www.publish.csiro.au/journals/env S. Foster et al., Environ. Chem. 2005, 2, 177–189. doi:10.1071/EN05061
Distribution and Speciation of Arsenic in Temperate MarineSaltmarsh Ecosystems
Simon Foster,A,B William Maher,A Anne Taylor,A Frank Krikowa,A and Kristy TelfordA
A Ecochemistry Laboratory, Institute of Applied Ecology, University of Canberra, Belconnen, 2601, Australia.B Corresponding author. Email: [email protected]
Environmental Context. The pathways by which arsenic is accumulated and transferred in aquatic ecosys-tems are relatively unknown. Examination of whole marine ecosystems rather than individual organismsprovides greater insights into the biogeochemical cycling of arsenic. Saltmarshes with low ecological diver-sity are an important terrestrial–marine interface about which little is known regarding arsenic concentrationsand species distribution. This study examines the cycling of arsenic within Australian saltmarsh ecosystemsto further understand its distribution and trophic transfer.
Abstract. This paper reports the distribution of total arsenic and arsenic species in saltmarsh ecosystems locatedin south-east Australia. We also investigated the relationship between arsenic, iron, and phosphorus concentrationsin saltmarsh halophytes and associated sediment.
Total mean arsenic concentrations in saltmarsh plants, S. quinqueflora and S. australis, for leaves rangedfrom 0.03 ± 0.05 to 0.67 ± 0.48 µg g−1 and 0.03 ± 0.02 to 0.08 ± 0.06 µg g−1, respectively, and for roots rangedfrom 2 ± 2 to 6 ± 12 µg g−1 and 0.39 ± 0.20 to 0.57 ± 1.06 µg g−1 respectively. Removal of iron plaque from theroots reduced the arsenic concentration variability to 0.40–0.79 µg g−1 and 0.95–1.05 µg g−1 for S. quinquefloraand S. australis roots respectively. Significant differences were found between locations for total arsenic con-centrations in plant tissues and these differences could be partially attributed to differences in sediment arsenicconcentrations between locations. For S. quinqueflora but not S. australis there was a strong correlation betweenarsenic and iron concentrations in the leaf and root tissues. A significant negative relationship between arsenic andphosphorus concentrations was found for S. quinqueflora leaves but not roots.
Total mean arsenic concentrations in salt marsh animal tissues (7 ± 2–21 ± 13 µg g−1) were consistent with thosefound for other marine animals. The concentration of total arsenic in gastropods and amphipods could be partiallyexplained by the concentration of total arsenic in the dominant saltmarsh plant S. quinqueflora.
Of the extractable arsenic, saltmarsh plants were dominated by arsenic(iii), arsenic(v) (66–99%), and glycerolarsenoribose (17–35%). Arsenobetaine was the dominant extractable arsenic species in the gastropods Salinatorsoilda (84%) and Ophicardelus ornatus (89%) and the crab Neosarmatium meinerti (89%). Amphipods containedmainly arsenobetaine (44%) with some phosphate arsenoribose (23%). Glycerol trimethyl arsonioribose was foundin both gastropods (0.7–0.8%) and the visceral mass of N. meinerti (0.1%).
These results show that arsenic uptake into plants from uncontaminated saltmarsh environments maybe dependenton plant iron uptake and inhibited by high phosphorus concentrations. Arsenic in saltmarsh plants is mainly presentas inorganic arsenic, but arsenic in animals that eat plant detritus is present as organo arsenic species, primarilyarsenobetaine and arsenosugars. The presence of glycerol trimethyl arsonioribose poses the question of whethertrimethylated arsonioriboses are transitory intermediates in the formation of arsenobetaine.
Keywords. arsenic — iron — phosphorus — speciation (nonmetals)
Manuscript received: 2 August 2005.Final version: 12 August 2005.
Introduction
Saltmarshes occur from the mean high tide to the upper springtide level.[1] Saltmarshes are common features of estuariesin many parts of the world and form the interface betweenmangrove forests and terrestrial ecosystems. They are depo-sitional zones that support large populations of crabs and,
when inundated, are thought to be important feeding groundsfor juvenile fish.[2] Saltmarshes in south-easternAustralia aredominated by only two species of halophyte plants (Sarco-cornia quinqueflora and Suaeda australis), two gastropodspecies (Salinator soilda and Ophicardelus ornatus), anda crab (Neosarmatium meinerti).[2–4] Saltmarsh food webs
Fig. 1. Arsenic cycling within a saltmarsh ecosystem.
are detrital based (Fig. 1) with both the gastropods andcrabs being detritovores.[5,6] Although little information isavailable on other invertebrates inhabiting these ecosystems itis known that supralittoral species of amphipods also occur ingreat numbers and are an important component of the detritalfood web.[5,7]
Marine and intertidal organisms have been shown to accu-mulate high concentrations of arsenic.[8] It is well establishedthat arsenobetaine (AB) is the dominant form of arsenic foundin marine animals,[8] while marine macroalgae[9] and marineherbivores often contain high concentrations of dimethylarsi-noriboses (see Fig. 2 for arsenic species structures).[10–12] Todate the occurrence of AB in angiosperms from uncontam-inated sites has not been conclusively demonstrated.[13–15]The presence of AB in temperate mangrove gastropods hasbeen reported by Kirby et al.,[12] who found that AB madeup 50–60% of the total arsenic. Similarly, Goessler et al.[16]found AB to be the major arsenic compound in several rockyintertidal gastropods. AB has also been found in the marineamphipod Allorchestes compressa.[17]
The proposed pathways for the formation of arsenobetaineinvolve the biotransformation of arsenoriboses toAB.[8,18–20]The main pathway for the formation of arsenobetaine isthought to be through the degradation of dimethylarsinori-boses to dimethylarsinylethanol (DMAE), which is furtherconverted into either dimethylarsinylacetic acid (DMAA)or arsenocholine.[8,18–20] Further oxidation of arsenocholineat the primary alcohol group results in the formation ofarsenobetaine.[21] However, the discovery of trimethylatedarsonioriboses in macroalgae[22,23] and gastropods,[10,24] andrecently in herbivorous fish,[10] has provided a simpler path-way for the formation of arsenobetaine.This pathway was firstproposed by Francesconi and Edmonds[18] with the view thattrimethylated arsonioriboses occurred in algae, but algae haveonly been shown to contain trace amounts of these arsenosugars.[22,23]
Iron and phosphorus are elements often thought to beassociated with arsenic uptake in wetland plants.[25] Iron
oxides (plaque) are commonly formed on the roots of wet-land/marsh plants.[12] This may reduce or enhance the uptakeof arsenic.[26] Arsenic and phosphate are structurally similarspecies.[25,27] They are therefore, thought to share the sameuptake pathways and their concentrations are often found tobe highly correlated in plant and algal tissues.[27–29]
In this study we report the distribution of total arsenicand arsenic species in saltmarsh ecosystems located in south-east Australia. We also investigated the relationship betweenarsenic, iron, and phosphorus concentrations in saltmarshhalophytes and associated sediments.
