Environmental hazard assessment of contaminated ...

Science of the Total Environment 574 (2017) 443–454

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Environmental hazard assessment of contaminated soils in Antarctica: Using a structured tier 1 approach to inform decision-making Joana Luísa Pereira a,⁎, Patrícia Pereira a, Ana Padeiro b, Fernando Gonçalves a, Eduardo Amaro b, Marcelo Leppe c, Sergey Verkulich d, Kevin A. Hughes e, Hans-Ulrich Peter f, João Canário b a

Department of Biology, CESAM, University of Aveiro, Aveiro, Portugal Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal INACH, Chilean Antarctic Institute, Punta Arenas, Chile d Arctic and Antarctic Research Institute, Saint-Petersburg, Russia e British Antarctic Survey, Natural Environment Research Council, High Cross, Madingley Road, Cambridge CB30ET, UK f Polar & Bird Ecology Group, Institute of Ecology, Friedrich Schiller University Jena, Germany b c

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Integrated environmental risks have been poorly assessed in Antarctica soils. • Anthropogenic trace element contamination was confirmed in soil samples from Fildes. • Adverse ecotoxicological effects were promoted by aqueous extracts of all tested soils. • Integrated risks were found high or uncertain, denoting contaminated sites of concern. • None of the soils can sustain natural functions, demanding better protective measures.

a r t i c l e

i n f o

Article history: Received 14 June 2016 Received in revised form 11 September 2016 Accepted 12 September 2016 Available online xxxx Editor: F. Riget Keywords: Antarctica soils Environmental Risk Assessment Trace-element contamination Elutriate ecotoxicity Ecotoxicological Line of Evidence Chemical Line of Evidence

a b s t r a c t Generally, Antarctica is considered to be an untouched area of the planet; however, the region's ecosystems have been subject to increased human pressure for at least the past half-century. This study assessed soils of Fildes Peninsula, where trace element pollution is thought to prevail. Four soil samples were collected from different locations and assessed following tier 1 methodologies for chemical and ecotoxicological lines of evidence (LoE) used in typical soil Environmental Risk Assessment (ERA). Trace element quantification was run on soil samples and sequential extracts, and elutriates were used to address their ecotoxicity using a standard ecotoxicological battery. The highest levels of trace elements were found for Cr, Cu, Ni and Zn, which were well above baseline levels in two sites located near previously identified contamination sources. Trace element concentrations in soils were compared with soil quality guidelines to estimate the contribution of the chemical LoE for integrated risk calculations; risk was found high, above 0.5 for all samples. Total concentrations in soil were consistent with corresponding sequentially extracted percentages, with Cu and Zn being the most bioavailable elements. Bacteria did not respond consistently to the elutriate samples and cladocerans did not respond at all. In contrast, the growth of microalgae and macrophytes was significantly impaired by elutriates of all soil samples, consistently to estimated trace element concentrations in the elutriate matrix. These results translated into lower risk values

⁎ Corresponding author at: Department of Biology and CESAM (Centre for Environmental and Marine Studies), University of Aveiro, Aveiro, Portugal. E-mail address: [email protected] (J.L. Pereira).

http://dx.doi.org/10.1016/j.scitotenv.2016.09.091 0048-9697/© 2016 Elsevier B.V. All rights reserved.

444

J.L. Pereira et al. / Science of the Total Environment 574 (2017) 443–454

for the ecotoxicological compared to the chemical LoE. Nevertheless, integrated risk calculations generated either an immediate recommendation for further analysis to better understand the hazardous potential of the tested soils or showed that the soils could not adequately sustain natural ecosystem functions. This study suggests that the soil ecosystem in Fildes has been inadequately protected and supports previous claims on the need to reinforce protection measures and remediation activities. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Due to the remoteness of Antarctica, it has often been considered a pristine environment, little impacted by human activities. The continent is governed by consultative parties of the Antarctic Treaty, which was signed in 1959 and entered into force in 1961. In 1991 the Protocol on Environmental Protection to the Antarctic Treaty (also known as the Madrid Protocol or Environment Protocol) was agreed and entered into force in 1998 designating the Antarctic Treaty area (the area south of latitude 60°S) as ‘a natural reserve, devoted to peace and science’. The Protocol stipulates the need to plan human activities in order to minimize impacts upon the Antarctic environment and ecosystem. However, mineral, petroleum, and natural gas deposits do exist in Antarctica, and, as discussed by Chapman and Riddle (2005, p. 202A), ‘it would be naïve to assume that these will always remain unexploited’; this launches concerns on the possibility of keeping ecosystem disturbance at manageable levels. Aside from any projection on future mineral resource activities, Antarctica is already a contaminant recipient via atmospheric transport from exogenous sources (Bargagli, 2008) and local human activities. Human pressure across the region has increased with the growth of research, touristic and transportation activities, mostly in the Antarctic Peninsula and offshore island groups (Corsolini, 2009; Kennicutt et al., 2010; Liggett et al., 2011; Peter et al., 2008). Large quantities of fossil fuel are used locally to support scientific stations, as well as related transportation by air, land and water; together with the consequent residues produced, this represents a major potential source of environmental contamination (Braun et al., 2012; Corsolini, 2009; Davis, 1999; Liggett et al., 2011; Peter et al., 2008; Sanchez-Hernandez, 2000). In part motivated by these threats, studies incorporating trace element quantification in different environmental compartments, such as water, soils, catchment sediments, plants, mosses and lichens have been carried out throughout Antarctica (e.g. see the review by Bargagli, 2008); these studies established natural background levels and/or identified human associated, past and current contamination events (e.g., Crockett, 1998; Lu et al., 2012; Malandrino et al., 2009; Mão de Ferro et al., 2013; Webster et al., 2003). In this context, specific concern has been raised recently with regard to King George Island, and in particular the Fildes Peninsula, which holds a major logistic hub, the Chilean airport, as well as a high density of scientific stations and field huts (Braun et al., 2012; Peter et al., 2013). Fildes Peninsula is one of the relatively few large ice-free areas in Antarctica, supports relatively high biodiversity (ATS, 2009; Braun et al., 2012, 2014; Peter et al., 2013; Zhao and Xu, 2000) and is considered to be of high ecological interest (ASOC, 2007). Although Fildes contains two Antarctic Specially Protected Areas (ASPAs; ATS, 2015), the extension of protective management measures in Fildes Peninsula following research that identified substantial environmental threats has been advocated (ASOC, 2007; Braun et al., 2012, 2014; Peter et al., 2008). Furthermore, studies addressing trace element contamination in soils of the Fildes Peninsula generally support these concerns (see below for details). Background values of trace elements in these soils are naturally high but widespread evidence of human-driven impact exists (Amaro et al., 2015; Lu et al., 2012). Since at least 1993, increased levels of several trace elements (Ba, Cr, Cu, Mn, Ni, Pb, Zn) were found in soils in the vicinities of human settlements at Fildes (Krzyszowska, 1993). Lu et al. (2012) identified human activities as the cause of elevated Cd, Hg and Pb levels in soils compared

to background levels. This link was recently confirmed in soils from Fildes Bay by Amaro et al. (2015), who found non-natural enrichment of Pb and, to a lesser degree of As, Cd, Cu, Hg and Zn. Still, it is worth remarking at this stage that, although chemical quantification, for example following sequential extraction procedures, can be used to estimate bioavailability of trace elements, such estimates do not directly translate into actual biological effects (Schultz et al., 2004). The way trace elements exert a biological effect is largely dependent on the environmental chemical conditions and on the organism itself (Campbell et al., 2006). While more than ten years have elapsed since Riddle and Chapman (2003) highlighted ecotoxicology as the ‘missing link’ in polar environments for a proper assessment of the environmental risk posed by contamination events, apart some notable exceptions (see e.g. Nydahl et al., 2015; Saul et al., 2005; Schafer et al., 2009), little ecotoxicology research has been done with Antarctic soils. Capitalising on ongoing effort within Fildes Peninsula for characterising trace element contamination in terms of sources, levels, distribution and availability, lines of evidence other than those provided by chemical analysis should be examined for an improved definition of the associated environmental hazard (i.e. possibility of impact). Such a definition may or may not demand for the completion of a structured Ecological Risk (probability of impact) Assessment (ERA) approach, in this case applied retrospectively, that can support adequate management actions towards environmental protection. ERA schemes rely on multiple lines of evidence (LoE) (Chapman et al., 2002). Typical soils ERAs consider the chemical, ecotoxicological and ecological LoE (Jensen and Pedersen, 2006) following the soil quality ‘Triad’ (Jensen and Mesman, 2006) that developed from the sediment quality ‘Triad’ (e.g. Chapman, 1990; Chapman et al., 1997). The Triad is organised in tiers of increasing complexity starting with a problem formulation stage based on a preliminary site characterisation and a screening assessment (Jensen and Pedersen, 2006; Weeks and Comber, 2005). In screening tiers of site-specific ERA, ecotoxicological effects can be addressed by testing contaminated environmental samples in batteries of bioassays using standard test species. Soil elutriate testing is a suitable alternative in this context. Since uptake from pore water is an important exposure route of soil organisms to contaminants (Janssen et al., 1997), testing soil elutriates with standard aquatic species allows a preliminary evaluation, or at least a comparative ranking (for example for prioritization purposes) of soil toxicity (Jensen and Mesman, 2006). Elutriates from contaminated soils are typically complex mixtures; their toxicity can be properly assessed by using a battery of test species where different functional levels, exposure routes and toxicity endpoints are covered, the essential requirement being the validity (relevance for the contaminated site conditions) and sensitivity of the bioassay (Swartjes et al., 2012). By integrating data on trace metal quantification in soils and the ecotoxicity of elutriates, this study aimed to provide, for the first time, an integrated hazard assessment (tier 1 risk assessment stage) of selected soils from Fildes Peninsula, Antarctica. The soils were selected on the basis of a previous, wider survey (Amaro et al., 2015; Padeiro et al., 2016) highlighting hotspots of trace element contamination deriving from human activities. Test species and endpoints were carefully selected to provide a meaningful battery covering different functional levels and bearing increasing complexity (Keddy et al., 1995; Maisto et al., 2011; Pandard et al., 2006), while also considering convenience criteria such as the amount of soil available for testing, viz.: (i) luminescence

J.L. Pereira et al. / Science of the Total Environment 574 (2017) 443–454

inhibition by the bacterium Vibrio fischeri (AE, 1998), representing decomposers; (ii) growth inhibition of the green microalgae Raphidocelis subcapitata (OECD, 2011), representing primary producers with contaminant uptake via surface contact; (iii) growth inhibition by the macrophyte Lemna minor (OECD, 2006), also a primary producer but where contaminants can be uptaken both via surface contact and systemically; and (iv) immobilisation of the microcrustacean consumer Daphnia magna (OECD, 2004). As far as we are aware, this is the first study endeavouring to start a structured, integrated ERA with Antarctic soils. Such data may be important in determining the environmental hazard posed by the studied soils, the protective level required and the need for any remediation action. But more than that, this study can pave the way for further, improved assessments of contamination events in Antarctica, which, it is hoped, will necessarily lead to more robust decision-making regarding effective environmental monitoring and protection programmes.

