Science of the Total Environment 649 (2019) 1381–1392 Contents lists available at ScienceDirect Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv Seagrass soil archives reveal centennial-scale metal smelter contamination while acting as natural filters A. Lafratta a,⁎, O. Serrano a, P. Masqué a,c,d, M.A. Mateo a,b, M. Fernandes e,g, S. Gaylard f, P.S. Lavery a,b a School of Science, Centre for Marine Ecosystems Research, Edith Cowan University, 270 Joondalup Drive, Joondalup, WA 6027, Australia b Centro de Estudios Avanzados de Blanes, Consejo Superior de Investigaciones Cientificas, Blanes 17300, Spain c The UWA Oceans Institute, School of Physics, University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia d Departament de Física, Institut de Ciència i Tecnologia Ambientals, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain e South Australian Water Corporation (SA Water), GPO Box 1751, Adelaide, SA 5001, Australia f South Australian Environment Protection Authority, GPO Box 2607, Adelaide, SA 5001, Australia g School of the Environment, Flinders University, GPO Box 2100, Adelaide, SA 5001, Australia 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 • Seagrass soils provide a reliable archive of metal contamination over centennial time scales. • Pb, Zn and Cd concentrations in soil in- creased up to 9-fold after the onset of the Port Pirie smelter. • Pb isotopes show that the Port Pirie smelter is the main source of Pb con- tamination. • Seagrass meadows act as significant fil- ters and sinks of metals in Port Pirie. • Seagrass soils are a possible source of remobilized pollution following disturbance. a r t i c l e i n f o a b s t r a c t Article history: The upper Spencer Gulf in South Australia hosts the world's largest single stream Pb-Zn smelter, which has caused en- Received 20 April 2018 vironmental and health issues related to elevated metal concentrations in the surrounding environment. The area also Received in revised form 26 August 2018 has extensive seagrass meadows, occupying N4000 km2. We reconstructed the fluxes of heavy metals over the last Accepted 27 August 2018 ~3000 years through a multi-parameter study of the soil archives formed by the seagrass Posidonia australis. Pb, Zn Available online 30 August 2018 and Cd concentrations increased up to 9-fold following the onset of smelter operations in the 1880s, and the stable Editor: Jay Gan Pb isotopic signatures confirmed the smelter has been the main source of lead pollution in the seagrass soils until pres- ent. Preliminary estimates suggest that over the past 15 years seagrass meadows within 70 km2 of the smelter accumu- Keywords: lated ~7–15% of the smelter emissions in their soils. Here we demonstrate that seagrass meadows act as pollution filters Heavy metals and sinks while their soils provide a record of environmental conditions, allowing baseline conditions to be identified Lead isotopes and revealing the time-course of environmental change. South Australia © 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license Seagrass soil deposits (http://creativecommons.org/licenses/by-nc-nd/4.0/). Holocene Southern Ocean ⁎ Corresponding author. E-mail address: a.lafratta@ecu.edu.au (A. Lafratta). https://doi.org/10.1016/j.scitotenv.2018.08.400 0048-9697/© 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1382 A. Lafratta et al. / Science of the Total Environment 649 (2019) 1381–1392 1. Introduction occasions, exceeded the maximum permitted levels defined in seafood standards (Corbin and Wade, 2004; Edwards et al., 2001; Gaylard Heavy metals are released to the environment mainly through emis- et al., 2011) and the collection of benthic molluscs has been banned sions to the atmosphere and can subsequently be transferred to ecosys- since 1996 in some areas of Spencer Gulf (Gaylard, 2014). This marine tems resulting in environmental and human health problems (Järup, pollution has been attributed to different sources of emissions from 2003). Mining and smelting of metal ores can be important sources of the smelter, including direct discharges, spillage and fugitive dust emis- metal contaminants, with studies undertaken worldwide in the vicinity sions during ship loading, direct atmospheric deposition, and indirect of smelting facilities to understand the impact of metal contamination deposition through stormwater run-off from the Port Pirie urban area on soil, water, sediment and biota (Corbin and Wade, 2004; Ettler, (Gaylard, 2014; Tonkin Consulting, 2009). 