Arsenic speciation and distribution in an arsenic hyperaccumulating plant

The Science of the Total Environment 300 (2002) 167–177
0048-9697/02/$ - see front matter ᮊ 2002 Elsevier Science B.V. All rights reserved.
PII: S0048-9697
Ž
02
.
00165-1
Arsenic speciation and distribution in an arsenic
hyperaccumulating plant
Weihua Zhang , Yong Cai *, Cong Tu , Lena Q. Ma *
aa, bb,1,
Department of Chemistry and Southeast Environmental Research Center, Florida International University, Miami, FL 33199,
a
USA
Soil and water Sciences Department, University of Florida, Gainesville, FL 32611, USA
b
Received 5 October 2001; accepted 25 March 2002
Abstract
Arsenic-contaminated soil is one of the major arsenic sources for drinking water. Phytoremediation, an emerging,
plant-based technology for the removal of toxic contaminants from soil and water, has been receiving renewed
attention. Although a number of plants have been identified as hyperaccumulators for the phytoextraction of a variety
of metals, and some have been used in field applications, no hyperaccumulator for arsenic had been previously
reported until the recent discovery of Brake fern (Pteris vittata), which can hyperaccumulate arsenic from soils. This
finding may open a door for phytoremediation of arsenic-contaminated soils. Speciation and distribution of arsenic in
the plant can provide important information helpful to understanding the mechanisms for arsenic accumulation,
translocation, and transformation. In this study, plant samples after 20 weeks of growth in an arsenic-contaminated
soil were used for arsenic speciation and distribution study. A mixture of methanolywater (1:1) was used to extract
arsenic compounds from the plant tissue. Recoveries of 85 to 100% were obtained for most parts of the plant
(rhizomes, fiddle heads, young fronds and old fronds) except for roots, for which extraction efficiency was
approximately 60%. The results of this study demonstrate the ability of Brake fern as an arsenic hyperaccumulator.
It transfers arsenic rapidly from soil to aboveground biomass with only minimal arsenic concentration in the roots.

The arsenic is found to be predominantly as inorganic species; and it was hypothesized that the plant uptakes arsenic
as arsenate
w
As(V)
x
and arsenate was converted to arsenite
w
As(III)
x
within the plant. The mechanisms of arsenic
uptake, translocation, and transformation by this plant are not known and are the objectives of our on-going research.
ᮊ 2002 Elsevier Science B.V. All rights reserved.
Keywords: Arsenic; Phytoremediation; Pteris vittata; Hyperaccumulator; Speciation
*Corresponding author. Tel.: q1-305-348-6210; fax: q1-
305-348-3772.
E-mail address: (Y. Cai),
(L.Q. Ma).
Co-corresponding author. Tel.: q1 352 392 9063
1
1. Introduction
Arsenic, ranking 20th in abundance in the
earth’s crust, is a toxic element widely encountered
in the environment and organisms (Cullen and
Reimer, 1989). Arsenic can enter terrestrial and
aquatic environments through both natural forma-
168 W. Zhang et al. / The Science of the Total Environment 300 (2002) 167–177
tion and anthropogenic activity. Natural pathways
of arsenic include weathering, biological activity,
and volcanic activity. The primary anthropogenic
input derives from combustion of municipal solid

waste, fossil fuels in coal- and oil-fired power
plants, release from metal smelters, and direct use
of arsenic-containing herbicides by industry and
agriculture. There are a number of ways by which
the human population can become exposed to
arsenic. The most important one is probably
through ingestion of arsenic in drinking water or
food (National Research Council, 1999; Le et al.,
2000; US EPA, 2001a).
Arsenic species are bioactive and toxic. Long-
term exposure to low concentrations of arsenic in
drinking water can lead to skin, bladder, lung, and
prostate cancer. Non-cancer effects of ingesting
arsenic at low levels include cardiovascular dis-
ease, diabetes, and anemia, as well as reproductive,
developmental, immunological and neurological
effects. Short-term exposure to high doses of
arsenic can cause other adverse health effects (US
EPA, 2001a). A recent report by the National
Academy of Sciences concluded that the previous
arsenic standard of 50 mgyl in drinking water does
not achieve US EPA’s goal of protecting public
health and should be lowered as soon as possible
(National Research Council, 1999). EPA has
recently decreased the drinking water standard to
10 mgyl in October 2001 to more adequately
protect public health (US EPA, 2001b). The
increase in the public awareness of the toxicity
and the environmental impact of arsenic contami-
nation and the possible implementation of new