445

2. Material and methods 2.1. Study area and sampling strategy Sampling sites were selected on the basis of a previous study (Amaro et al., 2015) and correspond to hotspots of trace metal contamination around Maxwell Bay in Fildes Peninsula, located in the southwest of King George Island, one of the South Shetland Islands in Antarctica (Fig. 1). Four sites were sampled in February 2014, three in the Ardley Cove area, representing contamination hotspots (AC1–3), and one in the Bellingshausen Dome area (BD), representing a putatively uncontaminated reference (no history of anthropogenic influence has been reported there). As depicted in Fig. 1, soils collected at Ardley Cove were in vicinity of reported sources of contamination, namely: fuel tanks, metal deposits and corroded buildings (Amaro et al., 2015). The four sites were sampled as part of the study by Padeiro et al. (2016)

Fig. 1. Location of Fildes Peninsula, King George Island, South Shetland Islands, Antarctica (a–b), where soil samples were taken. Soil sampling sites (AC1–3) in Ardley Cove (c) were marked as black circles; red pins indicate fuel storage tanks in the vicinities, brown areas denote waste disposal sites, the brick red area marks a fuel spill area, and black houses locate scientific station buildings. In the upper right corner (d) the location of the putative reference site (BD), in Bellingshausen Dome, is detailed. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

446

J.L. Pereira et al. / Science of the Total Environment 574 (2017) 443–454

that aimed to deliver a larger survey of trace element contamination in the area. The top soil layer (upper 10 cm) was sampled following ultra-clean protocol techniques using acid-decontaminated equipment and wearing powder-free latex gloves. Samples were collected with a decontaminated plastic spatula and stored in zip-locked plastic bags at −20 °C. At the laboratory, samples were oven-dried at 40 °C and then sieved (2 mm mesh size); only the fraction b 2 mm was used, to avoid the dilution of sample contents by coarse material (Mão de Ferro et al., 2013). Dry sieved soil samples were then stored in PTFE tubes until chemical analysis or in PTFE flasks until elutriate preparation for toxicity testing (see below), at 4 °C in the dark. 2.2. Trace elements analysis Trace element analyses of soils and aqueous extracts of soils were performed as briefly described below and detailed by Padeiro et al. (2016). For trace element analysis in soils, samples were dried, sieved and homogenized. Analyses were run in triplicate. Total mercury concentrations were determined by atomic absorption spectrometry (AAS) generally following Costley et al. (2000). For the quantification of As, Cd, Cr, Cu, Ni, Pb and Zn, soil samples were previously digested with Aqua Regia and hydrofluoric acid (HF). After cleansing samples for HF, total trace element concentrations were determined by ICP-MS. Sample preparation for ICP-MS quantification was based on mass dilution and two reagent blanks were prepared in a similar way for each set of 20 samples. In every ten samples a QC-S (quality control-solution) was measured in order to ensure the accuracy of the determinations (coefficient of variation b0.5%). In order to assess the availability of the trace elements in soils, sequential chemical extractions were performed in four steps with the most available fraction extracted in step (i) and the least available fraction extracted in step (iv). In step (i) water lixiviated elements were evaluated after extraction with ultrapure water. In step (ii) exchangeable elements were assessed through leaching the residue from (i) with sodium acetate. In step (iii) hydrous manganese and iron oxides were extracted by mixing the residue from (ii) with a hydroxylamine solution in 25% (v/v) acetic acid. Step (iv) consisted of the extraction of elements bound to sulphides by adding HCl to the residue from step (iii). After each successive extraction, the solutions were centrifuged at 3000 rpm for 10 min and the supernatant removed with a pipette. Trace element (As, Cd, Cr, Cu, Ni, Zn, Pb; Hg was not analysed given its very low levels in all soil samples) quantification in all extracts was performed by graphite furnace AAS. Blank reagents and quality control standards were used throughout the entire procedure. International certified reference materials MESS-3 (Marine Sediment, NRCC Canada) and SRM2710a (Montana Soil, NIST, USA) were used to ensure the accuracy of our procedure in the soil analysis. For all parameters, measured values were consistent within the ranges of certified values (P b 0.05). Analytical variability and homogeneity of soil samples (calculated as standard deviation (SD) of duplicates, with values below 10%) were considered satisfactory. 2.3. Toxicity testing 2.3.1. Organisms Bioluminescent bacteria (Vibrio fischeri) were supplied lyophilized as part of the Microtox® (AE, 1998) kit and reconstituted in the corresponding reconstitution solution immediately prior to testing. The microalga Raphidocelis subcapitata was obtained from unialgal cultures cyclically maintained in 150 mL Erlenmeyer glass vessels filled with 75 mL sterilized Woods Hole MBL culture medium (Stein, 1979). The macrophyte Lemna minor was collected in a pond and has been maintained in the laboratory as a successful long-term culture in Steinberg culture medium (OECD, 2006) renewed once a week. Monoclonal

cultures of Daphnia magna (clone A, sensu Baird et al., 1989a) were continuously reared in the laboratory in synthetic ASTM hardwater medium (ASTM, 1980) supplied with organic additive extracted from the algae Ascophyllum nodosum (Baird et al., 1989b). Cultures were renewed every other day, and the organisms were fed after renewal with R. subcapitata at a rate of 3.0 × 105 cells mL−1. All cultures were kept under a 16hL:8hD photoperiod and temperature of 20 ± 2 °C. 2.3.2. Toxicity tests Elutriates were prepared for toxicity testing within the following 8 weeks after collection, as recommended by USEPA (1998), in polypropylene vessels. Dry, sieved soils (see Section 2.1) were added to each toxicity test medium in a 1:4 (w/v) ratio. Vessels were shaken overnight in an orbital shaker (200 rpm), and left to settle for 8 h, all done at 4 °C in the dark. The overlying layer was centrifuged at 2500 ×g for 15 min. The supernatants (elutriate in each test medium) were stored at 4 °C in the dark until further use (48 h maximum holding period). Both pH and conductivity were measured (WTW-Multi3430 probe; n = 3) in neat elutriates for comparison with data for the corresponding blank media. The V. fischeri luminescence inhibition test was applied to elutriates following the instructions of the manufacturer through the 81.9% basic test protocol (AE, 1998). The growth inhibition of the green microalgae R. subcapitata following exposure to the sediment elutriates was assessed using a static bioassay conducted according to OECD (2011) and USEPA (2002), with adaptation to 24-well microplate use (Geis et al., 2000). The algae were exposed during 72 h under continuous illumination to serial dilutions of elutriates in MBL medium. Three replicates were established per treatment and each replicated well was filled with 990 μL test solution plus 10 μL microalgae inoculum adjusted so that the final nominal cell density at the beginning of the test was 104 cells mL−1. The test microplates were incubated as described above for algal cultures and the contents of each well were thoroughly mixed twice daily by repetitive pipetting to promote active gas exchange and prevent cell clumping. The microplates were kept under 23 ± 1 °C and continuous light supply as recommended by the guidelines for the whole 72-h exposure period. At the end of the bioassay, the biomass yield was calculated from microscopic cell density estimates based on a previously established calibration curve relating this parameter with spectrophotometric readings (440 nm). The growth inhibition tests with L. minor were performed according to OECD (2006) and USEPA (2002), with adaptation to 6-well microplate use (Kaza et al., 2007). Three colonies with three visible fronds each were harvested from the inoculum culture and randomly assigned to each well filled with 10 mL test solution. Three replicates were established per treatment, and treatments consisted of serial dilutions of elutriate in Steinberg medium plus a blank control: 0, 25, 50 and 100% elutriate. An additional replicated sample (n = 6) was collected from the inoculum culture and dried at 60 °C for 24 h to provide the initial dry weight for further biomass yield calculations. The test was kept under 23 ± 1 °C and continuous light supply as recommended in the guidelines for the whole 7-day exposure period. D. magna was exposed to increasing elutriate concentrations for 48 h, following the guidelines by OECD (2004) and USEPA (2002). Five newborn daphnids (b 24-h old; from the culture's 3rd–5th brood) were randomly assigned to each replicate and four replicates were established per treatment. The tests were carried out in glass vials filled with 10 mL test solution. No food or organic additives were provided during the test, and incubation conditions were kept as mentioned for rearing procedures (see above). At the end of the exposure period, each vial was checked for immobilized daphnids. 2.4. Data analysis Total trace element concentrations in soils, as well as baseline levels determined previously as typical from the Fildes peninsula in Antarctica

J.L. Pereira et al. / Science of the Total Environment 574 (2017) 443–454

(Amaro et al., 2015; Lu et al., 2012), were compared with different soil screening levels used widely as soil quality benchmarks: Ecological Soil Screening Levels (Eco-SSL) for soil invertebrates established under the United States framework for inorganic metals risk assessment (USEPA, 2005); Environment Canada Soil Quality Guidelines for environmental health considering the most protective land use, i.e. agriculture (CCME, 1999); Soil Values derived by Jänsch et al. (2007). The latter were used on risk calculations (see below). One-way ANOVA followed by the Tukey test was used to distinguish treatments within each R. subcapitata or L. minor bioassay. Whenever allowed by the dose-response curve, median effect concentrations (EC50), and respective 95% confidence intervals, were estimated as ecotoxicological references by non-linear regression, using the logistic equation fitted to the data through the least squares statistical method. Complying with the goal of assessing the environmental risk represented in the sampled sites at a tier 1 stage, a deterministic comparison of trace element concentrations and soil quality guidelines (chemical LoE), along with elutriate toxicity testing (ecotoxicological LoE) was carried out to produce risk benchmarks. Calculations followed the detailed guidance by Jensen and Mesman (2006) towards the establishment of risk values ranging between zero (no risk) and 1 (high risk) for both the chemical and the ecotoxicological LoE, hence comparable and allowing the calculation of an integrated risk benchmark. This uniform scaling is primarily achieved by comparing samples with a reference soil. In the present study the calculations were repeated assuming: (i) BD as a reference soil, because this sample was collected in an apparently low-impact area (see above); (ii) a model reference soil, considering baseline levels for trace elements found previously in Fildes soils (Amaro et al., 2015 for As; Lu et al., 2012 for Cd, Cr, Cu, Hg, Ni, Pb, Zn), and artificially setting reference ecotoxicological responses to 0% effect. For the chemical LoE, the total concentration in soils of each trace element was compared to the corresponding SSL, i.e. soil values by Jänsch et al. (2007). Then, assuming that the quantified elements may have dissimilar modes of toxic action in biological systems, the combined risk of all quantified elements was calculated through the response addition model (De Zwart and Posthuma, 2005). In the ecotoxicological LoE, effects induced by exposure to full-strength elutriate were scaled as proportions of luminescence inhibition for V. fischeri, yield inhibition for R. subcapitata and L. minor, and mortality for D. magna, all relative to the interim control first (stimulation effects were set to 0% inhibition for consistency within risk calculations) and then to the corresponding reference (see above). Combined risk deriving from different bioassays was calculated by averaging individual scaled effect. The risk derived from each LoE was integrated in the calculation of the integrated risk for each tested soil sample and the corresponding deviation parameter, the latter allowing the assessment of the consistency between risks represented by the involved LoE.