2015; Zhang et al., 2012). Metals in the marine environment are found Vegetated coastal ecosystems, such as mangroves, tidal marshes and in dissolved and particulate forms, the latter associated with lithogenic seagrasses, store and immobilize metals in their organic-rich soils for or biogenic particles. Some of these metals accumulate in marine sedi- millennia (Serrano et al., 2016a, 2011). This is partly through the bioac- ments, where biotic and abiotic processes can release them back to cumulation of metals in their tissues, which are subsequently buried in the water column or where they can be permanently stored, limiting their soils (Harbison, 1986; Serrano et al., 2011; Weis and Weis, 2004), their bioavailability (Zhang et al., 2014). and also through the action of their canopies, which enhance particle Australia is the second largest Pb and Zn producer in the world after trapping (Gacia et al., 1999) and reduce re-suspension (Gacia and China (Geoscience Australia, 2013). The smelter located in Port Pirie Duarte, 2001). The enhanced trapping of sediment also leads to the (Fig. 1), in the upper Spencer Gulf, South Australia, is the world's largest net accumulation of soil and, in response, the vertical growth of the single-stream Pb and Zn smelter (Jackson and Abbot, 2008; Nyrstar, plants to avoid burial. This suite of mechanisms can result in seagrass 2007; SACOME, 2016), and has been operating continuously since soils acting as environmental archives, which are otherwise rare in dy- 1889 (BHP, 1985). It has been estimated that in the last 15 years the namic coastal environments; the enhanced trapping of contaminants smelter emitted to the surrounding environment an annual average of in the soils magnifies environmental changes in ambient, water column 56,000 kg of Pb, 96,000 kg of Zn, 6200 kg of Mn, 3700 of Cr, 2600 kg of metal concentrations while the enhanced vertical accretion of the soil Cd and 2700 kg of As, and smaller amounts of Ni, Cu, Se, Mg, Hg, Co helps to preserve the environmental history, physically and and Be (in decreasing order), ranging from 30 to 2000 kg per year biogeochemically (Burdige, 2007), allowing the reconstruction of his- (Department of the Environment and Energy, 2016). torical contamination (Serrano et al., 2013). Recent studies in the Med- The effects of Pb and other heavy metals in the environment around iterranean Sea and Australia successfully used seagrass soils as archives Port Pirie have been extensively studied. Elevated Pb concentrations, to investigate changes of ecosystem condition and functioning over equal to or above 30 μg/dL, were found in the blood of children when time (Mateo et al., 2010; Serrano et al., 2016a, 2013, 2011). Archives, a survey was conducted in 1982 (Maynard et al., 2005). In 2004, 60% such as seagrass soils, offer the potential to reconstruct environmental of children in Port Pirie still had elevated Pb concentrations in blood histories and identify the sources of pollution, thereby improving the (N10 μg/dL) (Maynard et al., 2005), though since 2011 the average ability of environmental scientists and managers to understand, predict child blood lead level was below the critical value of 5 μg/dL used in and manage ecological change in coastal ecosystems. Australia to assess risk (South Australia Health, 2016). High concentra- The upper Spencer Gulf hosts one of the largest temperate seagrass tions of heavy metals (e.g. Pb, Zn, Cd, Mn, Cu and As) in animals, plants, meadows in Australia, occupying N4000 km2 (Tkalich and terrestrial soils, marine sediments, water, and air around Port Pirie have Sundarambat, 2008) and providing significant potential for capturing been attributed to the smelter's activities (Edwards et al., 2001; Gaylard and storing metals. Here we use seagrass soil archives to reconstruct et al., 2011; Lent et al., 1992; Ross et al., 2003; Ward and Young, 1981). the history of metal pollution at Port Pirie over a centennial time- In the marine environment, the concentrations of some metals have, on scale, identifying the magnitude and origin of metal contamination Fig. 1. Port Pirie, location of the study site in the North Spencer Gulf. A. Lafratta et al. / Science of the Total Environment 649 (2019) 1381–1392 1383 over time, and assessing the role of seagrasses as pollution filters and analytical precision was obtained, with an average relative standard de- sinks. We analyzed several biogeochemical components in seagrass viation (RSD) of 2% for duplicate analyses for each element. The recov- soil cores, including, for the first time, stable Pb isotopes to identify ery of the elements was satisfactory (average 99%). Certified and the potential sources of lead contamination. measured values of the standard reference material NIST 2702, percent- age of recovery and limits of quantification (LOQs) for each chemical el- 2. Material and methods ement have been reported in Supplementary Table S1. Some samples had concentrations of Cd and Zn below the LOQ (0.05 and 4 ppm respec- 2.1. Study site tively), in which case they have been reported as half of that value. Sta- ble lead isotope ratios were analyzed in both A12 and A15 cores to Port Pirie is located on Germein Bay in the upper Spencer Gulf decipher the origin of Pb and, thereby, the potential contamination (South Australia) (33″ 19′ S, 138″ E; Fig. 1). The shallow waters (up to sources. They were corrected