Assessing the Legacy of Red Mud Pollution in a Shallow Freshwater Lake: Arsenic Accumulation and Speciation in Macrophytes

: Little is known about long-term ecological responses in lakes following red mud pollution. Among red mud contaminants, arsenic (As) is of considerable concern. Determination of the species of As accumulated in aquatic organisms provides important information about the biogeo-chemical cycling of the element and transfer through the aquatic food-web to higher organisms. We used coupled ion chromatography and inductively coupled plasma mass spectrometry (ICP-MS) to assess As speciation in tissues of ﬁ ve macrophyte taxa in Kinghorn Loch, U.K., 30 years following the diversion of red mud pollution from the lake. Toxic inorganic As was the dominant species in the studied macrophytes, with As species concentrations varying with macrophyte taxon and tissue type. The highest As content measured in roots of Persicaria amphibia (L.) Gray (87.2 mg kg − 1 ) greatly exceeded the 3 − 10 mg kg − 1 range suggested as a potential phytotoxic level. Accumulation of toxic As species by plants suggested toxicological risk to higher organisms known to utilize macrophytes as a food source.


■ INTRODUCTION
The estimated global production of red mud, a byproduct of alumina production, is ∼120 million t a −1 . 1 In October 2010, failure of a containment reservoir in Ajka, Western Hungary, resulted in the release of ∼1 million m 3 of red mud waste, contaminating the Marcal River (catchment area 3078 km 2 ) and entering the Danube River. 2 The Hungarian red mud spill raised concerns about inappropriate storage of red mud and the impact of the waste on the receiving environment. However, little is known about the geochemical behavior of red mud in freshwaters 2 or of the short-and long-term ecological impacts and likelihood of recovery following its release into the environment. This is conspicuous given that many of its chemical constituents are redox sensitive and so likely to persist in aquatic depositional environments where they can represent an environmental and human health risk.
Red mud is highly alkaline due to the addition of sodium hydroxide (NaOH) during the production process and contains metal oxides and elevated concentrations of a range of minor trace elements 3−5 that can be toxic to aquatic organisms. Initial research after the Ajka accident focused on the short-term impact of the pollution on freshwaters. Mayes et al. 6 reported elevated concentrations of some contaminants, e.g., arsenic (As), vanadium (V), chromium (Cr), and nickel (Ni), in fluvial sediment downstream of the spill. The association of these elements mainly with residual phases suggests limited potential for their mobilization in the environment. However, in depositional zones, release of these elements may be expected. 6 Furthermore, river sediments contaminated with constituents of red mud can have pronounced toxic effects, for example, causing reduced bioluminescence of Vibrio f ischeri and growth of Lemna minor and Sinapis alba. 2 In aquatic ecosystems with short residence times, the long-term effects of pollution might be spatially limited to depositional zones, allowing system recovery. 2 However, the residence time of pollutants in higher retention time aquatic ecosystems can exceed those in fluvial systems. 7,8 Therefore, long-term monitoring of the red mud impact on aquatic life is particularly important in lakes that can have potentially longer time-scales of recovery.
Among red mud contaminants, As is of considerable concern. It is a toxic element, 9 with high concentrations in the aquatic environment posing an environmental and human health risk in many countries, e.g., in Pakistan, Bangladesh, India, Taiwan, 10 Japan, 11 China, 12 and the U.S.A. 13 Arsenic concentrations in contaminated groundwater were reported up to 1 mg L −1 in some areas of Bangladesh and 3 mg L −1 in Vietnam, 14 while an As concentration of 792 mg kg −1 was measured in the surface sediments of Lake Biwa, Japan. 11 In macrophytes, uptake of As can reduce phosphorus (P), nitrogen (N), potassium (K), chlorophyll a and protein content in tissues and, due to the chemical similarity of As to P, can interfere with some biochemical reactions. 15 Concentrations of contaminants accumulated by plants can be higher than in water, but lower than in bed sediments. 16 The bioavailability of trace elements to plants can be regulated by a range of site specific factors including their concentration in the environment, trace element speciation, exposure time, and absorption mechanism. 16 For example, uptake of the inorganic As species, arsenate, occurs through phosphate uptake pathways and has been shown to be negatively correlated with phosphate uptake by the macrophyte Spirodela polyrhiza L. 9 In addition, the capacity of aquatic plants to sequester and accumulate metals is affected by many factors, such as plant growth rate, biomass accumulation, and affinity for metal uptake. 15 Previous studies have shown that As bioaccumulation varies among macrophyte species and among aquatic plant tissues. 15−18 Arsenic can occur in inorganic and organic chemical forms, or species, which are associated with different levels of toxicity. 19 For example, the inorganic species arsenite (As(III)) is more toxic than organic arsenobetaine to plants and animals. 19 The uptake mechanism 20 and level of accumulation in plants 19,21 also vary between As species. Significant differences between concentrations of different As species have been reported in rice (Oryza sativa L.), with the predominant form being inorganic As. 21,22 Determination of As speciation in plants is important for assessing the plant As toxicity to consumers at higher levels of the food chain. 23−25 Most previous research on As species in plants has focused on crops and As speciation in plants in the aquatic environment has not been extensively studied, with most of the existing knowledge based on controlled laboratory experiments. 9,18,20 Consequently, little is known about variation of individual As species between tissues of aquatic plants. Inorganic and organic As species have been determined previously in whole macrophytes, 9,19,26 but concentrations of As species have not been examined in different plant tissues. Determination of the species of As accumulated in macrophytes allows for more accurate assessment of the environmental risk that increases with the occurrence of inorganic As and is particularly important at sites where fish and waterfowl are used for human consumption.
We assessed the impact of As on five macrophyte taxa three submerged species (Potamogeton pectinatus L., Elodea nuttallii (Planch) H. St. John, Myriophyllum spicatum L.), one rooted, floating-leaved species (Persicaria amphibia (L.) Gray), and one multicellular algae (Chara spp.) in Kinghorn Loch 30 years after the diversion of red mud leachate. Arsenic speciation analysis was applied to quantify inorganic As (arsenite, As(III) and arsenate, As(V)) and four organic As species: arsenobetaine, dimethylarsinic acid (DMA), monomethylarsonic acid (MMA), and tetramethylarsonium ions (Tetra). On the basis of the results of previous studies on total As accumulation