Materials and Methods
Sampling and Sample Preparation
Samples were collected from three locations (Tomago River, MoruyaRiver, and Congo Creek) on the south-east coast of New South Wales,Australia. Each of the saltmarshes was located on open estuaries andoccupied areas <0.5 km2. Plant samples collected were S. quinqueflora(commonly known as Samphire) and S. australis (commonly known asSea Blight). Gastropod species collected were S. soilda and O. orna-tus. The common saltmarsh crab N. meinerti was collected as well asan amphipod belonging to the family Talitridae.[30] On a separate occa-sion the roots of S. quinqueflora and S. australis were re-sampled todetermine the concentration of arsenic after the removal of iron plaque.
Plant samples were collected whole, including roots and associatedsediment, and placed into acid-washed plastic zip-lock bags. Gas-tropods, crabs, and amphipods were collected by hand and kept alive onice until arrival at the laboratory.
Plant samples were separated into sediment, root, and leaf tissueand rinsed in deionized water. Gastropods were cracked gently using asmall bench vice and extracted whole. Gastropod tissues were rinsedclean of shell grit and placed in individual acid-washed vials. Crabswere frozen and the top shell was removed to expose the visceral mass,which was removed and placed into acid-washed vials. Muscle tissuewas obtained from the claws of the crabs. Amphipods were rinsed indeionized water and 3–5 individuals placed into each acid washed vial.All samples were immediately frozen (∼−80◦C) then freeze-dried for∼48 h (Labconco) to a constant mass. Plant tissues were homogenizedusing a Retsch ZM100 mill (0.2 mm stainless steel mesh, Retsch), andstored in clean polyethylene vials in a desiccator until analyzed.All othersamples were homogenized using liquid nitrogen in an agate mortar andpestle due to very low sample masses (0.1–0.2 g).
178
RESEARCH FRONT
Distribution and Speciation of Arsenic in Temperate Marine Saltmarsh Ecosystems
AsHO OH
OH
As
O
OH
OHHO As
O
OH
CH3HO As
O
CH3HO
CH3
AsIII
ArseniteAsV
ArsenateMA
MethylarsonateDMA
Dimethylarsinate
As
O
CH3
CH3
CH3
CH3
O OR
HO OH
OH
OHR �
R �
R �
R �
OH
OP
O O�
O
OH
OH
OH
SO3�
OH
OSO3�
As O OR
HO OH
Dimethylated arsenoribosides:
Trimethylated arsonioribosides:
As�
As� As� As�
CH3H3C
H3C
H3C
H3C
H3C
H3C H3C
H3C H3C
CH3
CH3
TETRATetramethylarsonium ion
As
O
CH3
OH
DMAEDimethylarsinoylethanol
As
O
CH3
OH
O
DMAADimethylarsinoyl acetic acid
CH3
OH
CH3
ACArsenocholine
Glycerol
Phosphate
Sulfate
Sulfonate
CH3
CH3
ABArsenobetaine
OH
OCH3
CH3
TMAPTrimethylarsoniopropionate
OH
O
As
S
O OR
HO OH
Dimethylated thio-arsenoribosides:
Fig. 2. Chemical structures, names, and abbreviations of arsenic species.
Sample Analysis
Reagents and Standards
Nitric acid (HNO3; Aristar, BDH) was used for the determina-tion of total arsenic concentrations. Ammonium dihydrogen phos-phate (Suprapur, Merck) and pyridine (Extra Pure, Merck) wereused in the preparation of high-pressure liquid chromatography(HPLC) mobile phases. Formic acid (Extra Pure, Fluka) and ammo-nia solution (>99.9%, Aldrich) were used for the adjustment ofmobile-phase pH. Methanol (HiPerSolv, BDH), phosphoric acid(AR grade, BDH), acetone (Unichrom, Ajax Laboratory Chemicals),and deionized water (18.2 M� cm, Millipore) were used for theextraction of arsenic species. Hydrochloric acid (HCl; Trace Pur,Merck), l-cysteine (BioChemika, Fluka), and sodium tetrahydrobo-rate (NaBH4) (Laboratory Chem Chemicals, APS) were used for thereduction and derivatization of arsenic species. The NaBH4 solution
was stabilized by the addition of 0.01 M sodium hydroxide (Pronalys,Selby-Biolab) in deionized water. Dithionite-citrate-bicarbonate (DCB)solution contained 0.03 M sodium citrate, 0.125 M sodium bicarbon-ate (AR grade, BDH) with 1.5 g of sodium dithionite (Ajax LaboratoryChemicals) added to the solution.
Stock standard solutions (1000 mg L−1) of arsenous acid (As3+),arsenic acid (As5+), methylarsonic acid (MA), and dimethylarsinic acid(DMA) were prepared by dissolving sodium arsenite, sodium arsenateheptahydrate (Ajax Laboratory Chemicals), disodium methyl arsonate,and sodium dimethylarsenic (Alltech-Specialists), respectively, in0.01 M HCl–deionized water. Synthetic arsenobetaine (BCR-626, Insti-tute for Reference Materials and Measurements) was diluted withdeionized water to desired concentration.Arsenocholine (AC), trimethy-larsine oxide (TMAO), tetramethylarsonium ion (TETRA), and glyceroltrimethylated arsonioribose were kindly supplied by Professor Erik
179
RESEARCH FRONT
S. Foster et al.
Larsen (Danish Institute for Food and Veterinary Research, Departmentof Chemistry, Denmark), Professor Kevin Francesconi and Dr WalterGoessler (Institute of Food Chemistry, Karl-Franzens-University, Graz,Austria). Glycerol arsenoribose, sulfonate arsenoribose, and sulfatearsenoribose (OH-ribose, SO−
3 -ribose, OSO−3 -ribose respectively) were
isolated in-house from the marine macroalgae certified reference mate-rial Fucus 140 (IAEA). The phosphate arsenoribose (PO4-ribose) wasisolated in-house from the marine animal certified reference materialOyster 1566a (NIST). The identity of these arsenoriboses was pre-viously confirmed by high-performance liquid chromatography-massspectrometry (HPLC-MS).[31]
Trimethylarsoniopropionate (TMAP) was isolated in-house fromthe marine animal certified reference material lobster hepatopancreas(TORT-2; NRC-CNRC).[32]
Removal of Iron Plaque from Roots of Plant Material
Iron plaque was removed from the roots of S. quinqueflora and S. aus-tralis using dithionite-citrate-bicarbonate (DCB) extraction.[26] Briefly,1–2 g of previously deionized water washed root tissue was extractedwith 10 mL of DCB solution for 25 min at 60◦C, the supernatant wasnot further analyzed. After the extraction, plant root tissues were rinsedin deionized water, freeze-dried, and ground to a homogenous powder.Analysis for arsenic, iron, and phosphorus was as described below.