for instrument mass discrimination by 25 m) of the northern Spencer Gulf have limited exchange with the measuring the NIST SRM 981 Pb standard. Analytical precision was an open ocean (Kämpf et al., 2010), facilitating high rates of sediment de- average of ±0.5% and ±0.2% (RSD) for 206Pb/207Pb and 208Pb/207Pb ra- position. Spencer Gulf hosts extensive seagrass meadows, formed tios, respectively. mainly by Posidonia australis, P. sinuosa, Amphibolis antarctica and The geochronologies of both cores were determined through the Zostera mucronata (Tiller et al., 1989). This study was conducted in a analyses of 210Pb (for the last century) and 14C (for the last millennia). monospecific meadow of P. australis located about 1 km from the Concentrations of 210Pb were determined on the b125 μm soil fraction main smelter discharge point at First Creek, relatively sheltered from by alpha spectrometry through the measurement of its radioactive prevailing S-SE winds but exposed to W-NW winds, and about 3.5 km product 210Po using Passivated Implanted Planar Silicon (PIPS) detec- from the smelter, which is situated on the Port Pirie River. tors (CANBERRA, Meriden, CT, USA, Mod. PD-450.18 A.M.), assuming ra- dioactive equilibrium between the two radionuclides (Sanchez-Cabeza 2.2. Sampling and laboratory procedures et al., 1998). The concentrations of supported 210Pb in selected samples from both cores were determined through the measurement of 226Ra by In December 2014, two replicate soil cores (A12 and A15) were col- ultra-low background liquid scintillation counting (Quantulus 1220, lected from a seagrass meadow at Port Pirie (Fig. 1). The meadow was Wallac Oy, Finland) using a method adapted from Masque et al. located in a depth of 5 m and the replicate cores were collected within (2002). Excess 210Pb, used to estimate the sedimentation rates, was cal- 50 m of each other. The cores were sampled by manual percussion culated as the difference between measured total 210Pb concentrations and rotation using PVC pipes. Compression during coring was assessed and the concentration of supported 210Pb (Fig. 2). by measuring the length of the core protruding from the soil surface in- Both cores were radiocarbon dated using P. australis sheath debris side and outside the core (Glew et al., 2001). All results reported in this and shells (Supplementary Table S2). Sheaths were rinsed in ultrapure study refer to the decompressed depths (cm). The cores were sealed at MQ water and placed in a sonic bath to remove fine-grained, inorganic both ends, transported vertically and stored at 4 °C until processing in particles, and inspected under a stereomicroscope for attached contam- the laboratory. The cores were sliced at 0.5 cm-thick resolution for the inants before analyses. Shells were treated with 4% HCl to remove about top 20 cm, and at 1 cm intervals for the remainder. Each slice was 25% of the mass before being rinsed with Ultrapure MQ water. All sam- dried at 60 °C until constant dry weight (DW) to estimate dry bulk den- ples were dried at 60 °C to a constant weight before radiocarbon dating sity, homogenized and divided into sub-samples by quartering. One at AMS Direct Laboratory using standard procedures (ISO 17025 and ISO subsample was analyzed for soil grain-size, using a Mastersizer 2000 9001). laser-diffraction particle analyzer at the University of Barcelona (Spain), following digestion of the samples with 10% hydrogen peroxide 2.3. Numerical procedures to remove organic matter. Another sub-sample was ground in an agate mill and analyzed for: organic carbon; metals and other chemical ele- The Constant Flux:Constant Sedimentation rate model (CF:CS) ments; and stable Pb isotopes. (Krishnaswamy et al., 1971) was applied to obtain an age model and For organic carbon (Corg) analysis, about 1 g of ground sample was an average soil accumulation rates for the last few decades. Then ages acidified with 4% HCl to remove inorganic carbon, centrifuged for the last two centuries (200–0 yr BP) were extrapolated from the (3400 rpm during 5 min), and the supernatant with acid residues care- model. All raw radiocarbon dates were calibrated with CALIB software fully removed by pipette, avoiding resuspension. The sample was then v.4.4 (Stuiver et al., 1998) and corrected for the marine reservoir effect washed with Milli-Q water, centrifuged and the supernatant again re- (i.e. by subtracting 71 years (Bowman, 1985)), and are expressed as ra- moved. The residual samples were re-dried (60 °C until constant diocarbon dendrocalibrated years before present (BP, present taken as weight) and encapsulated in tin capsules. Corg was analyzed using a 2014 CE; Supplementary Table S2). Costech Elemental Analyzer interfaced to a Thermo-Finnegan Delta V Based on the 210Pb geochronology and radiocarbon dating, three dif- Isotope Ratio Mass Spectrometer at UH-Hilo Analytical Facilities. The ac- ferent periods were considered: the post-smelter (120–0 