Environmental Science & Technology
Article in macrophytes, 15−17 As speciation in rice 21 and terrestrial plants 14 and As species uptake by macrophytes in controlled laboratory experiments, 9,18,20 the following hypotheses were tested: (1) macrophytes in Kinghorn Loch accumulate higher concentrations of inorganic compared to organic As species; (2) accumulation of As species varies between above-sediment tissues (leaves and stems) of macrophytes and among macrophyte taxa; and (3) accumulation of As species in P. amphibia is greatest in the roots of the plant compared to the above-sediment tissues. This work provides the first comprehensive assessment of As accumulation and speciation in aquatic macrophytes following red mud pollution.

■ MATERIALS AND METHODS
Study Site. Kinghorn Loch is a small lake of surface area 11.3 ha, mean depth 4.5 m and maximum depth 12.8 m, situated in Fife, Scotland (56°10′N; 3°11′W). The lake received highly alkaline leachate from a nearby red mud landfill site from 1947 to 1983 when the discharge was diverted. Caustic liquor reaching the lake consisted of a highly alkaline solution of NaOH and carbonate, with a mean pH of 12.1. It contained high concentrations of dissolved aluminum (Al; 137 mg L −1 ), As (3.56 mg L −1 ), vanadium (V; 5.3 mg L −1 ), phosphate (PO 4 P; 2.95 mg L −1 ), sulfate (SO 4 S; 154 mg L −1 ), and chloride (Cl; 67.8 mg L −1 ). Red mud solids, released to the loch sporadically, accounted for most of the elevated input of iron (Fe). 4 The physicochemical pathways of pollutant transport within the bed sediments of the loch have been discussed in detail by Edwards (1985). 4 Long-term monitoring of water chemistry of the loch was initiated in 1980 by the Forth River Purification Board and continued by the Scottish Environment Protection Agency (SEPA). The pollution led to intense algal blooms, fish kills, and severe reduction of zooplankton, macroinvertebrate, and macrophyte species diversity and biomass. 4 By 1985, two years after the diversion of the leachate from the lake, a collapse in the phytoplankton activity, development of zooplankton, and an increase in abundance of benthic macroinvertebrates (although no increase in species richness) were observed. 4 Bioaccumulation studies of the plankton and macroinvertebrate populations in 1985 indicated no significant As bioaccumulation in Kinghorn Loch. 4 A pilot survey was conducted in 2013 to assess metal concentrations in surface waters and bed sediments. 27 Total As concentrations in lake surface water and bottom water (1 cm above sediment; sampled July 2013; Site 1; Figure 1) were 14.9 μg L −1 and 14.8 μg L −1 , respectively, and did not exceed Standards for Protection of Aquatic Life in the U.K. (50 μg L −1 ). 28 Concentrations of As in lake surface sediment at Site 1 (87 mg kg −1 ) and mean As concentration from 6 sites across the lake (160 mg kg −1 ) 27 in July 2013 exceeded the Canadian Sediment Quality Guidelines for Protection of Aquatic Life, which predicts toxicity from sediment concentrations ≥17 mg kg −1 of As. 29 We confirmed As release from bed sediments under reducing redox conditions in Kinghorn Loch in a laboratory intact-core incubation study. 