447

3. Results and discussion 3.1. Levels and availability of trace elements in soils Table 1 presents the levels of trace elements found in soils from the Bellingshausen Dome area (BD) and three sampling sites in the Ardley Cove area (AC1–AC3). The highest levels of trace elements were recorded in soils from Ardley Cove, particularly at AC2 for Cr and Ni and at AC3 for Cu and Zn, which was consistent with their proximity to putative contamination sources (a fuel spill area, station buildings and a waste disposal site; Fig. 1). Arsenic levels did not differ considerably between sampling sites, whereas levels of Pb, Cd and Hg were slightly higher at AC3 in comparison with the remaining sampling sites. Despite its proximity to another waste disposal site, soil AC1 contained lower trace element concentrations than AC2 and AC3, which made its contamination profile most similar to that of BD. Levels of As, Cr, Ni and Zn in BD were markedly above baseline levels established previously for Fildes Peninsula (Amaro et al., 2015; Table 1; Lu et al., 2012), which put in question the suitability of this site as a reference and led to our decision to use it as an additional test sample for that location to calculate related environmental risks (see Section 3.3). The highest deviation from what can be assumed as the background concentration in the study area was found for As, with all soil samples recording levels five times higher than common (baseline) levels found by Amaro et al. (2015). The trend for increased levels of particular trace elements found in Ardley Cove samples is common to many other sites across Antarctica, depending on the type of anthropogenic pressure at each location. Examples can be found in the literature for sites linked to noticeable human activity such as Doumer and King George Islands, Hut Point Peninsula on Ross Island, Seymour Island and Marie Byrd Land in West Antarctica (Carrasco and Prendez, 1991; Celis et al., 2015; Chaparro et al., 2007; Lupachev and Abakumov, 2013; Sheppard et al., 2000). Arsenic, Cr, Ni, Zn, and, to a lesser degree, Pb, were found in Ardley Cove at generally similar or higher levels than in areas elsewhere, with the trace elements likely to derive from specific contamination sources. This is supported by the differences between quantified levels and baseline levels (Amaro et al., 2015; Table 1; Lu et al., 2012), which ranged from two-fold for Pb at AC3 to more than five-fold for As within any of the sampled soils. Furthermore, the levels of trace elements in Ardley Cove were often above those found in non-impacted (see Amaro et al., 2015 for As and Cu; Krzyszowska, 1993) and even impacted soils in Fildes Peninsula (note the increase in 20 years by comparing to Krzyszowska, 1993; Lu et al., 2012); they were also generally above the levels found in other locations in Antarctica unaffected by anthropogenic contamination, such as some sites in McMurdo Station area and Dry Valleys, Victoria Land (Crockett, 1998; Webster et al., 2003), Admiralty Bay, King George Island (Santos et al., 2005) and Terra Nova Bay (Malandrino et al., 2009). Mercury stands out as an exception to

Table 1 Levels of trace elements (mg kg−1) in soils collected in the Bellingshausen Dome area (BD; putative reference site) and in Ardley Cove (AC1–AC3), Fildes Peninsula. Mean values (n = 3 analytical replicates) are presented, as well as the relative standard deviation (%) within brackets. The highest record found for each trace element was marked bold. References to baseline (or background) concentrations in Fildes soils and safety benchmarks (mg kg−1) were added to the table for comparative purposes.

As Cd Cr Cu Ni Pb Zn Hg

BD

AC1

AC2

AC3

Baselinea

Eco-SSLb

SQGc

Soil valuesd

18 (0.9) 0.2 (2.5) 38 (0.9) 69 (1.1) 19 (1.7) 5.3 (6.4) 83 (1.3) 0.007 (20)

17 (1.5) 0.1 (2.8) 54 (0.9) 62 (0.9) 23 (1.2) 7.6 (6.1) 74 (0.8) 0.007 (17)

19 (0.8) 0.1 (4.1) 95 (0.5) 72 (0.7) 35 (0.7) 8.0 (3.2) 81 (1.8) 0.006 (19)

18 (1.1) 0.2 (4.4) 35 (2.2) 111 (1.4) 17 (1.7) 11 (5.4) 116 (1.3) 0.012 (15)

3.57 0.37 22.56 89.45 10.37 5.44 51.41 0.013

– 140 – 80 – 1700 – –

17 3.8 64 63 50 70 200 0.012

5.63 6.78 5.02 55.0 64.0 163.5 160.3 0.001

a Baseline values for metal content in soils from Fildes Peninsula estimated using the method of relative cumulative frequency by Lu et al. (2012) for all elements but As, which was averaged from mean concentration values collected in groups of samples taken in Ardley Cove after excluding extreme values by Amaro et al. (2015). b US Ecological Soil Screening Levels (Eco-SSL) for soil invertebrates (USEPA, 2005). c Environment Canada Soil Quality Guidelines (CCME, 1999). d Soil values derived by Jänsch et al. (2007).

448

J.L. Pereira et al. / Science of the Total Environment 574 (2017) 443–454

this latter pattern: levels of geochemical origin (and not necessarily impacted) from McMurdo and Terra Nova Bay, in Victoria Land (Bargagli et al., 2005; Crockett, 1998), were almost one order of magnitude higher than those found in Ardley Cove soils; still Hg levels in Ardley Cove were found to be below general soil quality guidelines (SQG; Table 1). Although natural trace elements are typically present at high levels given the geochemistry of Antarctica (Malandrino et al., 2009), anthropogenic influence has also been demonstrated. High Cr levels may link to oil contamination (Alam and Sadiq, 1993; Caccia et al., 2003), and AC2, coincident with/adjacent to nearby fuel spill areas (Fig. 1), was the site with the highest levels of this trace element, with levels above the corresponding SQG (Table 1). Similarly, Cu and Pb have been associated with increased human impact relating to debris, runoff, fuel combustion, and particularly to aircraft transportation, shipping and sewage (Bargagli, 2008; Kennicutt et al., 2010; Kessler, 2013; Tin et al., 2009). Nevertheless, higher levels of these metals were found in AC3 (near waste disposal site) rather than in AC2, where stations are concentrated the most (and hence the theoretical likelihood of finding more contamination sources). Still, Pb was consistently found below the corresponding SQG (Table 1); As, Cd, Ni and Zn have generally similar sources (reviewed by Tin et al., 2009; Wuana and Okieimen, 2011) and were also found near or below SQGs. SQGs are used widely as protective benchmarks, but they were developed taking into account Canadian soils and soil processes (CCME, 1999), thus their applicability to other contexts should be assessed with care. For example, Antarctic soils are of low moisture, organic matter content and microbial biomass (Bargagli, 2008), resulting in slower rates of transport and transformation of persistent chemicals. The inconsistencies between SQG values and Eco-SSL values available or Soil Values (see Table 1), which are more directed by responses of eco-receptor representatives (Jänsch et al., 2007; USEPA, 2005), support further questioning of the protective ability of SQGs as far as Antarctica soils are concerned. Ideally, specific soil screening values for Antarctica should be set as reference concentration defining the threshold for triggering more detailed risk evaluations (Provoost et al., 2008; Snape et al., 2008). This would allow adequate site-specific decision making within the region by accounting for locally trivial chemical concentrations (e.g. trace element levels that are naturally high; Malandrino et al., 2009) and soil properties typical for the locality (e.g. metal toxicity towards plants and invertebrates depends on soil properties; Amorim et al., 2005; Rooney et al., 2007). Sequential extraction results (Fig. 2) also pointed towards an anthropogenic soil contamination. Metals deriving from anthropogenic enrichment of soils are typically more reactive or available (labile)

than those of geochemical origin, the former being the target of sequential extraction methods (Dung et al., 2013). In fact, there was spatial variation in the percentage of available trace elements, with a generally lower proportion extracted from the putative reference site at BD. Also, higher total concentrations in soils (Table 1) corresponded to higher extraction percentages (Fig. 2), further supporting an anthropogenic origin for element burden in Ardley Cove soils (see the rationale in e.g. Maiz et al., 2000). Cu, and to a much lesser extent, Zn are probably the elements of higher availability (notable in AC1–AC3) since their extractable percentages were the highest. The percentage extracted by water at the first stage was invariably small (max. 0.6% for Zn) or was below the detection limit, and both elements were mainly associated with oxides, carbonates and sulphides (step 1 to 3, respectively). However, these other exchangeable fractions (mainly oxides) were still likely to be available organisms in the soil or aquatic environment. Similarly, Pb in soils from AC2 and AC3 was probably bioavailable since a significant quantity was stored as Pb-oxides. Conversely, As and Cd were poorly or not leached at all, respectively, strongly suggesting that the soil burden of these elements is of geogenic origin and thus of low bioavailability. Moreover, mild levels of Ni were recorded in the whole extractable fractions (especially in AC2), although it was found mainly in the more refractory fractions such as carbonates and sulphides, thereby suggesting that this element is poorly available to organisms. Overall, Cu, Zn and Pb were more readily leached from Ardley Cove soils than the other elements and hence were more likely to be bioavailable and have a greater potential to exert toxic effects on exposed organisms. 3.2. Soil elutriates toxicity Elutriates from soils or sediments bearing high trace element burden may commonly have low pH values, generally in the range 3–5 (e.g., Antunes et al., 2007; Chapman et al., 2012; Loureiro et al., 2005a; Niemeyer et al., 2010; Soucek et al., 2000). This may, on the one hand, confound ecotoxicological assessments since the responses by sensitive organisms may be more immediately linked to pH than to the concentrations of contaminants per se; on the other hand, low pH is a promoter of metal mobilisation into the water column and thus a promoter of metal solubility and bioavailability (Chapman et al., 2012; de Paiva Magalhães et al., 2015; Dempsey et al., 1993; Soucek et al., 2000). However, this was not the case in the present study, where a mildly acidic lowest pH value was recorded in blank Steinberg media, with all full strength elutriates exhibiting slightly higher pH values (Table 2). Conversely, there was a marked increase in conductivity in elutriate

Fig. 2. Percentage of trace elements (Cr, Ni, Cu, Zn, As, Pb and Cd) sequentially extracted (step i–step iv) from soil samples collected in the Bellingshausen Dome area (BD; putative reference site) and in Ardley Cove (AC1–AC3), Fildes Peninsula, Antarctica.

J.L. Pereira et al. / Science of the Total Environment 574 (2017) 443–454

samples compared to the corresponding blank media, the most dramatic for elutriates prepared with distilled water and for BD elutriates (Table 2). While the former trend can be explained by the presence of several inorganic salts in synthetic test media, resulting in baseline levels higher than in distilled water (dissolution of natural salts in the soil only), the latter is more difficult to interpret given that BD soils were the least contaminated (Table 1). However, greater salinization of the BD soil sample was conceivable due to its closer proximity to seawater (Fig. 1). This particular BD feature may help to explain the toxic effects of the corresponding elutriate in tested freshwater species as depicted in Fig. 3. Results from the bioassays with D. magna were not graphically shown since the species was largely insensitive to the exposures over the short-term (12.5, 25, 50 and 100% of elutriates from each sample), despite daphnids being amongst the most metal-sensitive freshwater invertebrates (von der Ohe and Liess, 2004) and hence appropriate for integration in the battery used in the present study. No immobility (which is a surrogate used for mortality) was recorded except following exposure to 50% BD elutriate (5%) and 25% AC2 elutriate (also 5%); this inconsistency in the concentration-response relationship suggested that the observations should reflect random events rather than be related to elutriates' toxicity. The uptake of metals by daphnids occurs via dietary routes as well as directly from the aqueous phase, particularly for trace elements such as Hg, Cd, Ag, and Zn, or Se and MeHg (Tsui and Wang, 2007). It is therefore evident that trace element uptake from soil elutriates by D. magna in the present study can occur via both routes (both dissolved and particle-bound elements should be available), but their internal levels did not reach the detoxified binding capacity or excretion limits within the short exposure period, thereby leading to the observed absence of toxicity (for further information see discussion in e.g. Tsui and Wang, 2007; Wang, 2013). Such insensitivity of the cladoceran is unusual compared to the sensitivity found for Vibrio fischeri (Fig. 3). It is common to find elutriates from solid matrices contaminated with metallic elements (e.g. sludge, wastes and soils) that elicit comparatively higher or similar toxicity to cladocerans (Alvarenga et al., 2007, 2016; Antunes et al., 2008; Huguier et al., 2015; Maisto et al., 2011; Niemeyer et al., 2010; Pandard et al., 2006), although contrary examples can be found in the literature (Loureiro et al., 2005a). Moreover, there was no consistent trend in toxicity of Ardley Cove elutriates towards the bacteria. While AC1 and AC2 promoted a stimulation of bioluminescence, some inhibition occurred following exposure to elutriates from the most (AC3) and the least (BD) contaminated soil (Table 1; Fig. 1), the latter being that eliciting the highest toxicity. Comparison with the literature would suggest that none of these elutriates were likely to be toxic to the bacteria, although differences in soil properties or specific composition, interactions and overall strength within the trace element mixtures involved could act contrary to expectations. In fact, trace element burden of Ardley Cove soils is below that of soils with elutriates shown to impair V. fischeri luminescence (e.g. Maisto et al., 2011). In contrast, similar levels were found in soils with toxic elutriates where one or two