yr BP), recent curacy of the analysis of the Standard reference material NIST 8704 pre-smelter (200–120 yr BP) and pre-smelter (b3000 cal yr BP–120 yr (Buffalo River Sediment) was ≤1%. The Corg content (%) is reported for BP) periods. The impact of the smelter was assessed by comparing the the bulk (pre-acidified) samples. post-smelter metal concentrations against the recent pre-smelter pe- Concentrations of chemical elements (Pb, Cd, Zn, Cr, Co, Ni, Zr, Cu, riod, rather than the pre-smelter period, because the environmental Mn, As, Fe, Ti, S and Si) and stable lead isotopes were measured using in- conditions during the recent pre-smelter period are likely to have ductively coupled plasma mass spectrometry (ICP-MS) (Agilent, model been more similar to those of the post-smelter period, with the excep- 7500ce) at the Universitat Autònoma de Barcelona (Spain). For each tion of the smelter being operational and degree of coastal sample, 0.25 g of dried and homogenized bulk soil was digested at development. 240 °C using a mixture of concentrated HNO3, HF and HCl following The median concentrations of chemical elements during the recent EPA method 3052 (USEPA, 1996). After total digestion, the samples pre-smelter period (200–120 yr BP) were used to estimate enrichment were diluted with 1% HCl prior to injection into the ICP-MS. To control factors (EF) for the post-smelter period (120–0 yr BP) as follows: for possible spectral ICP-MS interferences due to the presence of HCl, the analyses were performed using a collision cell, using 5 mL/min of EF ¼ X max =X rps He. Reagent blanks and standard reference material NIST 2702 (Inor- ganics in Marine Sediment) were run in parallel to the samples. Good where Xmax is the maximum concentration of element X in the post- 1384 A. Lafratta et al. / Science of the Total Environment 649 (2019) 1381–1392 a median value of 0.7 g cm−3 in the layers corresponding to both the post-smelter and the recent pre-smelter periods (Supplementary Table S3). The median Corg concentration in the post-smelter period was 1.7–1.8%, while the median values for the recent pre-smelter period were 1.2–1.5%. The proportion of fine soil particles (b63 μm) in the post- smelter period was 30–40%, whereas in the recent pre-smelter period the median values were 30–33% (median values). Concentrations of Pb, Cd, Zn, Cu, Mn, Ti and Si increased from the be- ginning of the 20th century, reaching maxima in the 1970s and remain- ing approximately constant thereafter (Fig. 4). The concentrations of Co, Ni, Fe, As, S and Cr increased from the 1890s until 1950s, and decreased thereafter. Zr concentrations did not show clear trends along the cores. Major shifts in the concentrations of most chemical elements (i.e. Pb, Cd, Zn, Co, Ni, As, Fe, Cu and S) occurred in both cores between 130 and 100 yr BP, and again at around 60 to 40 yr BP (Fig. 4). In both cores there was a positive Enrichment Factor (EF) for all the elements during the post-smelter period. In particular, Pb, Cd and Zn had EFs ranging from 8 to 30, about an order of magnitude higher than for the other el- ements (Supplementary Table S3). In both post-smelter and recent pre- smelter periods, the range and median concentrations of Cr, Co, Ni, Zr, Cu, Mn, As, Fe, Ti, S and Si were similar in the two cores, while Pb, Cd and Zn concentrations were 2–3 times higher in A15 than in A12. Two principal components explained 75% of the total variance in the cores. PC1 accounted for 44% of the variance and was characterized by high positive loadings (N0.7) of chemical elements typically associated with anthropogenic activities (Pb, Cd, Zn, Cu and Cr), but also for some chemical elements usually related to lithogenic origin (Ti, Mn, Fe) in both cores (Supplementary Table S4). The second component (PC2) accounted for 31% of the variance and was characterized by high positive loadings (N0.5) for Ni, Fe, As, Co, Cr and S. PC1 and PC2 showed contrasting trends during the post-smelter period (Fig. 4): Fig. 2. Total (symbols) and supported (shaded area) 210Pb concentration profiles for cores both showed an initial increase in scores commencing between 130 A15 and A12. and 100 yr BP, followed by a second shift around 60 yr BP, when PC1 values became constant and PC2 values decreased. smelter period and Xrps is the median concentration of element X in the The inventories of Pb, Cd and Zn accumulated during the last recent pre-smelter period. 