27 Macrophyte Sampling. Macrophyte sampling was conducted on 18 July 2013 when whole plants were collected by hand and using a double-headed rake from a boat at three sites in Kinghorn Loch (56°40′26″N; 3°11′34″W (Site 1); 56°40′18″N; 3°11′33″W (Site 2); 56°40′22″N; 3°11′16″W (Site 3); and Supporting Information, SI, Figure S1). The sites were chosen to represent areas at different distances (approximately 100, 250, and 430 m for Sites 1, 2, and 3, respectively) and directions from the NW part of the lake, where the pollution entered the waterbody. Between three and ten plants per taxa (depending on the species specific volume of fresh plant material) were collected at each site. Samples represented the five most abundant macrophyte taxa in Kinghorn LochP. amphibia, P. pectinatus, E. nuttallii, M. spicatum, and Chara spp.as determined by a full lake survey conducted on 18 July 2013 following the procedure outlined in Gunn et al. 30 Sample Preparation. The macrophytes were stored in polyethylene bags at 4°C until processing in the laboratory within 24 h. The plant tissues were placed into bags containing distilled water and manually shaken, repeatedly, to remove sediment, macroinvertebrates, and epiphytic biofilms. We acknowledge that this approach may not have removed entirely epiphytic organisms, although others have reported near complete removal of this community following similar treatments, as well as plant tissue damage under severe agitation. 31 To test for any variation in As accumulation in the different parts of macrophytes, P. pectinatus, E. nuttallii, and M. spicatum samples were divided into leaves and stems, and samples of P. amphibia, the only collected species with welldeveloped roots, into leaves, stems and roots. Axes and branches of the multicellular algae Chara spp. were also separated and are referred to as stems and leaves, respectively, for the purpose of statistical analysis. Samples of each macrophyte tissue type for each species were combined by site from a number of individual plants to obtain enough dry material for As analyses. Samples were then oven-dried at 80°C and ground using a Retsch mixer mill MM200.
Chemical Analysis. Three replicate samples of each tissue type from each macrophyte species were analyzed. Approximately 100 ± 5 mg of each powdered sample was digested with 1% nitric acid using a microwave digester (CEM Mars 6 1800W). Arsenic speciation was determined by ion chromatography (Thermo Dionex IC5000 Ion Chromatograph) coupled to inductively coupled plasma mass spectrometry (Thermo Scientific iCap Q ICP-MS). Arsenic species in the solutions were identified by comparing their retention times with those of standards including inorganic As (As(III) and As(V)), arsenite, DMA, MMA, and Tetra and quantified by calibration curves with peak areas. Species with retention times not matching any of the standards were classified as unknown species. Total identified As species were calculated as the sum of identified inorganic and organic As species. The detection limit for all As species was 0.001 mg kg −1 . The As species concentrations measured in duplicate analyses of the Certified Reference Material (CRM; NIST 1568b Rice flour) conducted in the same manner were within 25% of the certified values. The certified and measured values of As species in the CRM are listed in Table 1. The CRM did not contain arsenobetaine and Tetra. The method for determination of As speciation is described in detail in the SI.
Statistical Analyses. The means of three replicate samples were used for all statistical analyses. Concentrations of different As species in P. amphibia tissues (leaves and stems) were analyzed together with data for the other macrophytes species to examine across species variation in the above-sediment parts of macrophytes, and separately (concentrations in leaves, stems, and roots) to test for differences between As concentrations in below-and above-sediment parts of P. amphibia.