regulations limiting arsenic in drinking water have
resulted in a growing interest in the study of the
biogeochemical cycling of arsenic and the devel-
opment of arsenic decontamination technologies.
Arsenic-contaminated soil is one of the major
sources of arsenic in drinking water (Nriagu, 1994;
National Research Council, 1999; Welch et al.,
2000; Kim and Nriagu, 2000). The concentration
of arsenic in cereals, vegetables and fruits is
directly related to the level of arsenic in contami-
nated soil. Although the remediation of arsenic-
contaminated soil is an important and timely issue,
cost-effective remediation techniques are not cur-
rently available. Phytoremediation, an emerging,
plant-based technology for the removal of toxic
contaminants from soil and water is a potentially
attractive approach (US EPA, 2000; Terry and
Banuelos, 1999; Raskin and Ensley, 2000; Dah-
mani-Muller et al., 2000). This technique has
received much attention lately as a cost-effective
alternative to the more established treatment meth-
ods used at hazardous waste sites. It is often the
only way to remediate soils contaminated with
metals without affecting their biological function.
A number of plants have been identified as hyper-
accumulators for the phytoextraction of a variety
of metals including Cd, Cr, Cu, Hg, Pb, Ni, Se
and Zn, and some of these plants have been used
in field applications (Dobson et al., 1997; Brooks,
1998; Terry and Banuelos, 1999; Reeves et al.,

2001).
Recently, Brake fern (Pteris vittata), an effica-
cious arsenic hyperaccumulating fern plant, has
been discovered in an abandoned wood-treatment
site in central Florida (Ma et al., 2001). This fern
can tolerate arsenic concentration as high as up to
1500 mgyg in soil, and has a bioconcentration
factor of 193. The arsenic concentration in the
plant can reach as high as 2.3% (dried weight).
The toxicity and bioavailability of arsenic are
closely associated with its oxidation state and
species. The determination of total arsenic in a
sample is insufficient to assess its environmental
risk (Koch et al., 2000). Speciation of arsenic in
plant samples can provide important information
helpful to understanding the mechanisms for arsen-
ic accumulation, translocation, transformation and
detoxification by Brake fern. It has been found
that a large amount of the arsenic in marine
organisms is in organic forms such as arsenosugars
in algae, and arsenobetaine and arsenocholine in
fish, mollusks, and crustaceans (Maeda, 1994;
Francesconi et al., 1994). The chemical structures
for these organoarsenic compounds in many
marine organisms have been reported (Maeda,
1994; Francesconi et al., 1994). However, little is
known about arsenic speciation in freshwater
aquatic plants or those in terrestrial environments
(Koch et al., 2000). Based on the limited infor-
mation available, it appears that, in contrast to

marine organisms, inorganic arsenic is the predom-
inant form of arsenic found in some freshwater
169W. Zhang et al. / The Science of the Total Environment 300 (2002) 167–177
and terrestrial plants (Helgesen and Larsen, 1998;
Koch et al., 1999, 2000).
Uptake, accumulation and translocation of
arsenic in both arsenic-tolerant and non-tolerant
plants have been studied although those plants are
not arsenic hyperaccumulators (Meharg and Mac-
nair, 1990, 1991a,b; Sneller et al., 1999; Schmoger
¨
et al., 2000; Pickering et al., 2000). Brake fern as
an arsenic hyperaccumulator not only has the
potential for phytoremediation of arsenic contam-
inated soil, but also provides an excellent oppor-
tunity to investigate plant detoxification mechan-
isms for arsenic. Following our identification of
this plant as an efficient hyperaccumulator of
arsenic, we investigated the speciation and distri-
bution of arsenic in the plant. This paper summa-
rizes the results from these studies.
2. Experimental
2.1. Reagents and standards
The inorganic arsenic standard and other indi-
vidual stock solutions of internal standards used
for inductively coupled plasma mass spectrometry
(ICPyMS) analysis (ICP grade, 1000 mgyl) were
purchased from GFS Chemicals, Inc. (Powell,
OH). Arsenic standards for speciation analysis
were obtained as sodium hydrogenarsenate heptah-

ydrate, Na HAsO Ø7H O (Aldrich, Milwaukee,
242
WI); sodium metaarsenite, NaAsO (Aldrich); and
2
cacodylic acid, (CH ) AsO(OH)(Sigma, St. Lou-
32
ise, MO). These standards were dissolved in dis-
tilled, deionized water to make 1000 ppm stock
solutions of arsenate
w
As(V)
x
, arsenite
w
As(III)
x
,
and dimethylarsinic acid (DMA), respectively. A
stock solution of monomethylarsonic acid (MAA)
(1000 mgyl) was provided by P.S. Analytical
(Kent, UK). The standards were used as received
without further purification. Fresh calibration stan-
dards were prepared every week or as needed by
diluting these commercial standards or stock solu-
tions either in 5% nitric acid (for total arsenic
analysis by ICPyMS) or in water (for speciation
analysis). Trace metal grade hydrochloric acid,
nitric acid, and HPLC grade methanol were
obtained from Fisher Scientific (Pittsburgh, PA).
All other chemicals used were of analytical grade