120 years in cores A15 and A12 ranged from 6.60 to 9.90 g Pb m−2, Principal Component Analysis (PCA) of chemical element concentra- 0.335 to 0.594 g Cd m−2 and 14.9 to 24.7 g Zn m−2 (Supplementary tions (Pb, Cd, Zn, Cr, Co, Ni, Zr, Cu, Mn, As, Fe, Ti, S and Si) was applied to Table S5). The accumulation rates of metals were greater in the post- explore the common variability of the chemical elements. Data were smelter period, in particular for Pb, Cd and Zn, which were 5– to 22– normalized using Z-scores (calculated as: [Xi − μ] / σ, where Xi is the fold higher than in the recent pre-smelter period (Supplementary concentration of a particular element in a particular sample, μ is the Table S5). mean of the population and σ is the standard deviation) to obtain Only some metals had concentrations that were significantly corre- average-centred distributions and avoid scaling effects (Martinez- lated with Corg: Co, Ni, and S were significantly correlated with Corg in Cortizas et al., 2013). Pearson correlation analysis was used to explore both cores; in addition, in core A15, Cr and Fe were significantly corre- the relationships between the chemical elements analyzed and the lated with Corg; and in core A12 As was correlated with Corg (Table 1). soil characteristics of the cores (Corg concentration and soil grain-size There were no significant correlations between the proportion of fine composition). Sequential Regime Shift Detection software (Rodionov, soil particles (b63 μm) and the concentrations of any element. 2004) was used to detect discontinuities in the temporal trends ob- For both cores, the stable lead isotope ratios increased with age of served in soil characteristics, chemical elements and PCA trends, with the samples. The oldest samples (N800 cal yr BP), the period representa- significant results indicating shift in the temporal trends (sig. level = tive of background Pb sources, had relatively high isotopic ratios 0.1). (208Pb:207Pb ratios of 2.430–2.382, and 206Pb:207Pb ratios 1.150–1.099) (Fig. 5). Samples with ages between 200 and 140 yr BP (1870s to 3. Results 1810s), representative of the recent pre-smelter period and the onset of European colonization, had lower 208Pb:207Pb ratios (2.349–2.337) The concentrations of excess 210Pb in cores A15 and A12 decreased and 206Pb:207Pb ratios (1.069–1.060). For the last 120 years (post- from the upper soil layers down to 7 cm depth (Fig. 2), allowing the ap- smelter period) the Pb isotopic ratios were even lower (2.335–2.322 plication of the CF:CS rate model. The average soil accumulation rates for 208Pb:207Pb and 1.061–1.045 for 206Pb:207Pb). over the last two centuries (200–0 yr BP) were 0.095 ± 0.016 cm yr−1 (or 0.069 ± 0.012 g cm−2 yr−1) for core A15 and 0.15 ± 0.04 cm yr−1 4. Discussion (or 0.16 ± 0.04 g cm−2 yr−1) for core A12. The dated samples had a temporal resolution of ~6 years per 0.5 cm. Radiocarbon analyses The analyses of a variety of parameters along the seagrass cores showed a chronological continuity along depth (Supplementary allowed us to show that seagrass meadows act as metal sinks and can Table S2) with respect to the 210Pb age-depth model, with the deepest be used as archives to successfully reconstruct the history of metal con- part of core A15 and A12 aged at 3100 and 3300 cal yr BP, respectively. tamination in the marine environment of Port Pirie. The variations in In both cores, soil density, Corg content and soil grain size composi- chemical element concentrations that we observed correspond to shifts tion varied with depth (Fig. 3). In both cores, soil dry bulk density had from pre-European baseline conditions to post-European settlement A. Lafratta et al. / Science of the Total Environment 649 (2019) 1381–1392 1385 Fig. 3. Bulk density, organic carbon content (%Corg) and percentage fine soil particles (% b63 μm) along cores A15 and A12 from Port Pirie. Organic carbon and fine soil particles contents are expressed as percentage of the total sample DW. The red horizontal lines indicate depths at which shifts were identified over the last 200 years using the Regime Shift Detection Software (see text). The last 120 years (yr BP ~1890s to 2014) are representing the post smelter period, indicated by the grey area. The y-axes in the upper panel show the years before present (yr BP) and the calendar years (calendar yr), while the lower panel show depths (cm) and calibrated years before present [cal yr BP]. and then to the post-smelter period (last 120 years). The magnitude of (Morelli et al., 2012), where they are considered to reflect the effects contamination, variation in the soil characteristics profiles and changes of European settlement (e.g. land-use change, agriculture and industry) in soil and chemical element accumulation rates during these periods increasing the loading of metals to coastal waters. However, the magni- indicate a history of anthropogenic contamination in the area, a conclu- tude of contamination at Port Pirie, which is several times greater than sion also supported by the PCA analyses. Finally, shifts in the Pb isotopic at Moreton Bay and Oyster Harbour (Fig. 7) despite all three sites having signatures confirm the smelter as the main source of Pb accumulated in similar baseline concentrations, and the greater enrichment factors the seagrass meadows surrounding Port Pirie. (Supplementary Table S3) for three main products of the smelter (Pb, The smelter began operating in 1889 (BHP, 1985), 40 years after the Zn