Environmental Science & Technology
Article A one-sided two sample t-test was used to test whether the mean concentration of inorganic As was greater than the mean concentration of total organic As in the above-sediment tissues of macrophytes and also whether below-sediment surface tissues of P. amphibia contained higher concentrations of As species than above-sediment tissues.
Mixed effect modeling was conducted to examine variations in concentrations of each As species across macrophyte species and tissue type (within shoots). Root samples were not included in these models, as they were only available for one species. Model validation was conducted on the final models. Homogeneity of variance was assessed by plotting residuals versus fitted values and the normality assumption was evaluated by plotting theoretical quantiles versus standardized residuals (Q−Q plots). In models with all macrophyte species, several response variables (i.e., inorganic As, arsenobetaine, total organic and the sum of all identified As species) were log 10 transformed to meet homogeneity and normality assumptions. The random effect part of the model allowed for variations among sites and incorporated a split plot design of the data. The analysis started with the model containing both nominal fixed explanatory variables (i.e., macrophyte species and macrophyte tissue) and the two-way interaction between these two terms. The model selection process followed the procedure described in Zuur et al. 32 Maximum likelihood ratio (ML) tests were performed, using the ANOVA function, to select the best-fit model by comparing AIC values and to determine the significance (P < 0.05) of dropped terms. The final model was rerun using restricted maximum likelihood (REML). Statistical analyses were conducted using R version 3.0.1, 33 using the "nlme" package 34 for the mixed modeling.

Difference between Inorganic and Organic As
Concentrations Across Macrophyte Taxa. Figure 1 shows variations in As species concentrations across the five macrophyte taxa and different plant tissues of each of the taxa. The above-sediment parts of the five macrophyte species in samples combined across sampling sites contained significantly higher concentrations of inorganic As species than total organic As, with mean values of 8.  Table 2). Higher concentrations of inorganic As were measured compared to total organic As concentrations in both tissues of all macrophytes species (Figure 1, Table S1). P. amphibia also contained significantly higher concentrations of inorganic As than total organic As, with mean values of 40.0 mg kg −1 d.w. (s.d. 65.7) for inorganic and 0.25 mg kg −1 d.w. (s.d. 0.11) for organic forms (Table 2), respectively, calculated as the arithmetic mean of above and below-sediment surface tissues As concentrations.
Differences in As Species Accumulation in Above-Sediment Macrophyte Tissues. The lowest mean concentrations of inorganic As in the above sediment-surface tissues were measured in the leaves and stems of P. amphibia (0.76 mg kg −1 ± 0.14 and 1.07 mg kg −1 ± 0.29, respectively) and the highest in leaves of E. nuttallii (45.1 mg kg −1 ± 20.7; Figure 1, Table S1). Two organic As species (MMA and Tetra) were below the limit of detection in all tissue types of Chara spp., with MMA also not detectable in leaves of M. spicatum. Mean concentrations of the four organic As species ranged from 0.01 mg kg −1 d.w. (DMA in Chara sp. branches, MMA in both tissues of P. amphibia, Tetra in P.amphibia leaves) to 1.16 mg kg −1 (arsenobetaine in Chara spp. branches). Unknown As species were present in the leaves of P. pectinatus and M. spicatum and in both tissues of Chara spp., with the highest mean concentration (1.04 mg kg −1 ± 0.47) measured in the latter species.
The model selection process indicated a model with plant tissue, plant species and the interaction term included as an optimal one for inorganic As, arsenobetaine, MMA, Tetra, total organic As, and all identified As species (Table S2). Due to a lack of a significant effect, the interaction term was omitted from the model for unknown As species. None of the explanatory variables was found to have a significant association with plant DMA concentration.
The model for Tetra indicated no significant difference in concentration between P. amphibia and Chara spp. (Table S3) The model showed that the four macrophyte species had lower total organic As concentrations than Chara spp., but the difference was only significant (estimate: −1.93, S.E.: 0.27, t: −7.22, P: < 0.001) in P. amphibia, E. nuttallii, M. spicatum, and P. pectinatus. For these species, the concentrations of total organic As in stems was significantly lower than observed in leaves (Table S3).
Differences in As Species Accumulation within P. amphibia. Inorganic As in roots was substantially higher (86.9 ± 75.8 mg kg −1 d.w.) than mean concentrations of any other identified As species measured in P. amphibia, which were ≤0.12 mg kg −1 d.w. (s.d. 0.07; Figure 1, Table S1). All analyzed organic species of As were measured in each tissue of P. amphibia, while unknown As species were only detected in plant roots (mean 0.17 mg kg −1 d.w. ± 0.22).
The below-sediment tissues of P. amphibia contained significantly higher concentrations of inorganic As, DMA, MMA and total organic As and identified As species than the above-sediment tissues (Table 3). There were no significant differences between concentrations of arsenobetaine and Tetra between tissues above and below the sediment surface.