or better. Distilled deionized water was prepared
using a Barnstead Fistream II Glass Still System
(Barnstead Thermolyne Corp., Dubuque, Iowa)
and was used in all standard and sample prepara-
tions. High purity grade (99.99%) argon for ICPy
MS was purchased from Air Products (Allentown,
PA).
All glass and plastic ware was cleaned prior to
use by soaking in 5% nitric acid overnight, rinsing
with water and storing clean. The procedural blank
produced after these cleaning steps has been found
to contain negligible amount of arsenic.
2.2. Sampling and sample preparation
Brake fern (Pteris vittata) samples used in this
study were collected from an arsenic contaminated
soil, following 20 weeks of growth in a green-
house. The surface layer (0–15 cm) of arsenic-
contaminated soil (sandy, siliceous, hyperthermic,
grossarenic paleudult), which contained 97 mgyg
of arsenic, was collected from an abandoned chro-
mated–copper–arsenate (CCA) wood preservation
site in Central Florida (Ma et al., 2001). Air-dried
soil of 1.5 kg was weighed into each plastic pot
with a diameter of 15 cm (2.5 l). The soil was
thoroughly mixed with 1.5 g of Osmocote
᭨
extended time-release fertilizer as a base fertilizer
(18-6-12)(Scotts–Sierra Horticultural Products
Co., Marysville, OH). A petri dish was placed
under each pot to collect potential leachate during

the experiment. After a one-week equilibrium
under moist conditions, each pot was transplanted
with one healthy fern with 5 to 6 fronds. The
plants were watered daily or as necessary. During
the experiment, the average temperature in the
greenhouse ranged from 14 (night) to 30 8C (day),
with an average photosynthetically active radiation
(PAR) of 825 mmol m Øs . After 12 weeks of
y2 y1
transplanting, additional fertilizers containing 50
mg N kg in the form of NH NO and 25 mg P
y1
43
kg of KH PO were applied to all ferns. After
y1
24
harvest, the plants was washed with water, and
then separated into 5 different groups for samples
collected from greenhouse (roots, rhizomes, fiddle
heads, young fronds, and old fronds). The samples
were freeze-dried, ground to fine powder using a
170 W. Zhang et al. / The Science of the Total Environment 300 (2002) 167–177
ceramic mortar and pestle, and stored in 20 ml
plastic vials at room temperature until use.
For total arsenic analysis, a digestion procedure
previously developed for arsenic analysis in sea-
grass by ICPyMS was adapted (Cai et al., 2000).
Briefly, 10 mg samples were digested in open
vessels with 10 ml of nitric acid for 1 h using a
sand bath (150 8C). Then, 1 ml hydrogen peroxide

was added into the sample vessel and the sample
was allowed to digest for an additional 30 min.
After cooling, the samples were transferred into a
100-ml volumetric flask, and brought to volume
with water. These solutions were diluted by a
factor of 10 with 5% nitric acid prior to analysis
using ICPyMS.
For arsenic speciation, 10 mg samples were
ultrasonically extracted with 5 ml 1:1 methanoly
water for 2 h. The samples were then centrifuged;
the supernatant was decanted into a 100-ml volu-
metric flask. The procedure was repeated with the
residual pellet and the two extracts were combined.
The residue was rinsed three times with 5 ml of
water (5ml=3), and all supernatants were com-
bined. The extract was then diluted to the 100 ml
mark with water and then filtered using 0.45 PTFE
syringe filters (Gelman). The filtrate was directly
subjected to HPLC, or diluted by a factor of 10
with water for speciation analysis. For total arsenic
analysis, the filtrate was diluted by a factor of 10
with 5% nitric acid.
2.3. Sample analysis
Total arsenic analysis was carried out on a
Model HP 4500 plus ICPyMS instrument (Hew-
lett–Packard Co., Wilmington, DE) equipped with
a Babington-type nebulizer and an ASX-500 auto-
sampler (Cetac Technologies Inc., Omaha, NE).
The instrumental configuration and general exper-
imental conditions can be found elsewhere (Cai et