449

elements were linked to the observed effects (e.g. Loureiro et al., 2005a; Niemeyer et al., 2010; Rodriguez-Ruiz et al., 2015). Given the observed inconsistencies, it is worth noting that interference can occur in Microtox® test readings due to the soil silt and clay content; suspended particles in the elutriates can promote bacteria adsorption, artificially decreasing luminescence, and hence overestimating toxicity (Ringwood et al., 1997). Although we did not characterise the particle size of our soils, they were roughly distinct as to texture and hence the possibility of such interferences in the results should not be immediately disregarded. Also, salinity correction and pH buffering made while testing with the Microtox® may have biased metal bioavailability and artificially constrained conclusions, as discussed by Abbondanzi et al. (2003). Producer species responded to all soil elutriates, with growth impairment in microalgae being considerably more severe than in the macrophytes (Fig. 3). Only the microalgae distinguished full strength elutriates, positioning BD as the least toxic but still inducing significant impairment of biomass yield at full strength (F4, 25 = 9.81, P b 0.001). Concentration-response curves obtained for microalgae exposed to elutriates from Ardley Cove samples were distinct (Fig. 3, middle panel). Therefore, although yield inhibition promoted by full-strength elutriates was high in all cases (above 69% relative to the internal control), different NOEC/LOEC (F4, 25 = 33.64, F4, 22 = 129.34, F4, 25 = 27.21 for AC1–3, respectively, all with P b 0.001) and EC50 values were retrieved; based on these latter (generally preferred for toxicity interpretation compared to NOEC/LOEC values; see e.g. Van der Hoeven, 1997), AC2 elutriate was the most toxic to microalgae compared to AC1 and AC3 whose EC50 cannot truly be distinguished (overlapping confidence intervals). Macrophytes were mildly, but significantly impaired by elutriates (F3, 10 = 11.37, P b 0.001 for BD; F3, 11 = 23.28, P b 0.001 for AC1; F3, 10 = 5.02, P = 0.002 for AC2; F3, 11 = 34.71, P b 0.001 for AC3), and similar maximum levels of effect were reached (27–35%; Fig. 3, right-hand panel) thus not distinguishing the toxicity of elutriates in practice. Microalgae have been frequently identified as key, sensitive organisms within testing batteries with metal-bearing soil elutriates (Huguier et al., 2015; Maisto et al., 2011) and their lower tolerance compared to Lemna sp. has been confirmed (Hentati et al., 2015). Although macrophytes exhibit additional intake routes for contaminants (via surface contact plus systemically), this increased tolerance is typical for Lemna sp., which has even been recommended as an effective bioremediation tool for metal-loaded effluents (Teixeira et al., 2014; Uysal and Tanerb, 2009; Uysal, 2013). The impact of trace element contaminants on organisms is dependent on their availability and consequent inter-compartment transferability; while the residual fraction (combined within the mineral lattice) should be non-labile, elements bound to carbonates, Fe/Mn oxides, organic matter and sulphides constitute the exchangeable fraction and are thus potentially bioavailable (Dung et al., 2013; Okoro et al., 2012). The toxicity of all elutriates towards microalgae and macrophytes may seem to be inconsistent with the availability profile provided by sequential chemical extractions (Fig. 2), since low cumulative

Table 2 Indicative parameters for the quality of the media used in toxicity testing. Mean pH and conductivity (n = 3; standard error within brackets) are shown for full-strength elutriates prepared with soil samples collected at Bellingshausen Dome area (BD) and Ardley Cove (AC1–AC3), as well as for blank media used for culturing and control treatments.

BD AC1 AC2 AC3 Blank media

pH Cond. (μS cm−1) pH Cond. (μS cm−1) pH Cond. (μS cm−1) pH Cond. (μS cm−1) pH Cond. (μS cm−1)

dH20

MBL

Steinberg

ASTM

7.30 (0.01) 1094 (9) 7.34 (0.02) 262.0 (2.9) 8.24 (0.20) 101.6 (7.1) 7.51 (0.10) 79.1 (3.5) 8.08 (0.08) 1.28 (0.21)

7.34 (0.01) 1740 (73) 7.77 (0.03) 896.0 (9.0) 7.81 (0.09) 765.7 (8.4) 7.01 (0.06) 721.0 (14) 7.27 (0.09) 515.7 (15)

6.67 (0.01) 1906 (17) 6.91 (0.01) 1070 (5) 7.02 (0.13) 937.3 (5.4) 6.61 (0.04) 854.3 (9.8) 6.00 (0.09) 950.3 (16)

7.60 (0.05) 1614 (7) 7.83 (0.04) 766.0 (3.1) 7.92 (0.06) 692.3 (4.4) 7.16 (0.07) 556.0 (4.6) 8.59 (0.03) 632.7 (15)

450

J.L. Pereira et al. / Science of the Total Environment 574 (2017) 443–454

30

50

25

40

20

30

15

20

10

10

5

0

0

-10

35 30 25 20 15 10 5 0

100

0

40 35 30 25 20 15 10 5 0

80

-5 60

-10 40

-15 20

-20

0 100

0 -5 -10 -15 -20 -25 -30

35 30 25 20 15 10 5 0

80 60 40 20 0 100

30 25 20 15 10 5 0

20 80

15 60

10 40

5

20

0

0 0 10 20 30 40 50 60 70 80 90 100

0 10 20 30 40 50 60 70 80 90 100

0 10 20 30 40 50 60 70 80 90 100

Fig. 3. Responses of Vibrio fischeri, Raphidocelis subcapitata and Lemna minor (organised in vertical panels) to a range of dilutions of full-strength elutriates prepared in the corresponding test media from Antarctica soils, sampled in Bellingshausen Dome (putative reference; BD) and Ardley Cove (AC1–AC3), Fildes Peninsula (organised horizontally). The circles indicate mean experimental records, including the corresponding error bars that represent the standard error. The lines were added for clarity, except for R. subcapitata where they represent adjusted logistic models for ECx estimation. Maximum effects recorded and EC50 values with the 95% confidence interval within brackets were added to the corresponding plot whenever applicable; statistically significant differences (Dunnet test; P b 0.05) between elutriate treatments and the corresponding control were marked with an asterisk. Different yy′ scaling was used to favour readability and maximum effects were clarified to avoid scaling-related misinterpretation.

percentages of extraction were obtained (max. 20–30% for Cu in AC1–3 in all the four extraction steps; see Section 3.1 above for detailed interpretation). However, on considering that the first two steps of extraction, representing water lixiviated and exchangeable elements, should reflect the contaminants readily available in elutriates (note the elutriate preparation protocol), the consistency between chemical and toxicological evidences improves. In Table A.1 (Appendix A, Supplementary Information), the elemental concentrations of elutriates have been estimated by integrating total element concentration in soils with sequentially extracted percentages. Trace element quantification in elutriate samples prepared in different testing media was not practicable mostly because we were extremely limited by the soil mass available, and consequently the suitability of producing elutriate volumes allowing sub-sampling for chemical quantification was impaired. The direct analysis of elutriates should be considered in further studies and although this approach would be ideal, the exercise exposed in Table A.1 is based in widespread methods for bioavailability evaluation hence becoming an alternative for discussion. The estimated Cu and Zn levels were above corresponding EC50 benchmarks from the literature (Drost et al., 2007; Naumann et al., 2007; Paixão et al., 2008), mostly with respect to toxicity towards microalgae

but also occasionally towards macrophytes (see the graphic interpretation in Table A.1). In addition, there could be synergic effects involving other trace elements present at lower levels within the elutriate mixture contributing to the overall toxicity records (O'Halloran, 2006; following the interpretation by e.g. Warne, 2003). Both Cr and Ni were found available (steps iii and iv; Table A.1) in AC soils at concentrations above EC50 benchmarks for microalgae (Deleebeeck et al., 2009; Paixão et al., 2008), and their potential contribution to the overall elutriates toxicity should not be disregarded. Additionally, analytical factors, including variability in extraction procedures and the interplay between peculiar features of the extraction methods and the source matrix, may have contributed as discussed previously (Campbell et al., 2006; Dung et al., 2013; Filgueiras et al., 2002; Rauret, 1998; Young et al., 2006). Moreover, it is important to emphasize that information gathered with non-biological systems should be translated into biological contexts with caution, because the particular conditions of the exposure matrix (e.g. pH, organic ligands, dissolved organic matter, ionic strength, presence of competing cations such as Ca2 +) influence the chemical interactions determining the partitioning of metals between the solid and the aqueous phase (Alkorta et al., 2006; Campbell et al., 2006; Peijnenburg and Jager, 2003; Zimmerman and Weindorf, 2010).

J.L. Pereira et al. / Science of the Total Environment 574 (2017) 443–454

All of these considerations reinforce the validity of actually performing tests on the biota before making conclusions on the bioavailability of contaminants when addressing a putatively hazardous scenario. 3.3. Antarctic soils ERA - tier I The individual contribution of the chemical and the ecotoxicological LoE and the integrated risk calculated for sampled Antarctica soils are shown in Table 3. As all samples were assessed using a hypothetical reference sample with baseline levels established for Fildes (Amaro et al., 2015 for As; Lu et al., 2012 for Cd, Cr, Cu, Ni, Pb, Zn and Hg), the hazardous potential order was invariably BD b AC1 b AC2 ≤ AC3. If using BD as a reference, AC1 was always found to be the soil at lowest risk, with AC2 and AC3 yielding similar risk values. If interpreted as suggested by Jensen and Mesman (2006; p. 53), the integrated risk for AC2 and AC3 bears a deviation term higher than 0.4 in both of the above scenarios, resulting in an immediate recommendation for further, more detailed studies to better understand the hazardous potential of these soils. Alternatively, if a risk estimation, even with high uncertainty, is required for immediate decision-making, then AC2 and AC3 should be presented as soils bearing an unacceptable risk for the ecosystem and unable to sustain a nature function. BD use should be restricted as well to industrial and residential use. Because BD was sourced as a control sample from a non-human-impacted area, this classification alone illustrates the importance of developing and employing site-specific environmental guidelines for Antarctica so that the feasibility of calculated integrated risks can be improved. The integrated risk calculated for AC1 was of lower uncertainty and, although bearing lower risk than AC2 and AC3, this soil is still assessed as inadequate to sustain nature functions, regardless of the reference used in calculations. Initially BD was collected to represent a reference sample, but the chemical and ecotoxicological assessment questioned such an attribute as detailed above. Thus, although calculations were also made using BD as a reference, we tend to recommend the alternative approach to assess the level of attention needed to properly manage these soils. This is consistent with both the classification of Fildes Peninsula as an area containing an Antarctic Specially Protected Area (ATS, 2015) and the awareness raised by numerous studies on the increasing contamination in Antarctica (see the Introduction for the detailed context). Another conservative and hence ultimately protective option was the selection of soil values to scale within the chemical LoE. The derivation of Canadian soil quality guidelines relies on LOEC or ECx data interpolated or extrapolated from dose-response curves, with arbitrary uncertainty (‘safety’) factors often being applied when these benchmarks are unavailable (CCME, 1999). Use of safety factors may or may not have a positive role in reducing uncertainty, but they greatly reduce the likelihood of underestimating risk; consequently, they may become

Table 3 Risk values calculated for the chemical LoE (combined risk, considering total concentrations of trace elements in soils compared to SSL) and the Ecotoxicological LoE (using effects of full strength elutriates compared to a reference), as well as the integrated risk with the corresponding deviation parameter. The guidance by Jensen and Mesman (2006) was followed in calculations either comparing all samples with an hypothetical reference sample containing the baseline levels of trace metals established for Ardley Cove soils (Amaro et al., 2015; Lu et al., 2012) or comparing AC1–AC3 with BD, thus assuming BD as a reference soil sample.