■ DISCUSSION
Macrophytes in Kinghorn Loch Accumulated Potentially Toxic Concentrations of As. The present study showed that lake macrophytes contained relatively high concentrations of As, 30 years following the diversion of red mud leachate from Kinghorn Loch. The As concentrations in shoots reported here were higher than those reported in two other lakes (i.e., 1.4 to 5.3 mg kg −1 in Lake Dianchi in China 35 and <1 mg kg −1 in Lake Velenjsko, Slovenia 36 ) characterized by elevated concentrations of As in surface sediment. In addition, the As concentration in E. nuttallii leaves in Kinghorn Loch was in excess of the maximum plant shoot content of As reported for wetlands affected by the Aznalcoĺlar mine spill in Spain (0.01−28.5 mg As kg −1 ). 37 However, the upper concentration range in plant roots documented by Taggart et al. 37 (1.39− 1089 mg kg −1 ) was greater than that measured in root samples collected from Kinghorn Loch (40.5−218 mg kg −1 ).
There are no widely accepted toxicity threshold concentrations for As in plant tissues. 17 Most of the mean concentrations measured in macrophyte tissues in Kinghorn Loch, however, were higher than the ranges 0.01−1.00 mg kg −1 and 3.00−10.0 mg kg −1 that were suggested by Chaney 38 as "normal" and potentially phytotoxic ranges of inorganic As concentrations in dry foliage, respectively. These results indicate that, although the total As concentrations measured

Environmental Science & Technology
Article in water in Kinghorn Loch 30 years after the pollution ceased entering the lake meet Standards for Protection of Aquatic Life, the high content of As in lake bed sediments remains a toxic hazard to macrophytes.