al., 2000). Arsenic standard solutions prepared in
5% nitric acid were used for the calibration curves.
Internal standard ( Y as internal standard) method
89
was used for quantitative determination of total
arsenic in the nitric acid-digested samples (Cai et
al., 2000) whereas the method of standard addi-
tions was applied to the methanolywater-extracted
samples.
Speciation analysis of arsenic in the methanoly
water-extracted samples was performed using both
high performance liquid chromatography (HPLC)
coupled with hydride generation atomic fluores-
cence spectrometry (HPLC-HG-AFS) and HPLC–
ICPyMS. The HPLC-HG-AFS instrument used
was a P S Analytical Millennium Excalibur system
(PSA 10.055, P.S Analytical, Kent, UK) coupled
to an HPLC system from Spectra-Physics Analyt-
ical, Inc. (Fremont, CA). The Millennium Excali-
bur system is an integrated atomic fluorescence
system incorporating vapor generation, gas–liquid
separation, moisture removal and atomic fluores-
cence stages. Data were acquired by a real-time
chromatographic control and data acquisition sys-
tem. The HPLC system is comprised of a P4000
pump and an AS 3000 autosampler with a 100-ml
injection loop. A strong anion exchange column
(PRP X-100, 250=4.6 mm, 10 mm particle size,
Hamilton) was used for separation. Potassium
phosphate (0.015 M for both K HPO and

24
KH PO ) at pH of 5.9 and flow rate of 1 mlymin
24
was used as mobile phase. For HPLC–ICPyMS,
the outlet of the analytical column was connected
to the nebulizer of the ICPyMS system by a 40
cm=0.25 mm i.d. PTFE tube. The HPLC condi-
tions used were the same as for HPLC-HG-AFS.
3. Results and discussion
3.1. Total arsenic concentration and distribution
The total amounts of arsenic and its distribution
in the Brake fern in the greenhouse experiments
are illustrated in Fig. 1. Brake fern rapidly and
efficiently accumulated a large amount of arsenic
from the moderately contaminated soil (97 ppm).
Arsenic distribution in the plant varied significant-
ly in different parts of the plant with concentrations
of 3894; 2610; 2336; 728; and 168 mgyg for old
fronds, young fronds, fiddle heads, rhizomes and
roots, respectively. It is interesting to note that
arsenic concentration in Brake fern roots was the
lowest (-168 mgyg), whereas those in fronds
were substantially greater with the old fronds
having the highest arsenic level. It is estimated
that )95% of arsenic taken up by the plant was
concentrated in the aboveground biomass. This is
171W. Zhang et al. / The Science of the Total Environment 300 (2002) 167–177
Fig. 1. Total concentrations of arsenic in different parts of the Brake fern (Pteris vittata) obtained using ICPyMS for plants from
greenhouse and field. The soils used for greenhouse experiments contained 97 mgyg of arsenic, while those in the field where fern
grew contained 153 mgyg. Rhizome apart from field samples was not analyzed.

in good agreement with our previous observations
(Ma et al., 2001). Arsenic concentrations in dif-
ferent parts of the plants collected from the field
are also included in Fig. 1 for comparison. Similar
distribution patterns can be found for plants from
both greenhouse experiments and field.
High arsenic tolerant plants have been reported
previously (Porter and Peterson, 1975; Meharg and
Macnair, 1991a,b; Bech et al., 1997; Helgesen and
Larsen, 1998; Brooks, 1998; Sneller et al., 1999;
Koch et al., 1999, 2000; Pickering et al., 2000;
Schmoger et al., 2000). Depending on plant spe-
¨
cies, arsenic tolerance may result from two strate-
gies: arsenic exclusion and arsenic accumulation
(Baker, 1987; Dahmani-Muller et al., 2000). The
exclusion strategy involves avoidance of arsenic
uptake or restriction of arsenic transport to the
shoots. Typha latifolia, found abundantly at arsen-
ic-contaminated sites, appears to be one example
of this (Dushenko et al., 1995; Koch et al., 1999).
The accumulation strategy consists of strong con-
centration of arsenic in plant tissue. Several terres-
trial plants found on mine tailings have been
observed to contain high levels of arsenic. Arsenic
concentrations of up to 3470 mgyg (dry weight)
have been reported for Agrostis tenuis (Porter and
Peterson, 1975). However, in order for the plant
to accumulate these high levels of arsenic, the soil
must contain an extremely high concentration of