Total Arsenic, Iron, and Phosphorus Analysis
Tissues were digested using a microwave digestion proceduredescribed previously by Baldwin et al.[33] Approximately 0.1 g of freeze-dried tissue was weighed into 7 mL Teflon polytetrafluroacetate diges-tion vessels (A. I. Scientific), and 1 mL of concentrated HNO3 (Aristar,BDH) added. Digestion was carried out using an MDS-81D microwaveoven (CEM) with a time programme consisting of three steps: 2 min at600 W, 2 min at 0 W, and 45 min at 450 W. After digestion, vessels wereallowed to cool at room temperature (∼25◦C) for ∼60 min and thendiluted with deionized water to 10 mL in polyethylene vials.Total arsenic(m/z 75), iron (m/z 56/57), and phosphorus (m/z 15) concentrationswere determined with a Perkin-Elmer Elan-6000 inductively coupledplasma mass spectrometer (ICP-MS). Internal standards (45Sc, 103Rh)were added on-line to compensate for any acid effects and instrumentdrift.[31] The potential interference to arsenic (m/z 75) from 40Ar35Cl+was determined by monitoring chloride at m/z 35, 35Cl16O+ at m/z 51,35Cl17O+ at m/z 52 and 40Ar37Cl+ at m/z 77. Selenium was monitoredat m/z 82 as a cross check on 40Ar37Cl+. Calibration standards (0, 1, 10,100, 1000 µg L−1) for the determination of total arsenic, iron and phos-phorus were prepared daily by appropriate dilution of the multi-elementcalibration standard (Accu Trace, Calibration Standard 2, 10 mg L−1).
Total arsenic in the terrestrial plant tissues was determined usinghydride generation-ICP-MS. Digests were diluted to 1% (v/v) HNO3,and 1 mL of 1% (w/v) l-cysteine was added to 9 mL of digest.The carriersolution was 1% (v/v) HCl, 2% NaBH4 stabilized with 0.01 M NaOH.Germanium(iv) was used as the internal standard and monitored to deter-mine instrument drift and hydride efficiency.[34] Calibration standardsfor the determination of arsenic by hydride generation were made bydiluting arsenic(v) to 0, 0.01, 0.1, 1, 10 µg L−1 with 1 mL of 1% (w/v)l-cysteine to 9 mL of standard.
Certified reference materials analyzed for arsenic were NISTSRM 1566a oyster tissue, NIST 1572 citrus leaves, NRCC DORM-2Dogfish muscle, and NRCC PACS-2 Marine sediment. Measuredarsenic concentrations (mean ± s.d.; n = 6) from the certified refer-ence material were for oyster tissue: measured, 14.5 ± 0.3 µg (As) g−1,certified 14.0 ± 1.2 µg (As) g−1; citrus leaves: measured, 3.6 ±0.1 µg (As) g−1, certified 3.1 ± 0.3 µg (As) g−1; DORM-2: measured,18.0 ± 1.4 µg (As) g−1, certified 18.0 ± 1.1 µg (As) g−1 and PACS-2:measured 27.2 ± 0.3 µg (As) g−1, certified 26.2 ± 1.5 µg (As) g−1.
Arsenic Speciation
Acetone Extraction
Approximately 0.1–0.2 g of homogenized freeze-dried whole-tissuewas added to 50 mL polypropylene vials and 10 mL of acetone added.
The samples were then agitated on a mixing wheel for 1 h and thesupernatant was removed after centrifuging at 3500 rpm for 15 min. Theextraction procedure was repeated twice, with the supernatant removedafter each centrifugation. After the final acetone extraction the residuepellet was dried to a constant mass in a fume cabinet at room temperature(∼25◦C).
The entire combined acetone supernatant was evaporated in a fumecabinet at room temperature (∼25◦C) to dryness. The residue was resus-pended in 0.5 mL concentrated HNO3, and digestion was undertaken ina hot water bath (90◦C) for 1 h. Digested acetone extracts were allowedto cool at room temperature (∼25◦C) and then diluted with deionizedwater to 5 mL in 10 mL polyethylene vials. Total arsenic concentrationwas determined by ICP-MS.
Methanol/Water Extraction
Water-soluble arsenic species were extracted from biological mate-rial by a microwave extraction procedure developed by Kirby andMaher.[32] Approximately 0.1 g of the acetone extracted pellet wasweighed into 50 mL polypropylene vials and 10 mL of 50% (v/v)methanol/deionized water added. Mixtures were loaded into the carouselof an MDS-200 microwave oven (CEM) and heated to 70–75◦C for15 min. The supernatant was removed after centrifuging at 3500 rpmfor 15 min. The procedure was repeated twice and the supernatants werecombined.
The methanol/water supernatant (25 mL) was evaporated to drynessusing an RVC 2-18 rotational vacuum concentrator (50◦C; Christ, Aus-tralia) and stored in a refrigerator (∼4◦C) until speciation analysis. Theremaining 5 mL of the methanol/water supernatant was evaporated todryness using a RVC 2-18 rotational vacuum concentrator (50◦C), andresuspended in 0.5 mL 10% HNO3, and then further diluted to 5 mL fortotal arsenic analysis by ICP-MS.
Phosphoric Acid Extraction
Water-soluble arsenic species were extracted from sediments usingphosphoric acid.[35] Sediment was extracted in 50 mL polyethylene cen-trifuge tubes. Approximately 0.1 g of sediment was weighed into eachtube, to which 10 mL of 0.5 M of phosphoric acid was added. The sam-ples were then agitated on a mixing wheel for 1 h and the supernatantremoved after centrifuging at 3500 rpm for 15 min.
Water-Soluble Arsenic Speciation
Prior to chromatography, previously stored methanol/water extractedresidues were resuspended in 0.5–10 mL deionized water to an arsenicconcentration ∼500 µg L−1. All extracts were filtered through a0.20 µm RC syringe filter (Millipore). Sediment digests were diluted1 : 1 with deionized water prior to analysis. Aliquots of 100 µL wereinjected onto a HPLC system consisting of a Perkin-Elmer Series 200mobile phase delivery and auto sampler system (Perkin-Elmer). Theeluent from HPLC columns was directed by PEEK (polyether–ether–ketone; i.d. 0.02 mm; Supelco) capillary tubing into a Rhyton crossflownebulizer of a Perkin-Elmer Elan-6000 ICP-MS, which was used as anarsenic-selective detector by monitoring the signal intensity at m/z 75.Potential polyatomic interferences were checked by monitoring for otherions as described for total arsenic analysis.
The column conditions used for the separation of arsenic species[34]were:
(1) A PEEK Hamilton PRP-X100 anion-exchange column (250 by4.6 mm, 10 µm, Phenomenex) and an aqueous 20 mM NH4H2PO4mobile phase adjusted to pH 5.6 with aqueous ammonia (flowrate: 1.5 mL min−1; temperature: 40◦C) were used for the identi-fication and quantification of arsenic(v), DMA, MA, PO−
4 , SO−3 ,
and OSO−3 arsenoribosides.