and Cd), point to a specific industrial source of contamination at first European settlement in Port Pirie (The Register, 1921), producing Port Pirie. Furthermore, while there was a slight increase in the contents mainly Pb, Zn, Cd, sulphuric acid, Cu, Au, Ag and antimonial alloys of fine particles and Corg in the soil accumulated after European settle- (Ward and Hutchings, 1996). Production of Pb by the smelter increased ment, the absence of significant correlations between chemical ele- from 60,000 tonnes yr−1 in 1915 to 220,000 tonnes yr−1 in 1954, and ments and fine soil particles b 63 μm together with the general lack of thereafter remained relatively constant until 2011 (average of relationship between Corg content and most of the metals analyzed (es- 200,000 tonnes yr−1) (Fig. 6; Department of Mines, 1966; Department pecially Pb, Cd and Zn) suggest that the shifts in chemical element con- of Mines and Energy South Australia, 1988; Nyrstar, 2014, 2012, 2010, centrations over the last 100 to 130 years were not related to changes in 2008, 2007, Pasminco Ltd, 1990, 2003, 1999, 1994). soil composition (e.g. catchment erosion) but, rather, were caused by Our findings provide further evidence that, since opening, the increased metal loading to the system. In fact, it would have required smelter has been a point source of metal contamination in the area between 500 and 1000 years of the pre-smelter background conditions (Cartwright et al., 1976; Corbin and Wade, 2004; Edwards et al., 2001; to accumulate the same stocks of Pb, Cd and Zn that were accumulated Ross et al., 2003; Taylor et al., 2014). Concentrations of Pb, Cd, Zn, Cu during the last 120 years. The increase in metal fluxes after the and Mn in surface marine sediments were reported to increase with onset of the smelting operations which we observed in our seagrass proximity to the smelter (Ward and Young, 1981), with the highest con- cores is in agreement with a study of historical metal fluxes in centrations measured near First Creek (Fig. 1). Our application of Re- supratidal soils at Port Pirie, which reported a dramatic increase gime Shift Detection Limit Software to the Posidonia soil archives in Pb, Cd and Zn concentrations, reaching levels of concern (Lent shows that metal concentrations began to increase between 100 and et al., 1992). While the two seagrass cores showed differences in 130 yr BP, rapidly departing from the pre-smelter background concen- metal accumulation rates (Supplementary Table S5), likely due to trations. Increases in metal concentrations of seagrass soils and inter- spatial variability in depositional dynamics within the seagrass tidal sediments have been reported for this same time period in other meadow, both showed the same qualitative trends and the coastal areas around Australia, such as Oyster Harbour (Western quantitative differences fall within the uncertainty associated Australia) (Serrano et al., 2016a) and Moreton Bay (Queensland) with sedimentation rates. 1386 A. Lafratta et al. / Science of the Total Environment 649 (2019) 1381–1392 Fig. 4. Element concentrations (ppm) in the bulk soil of cores A15 and A12 in Port Pirie during the last ~3000 years and records of factor scores (PC1 and PC2) versus age (yr BP/calendar yr) for the last 200 years and versus soil depth (cm) afterwards. Calibrated years before present (cal yr BP) are indicated for A15 in plain text and for A12 in bold text at the corresponding depth. The red horizontal lines indicate significant shifts in temporal trends over the last 200 years (Regime Shift Detection Software). The last 120 years (yr BP ~1890s to 2014) are representing the post smelter period, indicated by the grey area. The y-axes in the upper panel show the years before present (yr BP) and the calendar years (calendar yr), while the lower panel show depths (cm) and calibrated years before present [cal yr BP]. A. Lafratta et al. / Science of the Total Environment 649 (2019) 1381–1392 1387 Fig. 4 (continued). 1388 A. Lafratta et al. / Science of the Total Environment 649 (2019) 1381–1392 Table 1 Pearson correlation coefficients between elements and organic carbon (%Corg) and fine sediment (b63 μm) determined in entire cores A15 and A12. Pb Cd Zn Cr Co Ni Zr Cu Mn As Fe Ti S Si %Corg (A15) 0.299 0.272 0.261 0.552 0.521 0.440 0.171 0.103 0.110 0.238 0.424 0.344 0.538 0.073 NS NS NS ⁎⁎ ⁎⁎ ⁎ NS NS NS NS ⁎ NS ⁎⁎ NS %Corg (A12) −0.112 −0.274 −0.261 0.326 0.701 0.746 0.066 0.125 −0.354 0.585 0.424 0.275 0.726 −0.325 NS NS NS NS ⁎⁎ ⁎⁎ NS NS NS ⁎ NS NS ⁎⁎ NS b63 μm (A15) 0.387 0.314 0.334 0.218 0.2 −0.002 0.149 0.270 0.383 −0.052 0.191 0.341 −0.101 0.111 NS NS NS NS NS NS NS NS NS NS NS NS NS NS b63 μm (A12) 0.324 0.255 0.251 0.361 0.308 0.113 0.433 0.352 −0.17 −0.026 0.095 0.358 0.38 −0.423 NS NS NS NS NS NS NS NS NS NS NS NS NS NS Significant correlations are reported in bold. ⁎⁎ Correlation is significant at the 0.01 level (2-tailed). ⁎ Correlation is significant at the 