arsenic (as high as 26500 mgyg). The metal
accumulation efficiency in plants can be evaluated
using the bioconcentration factor (BF), which is
defined as the ratio of metal concentration in the
plant biomass to metal concentration in the soil.
Hyperaccumulating plants are those that have a
BF)1 (Brooks, 1998). Although Agrostis tenuis
can accumulate arsenic to concentrations as high
as 3470 mgyg, it has a BF much less than one.
The BF for Brake fern can be as high as 193 (Ma
et al., 2001), indicating an efficient accumulation
of arsenic from soil by this plant.
Plants that accumulate arsenic may either store
arsenic in the roots or translocate it to the above-
ground biomass. These differences in storage of
172 W. Zhang et al. / The Science of the Total Environment 300 (2002) 167–177
arsenic suggest different processes for arsenic
accumulation and transport mechanisms within
different plants. Arsenic accumulation in root cells,
such as those observed in tomato root systems can
be related to an exclusion strategy (Carbonell-
Barrachina et al., 1997; Dahmani-Muller et al.,
2000). When high arsenic concentrations are pres-
ent in shoots but not in roots an efficient root-to-
shoot transport system may be important for
arsenic tolerance and account for hyperaccumula-
tion as in Brake fern. The results shown in Fig. 1
for plants grown under greenhouse conditions and
those of fern samples taken from the field indicate
arsenic concentration in old fronds is greater than

those in young fronds. The transport of arsenic
from roots to fronds is most likely carried out
through the xylem sap. The differences in arsenic
concentration in young and old fronds may suggest
that a larger cumulative amount of transpiration
stream has been probably passed through the old
fronds over time. Translocation of metals from
roots to the aging leaves has been considered as a
detoxification process to assist removal of arsenic
from the plant as the old leaves senesce and
eventually fall off the plant (Dahmani-Muller et
al., 2000; Perronnet et al., 2000). In an experiment
with Brake fern taken from the CCA site, arsenic
content in a naturally dried (dead) frond was found
to be relatively low (84 and 428 mgyg for the
young and old fronds, respectively), whereas that
in the living fronds taken from the same plant was
found to be high (4893 and 7575 mgyg, for the
young and old frond parts, respectively). It is
postulated that low concentrations of arsenic in
the dead leaves resulted either from being washed
out by rain after break up of the plant cell or by
translocation of arsenic to the living parts before
abscission in a manner similar to that of plant
nutrients (Goodwin and Mercer, 1983).
3.2. Speciation and transformation of arsenic
In order to obtain speciation information on
arsenic present in the plant, arsenic was extracted
with a 1:1 mixture of methanolywater. The recov-
ery of arsenic using this extraction procedure with

respect to the total arsenic concentration obtained
using nitric acid digestion was evaluated. Recov-
eries ranged from 85 to 100% for most parts of
the plant (rhizomes, fiddle heads, young fronds
and old fronds) except for the roots, where extrac-
tion efficiency was approximately 60%. These
recoveries were higher than those reported for
other plants using the same extraction method
(Koch et al., 1999, 2000).
Chromatograms of arsenic species in living plant
parts obtained using HPLC-HG-AFS and HPLC-
ICPyMS are shown in Fig. 2. The HPLC-HG-AFS
technique determines only the hydride-forming
arsenic species (As(III),As(V), MMA, and
DMA), whereas the HPLC-ICPyMS method can
provide extra information on other species of
arsenic. The results shown in Fig. 2 clearly indicate
that the extractable arsenic species in the Brake
fern consisted of only inorganic arsenic species,
As(III) and As(V). In order to confirm the absence
of other arsenic species, which may not be deter-
mined by either HPLC-HG-AFS or HPLC-ICPy
MS, the methanolywater extract was directly
analyzed by ICPyMS without HPLC separation.
The method of standard addition was used for
quantification in order to compensate for matrix
effects. It was found that the total concentration
of arsenic obtained in each part of the plant with
ICPyMS was in good agreement with the sum of
As(III) and As(V) obtained using HPLC-HG-AFS

(Fig. 3). This result suggests that the stable orga-
noarsenic compounds (e.g. methylated species,
arsenosugars) are not present in the living plant in
any significant quantity. This, however, does not
rule out the presence of intermediary organoarsenic
compounds such as arsenic-biomolecule complex-
es, which may decompose into simple inorganic
arsenic species during the course of the extraction
andyor separation. In fact, such complexation may
be necessary to enable the plant to accumulate
extremely high concentrations of arsenic while at
the same time avoiding high concentration of free
arsenic in cytoplasm, which cause disruption of
cell function and even cellular death. Phytochela-
tins (PCs), a family of peptides with the general
structure (g-GluCys) -Gly, have been reported to
n
be induced upon exposure to arsenic in some
plants (Grill et al., 1987; Maitani et al., 1996;
Sneller et al., 1999, 2000; Schmoger et al., 2000).
¨
Complexation and detoxification of arsenic by the
173W. Zhang et al. / The Science of the Total Environment 300 (2002) 167–177
Fig. 2. Chromatograms of arsenic species extracted with a 1:1 methanolywater mixture. a: analyzed by HPLC-HG-AFS, and b:
analyzed by HPLC-ICPyMS.
induced PCs has been confirmed using different
techniques in some research (Schmoger et al.,
¨
2000). However, other research failed to demon-
strate the formation of arsenic-PC complexes