Chemical LoE – combined risk All samples BD as reference Ecotox LoE – risk All samples BD as reference Integrated risk (deviation) All samples BD as reference

BD

AC1

AC2

AC3

0.60 0.00

0.70 0.23

0.85 0.62

0.85 0.59

0.24 0.00

0.47 0.30

0.32 0.11

0.39 0.20

0.45 (0.45) 0.60 (0.28) 0.68 (0.64) 0.68 (0.54) 0.00 0.27 (0.09) 0.41 (0.62) 0.43 (0.46)

451

overprotective, often leading to unrealistic conclusions in hazard and risk assessment (Chapman et al., 1998). The US Eco-SSL benchmarks for soil invertebrates are protective of ecological receptors for soils with pH ranging 4.0–8.5 and organic matter content below 10%, which is the case with the soils in the present study; also, they are based upon feasible ECx estimates or calculated NOAL-based MATC values (USEPA, 2005). However, Eco-SSL values were not available for all of the trace elements addressed in the present study (Table 1), which ultimately led to the option of using soil values by Jänsch et al. (2007) for consistency within the analysis; these correspond to the HC5 yield from Species Sensitivity Distributions fed with EC50 data retrieved in validated toxicity tests with a large range of soil organisms. Eventually a less conservative option was the use of elutriate testing to compose the ecotoxicological LoE, although this and equivalent approaches (e.g. Whole Effluent Toxicity testing) are commonly accepted as adequate for screening stages of ERA (Chapman, 2000; Jensen and Mesman, 2006). Elutriate testing has been found to be less sensitive than direct contact testing (e.g. Antunes et al., 2008) although the opposite has also been observed (e.g. Loureiro et al., 2005a). Pandard et al. (2006) found with a wide range of samples that the comparative sensitivity of elutriate testing is related to the species used. Still, soil contact toxicity tests could be applied in addition to better feed risk calculations, and eventually account for the high level of uncertainty observed. Standard solid-phase V. fischeri testing and behavioral testing with collembolans, isopods or earthworms have been purposed as adequate platforms for screening stages of ERA (Antunes et al., 2008; Loureiro et al., 2005b; Niemeyer et al., 2010; Mooney et al., 2013; Wasley et al., 2016). However, these were not straightforward to implement in the present study for two main reasons. First, considerable amounts of soil are required for these bioassays, which is logistically difficult with samples from Antarctica. Furthermore, the benefits gained through directly testing actual soil samples could be impaired by the use of standard species and standard test conditions that would very likely bias the conclusions through under- or overestimation of the hazard. This has already been evidenced for polar freshwater and marine environments (Chapman and Riddle, 2005; Chapman, 2016), and it is very likely to apply to soils as well. A potential way to overcome such limitations is to develop adapted methodologies to culture and test using organisms that can represent Antarctica soil biota in future environmental risk assessments. Suitable options may exist, especially in coastal ice-free areas, where the proximity of the sea buffers climate severity, increases water availability and supports the increase in soil organic matter content (Bargagli, 2008). Developed soils can be found in Antarctica sustaining an appreciable abundance and variety of organisms, including bacteria (e.g. Chong et al., 2012); fungi (Farrell et al., 2011); lichens and mosses (e.g. Green et al., 2015); algae (Garraza et al., 2011) and two species of flowering plants, the grass Deschampsia antarctica (Park et al., 2012) and cushion plant Colobanthus quitensis; ciliates, rotifers, mites, tardigrades, springtails and nematodes (Caruso et al., 2007; Lee et al., 2013a; Velasco-Castrillón and Stevens, 2014; Velasco-Castrillón et al., 2014); and non-native midges and enchytraeids (Hughes and Worland, 2010). Knowledge has been developed on the ecological preferences, physiology and abiotic constraints of several of these organisms in Antarctica (e.g. Caruso et al., 2007; Chong et al., 2012; Farrell et al., 2011; Garraza et al., 2011; Lee et al., 2013a; McGaughran et al., 2010; Nielsen et al., 2011; Park et al., 2012), which constitutes an important basis for the development of meaningful culturing and testing protocols. Also encouraging is the recent adaptation of culturing and/or testing methodologies with Antarctic bacteria (Schafer et al., 2009), moss and terrestrial algae (Nydahl et al., 2015), and plants (Lee et al., 2013b), although the range of stressors already tested is naturally limited and further work needs to be carried out to properly frame the sensitivity of the species to environmental pollutants. Finally, it is important to mention that the integrated risk parameter was calculated considering only the two available LoE (Chemical and

452

J.L. Pereira et al. / Science of the Total Environment 574 (2017) 443–454

Ecotoxicological; Table 3). This would have introduced some bias into both this parameter and the deviation, which are ideally established using the full triad including the Ecological LoE. At the time of the sampling it was logistically unfeasible (field constraints, limited laboratory facilities and sample transportation could compromise the results) to carry out studies that could configure the Ecological LoE. Floristic surveys are recommended in tier I (Jensen and Mesman, 2006) but the plant cover in Fildes is very limited (only two sparsely distributed plant species; Bargagli, 2008). Analysis of the spatial distribution of lichen and moss biomass could provide a useful alternative, especially given that previous studies already exist for the region (Bohuslavová et al., 2012). However, relevant taxonomic expertise or sample preservation for later molecular identification would be required, which is inconsistent with the typically low requirements of a screening stage within ERA. Also, and more generally, relevant ecological indicators such as soil fauna feeding activity (using e.g. bait-lamina in the field) or basal soil respiration can be used to complete the ecological LoE in tier I (Niemeyer et al., 2010), since the logistic and technical requirements involved are not highly demanding. 4. Conclusion Although putative environmental threats over Antarctica are an awareness flag for the general public concerning the state of the Antarctic ecosystem, controversial arguments have been placed by the scientific community regarding soil contamination and the magnitude of its environmental impact. These mostly focus on whether trace metal contamination in soils is of geogenic or anthropogenic origin, with some authors suggesting that the contribution of human activities in Antarctica is negligible apart from some notable exceptions (e.g. Bargagli, 2008). This offers a parallel to perhaps the most basic question concerning the establishment of ERA frameworks worldwide to address contaminated environments and eventually assess pollution, i.e. “How much contamination is ‘too much’?” for a given ecosystem (Campbell et al., 2006); as well as it reflects the need to discuss “how clean is clean enough?” (Bowen, 1972), towards environmental remediation in Antarctica (see e.g. the reasoning and discussion in Nydahl et al., 2015). By applying an adapted screening stage of ERA to four soil samples from Ardley Cove in Antarctica, this study integrated chemical quantification and its scaling to known background levels with ecotoxicological data on the responses of standard organisms to the bioavailable fraction of trace elements burdening the soils. While there were several shortcomings within the approach (related to technical and logistical difficulties faced in the field) that may bias a definitive conclusion, the assessment indicated the unsuitability of all soils to support natural ecosystem functions. Together with high levels of uncertainty in the conclusions of this screening stage, this highlights the need for further, more detailed studies to allow a more reliable determination of the levels of environmental risk involved, and better inform a decision to proceed with remediation or not. Besides the challenging exercise of adapting, for the first time, a typical soil ERA approach to inhospitable Antarctica, more specifically, this study suggests that the current level of environmental management in Fildes Peninsula (ATS, 2015) needs improvement to ensure proper protection of the local soil ecosystem functions. Acknowledgements Authors would like to thank the Portuguese Polar Programme and the Chilean Antarctic Institute for logistic support. Thanks are also due for the financial support to CQE (UID/QUI/00100/2013) and CESAM (UID/AMB/50017), to FCT/MEC through national funds, and the cofunding by the FEDER, within the PT2020 Partnership Agreement and Compete 2020. This study received financial support through the Portuguese Polar Programme (PROPOLAR) by the Portuguese Foundation for Science and Technology (FCT). Joana L. Pereira and Patrícia Pereira were recipients of grants from the FCT (SFRH/BPD/01971/2014 and SFRH/