(2) A Supelocosil LC-SCX cation-exchange column (250 by 4.6 mm,5 µm, Supelco, USA) and an aqueous 20 mM pyridine mobilephase adjusted to pH 2.6 with formic acid (flow rate: 1.5 mL min−1;temperature: 40◦C) were used for the identification of AB, glycerolarsenoriboside, glycerol trimethyl arsonioribose, TETRA, AC, andTMAP.
180
RESEARCH FRONT
Distribution and Speciation of Arsenic in Temperate Marine Saltmarsh Ecosystems
Arsenic(iii) was measured using identical ICP-MS conditions with theuse of an in-line hydride generation system.[34]
External calibration curves for quantification of arsenic specieswere prepared by diluting arsenic(v) for anionic species and AB forcationic species to 0, 1, 10, 100 µg L−1 daily. Purity of arsenic specieswas periodically determined by HPLC-ICP-MS. Typical precision forreplicate analysis of arsenic species is: AB (10 µg L−1, coefficient ofvariance (CV) <1%); TMAP (0.2 µg L−1, CV <6%); AC (0.05 µg L−1,CV <8%); TETRA (0.2 µg L−1, CV <8%); arsenic(iii) (0.5 µg L−1,CV <5%); DMA (0.2 µg L−1, CV <10%); MA (0.2 µg L−1, CV<5%); arsenic(v) (0.5 µg L−1, CV <6%); arsenoriboses (1 µg L−1, CV<15%).
The chromatography package Total Chrom (Perkin-Elmer) was usedto quantify arsenic species by peak areas.Arsenic species were identifiedby spiking with known standards and comparisons of retention times.
The accuracy of arsenic speciation procedure was determinedby the analysis of the certified reference material, DORM-2. Theconcentrations (mean ± s.d.) of AB (16.1 ± 0.2 µg g−1) and TETRA(0.26 ± 0.04 µg g−1) measured in DORM-2 tissues were similar to cer-tified values (AB, 16.4 ± 1.1 µg g−1; TETRA, 0.248 ± 0.054 µg g−1).
Table 1. Total arsenic concentrations in saltmarsh sediment and plant and animal tissues
Sediment 0.54 ± 0.16 1.3 ± 0.3 1.3 ± 1.4 1.0 ± 0.9a b b
S. australis Leaves 0.08 ± 0.06 0.03 ± 0.02 0.06 ± 0.12 0.06 ± 0.08a a a
Roots 0.39 ± 0.26 0.45 ± 0.28 0.57 ± 1.06 0.47 ± 0.63a a a
Sediment 1 ± 2 1.2 ± 0.3 1.3 ± 0.8 1.3 ± 1.1a a a
S. soilda Whole 35 ± 15 15 ± 8 14 ± 2 21 ± 13a b b
O. ornatus Whole 31 ± 16 20 ± 9 9 ± 2 20 ± 14a a b
N. meinerti Muscle 13 ± 10 14 ± 5 9 ± 4 12 ± 6a a a
Visceral 16 ± 6 13 ± 5 9 ± 1 13 ± 5a a a
Talitrid amphipod Whole 8 ± 2 4.9 ± 0.3 6.3 ± 0.2 7 ± 2a c b
TR: Tomago River; MR: Moruya River; CR: Congo Creek. 1 n = 10; 2 n = 30; 3 Turkey’s post hoc analysis ofsignificance (<0.05) between locations. Different letters indicates significant difference (<0.05) where a > b > c.
Table 2. One-way analysis of variance (ANOVA) for total arsenic concentrations of saltmarsh plant and animal tissues
Common name Species Tissue VariableA d.f. As
MS F PC
Samphire S. quinqueflora Leaves Loc 2 27.44 40.64 <0.001Roots Loc 2 3.36 2.17 NSB
Sediment Loc 2 1.99 4.46 <0.05Sea blight S. australis Leaves Loc 2 4.64 3.15 NS
Roots Loc 2 0.36 0.51 NSSediment Loc 2 0.69 1.54 NS
Solid air-breather S. soilda Whole Loc 2 2.39 16.32 <0.001Mangrove air-breather O. ornatus Whole Loc 2 3.37 18.14 <0.001Mangrove crab N. meinerti Muscle Loc 2 0.20 0.88 NS
Visceral Loc 2 0.30 3.28 NSLand/sand hopper Talitrid amphipod Whole Loc 2 0.32 21.46 <0.001
A Loc = location. B NS = not significant. C Significance level was set at 0.05.
Data Analysis
Significant differences of total arsenic concentrations between locationswere determined by single-factorANOVA (factor = location) with a sig-nificance level of P = 0.05 applied to normalized data (SPSS 11.5).Linear regression analysis was carried out using SPSS 11.5 with asignificance level of P = 0.05.
Cluster analysis and principal component analysis (PCA) were usedto classify groups with similar arsenic species proportions (Primer 5;PRIMER-E, Plymouth, UK).[36]
Results
Total Arsenic Concentrations
Plants and Associated Sediments
In general, arsenic concentrations in plant tissues werehighest at Tomago River, intermediate at Moruya River andlowest at Congo Creek (Table 1). Significant differences
Fig. 3. Total arsenic (a), iron (b), and phosphorus (c) concentrations insaltmarsh sediment, plants, and animal tissues. Box = mean, 25th and75th percentiles; whiskers = 5th and 95th percentiles.
in arsenic concentrations between locations were found forS. quinqueflora leaves and associated sediment (Table 2).
No linear relationship was found between sediment androot (r2 = 0.001, P > 0.05) or sediment and leaf arsenicconcentrations (r2 = 0.06, P > 0.05) of S. quinqueflora. Theconcentration of arsenic in the roots of S. quinqueflora was
As
(µg
g�1 )
0
2
4
6
8
(a)
(b)
As
(µg
g�1 )
0
1
2
3
4
TR MR CR
Fig. 4. Box plot showing the difference in arsenic concentration vari-ability before (grey boxes) and after (white boxes) removal of ironplaque from the roots of (a) S. quinqueflora; and (b) S. australis(n = 5). TR =Tomago River; MR = Moruya River; CR = Congo Creek.Box = mean, 25th and 75th percentiles; whiskers = 5th and 95thpercentiles.
highly variable (Figs 3 and 4a); however, after the removal ofiron plaque from the roots surface of these plants, variabilitywas reduced although the overall mean arsenic concentrationremained similar (Fig. 4a).
No significant difference in arsenic concentrationsbetween locations was found for S. australis roots and leaves(Table 2). Only a weak linear relationship was found betweensediment and root (r2 = 0.22, P < 0.05) and sediment andleaf (r2 = 0.18, P < 0.05) arsenic concentrations. S. australisroots have less iron plaque, and removal of iron plaque hada lesser effect on the variability of arsenic concentrations inthe roots (Fig. 4b).