(Maitani et al., 1996) although PCs were indeed
induced in plants upon exposure to arsenic. The
174 W. Zhang et al. / The Science of the Total Environment 300 (2002) 167–177
Fig. 3. Total concentration of arsenic obtained by ICPyMS and the sum of As(III) and As(V) obtained by HPLC-HG-AFS in
different parts of the Brake fern ( Pteris vittata).
role played by PCs in the accumulation and detox-
ification of arsenic and other metals in plants is
still in debate (De Knecht et al., 1992; Leopold et
al., 1999). Our results suggest that the formation
of stable arsenic-PC complexes does not occur in
the Brake fern in any significant quantity. Research
on heavy metal hyperaccumulating plants indicates
that some organic and amino acids (histine and
proline) and polyhydroxy phenolic compounds
may also be involved in heavy metal detoxification
in plants (Kramer et al., 1996). The role played
¨
by organic and amino acids in the accumulation
and detoxification of arsenic in plants is currently
unknown.
The conversion of As(V) to As(III) within the
plant is interesting to note. Fig. 4 shows the
percentages of As(III) with respect to the total
arsenic content obtained by ICPyMS in different
parts of Brake fern. Approximately 60–74% of
the arsenic in the fronds was present as As(III)
compared to only 8.3% in the roots. Note that
As(V) is the predominant species in the roots. In
a recent study, soils were spiked with 50 mgAs
gasAs(III),As(V), dimethylarsinic acid

y1
(DMA), or methylarsonic acid (MMA). After 18
weeks, arsenic in soil was mainly present as
arsenate with little detectable organic species or
arsenite regardless of arsenic species added to the
soil (Tu et al., 2002). It is conceivable from these
results that arsenic was taken up by Brake fern
roots primarily as arsenate from soil using the
phosphate uptake system. Arsenic competes with
phosphate as a substrate for the phosphate uptake
system in a wide variety of species (e.g. Wells
and Richardson, 1985; Macnair and Cumbes, 1987;
Meharg and Macnair, 1990, 1991a). It has been
reported that in both arsenic tolerant and non-
tolerant Holcus lanatus L., arsenic uptake uses
phosphate uptake system (Macnair and Cumbes,
1987; Meharg and Macnair, 1990, 1991a). It was
further proposed that the uptake of arsenic in the
arsenic-tolerant H. lanatus is restricted by the
altered phosphate uptake system, yet the tolerant
plants were capable of accumulating arsenic to
high concentration over longer time periods
(Meharg and Macnair, 1991a). It seems unlikely
175W. Zhang et al. / The Science of the Total Environment 300 (2002) 167–177
Fig. 4. Percentage of As(III) found in different parts of the plant with respect to the total As content obtained by ICPyMS.
that the restricted uptake of arsenic by plant roots
is a proper hypothesis for Brake fern since arsenic
is taken up by this hyperaccumulator to an
extremely high level in a very short time period
(Ma et al., 2001). The effect of phosphate on the

arsenic uptake by Brake fern is a topic of further
study.
In the fronds, As(III) is the major species.
Consider that total arsenic level in the roots is
much smaller that that in fronds, As(III) is the
predominant species in the Brake fern. Terrestrial
plants do not have arsenic detoxification system
of algae by methylation of arsenic, and this is
perhaps the reason why inorganic arsenics species
are predominant in terrestrial plants (Helgesen and
Larsen, 1998; Koch et al., 2000; Mattusch et al.,
2000). It seems likely that reduction of As(V) to
As(III) is an essential process for arsenic detoxi-
fication in Brake fern, although As(III) is generally
believed to be more toxic than As(V) to organisms.
Under the reducing environment of plant cells, it
is postulated that As(V) is readily reduced to
As(III). Organic ligands such as thiols, induced
probably by the exposure of the plant to arsenic,
should be able to complex arsenic to avoid the
damage of the plant cells by free As(III). The
presence of this type of organic ligands (chelators)
and their role in the arsenic accumulation and
tolerance by Brake fern is currently being
investigated.
In summary, the discovery of the arsenic hyper-
accumulating plant opens a door for phytoreme-
diation of arsenic-contaminated soils (Ma et al.,
2001) and also provides a unique research oppor-
tunity to understand arsenic uptake, translocation,