BPD/69563/2010, respectively). Ana Padeiro and Eduardo Amaro were supported by the programme “PROPOLAR, Bolsas de Mobilidade de Jovens Cientistas Polares. This paper contributes to the ‘State of the Antarctic Ecosystem’ research programme (AntEco) of the Scientific Committee on Antarctic Research (SCAR). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.scitotenv.2016.09.091. References Abbondanzi, F., Cachada, A., Campisi, T., Guerra, R., Raccagni, M., Iacondini, A., 2003. Optimisation of a microbial bioassay for contaminated soil monitoring: bacterial inoculum standardisation and comparison with Microtox® assay. Chemosphere 53, 889–897. http://dx.doi.org/10.1016/S0045-6535(03)00717-3. AE, 1998. Microtox Acute Toxicity Test. Azur Environmental, Carlsbad. Alam, I.A., Sadiq, M., 1993. Metal concentrations in Antarctic sediment samples collected during the Trans-Antarctica 1990 Expedition. Mar. Pollut. Bull. 26, 523–527. http:// dx.doi.org/10.1016/0025-326X(93)90472-V. Alkorta, I., Epelde, L., Mijangos, I., Amezaga, I., Garbisu, C., 2006. Bioluminescent bacterial biosensors for the assessment of metal toxicity and bioavailability in soils. Rev. Environ. Health 21, 139–152. http://dx.doi.org/10.1515/REVEH.2006.21.2.139. Alvarenga, P., Palma, P., Gonçalves, A.P., Fernandes, R.M., Cunha-Queda, A.C., Duarte, E., Vallini, G., 2007. Evaluation of chemical and ecotoxicological characteristics of biodegradable organic residues for application to agricultural land. Environ. Int. 33, 505–513. http://dx.doi.org/10.1016/j.envint.2006.11.006. Alvarenga, P., Mourinha, C., Farto, M., Palma, P., Sengo, J., Morais, M.C., Cunha-Queda, C., 2016. Ecotoxicological assessment of the potential impact on soil porewater, surface and groundwater from the use of organic wastes as soil amendments. Ecotoxicol. Environ. Saf. 126, 102–110. http://dx.doi.org/10.1016/j.ecoenv.2015.12.019. Amaro, E., Padeiro, A., Mão de Ferro, A., Mota, A.M., Leppe, M., Verkulich, S., Hughes, K.a., Peter, H.-U., Canário, J., 2015. Assessing trace element contamination in Fildes Peninsula (King George Island) and Ardley Island. Antarctic. Mar. Pollut. Bull. 97, 523–527. http://dx.doi.org/10.1016/j.marpolbul.2015.05.018. Amorim, M.J.D.B., Römbke, J., Schallnass, H.-J., Soares, A.M.V.M., 2005. Effect of soil properties and aging on the toxicity of copper for Enchytraeus albidus, Enchytraeus luxurious, and Folsomia candida. Environ. Toxicol. Chem. 24, 1875–1885. http://dx.doi.org/ 10.1897/04-505R.1. Antunes, S.C., Pereira, R., Gonçalves, F., 2007. Evaluation of the potential toxicity (acute and chronic) of sediments from abandoned uranium mine ponds. J. Soils Sediment. 7, 368–376. http://dx.doi.org/10.1065/jss2007.08.247. Antunes, S.C., Castro, B.B., Pereira, R., Gonçalves, F., 2008. Contribution for tier 1 of the ecological risk assessment of Cunha Baixa uranium mine (Central Portugal): II. Soil ecotoxicological screening. Sci. Total Environ. 390, 387–395. http://dx.doi.org/10.1016/j. scitotenv.2007.07.053. ASOC, 2007. Implementing the Madrid protocol: a case study of Fildes Peninsula, King George Island. XXX Antarctic Treaty Consultative Meeting, p. 10 New Delhi. ASTM, 1980. Standard Practice for Conducting Acute Toxicity Tests With Fishes, Macroinvertebrates and Amphibians - Vol. Report E 729-80. Philadelphia. ATS, 2009. Management plan for Antarctic Specially Protected Area No. 125 - Fildes Peninsula, King George Island. Antarct. Treaty Syst. (URL) http://www.ats.aq/ documents/recatt/Att424_e.pdf. ATS, 2015. Status of Antarctic Specially Protected Area and Antarctic Specially Managed Area Management Plans [WWW Document]. Antarct. Treaty Secr. URL http://ats. aq/documents/ATCM38/WW/atcm38_ww005_e.pdf (accessed 2.9.16). Baird, D.J., Barber, I., Bradley, M., Calow, P., Soares, A.M.V.M., 1989a. The Daphnia bioassay: a critique. Hydrobiologia 188 (189), 403–406. http://dx.doi.org/10.1007/ BF00027806. Baird, D.J., Soares, A.M.V.M., Girling, A., Barber, I., Bradley, M.C., Calow, P., 1989b. The longterm maintenance of Daphnia magna Straus for use in ecotoxicity tests: problems and prospects. In: Lokke, H., Tyle, H., Bro-Rasmussen, F. (Eds.), Proceedings of the First European Conference on Ecotoxicology. Lyngby, pp. 144–148. Bargagli, R., 2008. Environmental contamination in Antarctic ecosystems. Sci. Total Environ. 400, 212–226. http://dx.doi.org/10.1016/j.scitotenv.2008.06.062. Bargagli, R., Agnorelli, C., Borghini, F., Monaci, F., 2005. Enhanced deposition and bioaccumulation of mercury in Antarctic terrestrial ecosystems facing a coastal polynya. Environ. Sci. Technol. 39, 8150–8155. http://dx.doi.org/10.1021/es0507315. Bohuslavová, O., Šmilauer, P., Elster, J., 2012. Usnea lichen community biomass estimation on volcanic mesas, James Ross Island. Antarctica. Polar Biol. 35, 1563–1572. http://dx. doi.org/10.1007/s00300-012-1197-0. Bowen, D.H.M., 1972. How clean is clean enough? Environ. Sci. Technol. 6 (2), 103. http:// dx.doi.org/10.1021/es60061a600 Editorial. Braun, C., Mustafa, O., Nordt, A., Pfeiffer, S., Peter, H.-U., 2012. Environmental monitoring and management proposals for the Fildes Region, King George Island. Antarctica. Polar Res. 18206. http://dx.doi.org/10.3402/polar.v31i0.18206. Braun, C., Hertel, F., Mustafa, O., Nordt, A., Pfeiffer, S., Peter, H., 2014. Environmental assessment and management challenges of the Fildes Peninsula region. In: Tin, T., Liggett, D., Maher, P.T., Lamers, M. (Eds.), Antarctic Futures. Springer Science + Business Media, Dordrecht, pp. 169–191 http://dx.doi.org/10.1007/97894-007-6582-5.

J.L. Pereira et al. / Science of the Total Environment 574 (2017) 443–454 Caccia, V.G., Millero, F.J., Palanques, A., 2003. The distribution of trace metals in Florida Bay sediments. Mar. Pollut. Bull. 46, 1420–1433. http://dx.doi.org/10.1016/S0025326X(03)00288-1. Campbell, P.G.C., Chapman, P., Hale, B.A., 2006. Risk assessment of metals in the environment. In: Harrison, R.M., Hester, R.E. (Eds.), Chemicals in the Environment: Assessing and Managing Risk. The Royal Society of Chemistry, pp. 102–130. Carrasco, M.A., Prendez, M., 1991. Element distribution of some soils of continental Chile and the Antarctic Peninsula. Projection to atmospheric-pollution. Water Air Soil Pollut. 57–8, 713–722. http://dx.doi.org/10.1007/BF00282935. Caruso, T., Borghini, F., Bucci, C., Colacevich, A., Bargagli, R., 2007. Modelling local-scale determinants and the probability of microarthropod species occurrence in Antarctic soils. Soil Biol. Biochem. 39, 2949–2956. http://dx.doi.org/10.1016/j.soilbio.2007.06. 010. CCME, 1999. Canadian Soil Quality Guidelines for the Protection of Environmental and Human Health - Canadian Environmental Quality Guidelines. Celis, J., Barra, R., Espejo, W., González-Acuna, D., Jara, S., 2015. Trace element concentrations in biotic matrices of Gentoo penguins (Pygoscelis papua) and coastal soils from different locations of the Antarctic Peninsula. Water Air Soil Pollut. 226. http://dx.doi. org/10.1007/s11270-014-2266-5. Chaparro, M.A.E., Nuñez, H., Lirio, J.M., Gogorza, C.S.G., Sinito, A.M., 2007. Magnetic screening and heavy metal pollution studies in soils from Marambio Station, Antarctica. Antarct. Sci. 19, 379. http://dx.doi.org/10.1017/S0954102007000454. Chapman, P.M., 1990. The sediment quality triad approach to determining pollution-induced degradation. Sci. Total Environ. 97-98, 815–825. http://dx.doi.org/10.1016/ 0048-9697(90)90277-2. Chapman, P.M., 2000. Whole effluent toxicity testing—usefulness, level of protection, and risk assessment. Environ. Toxicol. Chem. 19, 3–13. http://dx.doi.org/10.1002/etc. 5620190102. Chapman, E.E.V., Hedrei Helmer, S., Dave, G., Murimboh, J.D., 2012. Utility of bioassays (lettuce, red clover, red fescue, Microtox, MetSTICK, Hyalella, bait lamina) in ecological risk screening of acid metal (Zn) contaminated soil. Ecotoxicol. Environ. Saf. 80, 161–171. http://dx.doi.org/10.1016/j.ecoenv.2012.02.025. Chapman, P.M., 2016. Toxicity delayed in cold freshwaters? J. Great Lakes Res. 42, 286–289. http://dx.doi.org/10.1016/j.jglr.2015.03.018. Chapman, P.M., Riddle, M.J., 2005. Toxic effects of contaminants in polar marine environments. Environ. Sci. Technol. 39, 200A–207A. http://dx.doi.org/10.1016/j. chemosphere.2010.02.012. Chapman, P.M., Anderson, B., Carr, S., Engle, V., Green, R., Hameedi, J., Harmon, M., Haverland, P., Hyland, J., Ingersoll, C., Long, E., Rodgers, J., Salazar, M., Sibley, P.K., Smith, P.J., Swartz, R.C., Thompson, B., Windom, H., 1997. General guidelines for using the Sediment Quality Triad. Mar. Pollut. Bull. 34, 368–372. http://dx.doi.org/ 10.1016/S0025-326X(96)00138-5. Chapman, P.M., Fairbrother, A., Brown, D., 1998. A critical evaluation of safety (uncertainty) factors for ecological risk assessment. Environ. Toxicol. Chem. 17, 99–108. http:// dx.doi.org/10.1002/etc.5620170112. Chapman, P.M., McDonald, B.G., Lawrence, G.S., 2002. Weight-of-evidence issues and frameworks for sediment quality (and other) assessments. Hum. Ecol. Risk Assess. An Int. J. 8, 1489–1515. http://dx.doi.org/10.1080/20028091057457. Chong, C.W., Pearce, D.A., Convey, P., Yew, W.C., Tan, I.K.P., 2012. Patterns in the distribution of soil bacterial 16S rRNA gene sequences from different regions of Antarctica. Geoderma 181–182, 45–55. http://dx.doi.org/10.1016/j.geoderma.2012.02.017. Corsolini, S., 2009. Industrial contaminants in Antarctic biota. J. Chromatogr. A 1216, 598–612. http://dx.doi.org/10.1016/j.chroma.2008.08.012. Costley, C.T., Mossop, K.F., Dean, J.R., Garden, L.M., Marshall, J., Carroll, J., 2000. Determination of mercury in environmental and biological samples using pyrolysis atomic absorption spectrometry with gold amalgamation. Anal. Chim. Acta 405, 179–183. http://dx.doi.org/10.1016/S0003-2670(99)00742-4. Crockett, A.B., 1998. Background levels of metals in soils, McMurdo Station, Antarctica. Environ. Monit. Assess. 50, 289–296. http://dx.doi.org/10.1023/A:1005837100718. Davis, P.B., 1999. Beyond guidelines: a model for Antarctic tourism. Ann. Tour. Res. 26, 516–533. http://dx.doi.org/10.1016/S0160-7383(98)00106-6. de Paiva Magalhães, D., da Costa Marques, M.R., Baptista, D.F., Buss, D.F., 2015. Metal bioavailability and toxicity in freshwaters. Environ. Chem. Lett. 13, 69–87. http://dx.doi. org/10.1007/s10311-015-0491-9. De Zwart, D., Posthuma, L., 2005. Complex mixture toxicity for single and multiple species: proposed methodologies. Environ. Toxicol. Chem. 24, 2665–2676. http://dx. doi.org/10.1897/04-639r.1. Deleebeeck, N.M.E., De Laender, F., Chepurnov, V.A., Vyverman, W., Janssen, C.R., De Schamphelaere, K.A.C., 2009. A single bioavailability model can accurately predict Ni toxicity to green microalgae in soft and hard surface waters. Water Res. 43, 1935–1947. http://dx.doi.org/10.1016/j.watres.2009.01.019. Dempsey, B.A., Tai, Y.L., Harrison, S.G., 1993. Mobilization and removal of contaminants associated with urban dust and dirt. Water Sci. Technol. 28, 225–230. Drost, W., Matzke, M., Backhaus, T., 2007. Heavy metal toxicity to Lemna minor: studies on the time dependence of growth inhibition and the recovery after exposure. Chemosphere 67, 36–43. http://dx.doi.org/10.1016/j.chemosphere.2006.10.018. Dung, T.T.T., Cappuyns, V., Swennen, R., Phung, N.K., 2013. From geochemical background determination to pollution assessment of heavy metals in sediments and soils. Rev. Environ. Sci. Biotechnol. 12, 335–353. http://dx.doi.org/10.1007/s11157-013-9315-1. Farrell, R.L., Arenz, B.E., Duncan, S.M., Held, B.W., Jurgens, J.A., Blanchette, R.A., 2011. Introduced and indigenous fungi of the Ross Island historic huts and pristine areas of Antarctica. Polar Biol. 34, 1669–1677. http://dx.doi.org/10.1007/s00300-011-1060-8. Filgueiras, A.V., Lavilla, I., Bendicho, C., 2002. Comparison of the standard SM&T sequential extraction method with small-scale ultrasound-assisted single extractions for metal partitioning in sediments. Anal. Bioanal. Chem. 374, 103–108. http://dx.doi.org/10. 1007/s00216-002-1427-3.