Concentrations of Inorganic As in Macrophytes
Were Significantly Higher Compared to Organic As. Inorganic As was the dominant species in the studied macrophytes in Kinghorn Loch. Although As speciation in plants in the aquatic environment has not been extensively studied, these field observations are in agreement with the results from controlled laboratory experiments that have shown much higher uptake of arsenate and arsenite, compared to DMA and MMA, by Spirodela polyrhiza L., 9 and by Azolla caroliniana Willd. and Azolla f iliculoides Lam., 18 and higher arsenate uptake in comparison to DMA by Salvinia natans L. 20 The present study also confirms the results of the survey investigating As species accumulation in a range of aquatic plants sampled in a wetland affected by As from mining activities in Yellowknife, Canada. 19 These observations made on aquatic plants are generally consistent with the results of wider research, which have investigated As speciation in several plants used for human consumption, e.g., rice Oryza sativa, 21,39 as well as other terrestrial plants. 14 It is not known if the presence of various organic As compounds in tissues is a result of uptake or transformation of As by plants. 14 The present study identified four organic As species within macrophyte tissues, which except for arsenobetaine in Chara spp., were present at concentrations not indicated as potentially phytotoxic. 38 DMA, MMA, and Tetra, but not arsenobetaine, were previously identified in aquatic plants from Yellowknife, Canada. 19 A review of As speciation in terrestrial plants showed that methylated As forms were also reported (as a minor fraction of the total accumulated As) in a wide number of plant species. 14 The presence of arsenobetaine has been previously reported in terrestrial plants, however, in a smaller number of species compared to methylated As compounds. 14 In addition to the As compounds discussed above, the results of the present study indicate that Chara spp., P. pectinatus and M. spicatum all accumulated a small fraction of As species that were not identified by the analysis procedure employed. A few aquatic plant species from the Yellowknife study, such as Myriophyllum sp., contained arseno-sugars. 19 Koch et al. 19 suggested that the presence of arseno-sugars in submerged plant tissues could be explained by contamination of samples with organisms containing this form of As, e.g., epiphytic algae growing in physical contact with some macrophytes. Thus, it is possible that arseno-sugars accounted for some of the unidentified As species present in macrophytes at Kinghorn Loch, despite the washing procedure, although the effect of this procedure was not quantified for this study.
Accumulation of As Species Varied between Above-Sediment Tissues of Macrophytes and among Macrophyte Species. Most studies have examined variation in As accumulation only for total As. The extent of metal(loid) accumulation investigated at the total concentration level has been shown to differ between macrophyte species and tissues 40−42 and thus, it was hypothesized that the concentrations of different As chemical forms within plants in Kinghorn Loch would also vary depending on plant tissue and species. Variation in total identified As, and consequently variation of inorganic As, which had the highest contribution to the total As concentrations in this study, confirmed that hypothesis. With the exception of DMA, organic As species concentrations were also reported to vary significantly with plant species and tissue type. Across-species As variation observed in the present study is in agreement with the differences in the uptake of inorganic and organic As forms by different macrophyte species previously reported. 18,19 Variation in metal(loid) uptake may be related to differences in plant growth rate and absorption efficiency for metals. 15 In addition, some macrophytes can accumulate greater concentrations of As than others, but might show fewer signs of damage due to more effective mitigation mechanisms. 43 In the present study, the highest shoot concentrations of inorganic As were observed in E. nuttallii. Previous estimates of metal (Hg and Cd) uptake by E. nutallii shoots indicated high and rapid bioaccumulation of metals, with shoots having a high tolerance due to the protection of cellular machinery by highly effective tolerance mechanisms. 44 This might explain the capability of this species to accumulate high concentrations of As in Kinghorn Loch. Lower content of inorganic As in above-sediment tissues of other taxa in the present study might indicate different plant defense mechanisms in comparison to E. nuttallii, such as limiting As absorption. In order to decrease As uptake and enhance their survival in the presence of arsenate, plants can suppress phosphate uptake pathways used for arsenate absorption, 14 which may be the defense process occurring in some of the macrophyte species in Kinghorn Loch. Acrosstissue variation observed in As species may be a result of differences in translocation of As species between tissues by plants or reflect differences in tissue-specific metabolism, although knowledge of the ability of plants to transform inorganic As into methylated species is limited. 14 Concentrations of all detected As species in macrophytes in Kinghorn Loch were higher (apart from DMA in Chara spp.) in leaves compared to stems of the four macrophyte taxa sampled. The only exception was P. amphibia, in which total As concentrations were significantly higher in stems than leaves. The significantly higher concentrations of inorganic and organic As in leaves compared to stems in M. spicatum in the present study were in agreement with total As distribution in shoots of this species in Kadin Creek, Turkey. 45 Roots of P. amphibia Accumulated Significantly Higher Concentrations of Most Identified As Species Compared to the Above-Sediment Tissues. Most previous studies have shown that a major fraction of total As is accumulated in the roots of aquatic plants. [15][16][17]42,46 To date, few plants have been identified with the ability to translocate high amounts of As from roots to above-sediment tissues. 46 This was also demonstrated for P. amphibia, in the present study. Chou et al. 47 suggested that minor translocation of As from roots to shoots in rice plants was indicative of adaptive survival behavior in As contaminated sediments. The results for inorganic As from Kinghorn Loch indicate that this might also be the strategy of P. amphibia. Favas et al. 42 suggested that As speciation appears to play an important role in both the uptake mechanism and further translocation of As between plant tissues. Arsenate is an analogue of phosphate 14 and uptake, accumulation and translocation of inorganic As is thought to be affected by its relationship with phosphate transporters 42 and the presence of Fe and sulfur (S). 25 Fe plaque on plant roots can act either as a sink or a source of As to plants and S can help detoxify As through complexation and restricting its translocation to shoots. 25 However, the role of the Fe plaque in the translocation of As into the macrophytes in Kinghorn Loch

