
1
1
0. Broadley MR, Wiley NJ, Mead A. 1999. A method to assess
taxonomic variation in shoot caesium among flowering plants.
Environ Pollut 106:341–349.
cophytes typically employed in most phytoremediation strat-
egies for some sites [1–5,7,23]. These data also clearly dem-
onstrate the need to identify or design chelators that are
environmentally compatible but that effectively complex met-
als from soil and facilitate their translocation to the above-
ground plant material.
1. Kohl K. 1996. Population-specific traits and their implication for
the evolution of a drought-adapted ecotype in Armeria maritima
.
Botanica ACTA 109:206–215.
2. Williams TP, Bubb JB, Lester JN. 1994. The occurrence and
distribution of trace metals in halophytes. Chemosphere 28:1189–
1
199.
Acknowledgement—This research was funded in part by a research
contract with the U.S. Postal Service and by Grant P42 ES04940 from
the National Institute of Environmental Health Sciences.
13. Laurie SH, Tancock NP, McGrath SP, Sanders JR. 1995. Influence
of EDTA complexation on plant uptake of manganese (II). Plant
Sci 109:231–235.
1
4. Maier RM, Neilson JW, Artiola JF, Jordan FL, Glenn EP, Descher
SM. 2001. Remediation of metal-contaminated soil and sludge
using biosurfactant technology. J Occup Med Environ Health 14:
241–248.
REFERENCES
1
. Blaylock MJ, Salt DE, Dushenkov S, Zakharova O, Gussman C,
Kapulnik Y, Ensley BD, Raskin I. 1997. Enhanced accumulation
of Pb in Indian mustard by soil-applied chelating agents. Environ
Sci Technol 31:860–865.
. Huang JW, Chen J, Berti WR, Cunningham SD. 1997. Phytore-
mediation of lead-contaminated soils: Role of synthetic chelates
in lead phytoextraction. Environ Sci Technol 31:800–805.
. Wu J, Hsu FC, Cunningham SD. 1999. Chelate-assisted Pb phy-
toextraction: Pb availability, uptake and translocation constraints.
Environ Sci Technol 33:1898–1904.
15. Torrens JL, Herman DC, Miller-Maier RM. 1998. Biosurfactant
rhamnolipid) sorption and the impact on rhamnolipid-facilitated
(
removal of cadmium from various soils. Environ Sci Technol 32:
776–781.
2
3
4
5
16. Sillanpaa M, Pirkanniemi K. 2001. Recent developments in che-
late degradation. Environ Technol 22:791–801.
17. Nortemann B. 1999. Biodegradation of EDTA. Appl Microbiol
Biotechnol 51:751–759.
. Vassil AD, Kapulnik Y, Raskin I, Salt DE. 1998. The role of
EDTA in lead transport and accumulation by Indian mustard.
Plant Physiol 117:447–453.
. Epstein AL, Gussman CD, Blaylock MJ, Yermiyahu U, Huang
JW, Kapulnik Y, Orser CS. 1999. EDTA and Pb-EDTA accu-
mulation in Brassica juncea grown in Pb-amended soil. Plant
Soil 208:87–94.
. Glenn EP, Brown JJ, Blumwald E. 1999. Salt tolerance and crop
potential of halophytes. Crit Rev Plant Sci 18:227–255.
. Thomas JC, Malick FK, Endreszl C, Davies EC, Murray KS.
18. Barona A, Aranguiz I, Elias A. 2001. Metal associations in soils
before and after EDTA extractive decontamination: Implications
for the effectiveness of further clean-up procedures. Environ Pol-
lut 113:79–85.
19. Ochoa-Loza FJ, Artiola JF, Maier RM. 2001. Stability constants
for the complexation of various metals with a rhamnolipid bio-
surfactant. J Environ Qual 30:479–485.
6
7
20. Maslin P, Maier RM. 2000. Rhamnolipid-enhanced mineralization
of phenanthrene in organic-metal co-contaminated soils. Biorem
J
4:295–308.
1
998. Distinct responses to copper stress in the halophyte Me-
sembryanthemum crystallinum Physiologia plantarum 102:360–
68.
21. Vernon DM, Bohnert HJ. 1992. Novel methyl transferase induced
.
by osmotic stress in the faculative halophyte Mesembryanthemum
crystallinum. EMBO J 11:2077–2085.
3
8
9
. Kahn MA, Ungar IA, Showalter AM. 2000. Effects of sodium
chloride treatments on growth and ion accumulation of the hal-
22. Sposito G. 1989. The Chemistry of Soils. Oxford University Press,
Oxford, UK.
23. Williams TP, Bubb JB, Lester JN. 1994. Metal accumulation with-
in salt-marsh environments—A review. Mar Pollut Bull 28:277–
290.
24. Hasegawa PM, Bressan RA, Zhu JK, Bohnert HJ. 2000. Plant
cellular and molecular responses to high salinity. Plant Mol Biol
51:463–499.
ophyte Haloxylon recurvum
.
Comm Soil Sci Plant Anal 31:2763–
2
774.
. Lunque CJ, Castellanos EM, Castillo JM, Gonzalez M, Gonzalez-
Vilches MC, Figueroa ME. 1999. Metals in halophytes of a con-
taminated estuary (Odiel Saltmarshes, SW Spain). Mar Pollut
Bull 38:49–51.