453

Garraza, G.G., Mataloni, G., Fermani, P., Vinocur, A., 2011. Ecology of algal communities of different soil types from Cierva Point, Antarctic Peninsula. Polar Biol. 34, 339–351. http://dx.doi.org/10.1007/s00300-010-0887-8. Geis, S.W., Fleming, K.L., Korthals, E.T., Searle, G., Reynolds, L., Karner, D.A., 2000. Modifications to the algal growth inhibition test for use as a regulatory assay. Environ. Toxicol. Chem. 19, 36–41. http://dx.doi.org/10.1002/etc.5620190105. Green, T.G.A., Seppelt, R.D., Brabyn, L.R., Beard, C., Türk, R., Lange, O.L., 2015. Flora and vegetation of Cape Hallett and vicinity, northern Victoria Land, Antarctica. Polar Biol. 38, 1825–1845. http://dx.doi.org/10.1007/s00300-015-1744-6. Hentati, O., Abrantes, N., Caetano, A.L., Bouguerra, S., Gonçalves, F., Römbke, J., Pereira, R., 2015. Phosphogypsum as a soil fertilizer: ecotoxicity of amended soil and elutriates to bacteria, invertebrates, algae and plants. J. Hazard. Mater. 294, 80–89. http://dx. doi.org/10.1016/j.jhazmat.2015.03.034. Hughes, K.A., Worland, M.R., 2010. Spatial distribution, habitat preference and colonization status of two alien terrestrial invertebrate species in Antarctica. Antarct. Sci. 22, 221–231. http://dx.doi.org/10.1017/S0954102009990770. Huguier, P., Manier, N., Chabot, L., Bauda, P., Pandard, P., 2015. Ecotoxicological assessment of organic wastes spread on land: towards a proposal of a suitable test battery. Ecotoxicol. Environ. Saf. 113, 103–111. http://dx.doi.org/10.1016/j.ecoenv.2014.11. 017. Jänsch, S., Roembke, J., Schallinass, H.-J., Terytze, K., 2007. Derivation of soil values for the path “soil-soil organisms” for metals and selected organic compounds using species sensitivity distributions. Environ. Sci. Pollut. Res. 14, 308–318. http://dx.doi.org/10. 1065/espr2006.06.310. Janssen, R.P.T., Peijnenburg, W., Posthuma, L., VandenHoop, M., 1997. Equilibrium partitioning of heavy metals in Dutch field soils. I. Relationship between metal partition coefficients and soil characteristics. Environ. Toxicol. Chem. 16, 2470–2478. http://dx.doi.org/10.1002/etc.5620161206. Ecological risk assessment of contaminated land - decision support for site specific investigation. In: Jensen, J., Mesman, M. (Eds.), RIVM Report 711701047. Dutch National Institute for Public Health and Environment (RIVM). Jensen, J., Pedersen, M.B., 2006. Ecological risk assessment of contaminated soil. Rev. Environ. Contam. Toxicol. 186, 73–105. http://dx.doi.org/10.1180/0026461056950274. Kaza, M., Mankiewicz-Boczek, J., Izydorczyk, K., Sawicki, J., 2007. Toxicity assessment of water samples from rivers in Central Poland using a battery of microbiotests – a pilot study. Polish J. Environ. Stud. 16, 81–89. Keddy, C.J., Greene, J.C., Bonnell, M.A., 1995. Review of whole-organism bioassays: soil, freshwater sediment and freshwater assessment in Canada. Ecotoxicol. Environ. Saftey 30, 221–251. http://dx.doi.org/10.1006/eesa.1995.1027. Kennicutt II, M.C., Klein, A., Montagna, P., Sweet, S., Wade, T., Palmer, T., Sericano, J., Denoux, G., 2010. Temporal and spatial patterns of anthropogenic disturbance at McMurdo Station. Antarctica. Environ. Res. Lett. 5, 1–10. http://dx.doi.org/10.1088/ 1748-9326/5/3/034010. Kessler, R., 2013. Sunset for leaded aviation gasoline. Environ. Health Perspect. 121 http:// dx.doi.org/10.1289/ehp.121-a54. Krzyszowska, A., 1993. Human impact around polar stations on Fildes peninsula (King George island, Antarctica). XX Polar Symposium - Man Impact on Polar Environment, pp. 203–208 Lublin. Lee, J.E., Le Roux, P.C., Meiklejohn, K.I., Chown, S.L., 2013a. Species distribution modelling in low-interaction environments: insights from a terrestrial Antarctic system. Aust. Ecol. 38, 279–288. http://dx.doi.org/10.1111/j.1442-9993.2012.02401.x. Lee, J., Noh, E.K., Choi, H.S., Shin, S.C., Park, H., Lee, H., 2013b. Transcriptome sequencing of the Antarctic vascular plant Deschampsia antarctica Desv. under abiotic stress. Planta 237, 823–836. http://dx.doi.org/10.1007/s00425-012-1797-5. Liggett, D., McIntosh, A., Thompson, A., Gilbert, N., Storey, B., 2011. From frozen continent to tourism hotspot? Five decades of Antarctic tourism development and management, and a glimpse into the future. Tour. Manag. 32, 357–366. http://dx.doi.org/ 10.1016/j.tourman.2010.03.005. Loureiro, S., Ferreira, A.L.G., Soares, A.M.V.M., Nogueira, A.J.A., 2005a. Evaluation of the toxicity of two soils from Jales Mine (Portugal) using aquatic bioassays. Chemosphere 61, 168–177. http://dx.doi.org/10.1016/j.chemosphere.2005.02.070. Loureiro, S., Soares, A., Nogueira, A.J.A., 2005b. Terrestrial avoidance behaviour tests as screening tool to assess soil contamination. Environ. Pollut. 138, 121–131. http://dx. doi.org/10.1016/j.envpol.2005.02.013. Lu, Z., Cai, M., Wang, J., Yang, H., He, J., 2012. Baseline values for metals in soils on Fildes Peninsula, King George Island, Antarctica: the extent of anthropogenic pollution. Environ. Monit. Assess. 184, 7013–7021. http://dx.doi.org/10.1007/ s10661-011-2476-x. Lupachev, A.V., Abakumov, E.V., 2013. Soils of Marie Byrd Land, West Antarctica. Eurasian Soil Sci. 46, 994–1006. http://dx.doi.org/10.1134/S1064229313100049. Maisto, G., Manzo, S., De Nicola, F., Carotenuto, R., Rocco, A., Alfani, A., 2011. Assessment of the effects of Cr, Cu, Ni and Pb soil contamination by ecotoxicological tests. J. Environ. Monit. 13, 3049–3056. http://dx.doi.org/10.1039/C1em10496a. Maiz, I., Arambarri, I., Garcia, R., Millán, E., 2000. Evaluation of heavy metal availability in polluted soils by two sequential extraction procedures using factor analysis. Environ. Pollut. 110, 3–9. http://dx.doi.org/10.1016/S0269-7491(99)00287-0. Malandrino, M., Abollino, O., Buoso, S., Casalino, C.E., Gasparon, M., Giacomino, A., La Gioia, C., Mentasti, E., 2009. Geochemical characterisation of Antarctic soils and lacustrine sediments from Terra Nova Bay. Microchem. J. 92, 21–31. http://dx.doi.org/10.1016/ j.microc.2008.09.003. Mão de Ferro, A., Mota, A.M., Canário, J., 2013. Sources and transport of As, Cu, Cd and Pb in the environmental compartments of Deception Island, Antarctica. Mar. Pollut. Bull. 77, 341–348. http://dx.doi.org/10.1016/j.marpolbul.2013.08.037. McGaughran, A., Convey, P., Redding, G.P., Stevens, M.I., 2010. Temporal and spatial metabolic rate variation in the Antarctic springtail Gomphiocephalus hodgsoni. J. Insect Physiol. 56, 57–64. http://dx.doi.org/10.1016/j.jinsphys.2009.09.003.