Environmental Science & Technology
Article cannot be verified as the As content of the Fe plaque observed was not determined. In contrast, Rahman et al. 21 reported that uptake of the organic As species, DMA, through roots and shoots of rice plants was not affected by Fe. Moreover DMA and arsenate are suggested to be taken up by different pathways. 9 Despite these differences, concentrations of DMA in the present study were, similarly to inorganic As species, significantly higher in the plant roots than in shoots. Knowledge of the mechanisms of translocation and accumulation of organic As species in plants is, however, limited.
Toxicological Implications of Measured As Species Concentrations in Macrophytes in Kinghorn Loch. Information about the relative concentrations of As species present in macrophytes can provide important toxicological context. Organic As species, recorded in the present study at concentrations comparable to background levels of total As, are generally less toxic to organisms, including aquatic plants, than inorganic As compounds. 48 Moreover, arsenobetaine, the organic As species present in the highest concentration of those analyzed in the present study is reported to have nontoxic properties. 49 The dominance of inorganic As forms, which are present in elevated concentrations in all macrophyte taxa in Kinghorn Loch indicates that, even following 30 years of recovery, red mud pollution still poses a risk to the macrophytes of this lake, and potentially to higher organisms in the food web. Inorganic As species are highly toxic to plants. 14 It is possible that the macrophyte community that has developed in Kinghorn Loch is tolerant to elevated As concentrations. For example, two of the dominant macrophytes species included in the study, M. spicatum and P. pectinatus, have been reported to accumulate high metal concentrations, persisting under conditions of concentrations in tissues higher than those regarded as toxic for plants. 50 In addition, P. pectinatus can quickly colonize polluted waters and environments unsuitable for other species 50 and Elodea sp. is known to be able to adapt to a wide range of environmental conditions. 51 These traits have led researchers to consider the use of M. spicatum, P. pectinatus, and E. nuttallii for phytoremediation. 44,52,53 The presence of species that have been shown to be persistent in heavily polluted waters suggests that the elevated As concentrations in the sediment may still influence the macrophyte community structure in Kinghorn Loch.

* S Supporting Information
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.6b00942.
Macrophyte sampling sites, a detailed description of the method for determination of As speciation, As species concentrations in macrophytes, ranking of mixed models, and coefficient estimates of the optimal mixed models explaining variation in arsenic species concentrations in macrophytes (PDF)

Notes
The authors declare no competing financial interest.

■ ACKNOWLEDGMENTS
We would like to acknowledge Dr. Ron Edwards from Craigencalt Rural Community Trust and Ken Evans from Rio Tinto Alcan for their support of this research. We would also like to thank James Watt for his assistance with field work and Dr. Ron Smith and Dr. Lindsey Banin (CEH Edinburgh) for support with statistical analysis. The PhD studentship to J.P.O. was funded by the U.K. Natural Environment Research Council (NERC).