0.05 level (2-tailed). The temporal trends in PC1 and PC2 also support the hypothesis of the Port Pirie River (Fig. 1), which flows into the Spencer Gulf the smelter as the main source of contamination in Port Pirie (Fig. 4), (Edwards et al., 2001). Subsequently, most liquid effluents from the with major contamination events occurring 100–130 yr BP and smelter were discharged directly to First Creek (Fig. 1) (Ward and 40–60 yr BP. PC1 explained 44% of the variance in metal concentrations, Young, 1981). Since 1972, the improvements to the drainage system in- and was primarily characterized by elements (Pb, Cd, Zn, Mn, Cu, Cr) creased the opportunity for metal deposition in the effluent channel typically linked to anthropogenic activities, such as mining, smelting rather than in the sea thanks to a system of weir and settling ponds and metals refining (Nriagu, 1990), and Ti, which indicates terrestrial (Ferguson, 1983; Sims, 1973). The introduction of a process effluent inputs possibly related to changes in land-use and associated run-off treatment system (PETS) in 2003 led to a significant reduction in Pb (Koinig et al., 2003). This hypothesis is supported by the higher metal and Zn emissions to the liquid effluent stream discharged to First accumulation rates and the slight increase in fine particles during the creek (Environmental Protection Authority, 2003). Air emissions were post-smelter period, as has been reported for other locations following further dispersed with the construction of a 205 m tall stack in 1979 land-use change (Macreadie et al., 2012; Serrano et al., 2016b) (Supple- and a new baghouse between 1982 and 1992 (Maynard et al., 1993). mentary Table S5; Fig. 3, Supplementary Table S3). PC2, on the other However, none of these pollution reduction measures were reflected hand, was associated with elements that had a positive correlation in the Pb concentration profiles of the seagrass soils studied, which with Corg (Ni, Fe, S, As, Co and Cr). The increase in PC2 values commenc- remained constant between the late 1970s and the early 2000s. In con- ing 100–130 yr BP, together with the increase in organic carbon, likely trast, other studies have reported a significant reduction in metal con- reflects the onset of land clearing and agricultural run-off from follow- centrations of aerial dust deposition since the 1970s (Maynard et al., ing European Settlement rather than increasing industrial metal inputs. 1993) and in the marine surface waters following the installation of The temporal trends in Pb concentrations in the seagrass cores tend the PETS (Corbin and Wade, 2004; Gaylard, 2014). The lack of decrease to follow the trends in Pb production by the smelter since 1915 (Fig. 6) in metal concentrations in seagrass archives following mitigation mea- but, surprisingly, show no changes that might reflect the environmental sures could have several, not necessarily mutually exclusive, explana- management history of the smelter. Several measures have been imple- tions. It is possible that the measures taken to reduce emissions may mented since the 1940s in an effort to reduce or disperse metal emis- not have been effective or other processes, such as winds or currents, sions from the smelter. Before 1939, liquid effluents were directed to are driving the deposition of metals at specific spatial or temporal scales Fig. 5. 208Pb/207Pb vs 206Pb/207Pb bi-plot. Distribution of Pb isotope signatures along core A15 and A12 and comparison with possible sources of contamination in Port Pirie. Sources of data for isotope values of potential sources are provided in Supplementary Table S6. A. Lafratta et al. / Science of the Total Environment 649 (2019) 1381–1392 1389 Fig. 6. Pb annual production from 1915 to 2014 at the Port Pirie smelter; Pb concentrations in cores A15 and A12 during the post-smelter period. The red dotted line and the red vertical lines are indicating shifts in the Pb production and in the seagrass cores respectively, according to the Regime Shifts Detector Software (Department of Mines, 1966; Department of Mines and Energy South Australia, 1988; Nyrstar, 2014, 2012, 2010, 2008, 2007; Pasminco Ltd, 1990, 2003, 1999, 1994). (Lent et al., 1992; Maynard et al., 2005). Since our cores were collected concentrations and may prevent us from seeing the effects of the im- about 1 km away from First Creek (Fig. 1), which receives waste waters provements of the last few decades. from the smelter, concentration of metals in the seagrass soils could be The isotopic composition of Pb in core A15 and A12 provides evi- reflecting the contamination of this specific area, rather than the re- dence that the smelter is the source of the Pb contamination to seagrass gional pattern in contamination. Also, dredging and shipping operations meadows around Port Pirie since the onset of smelter operations. The occurring since European settlement, especially in the Port Pirie River, major sources of Pb