Animals
Arsenic concentrations in animals were in general highestat Tomago River, intermediate at Moruya River, and lowestat Congo Creek (Table 1). Significant differences in arsenicconcentrations between locations were found for S. soilda,O. ornatus, and amphipods (Table 2). Arsenic concentrationsincreased from producers to consumers (Fig. 3). Large vari-ability in arsenic concentrations occurred for both gastropodspecies (Fig. 3) with significant differences between locations
182
RESEARCH FRONT
Distribution and Speciation of Arsenic in Temperate Marine Saltmarsh Ecosystems
S. quinqueflora (As �g g�1) S. quinqueflora (As �g g�1)
Fig. 5. Linear relationship between total arsenic concentration in saltmarsh plant leaves and animal tissues (n = 30). Dotted lines represent 95%confidence limits. (a) S. soilda; (b) O. ornatus; (c) amphipod.
Fig. 6. Linear relationship between arsenic and iron concentrations in saltmarsh plant tissues (n = 30). Dotted lines represent 95% confidence limits.(a) S. quinqueflora leaves; (b) S. quinqueflora roots; (c) S. australis leaves; (d) S. australis roots.
183
RESEARCH FRONT
S. Foster et al.
0 500 1000 1500 2000 2500
As
(�g
g�1 )
As
(�g
g�1 )
As
(�g
g�1 )
As
(�g
g�1 )
�1.0
�0.5
0.0
0.5
1.0
1.5
2.0
P (�g g�1) P (�g g�1)
500 1000 1500 2000 2500 3000�2
0
2
4
6
8
10
1000 1500 2000 2500 3000 3500 4000 4500 5000�0.1
0.0
0.1
0.2
0.3
0.4
0.5
0 500 1000 1500 2000 25000.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
(a) (c)
(d)(b)
r 2 � �0.27P � �0.01
r 2 � 0.12 P � �0.05
r 2 � �0.03 P � NS
r 2 � 0.16 P � �0.05
Fig. 7. Relationship between arsenic and phosphorus concentrations in saltmarsh plant tissues (n = 30). Dotted lines represent 95% confidencelimits. (a) S. quinqueflora leaves; (b) S. quinqueflora roots; (c) S. australis leaves; (d) S. australis roots.
(Table 1). No significant relationships were found betweenthe tissue arsenic concentration of S. australis and herbivoresand detritivores. Significant linear relationships were foundbetween the tissue arsenic concentrations of S. quinquefloraand S. soilda, O. ornatus and the amphipods (Fig. 5a–c).
Relationship Between Arsenic, Iron, and PhosphorusConcentrations in Plants
Significant linear relationships between arsenic and iron con-centrations were found for both S. quinqueflora leaves androots (Fig. 6a,b) and S. australis roots (Fig. 6d), but not forS. australis leaves (Fig. 6c). The linear relationship betweenarsenic and iron concentrations in the roots of S. quinqueflorawas still evident after the removal of iron plaque from the rootsurface (r2 = 0.86, P < 0.05).The linear relationship betweenarsenic and iron in the roots of S. australis was enhanced afterthe removal of iron plaque from the root surface (r2 = 0.93,P < 0.01).
A significant negative linear relationship was foundbetween arsenic and phosphorus concentrations for S. quin-queflora leaves (Fig. 7a); however, no significant relationshipwas found for arsenic and phosphorus concentrations forS. quinqueflora roots (Fig. 7b). Although significant linearrelationships between arsenic and phosphorus concentrationswere found for the leaves and roots of S. australis, these rela-tionships were weak and much of the data lie outside the 95%confidence intervals (Fig. 7c,d).
Water-Soluble Arsenic Species
Arsenic concentrations in acetone extracts were low(<0.18 ± 0.21 µg g−1; 0.5–3%) and not analyzed further.Thewater-soluble arsenic species in sediments and plant materialwere dominated by inorganic arsenic (Table 3). However, inleaves and roots, glycerol arsenoribose was also present atappreciable concentrations ranging from 17 to 35% of thetotal extractable arsenic (Table 3). Both gastropods containedmostly AB (Table 3). Glycerol trimethyl arsonioribose waspresent in both the gastropods and N. meinerti visceral massat low concentrations (0.1–0.8%) (Table 3, Fig. 8). The pres-ence of glycerol trimethyl arsonioribose was confirmed byspiked addition of a synthetic standard to the original samplewith no distortion in peak shape found (Fig. 8).The amphipodwas dominated by AB (44%), with phosphate arsenoribosemaking up 23% of the total extracted arsenic (Fig. 9). Anarsenic peak at a retention time of 25 min was found in bothO. ornatus (0.4%) and N. meinerti visceral mass (2.2%),which did not match any of our arsenic standards.
Principal component analysis of arsenic species propor-tions separated the saltmarsh organisms into three groups(Fig. 10). Group one consisted of both gastropods andN. meinerti muscle tissue, which were grouped togetherbased on the similarity of AB proportion within these sam-ples (Table 4). Although in the same group, both gastropodspecies are separated from N. meinerti muscle tissue owingto the presence of glycerol trimethyl arsonioribose (Fig. 10).Group two consisted of the amphipod and N. meinerti visceral
184
RESEARCH FRONT
Distribution and Speciation of Arsenic in Temperate Marine Saltmarsh Ecosystems
Tab
le3.
Wat
er-s
olub
lear
seni
csp
ecie
sin
salt
mar
shse
dim
ent
and
plan
tan
dan
imal
tiss
ues
from
Tom
ago
Riv
er
Spe
cies
Tis
sue
Tota
lAs
Ext
ract
edC
olum
nA
BB
OH
-T
riO
H-
TM
AP
AC
TE
TR
AIn
orga
nic
DM
AM
AP
O− 4
OS
O− 3
Unk
(µg
g−1)
As
(%)
reco
very
ribo
seri
bose
As
ribo
seri
bose
As
anio
n(%
)A
S.qu
inqu
eflo
raL
eave
s1.
739
65n.
d.C
0.14
(32)
n.d.
n.d.
n.d.
n.d.
0.29
(66)
0.01
(2.1
)n.
d.n.
d.n.
d.n.
d.R
oots
5.4
1071
n.d.
0.07
(17)
n.d.
n.d.
n.d.
n.d.
0.31
(76)
0.02
(5.4
)n.
d.0.
01(1
.5)
n.d.
n.d.
Sed
imen
t7.
895
53n.
d.n.
d.n.
d.n.
d.n.
d.n.
d.3.
9(9
9)n.
d.0.
07(1
.1)
n.d.
n.d.
n.d.
S.au
stra
lis
Lea
ves
0.33
9984
n.d.
0.07
(27)
n.d.
n.d.
n.d.
n.d.
0.19
(68)
0.01
(4.2
)n.
d.n.
d.n.
d.n.
d.R
oots
0.73
3511
1n.
d.0.
10(3
5)n.
d.n.
d.n.
d.n.
d.0.
18(6
5)n.
d.n.
d.n.
d.n.
d.n.
d.S
edim
ent
5.8
5585
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
2.7
(100
)n.
d.n.
d.n.
d.n.
d.n.
d.S.
soil
daW
hole
4410
499
38(8
4)n.
d.0.
36(0
.8)
0.44
(1.0
)n.
d.1.
0(2
.3)
0.45
(1.0
)0.
69(1
.5)
n.d.