454

J.L. Pereira et al. / Science of the Total Environment 574 (2017) 443–454

Mooney, T.J., King, C.K., Wasley, J., Andrew, N.R., 2013. Toxicity of diesel contaminated soils to the subantarctic earthworm Microscolex macquariensis. Environ. Toxicol. Chem. 32, 370–377. http://dx.doi.org/10.1002/etc.2060. Naumann, B., Eberius, M., Appenroth, K.J., 2007. Growth rate based dose-response relationships and EC-values of ten heavy metals using the duckweed growth inhibition test (ISO 20079) with Lemna minor L. clone St. J. Plant Physiol. 164, 1656–1664. http://dx.doi.org/10.1016/j.jplph.2006.10.011. Nielsen, U.N., Wall, D.H., Adams, B.J., Virginia, R.A., 2011. Antarctic nematode communities: observed and predicted responses to climate change. Polar Biol. 34, 1701–1711. http://dx.doi.org/10.1007/s00300-011-1021-2. Niemeyer, J.C., Moreira-Santos, M., Nogueira, M.A., Carvalho, G.M., Ribeiro, R., Da Silva, E.M., Sousa, J.P., 2010. Environmental risk assessment of a metal-contaminated area in the tropics. Tier I: screening phase. J. Soils Sediment. 10, 1557–1571. http://dx. doi.org/10.1007/s11368-010-0255-x. Nydahl, A.C., King, C.K., Wasley, J., Jolley, D.F., Robinson, S.A., 2015. Toxicity of fuel-contaminated soil to Antarctic moss and terrestrial algae. Environ. Toxicol. Chem. 34, 2004–2012. http://dx.doi.org/10.1002/etc.3021. OECD, 2004. Daphnia sp. Acute Immobilisation Test, OECD Guideline for the Testing og Chemicals. Test no. 202. http://dx.doi.org/10.1787/9789264069947-en. OECD, 2006. Lemna sp. Growth Inhibition Test, OECD Guidelines for the Testing of Chemicals. Guideline no. 221. http://dx.doi.org/10.1787/9789264016194-en. OECD, 2011. Freshwater Alga and Cyanobacteria, Growth Inhibition Test. OECD Guidelines for the Testing of Chemicals. Guideline no. 221. http://dx.doi.org/10.1787/ 9789264069923-en. O'Halloran, K., 2006. Toxicological considerations of contaminants in the terrestrial environment for ecological risk assessment. Hum. Ecol. Risk. Assess. 12, 74–83. http://dx. doi.org/10.1080/10807030500428603. Okoro, H.K., Fatoki, O.S., Adekola, F.A., Ximba, B.J., Snyman, R.G., 2012. A review of sequential extraction procedures for heavy metals speciation in soil and sediments. Open Access Sci. Reports 1, 181. http://dx.doi.org/10.4172/scientificreports.181. Padeiro, A., Amaro, E., Correia Dos Santos, M.M., Araújo, M.F., Gomes, S.S., Leppe, M., Verkulich, S., Hughes, K.A., Peter, H.-U., Canário, J., 2016. Trace element contamination and availability in Fildes Peninsula, King George Island, Antarctica. Environ. Sci. Process. Impacts 18, 648–657. http://dx.doi.org/10.1039/C6EM00052E. Paixão, S.M., Silva, L., Fernandes, A., O'Rourke, K., Mendonça, E., Picado, A., 2008. Performance of a miniaturized algal bioassay in phytotoxicity screening. Ecotoxicology 17, 165–171. http://dx.doi.org/10.1007/s10646-007-0179-4. Pandard, P., Devillers, J., Charissou, A.M., Poulsen, V., Jourdain, M.J., Férard, J.F., Grand, C., Bispo, A., 2006. Selecting a battery of bioassays for ecotoxicological characterization of wastes. Sci. Total Environ. 363, 114–125. http://dx.doi.org/10.1016/j.scitotenv. 2005.12.016. Park, J.S., Ahn, I.Y., Lee, E.J., 2012. Influence of soil properties on the distribution of Deschampsia antarctica on King George Island, Maritime Antarctica. Polar Biol. 35, 1703–1711. http://dx.doi.org/10.1007/s00300-012-1213-4. Peijnenburg, W.J.G.M., Jager, T., 2003. Monitoring approaches to assess bioaccessibility and bioavailability of metals: matrix issues. Ecotoxicol. Environ. Saf. 56, 63–77. http://dx.doi.org/10.1016/S0147-6513(03)00051-4. Peter, H.-U., Buesser, C., Mustafa, O., Pfeiffer, S., 2008. Risk Assessment for the Fildes Peninsula and Ardley Island, and Development of Management Plans for their Designation as specially Protected or Specially Managed Areas - Research Report 203 13 124 UBA-FB 001155e Dessau-Rosslau. Peter, H., Braun, C., Janowski, S., Nordt, A., Stelter, M., 2013. The Current Environmental Situation and Proposals for the Management of the Fildes Peninsula Region DessauRoßlau Germany. Provoost, J., Reijnders, L., Swartjes, F., Bronders, J., Carlon, C., D'Alessandro, M., Cornelis, C., 2008. Parameters causing variation between soil screening values and the effect of harmonization. J. Soils Sediment 8, 298–311. http://dx.doi.org/10.1007/s11368-0080026-0. Rauret, G., 1998. Extraction procedures for the determination of heavy metals in contaminated soil and sediment. Talanta 46, 449–455. http://dx.doi.org/10.1016/S00399140(97)00406-2. Riddle, M.J., Chapman, P.M., 2003. Polar ecotoxicology – a missing link. Antarct. Sci. 15, 317. http://dx.doi.org/10.1017/S0954102003001433. Ringwood, A.H., DeLorenzo, M.E., Ross, P.E., Holland, A.F., 1997. Interpretation of Microtox® solid-phase toxicity tests: the effects of sediment composition. Environ. Toxicol. Chem. 16, 1135–1140. http://dx.doi.org/10.1002/etc.5620160607. Rodriguez-Ruiz, A., Etxebarria, J., Boatti, L., Marigómez, I., 2015. Scenario-targeted toxicity assessment through multiple endpoint bioassays in a soil posing unacceptable environmental risk according to regulatory screening values. Environ. Sci. Pollut. Res. 22, 13344–13361. http://dx.doi.org/10.1007/s11356-015-4564-x. Rooney, C.P., Zhao, F.J., McGrath, S.P., 2007. Phytotoxicity of nickel in a range of European soils: influence of soil properties, Ni solubility and speciation. Environ. Pollut. 145, 596–605. http://dx.doi.org/10.1016/j.envpol.2006.04.008. Sanchez-Hernandez, J.C., 2000. Trace elements contamination in Antarctic ecosystems. Rev. Environ. Contam. Toxicol. 166, 83–127. Santos, I.R., Silva-Filho, E.V., Schaefer, C.E.G.R., Albuquerque-Filho, M.R., Campos, L.S., 2005. Heavy metal contamination in coastal sediments and soils near the Brazilian Antarctic Station, King George Island. Mar. Pollut. Bull. 50, 185–194. http://dx.doi. org/10.1016/j.marpolbul.2004.10.009. Saul, D.J., Aislabie, J.M., Brown, C.E., Harris, L., Foght, J.M., 2005. Hydrocarbon contamination changes the bacterial diversity of soil from around Scott Base, Antarctica. FEMS Microbiol. Ecol. 53, 141–155. http://dx.doi.org/10.1016/j. femsec.2004.11.007. Schafer, A.N., Snape, I., Siciliano, S.D., 2009. Influence of liquid water and soil temperature on petroleum hydrocarbon toxicity in Antarctic soil. Environ. Toxicol. Chem. 28, 1409–1415. http://dx.doi.org/10.1897/08-434.1.

Schultz, E., Joutti, A., Räisänen, M.L., Lintinen, P., Martikainen, E., Lehto, O., 2004. Extractability of metals and ecotoxicity of soils from two old wood impregnation sites in Finland. Sci. Total Environ. 326, 71–84. http://dx.doi.org/10.1016/j.scitotenv.2003.12. 008. Sheppard, D.S., Claridge, G.G.C., Campbell, I.B., 2000. Metal contamination of soils at Scott Base, Antarctica. Appl. Geochemistry 15, 513–530. http://dx.doi.org/10.1016/S08832927(99)00055-4. Snape, I., Acomb, L., Barnes, D.L., Bainbridge, S., Eno, R., Filler, D.M., Plato, N., Poland, J.S., Raymond, T.A., Rayner, J.L., Riddle, M.J., Rike, A.G., Rutter, A., Schafer, A.N., Siciliano, S.D., Walworth, J.L., 2008. Contamination, regulation, and remediation: an introduction to bioremediation of petroleum hydrocarbons in cold regions. In: Filler, D.M., Snape, I., Barnes, D.L. (Eds.), Bioremediation of Petroleum Hydrocarbons in Cold Regions. Cambridge University Press, New York, pp. 1–37. Soucek, D.J., Cherry, D.S., Currie, R.J., Latimer, H.A., Trent, G.C., 2000. Laboratory to field validation in an integrative assessment of an acid mine drainage – impacted watershed. Environ. Toxicol. Chem. 19, 1036–1043. http://dx.doi.org/10.1002/etc. 5620190433. Stein, J.R., 1979. Handbook of Phycological Methods: Culture Methods and Growth Measurements, Handbook of Phycological Methods. Cambridge University Press, Cambridge. Swartjes, F.A., Rutgers, M., Lijzen, J.P.A., Janssen, P.J.C.M., Otte, P.F., Wintersen, A., Brand, E., Posthuma, L., 2012. State of the art of contaminated site management in The Netherlands: policy framework and risk assessment tools. Sci. Total Environ. 427–428, 1–10. http://dx.doi.org/10.1016/j.scitotenv.2012.02.078. Teixeira, S., Vieira, M.N., Espinha Marques, J., Pereira, R., 2014. Bioremediation of an ironrich mine effluent by Lemna minor. Int. J. Phytoremediation 16, 1228–1240. http://dx. doi.org/10.1080/15226514.2013.821454. Tin, T., Fleming, Z.L., Hughes, K.a., Ainley, D.G., Convey, P., Moreno, C.A., Pfeiffer, S., Scott, J., Snape, I., 2009. Impacts of local human activities on the Antarctic environment. Antarct. Sci. 21, 3–33. http://dx.doi.org/10.1017/S0954102009001722. Tsui, M.T.-K., Wang, W.-X., 2007. Biokinetics and tolerance development of toxic metals in Daphnia magna. Environ. Toxicol. Chem. 26, 1023–1032. http://dx.doi.org/10.1897/ 06-430r.1. USEPA, 1998. Great Lakes Dredged Material Testing and Evaluation Manual -Appendix G: Biological Effects Testing Procedures. USEPA, 2002. Short-term Methods for Estimating the Chronic Toxicity of Effluents and Receiving Waters to Freshwater Organisms - EPA-821-R-02-013 (Washington DC). USEPA, 2005. Guidance for Developing Ecological Soil Screening Levels, OSWER Directive 9285.7-55. Washington, D.C.. Uysal, Y., 2013. Removal of chromium ions from wastewater by duckweed, Lemna minor L. by using a pilot system with continuous flow. J. Hazard. Mater. 263, 486–492. http://dx.doi.org/10.1016/j.jhazmat.2013.10.006. Uysal, Y., Tanerb, F., 2009. Effect of pH, temperature, and lead concentration on the bioremoval of lead from water using Lemna minor. Int. J. Phytoremediation 11, 591–608. http://dx.doi.org/10.1080/15226510902717648. Van der Hoeven, N., 1997. How to measure no effect. Part III: statistical aspects of NOEC, ECx and NEC estimates. Environmetrics 8, 255–261. http://dx.doi.org/10.1002/ (SICI)1099-095X(199705)8:3b255::AID-ENV246N3.0.CO;2-P. Velasco-Castrillón, A., Stevens, M.I., 2014. Morphological and molecular diversity at a regional scale: a step closer to understanding Antarctic nematode biogeography. Soil Biol. Biochem. 70, 272–284. http://dx.doi.org/10.1016/j.soilbio.2013.12.016. Velasco-Castrillón, A., Schultz, M.B., Colombo, F., Gibson, J.A.E., Davies, K.A., Austin, A.D., Stevens, M.I., 2014. Distribution and diversity of soil microfauna from East Antarctica: assessing the link between biotic and abiotic factors. PLoS One 9. http://dx.doi.org/10. 1371/journal.pone.0087529. von der Ohe, P.C., Liess, M., 2004. Relative sensitivity distribution of aquatic invertebrates to organic and metal compounds. Environ. Toxicol. Chem. 23, 150–156. http://dx.doi. org/10.1897/02-577. Wang, W.X., 2013. Prediction of metal toxicity in aquatic organisms. Chinese Sci. Bull. 58, 194–202. http://dx.doi.org/10.1007/s11434-012-5403-9. Warne, M.S., 2003. A review of the ecotoxicity of mixtures, approaches to, and recommendations for, their management. In: Langley, A., Gilbey, M., Kennedy, B. (Eds.), Proceedings of the Fifth National Workshop on the Assessment of Site Contamination. National Environment Protection Council (NEPC), Adelaide, Australia, pp. 253–276. Wasley, J., Mooney, T.J., King, C.K., 2016. Soil invertebrate community change over fuelcontaminated sites on a subantarctic island: an ecological field-based line of evidence for site risk assessment. Integr. Environ. Assess. Manag. 12, 306–314. http://dx.doi. org/10.1002/ieam.1674. Webster, J., Webster, K., Nelson, P., Waterhouse, E., 2003. The behaviour of residual contaminants at a former station site. Antarctica. Environ. Pollut. 123, 163–179. http:// dx.doi.org/10.1016/S0269-7491(02)00403-7. Weeks, J.M., Comber, S.D.W., 2005. Ecological risk assessment of contaminated soil. Mineral. Mag. 69, 601–613. http://dx.doi.org/10.1180/0026461056950274. Wuana, R.a., Okieimen, F.E., 2011. Heavy metals in contaminated soils: a review of sources, chemistry, risks and best available strategies for remediation. ISRN Ecol. 2011, 1–20. http://dx.doi.org/10.5402/2011/402647. Young, S.D., Zhang, H., Tye, A.M., Maxted, A., Thums, C., Thornton, I., 2006. Characterizing the availability of metals in contaminated soils. I. The solid phase: sequential extraction and isotopic dilution. Soil Use Manag. 21, 450–458. http://dx.doi.org/10.1079/ sum2005348. Zhao, Y., Xu, C.-H., 2000. Human impacts on the terrestrial ecosystem of Fildes Peninsula of King George Island, Antarctica. J. Environ. Sci. 12, 12–17. Zimmerman, A.J., Weindorf, D.C., 2010. Heavy metal and trace metal analysis in soil by sequential extraction: a review of procedures. Int. J. Anal. Chem. 2010, ID387803. http:// dx.doi.org/10.1155/2010/387803 7 pages.