transformation, and detoxification. The present
study demonstrates that (1) the plant can accu-
mulate a large amount of arsenic from soils and
transfer it to the aboveground biomass; (2) the
plant contains predominately inorganic arsenic spe-
cies; and (3) conversion of As(V) to As(III)
occurs during the course of arsenic translocation
with 60–74% of arsenic in the fronds as As (III)
compared to only 8.3% in the roots. Further studies
are currently underway to address the mechanisms
of arsenic uptake and transformation.
176 W. Zhang et al. / The Science of the Total Environment 300 (2002) 167–177
Acknowledgments
This research is partially supported by NSF
grant (BES-0086768). We thank John T. Landrum
and Anita Holloway for their assistance for the
preparation of this manuscript. We would also like
to thank the Advanced Mass Spectrometry Facility
(AMSF) at FIU for our access to the ICPyMS.
This is contribution ࠻ 175 of the Southeast Envi-
ronmental Research Center at FIU.
References
Baker AJM. Metal tolerance. New Phytol 1987;106:93–111.
Bech J, Poschenrieder C, Llugany M, Barcelo J, Tume P,
Tobias FJ, Barranzuela JL, Vasques ER. Arsenic and heavy
metal contamination of soil and vegetation around a copper
mine in Northern Peru. Sci Total Environ 1997;203:83–91.
Brooks RR, editor. Plants that hyperaccumulate heavy metals
University Press: Cambridge, 1998.
Cai Y, Georgiadis M, Fourqurean JW. Determination of arsenic

in seagrass using inductively coupled plasma mass spec-
trometry. Spectrochimica Acta Part B 2000;55:1411–1422.
Carbonell-Barrachina AA, Burlo F, Burgos-Hernandez A,
Lopez E, Mataix J. The influence of arsenic concentration
on arsenic accumulation in tomato and bean plants. Scientia
Horticulturae 1997;71:167–176.
Cullen WR, Reimer KJ. Arsenic speciation in the environment.
Chem Rev 1989;89:713–764.
Dahmani-Muller H, van Oort F, Gelie B, Balabane M. Strate-
gies of heavy metal uptake by three plant species growing
near a metal smelter. Environ Pollu 2000;109:231–238.
De Knecht JA, Koevoets LM, Verkleij JAS, Ernst WHO.
Evidence against a role for phytochelatins in naturally
selected increased cadmium tolerance in Silene vulgaris
(Moench) Garcke. New Phytol 1992;122:681 –688.
Dobson AP, Bradshaw AD, Baker AJM. Hopes for the future:
restoration ecology and conservation biology. Science
1997;277:515–522.
Dushenko WT, Bright DA, Reimer KJ. Arsenic bioaccumula-
tion and toxicity in aquatic macrophytes to gold-mine
effluent: relationship with environmental partitioning, metal
uptake and nutrients. Aquatic Botany 1995;50:141–158.
Francesconi KA, Edmonds JS, Morita M. Determination of
arsenic and arsenic species in marine environmental sam-
ples. In: Nriagu JO, editor. Arsenic in the Environment Part
I: Cycling and Characterization Canada: John Wiley and
Sons, 1994:189–219.
Goodwin TW, Mercer EI. Introduction to Plant Biochemistry.
3rd edition Pergamon Press: New York, 1983:1–677.
Grill E, Winnacker EL, Zenk MH. Phytochelatins, a class of

heavy-metal-binding peptides from plants are functionally
analogous to metallothioneins. Proc Natl Acad Sci USA
1987;84:439–443.
Helgesen H, Larsen EH. Bioavailability and speciation of
arsenic in carrots grown in contaminated soil. Analyst
1998;13:791–796.
Kim MJ, Nariagu J. Oxidation of arsenite in groundwater
using ozone and oxygen. Sci Total Environ 2000;247:71–
79.
Kramer U, Cotter-Howells JD, Charnock JM, Baker AJM,
¨
Smith JAC. Free histidine as a metal chelators in plants that
accumulate nickel. Nature 1996;379:635–638.
Koch I, Feldmann J, Wang L, Andrewes P, Reimer KJ, Cullen
WR. Arsenic in the Meager Creek hot springs environment,
British Columbia, Canada. Sci Total Environ 1999;236:101–
117.
Koch I, Wang L, Ollson C, Cullen WR, Reimer KJ. The
predominance of inorganic arsenic species in plants from
Yellowknife, Northwest Territories, Canada. Environ Sci
Technol 2000;34:22–26.
Le XC, Yalcin S, Ma M. Speciation of submicrogram per liter
levels of arsenic in water: on-site species separation inte-
grated with sample collection. Environ Sci Technol
2000;34:2342–2347.
Leopold I, Gunther D, Schmidt J, Neumann D. Phytochelatins
and heavy metal tolerance. Phytochemistry 1999;50:1323–
1328.
Ma LQ, Komart KM, Tu C, Zhang W, Cai Y, Kennelly ED.
A fern that hyperaccumulates arsenic. Nature 2001;409:579.