4.0
(9.0
)n.
d.n.
d.O
.orn
atus
Who
le51
104
9344
(89)
n.d.
0.36
(0.7
)0.
22(0
.5)
n.d.
1.0
(2.1
)1.
4(3
.0)
0.14
(0.3
)n.
d.2.
0(4
.2)
n.d.
0.19
(0.4
)N
.mei
nert
iM
uscl
e14
7410
09.
2(8
9)n.
d.n.
d.0.
22(2
.1)
0.00
7(0
.1)
0.05
(0.5
)0.
58(6
.0)
n.d.
n.d.
0.23
(2.3
)n.
d.n.
d.V
isce
ral
1379
896.
0(6
8)n.
d.0.
01(0
.1)
0.12
(1.4
)0.
01(0
.1)
0.04
(0.4
)1.
5(1
6)n.
d.n.
d.0.
63(7
.0)
0.41
(4.6
)0.
19(2
.2)
Tali
trid
amph
ipod
Who
le6.
372
581.
2(4
4)n.
d.n.
d.0.
03(1
.1)
0.00
5(0
.2)
0.01
(0.4
)0.
45(1
6.5)
0.01
(0.3
)n.
d.0.
6(2
3)0.
39(1
4.7)
n.d.
AC
olum
nre
cove
ries
are
the
tota
lsu
mof
spec
ies
elut
edfr
omth
eco
lum
nv.
the
arse
nic
extr
acte
dfr
omth
esa
mpl
e.B
Ars
enic
quot
edin
µg
g−1
(dry
mas
s)w
ith
the
perc
enta
geof
the
tota
lsu
mof
spec
ies
inbr
acke
ts.C
n.d.
=no
tqua
ntif
iabl
e<
0.00
5µ
gg−
1fo
ral
lspe
cies
exce
ptin
orga
nic
arse
nic<
0.00
03µ
gg−
1.
mass, which both contained sulfate arsenoribose. Althoughnot shown on the PCA, amphipods are discriminated fromN. meinerti visceral mass in the third dimension wherephosphate arsenoribose is influencing its position (Table 4).Group three contained the plants and sediment based on thedominance of inorganic arsenic (Fig. 10).
Discussion
Total Arsenic Concentrations
Plant and Associated Sediment
Sediment arsenic concentrations are similar to those foundby Kirby et al.[12] in Australian mangrove ecosystems, but atthe lower range of those found by other authors who haveanalyzed uncontaminated marine sediments.[37,38] The con-centration of arsenic in S. quinqueflora and S. australis arealso similar to those measured in wetland plants from othernon-contaminated sites.[15,39] The variability in leaf and rootarsenic concentrations in S. quinqueflora between locationscould not be explained by the arsenic concentrations in theassociated sediments. In contrast, tissue arsenic concentra-tions in S. australis roots and leaves were found to be lowerand correlated (r2 = 0.22, P < 0.05 and r2 = 0.18, P < 0.05respectively) with the sediment arsenic concentrations.
Differences in total arsenic concentrations of S. quin-queflora root tissues appear to be due to the formation ofiron plaque on the roots and adsorption of arsenic (Fig. 6b).This does not appear to occur as strongly for S. australis(Fig. 6d). This may be due to differences in the redox condi-tions between the rhizospheres of these two species or relatedto plant root physiology. As previously found for mangroveleaves and roots,[12] plant roots analyzed after only washingwith deionized water contained highly variable arsenic con-centrations, which were much higher than the plants leaves.Removal of iron plaque from plant roots analyzed in thisstudy reduced this variability while the overall mean arsenicconcentrations remained similar (Fig. 4a,b), indicating mostof this additional arsenic was associated with iron plaquethat forms in the oxidized root environment.[25] The extent ofiron plaque formation is probably the important environmen-tal mechanism controlling arsenic accumulation in saltmarshplants.
Animals
The range of arsenic concentrations found in gastropods,crabs, and amphipods[17] are similar to arsenic concentra-tions reported for marine/terrestrial herbivorous gastropodsand crabs,[12] but lower than the concentrations normallyfound for carnivorous gastropods (112–339 µg g−1).[24] Thelarge variability in arsenic concentrations of gastropods andamphipods between locations (Table 1) could be partiallyexplained by the relationship between arsenic concentrationsin S. quinqueflora leaves and total arsenic concentrationsin gastropods and amphipods (Fig. 5a–c). S. quinqueflorais the dominant plant in the three saltmarshes investigated(∼70–80% coverage) and is the primary source of detritus(food) for the gastropods and amphipods, analyzed in thisstudy, which tend to live under the canopy of S. quinqueflora
Fig. 8. Overlay of glycerol trimethyl arsonioribose in O. ornatus; Supelocosil SCX cation exchange chromatography. Solid line original 100 µLsample injection, and dotted line 100 µL injection spiked with ∼20 µg L−1 synthetic glycerol trimethyl arsonioribose.
Fig. 9. Phosphate arsenoribose in amphipod; Hamilton PRP-X100 anion exchange chromatography. Solid line original 20 µL sample injection,and dotted line 20 µL injection spiked with ∼10 µg L−1 phosphate arsenoribose.
186
RESEARCH FRONT
Distribution and Speciation of Arsenic in Temperate Marine Saltmarsh Ecosystems
Table 4. Principal components analysis of arsenic species proportions in saltmarsh sediment and plant and animaltissues
Factor loadings in bold have more influence on the samples location in three-dimensional space
A PC1 = x-axis; PC2 = y-axis; PC3 = z-axis. B Percentage of variation explained by each axis. C Eigenvector loadingscontributing to the distribution of points on each axis.
(personal observation). For N. meinerti there is no relation-ship between S. quinqueflora leaf arsenic concentrations andN. meinerti total arsenic concentration. This may be due todifferences in diet and/or consumption patterns. Gut contentsanalysis of N. meinerti has shown that this species tends toconsume mainly aged leaves from its environment (which itstores in its burrows); however, animal tissues, sediments,and algae have also been found in its gut.[40,41]
Relationship Between Arsenic, Iron, and PhosphorusConcentrations in Plants
As plants have no active uptake system for arsenic, unlike cop-per and zinc,[42] arsenic must be taken up during the transportof other essential elements such as phosphorus or another ele-ment to which it is exposed, i.e., iron from plaque coating theroots. Strong correlations were found between arsenic andiron concentrations for S. quinqueflora in leaf tissues, sug-gesting that the uptake of arsenic is clearly associated with theuptake of iron in plants leaf tissues (Fig. 6b). In some plants,iron plaque on roots has been shown to promote the uptakeof iron and associated trace elements.[26] The mechanisms ofhow this occurs is unclear but may be related to the inclusionof iron minerals within the root xylem. The arsenic concen-trations in this plants leaf material is appreciably higher thanin S. australis, which does not have as much iron plaque anddoes not demonstrate the same relationship between arsenicand iron concentrations (Fig. 6d). Iron and arsenic concentra-tions are clearly correlated in the root material (both with andwithout plaque) of both plants, reflecting the ability of ironoxides to bind and immobilize arsenic around root surfaces.Sediments in mashes and bogs outside the oxidized root zoneare often water logged and organic rich, and thus, anoxic.Arsenic that is mobile under anoxic conditions can migrateto the oxidized root zone of plants where it is immobilizedby iron oxides.[27]
In S. australis there is little evidence that arsenic andphosphorus concentrations are correlated in either the leafor root tissues, suggesting that if arsenic and phosphorus do
�3 �2 �1 0 1 2 3 4�4
�3
�2
�1
0
1
2
3P
C2
PC1
Group 1
Group 2
Group 3
ABInorganicAs
Glyceroltrimethyl
arsonioribose
OSO3� ribose
Fig. 10. Principal component analysis (PCA) of arsenic species pro-portions in saltmarsh sediment and plant and animal tissues. Arrowsindicate the factor contributing to the pattern in two-dimensional space.Group 1: O. ornatus, S. soilda and contained similar AB percent-ages; Group 2: N. meinerti muscle and visceral mass contained sulfatearsenoribose; Group 3: S. quinqueflora and S. australis leaves, roots,and sediment were dominated by inorganic arsenic.
share a common pathway, at these concentrations there isno effect on the uptake of one element by the other in thisplant species (Fig. 7c,d). In contrast S. quinqueflora leaveshave a significant negative correlation between leaf arsenicand phosphorus concentrations (Fig. 7b). Some workers havereported that phosphorus inhibits the uptake of arsenic[43]and this may be occurring; however, the regression relation-ship (r2 = 0.27, P < 0.05) explains only a small proportionof the data and does not justify a conclusion that inhibitionof arsenic uptake by phosphorus is occurring.
187
RESEARCH FRONT
S. Foster et al.
Water-Soluble Arsenic Species
The PCA clearly identified three groups of organisms withmarkedly different arsenic species proportions (Fig. 10).The plant leaves and roots (Group 3) were dominated byinorganic arsenic and glycerol arsenoribose with traces ofDMA (Table 4). This is consistent with other studies thathave reported arsenic species in terrestrial plants.[14,44] Stud-ies that have examined the uptake and accumulation ofarsenic in plants have reported that arsenic(v) is reduced toarsenic(iii) and sequestered in vacuoles by phytochelatins([γ-glutamate-cysteine]n-glycine).[45] Thus, plants have amechanism to immobilize inorganic arsenic at normal arsenicconcentrations found in uncontaminated environments.
All animals (Groups 1 and 2) contained appreciable quan-tities of AB (Fig. 10, Table 3). This is similar to the majorityof marine animals that have been analyzed and reported inthe published literature.[8] The amphipod (Group 2) containslower concentrations of AB (44%) than normally found inmarine animals and in another study that analyzed marineamphipods.[17] Amphipods also contained high concentra-tions of phosphate arsenoribose (23%) and sulfate arsenori-bose (15%). These species were not found to be presentin S. quinqueflora or S. australis tissues, which containedglycerol arsenoribose (17–35%; Table 3). Thus, it is unclearwhether amphipods are either concentrating and transform-ing glycerol arsenoribosides from plant detritus or obtainingphosphate and sulfate arsenoriboses from other sources, suchas microalgae, bacteria, or faecal material. Group one (gas-tropods and crab muscle) were separated from Group twoorganisms by the presence of glycerol trimethylated arso-nioribose (Fig. 10). The presence of glycerol trimethylatedarsonioribose together with high concentrations of AB inS. soilda, O. ornatus, and N. meinerti compared to theamphipod suggests that at least some of the AB presentmay be derived from these sugars. The gastropod O. orna-tus (Group 1) and N. meinerti viseral tissue (Group 2)contain an unidentified anionic arsenic compound eluting ina broad peak from ∼20 to 28 min. Based on the retentiontimes for thio-arsenic sugars,[46] this unknown arsenic com-pound may be a thio-arsenoribose, which would be a possibleintermediate in the formation of trimethyl arsonioriboses.
Concluding Comments
The results from this study show that arsenic uptake intosaltmarsh plants from uncontaminated environments maybe dependent on iron uptake and inhibited by high phos-phorus concentrations. The presence of glycerol trimethylarsonioribose found in gastropods of this study and in otherpublished studies of gastropods, macroalgae-eating fish,and abalone[10,24] raises the question of its importance inthe formation of AB.
Trimethyl arsonioriboses have been found to degradein anaerobic environments to quantitatively producearsenocholine,[18] which can then be oxidized to AB.[21] Itis easier to envisage the synthesis of AB in marine animalsfrom the trimethyl arsonioriboses than dimethyl arsinoribo-sides as trimethyl arsonioriboses only require degradation
to AC and further oxidation to AB. The recent isolationof dimethyl thio-arsinoriboses (Fig. 2),[46] which are lesspolar than dimethyl arsenoribosides and probably form inanaerobic, sulfide-rich gut environments, provides a plausi-ble transformation pathway for the formation of trimethylarsonioriboses. Trimethyl arsonioriboses in marine organ-isms are possible transitory intermediates and their role in theformation of AB may be underestimated. Most studies depu-rate the guts of animals before analysis, often for longperiods (24–48 h), and this may result in the loss of thesearsenic species.[47] As well, these arsenosugars may only bepresent in significant quantities while algal material is beingdigested. We believe gastropods (and other molluscs) containmeasurable trimethyl arsonioriboses as they graze for longperiods and have long gut retention times.[48] Perhaps otherherbivores and detritivores do contain significant transientconcentrations of trimethyl arsonioriboses, but samplingmust be timed to ensure maximum likelihood of measurementof these species.
References
[1] D. D. Swinbanks, J. Exp. Mar. Biol. Ecol. 1982, 62, 69.doi:10.1016/0022-0981(82)90217-9
[2] D. Morrisey, in Coastal Marine Ecology of Temperate Aus-tralia (Eds A. J. Underwood, M. G. Chapman) 1995, pp. 205(University of New South Wales Press LTD: Sydney).
[3] I. T. Hyman, W. F. Ponder, G. W. Rouse, J. Nat. Hist. 2004, 38,2377. doi:10.1080/00222930310001647370
[4] P. J. F. Davie, Mem. Queensl. Mus. 1994, 35, 35.[5] M. A. Graça, S. Y. Newell, R. T. Kneib, Mar. Biol. 2000, 136,
281. doi:10.1007/S002270050686[6] M. Zimmer, S. C. Pennings, T. L. Buck, T. H. Carefoot, Estuaries
2004, 27, 753.[7] A. M. M. Richardson, M. E. Mulcahy, Estuarine Coastal and
Shelf Science 1996, 43, 801. doi:10.1006/ECSS.1996.0104[8] J. S. Edmonds, K. A. Francesconi, in Organometallic Compounds
in the Environment (Ed. P. J. Craig) 2003, pp. 196 (Wiley: NewYork, NY).