Macnair MR, Cumbes Q. Evidence that arsenic tolerance in
Holcu lanatus L. is cansed by an altered phosphate uptake
system. New Phytol 1987;107:387–394.
Maeda S. Biotransformation of arsenic in the freshwater
environment. In: Nriagu JO, editor. Arsenic in the Environ-
ment Part I: Cycling and Characterization Canada: John
Wiley and Sons, 1994:155–187.
Maitani T, Kubota H, Sato K, Yamada T. The composition of
metals bound to class III methallothionein (phytochelatins
and its desglucyl peptide) induced by various metals in root
cultures of Rubia tinctorum. Plant Physiol 1996;110:1145–
1150.
Mattusch J, Wennrich R, Schmidt AC, Reisser W. Determina-
tion of arsenic species in water, soils and plants. Fresenius
J Anal Chem 2000;366:200–203.
Meharg AA, Macnair MR. An altered phosphate uptake system
in arsenate-tolerant Holcus lanatus L. New Phytol
1990;116:29–35.
Meharg AA, Macnair MR. Uptake, accumulation and translo-
cation of arsenate in arsenate-tolerant and non-tolerant
Holcus lanatus L. New Phytol 1991a;117:225–231.
Meharg AA, Macnair MR. The mechanisms of arsenate toler-
ance in Deschampsia cespitosa (L.) Beauv. And Agrostis
capillaries L. New Phytol 1991b;119:291–297.
National Research Council. Arsenic in Drinking Water, Nation-
al Academy Press: Washington, DC, 1999.
Nriagu JO, editor. Arsenic in the Environment Part I: Cycling
and Characterization Canada: John Wiley and Sons, 1994.
Perronnet K, Schwarz C, Gerard E, Morel JL. Availability of
cadmium and zinc accumulated in the leaves of Thlaspi

177W. Zhang et al. / The Science of the Total Environment 300 (2002) 167–177
caerulescens incorporated into soil. Plant Soil
2000;227:257–263.
Pickering IJ, Prince RC, George MJ, Smith RD, George GN,
Salt DE. Reduction and coordination of arsenic in Indian
mustard. Plant Physiol 2000;122:1171–1177.
Porter EK, Peterson PJ. Arsenic accumulation by plants on
mine waste (United Kingdom). Sci Total Environ
1975;4:365–371.
Reeves RD, Schwarz C, Morel JL, et Edmondson J. Distribu-
tion and metal-accumulating behavior of Thlaspi caerules-
cens and associated metallophytes in France. International J
Phytoremediation 2001;3:145–172.
Raskin I, Ensley BD, editors. Phytoremediation of Toxic
Metals New York: John Wiley and Sons, Inc, 2000.
Schmoger MEV, Oven M, Grill E. Detoxification of arsenic
¨
by phytochelatins in plants. Plant Physiol 2000;122:793 –
801.
Sneller FEC, Van Heerwaarden LM, Kraaijeveld-Smit FJL,
Ten Bookum WM, Koevoets PLM, Schat H, Verkleij JAC.
Toxicity of arsenate in Silene Vulgaris, accumulation and
degradation of arsenate-induced phytochelatins. New Phytol
1999;144:223–232.
Sneller FEC, Van Heerwaarden LM, Kraaijeveld-Smit FJL,
Koevoets PLM, Schat H, Verkleij JAC. Derivatization of
phytochelatins from Silene vulgaris, induced upon exposure
to arsenate and cadmium: comparison of derivatization with
Ellman’s reagent and monobromobimane. J Agric Food
Chem 2000;48:4014–4019.

Terry N, Banuelos G, editors. Phytoremediation of Contami-
nated Soil and Water Boca Raton: Lewis Publishers, 1999.
Tu C, Ma LQ, Zhang W, Cai Y, Harris WG. 2002 Speciation,
each ability and elemental association of arsenic in Chinese
Brake fern (p. vitta L.). Environ Pollution Submitted.
US EPA, Office of Water. Proposed Revision to Arsenic
Drinking Water Standard, 815-F-00-012, 2001a (http:yy
www.epa.govysafewateryarsyproposalfs.html).
US EPA, announces arsenic standard for drinking water of 10
parts per billion, EPA headquarters press release, Washing-
ton, DC, October 31, 2001b.
US EPA, Office of Research and Development. Introduction
to Phytoremediation, EPAy600yR-99y107. 2000.
Welch AH, Westjohn DB, Helsel DR, Wanty RB. Arsenic in
ground water of the United States: occurrence and geochem-
istry. Ground Water 2000;38:589–604.
Wells JM, Richardson DHS. Anion accumulation by the moss
Hylocomium splendens: uptake and competition studies
involving arsenate, selenate, selenite, phosphate, sulfate and
sulfite. New Phytol 1985;101:571–583.