[9] R. Tukai, W. A. Maher, I. J. McNaught, M. J. Ell-wood, M. Coleman, Mar. Freshwater Res. 2002, 53, 971.doi:10.1071/MF01230
[10] J. Kirby, W. Maher, D. Spooner, Environ. Sci. Technol. in press,[11] J. Kirby, W. Maher, Appl. Organomet. Chem. 2002, 16, 108.
doi:10.1002/AOC.268[12] J. Kirby, W. Maher, A. Chariton, F. Krikowa, Appl. Organomet.
Chem. 2002, 16, 192. doi:10.1002/AOC.283[13] V. M. Dembitsky, T. Rezanka, Plant Sci. 2003, 165, 1177.
doi:10.1016/J.PLANTSCI.2003.08.007[14] A. M. de Bettencourt, M. F. Duarte, S. Facchetti,
M. H. Florencio, M. L. Gomes, H. A. van’t Klooster,L. Montanarella, R. Ritsema, L. F. Vilas-Boas, Appl.Organomet. Chem. 1997, 11, 439. doi:10.1002/(SICI)1099-0739(199705)11:5<439::AID-AOC609>3.0.CO;2-H
[15] Y. Bohari, G. Lobos, H. Pinochet, F. Pannier, A. Astruc, M. Potin-Gautier, J. Environ. Monit. 2002, 4, 596. doi:10.1039/B203988P
[16] W. Goessler, W. Maher, K. J. Irgolic, D. Kuehnelt, C. Schla-genhaufen, T. Kaise, Fresenius J. Anal. Chem. 1997, 359, 434.doi:10.1007/S002160050605
[17] J. S. Edmonds, Y. Shibata, K. A. Francesconi, R. J. Rippingale,M. Morita, Appl. Organomet. Chem. 1997, 11, 281. doi:10.1002/(SICI)1099-0739(199704)11:4<281::AID-AOC581>3.0.CO;2-S
[18] K. Francesconi, J. S. Edmonds, Appl. Organomet. Chem. 1992,6, 247. doi:10.1002/AOC.590060220
[19] J. S. Edmonds, Bioorg. Med. Chem. Lett. 2000, 10, 1105.doi:10.1016/S0960-894X(00)00176-1
188
RESEARCH FRONT
Distribution and Speciation of Arsenic in Temperate Marine Saltmarsh Ecosystems
[20] A. W. Ritchie, J. S. Edmonds, W. Goessler, R. O. Jenkins,FEMS Microbiol. Lett. 2004, 235, 95. doi:10.1016/J.FEMSLE.2004.04.016
[21] A. Christakopoulos, H. Norin, M. Sandstrom, H. Thor,P. Moldeus, R. Ryhage, J. Appl. Toxicol. 1988, 8, 119.
[22] S. McSheehy, P. Pohl, D. Velez, J. Szpunar, Anal. Bioanal.Chem. 2002, 372, 457. doi:10.1007/S00216-001-1182-X
[23] Y. Shibata, M. Morita, Agric. Biol. Chem. 1988, 52, 1087.[24] K. A. Francesconi, W. Goessler, S. Panutrakul, K. J. Irgolic,
Sci. Total Environ. 1998, 221, 139. doi:10.1016/S0048-9697(98)00272-1
[25] W. Liu, Y.-G. Zhu, F. A. Smith, S. E. Smith, New Phytol. 2004,162, 481. doi:10.1111/J.1469-8137.2004.01035.X
[26] M. L. Otte, J. Rozema, L. Koster, M. S. Haarsma, R. A.Broekman, New Phytol. 1989, 111, 309.
[27] M. L. Otte, J. Rozema, M. A. Beek, B. J. Kater,R. A. Broekman, Sci. Total Environ. 1990, 97/98, 879.doi:10.1016/0048-9697(90)90279-4
[28] M. Quaghebeur, Z. Rengel, Plant Physiol. 2003, 132, 1600.doi:10.1104/PP.103.021741
[29] M. O. Andreae, D. W. Klumpp, Environ. Sci. Technol. 1979, 13,738. doi:10.1021/ES60154A001
[30] J. A. Friend, A. M. M. Richardson, Annu. Rev. Entomol. 1986,31, 25. doi:10.1146/ANNUREV.EN.31.010186.000325
[31] W. Maher, F. Krikowa, J. Kirby, A. T. Townsend, P. Snitch, Aust.J. Chem. 2003, 56, 103. doi:10.1071/CH02203
[32] J. Kirby, W. Maher, J. Anal. At. Spectrom. 2002, 17, 838.doi:10.1039/B202276C
[33] S. Baldwin, M. Deaker, W. Maher, Analyst 1994, 119, 1701.doi:10.1039/AN9941901701
[34] J. Kirby, W. Maher, M. J. Ellwood, F. Krikowa, Aust. J. Chem.2004, 57, 957. doi:10.1071/CH04094
[35] M. J. Ellwood, W. A. Maher, Anal. Chim. Acta 2003, 477, 279.[36] K. R. Clarke, R. M. Warwick, Changes in Marine Communities:
An Approach to Statistical Analysis and Interpretation 1994(Plymouth Marine Laboratory: Plymouth, UK).
[37] A. A. Meharg, J. Hartley-Whitaker, New Phytol. 2002, 154, 29.doi:10.1046/J.1469-8137.2002.00363.X
[38] M. Bissen, F. H. Frimmel, Acta Hydrochim. Hydrobiol. 2003,31, 9. doi:10.1002/AHEH.200390025
[39] M. Krachler, H. Emons, Fresenius J. Anal. Chem. 2000, 368,702. doi:10.1007/S002160000578
[40] S. Y. Lee, Mar. Freshwater Res. 1998, 49, 335. doi:10.1071/MF97179
[41] F. Dahdouh-Guebas, M. Verneirt, J. F. Tack, N. Koedam,Hydrobiologia 1997, 347, 83. doi:10.1023/A:1003015201186
[42] M. L. Otte, M. S. Haarsma, R. A. Broekman, J. Rozema, Environ.Pollut. 1993, 82, 13. doi:10.1016/0269-7491(93)90157-J
[43] C. Tu, L. Q. Ma, Plant Soil 2003, 249, 373–382. doi:10.1023/A:1022837217092
[44] J. Zheng, H. Hintelmann, B. Dimock, M. S. Dzurko, Anal.Bioanal. Chem. 2003, 377, 14. doi:10.1007/S00216-003-1920-3
[45] M. E. V. Schmoger, M. Oven, E. Grill, Plant Physiol. 2000,122, 793. doi:10.1104/PP.122.3.793
[46] E. Schmeisser, R. Raml, K. Francesconi, D. Kuehnelt,A.-L. Lindberg, C. Sörös, W. Goessler, Chem. Commun. 2004,16, 1824. doi:10.1039/B406917J
[47] A. E. Geiszinger, W. Goessler, K. A. Francesconi, Mar. Environ.Res. 2002, 53, 37. doi:10.1016/S0141-1136(01)00106-4
[48] P. J. Britz, T. Hecht, J. Knauer, S.A. J. Marine Sci. 1996, 17, 297.