pollution in Australia are leaded petrol, smelters may have resuspended metals trapped in the sediment into the water and mines, pesticides, plumbing and painting products (Gulson et al., column (Bell, 1983; Kalnejais et al., 2010), contributing metal loads to 1994, 1981; Kristensen et al., 2017). In Port Pirie, additional possible the water column not associated with contemporary smelter emissions. sources of metals to the coastal waters include: stormwater run-off; at- Furthermore, the resolution of our cores is about 6 years per sample on mospheric deposition from steelworks and mills at Whyalla and the average, which is not sufficient to capture annual changes in metal Power Stations at Port Augusta; vehicle exhausts (Corbin and Wade, 2004; Gaylard, 2014); and the smelter. The smelter has used mainly Pb ores from Broken Hill in New South Wales (NSW) and Cannington in Queensland (QLD) and lesser amounts of ores from Lawn Hill (QLD), Rosebery Tasmania (TAS), Cobar (NSW) and, more recently, Angas (South Australia) (Jackson and Abbot, 2008; Ohmsen, 2001). The shift in the Pb isotope ratios of the seagrass soils following the European settlement (1800s–1870s) and commissioning of the smelter (1890s) towards the values of the main ores used in the smelter (Fig. 5) is strong evidence that the smelter is the main source of the increased Pb concentrations in the marine environment during this period. The disparity between the seagrass soil signatures and those of the Rose- bery, Cobar and Angas ores reflects their small contribution to the smelter ore supply (Kristensen et al., 2016; Ohmsen, 2001). Some of the seagrass soil samples have stable Pb isotope signature similar to aerosol collected in the 1990s from Adelaide (Fig. 5) (Bollhöfer and Rosman, 2000), the largest city in South Australia and located about 200 km south of Port Pirie (Fig. 1). However, we discount that aerosols are the primarily source for the lead pollution in the soils (Gulson et al., 1981; Kristensen et al., 2016). The similarity in values is likely because of the composition of the Australian leaded petrol. Australian leaded petrol, which was an important source of Pb emissions in Australia from the 1930s until the early 2000s (Kristensen, 2015), used a mixture of different Australian (mainly Broken Hill and Mount Isa) and American ores (Gulson et al., 1983), and Broken Hill has been continu- ously used by the Port Pirie smelter. However, the clear change in the stable Pb isotopic signature of the seagrass soils that commenced in the post-smelter period cannot be related to the use of leaded petrol, be- cause those shifts pre-dates the use of petrol, which commenced in the Fig. 7. Pb concentrations (ppm) over time (last 200 years) and along depth afterwards in 1930s. the seagrass soil of Port Pirie (A15 and A12), Moreton bay (QLD; Morelli et al., 2012) and Despite the effort to reduce the metal emission in the environment Oyster Harbor (WA; Serrano et al., 2016a). of Port Pirie there is still concern about the concentrations of metals in 1390 A. Lafratta et al. / Science of the Total Environment 649 (2019) 1381–1392 seawater and shellfish in some areas, such as Port Pirie Harbour and multi-coring approach are warranted to provide a more accurate esti- Germein Bay (Gaylard, 2014). The concentrations of Pb, Zn and Cd in mate of the metals stock in the seagrasses and possibly to determine seagrass soils during the post-smelter period fall between the two In- the fate of metals following disturbance of seagrass soils. terim Sediment Quality Guidelines provided by ANZECC/ARMCANZ (2000): ISQG-Low (Pb: 50 mg/kg dry wt; Cd: 1.5 mg/kg dry wt; Zn: 5. Conclusion 200 mg/kg dry wt); and ISQG-High values (Pb: 220 mg/kg dry wt; Cd: 10 mg/kg dry wt; Zn: 410 mg/kg dry wt) (Fig. 4, Supplementary Seagrasses can provide a rare archive in shallow coastal waters that Table S3). According to those Guidelines, for metal concentrations fall- are often disturbed and are, therefore, unfavourable for the coherent ac- ing between the two ISQGs the effect on biota is poorly known and cumulation of materials over time. The seagrass archives at Port Pirie re- highly variable. This indicates the need for further investigations into veal a series of anthropogenic contamination events, commencing with factors such as the bioavailability and toxicity of the metals to help un- European settlement of the region and followed by commissioning of derstand the possible effects of the contamination. the Port Pirie smelter. The multi-parameter approach used to analyze As has been shown elsewhere (Serrano et al., 2011), our findings in- these archives indicates that the major source of Pb pollution since dicate that the soils of seagrass meadows around Port Pirie may act as a 1890s in seagrass soils is the smelter. Furthermore, our preliminary as- significant sink for metal pollution, preventing the dispersion of metals sessment of the surficial soils suggests that the seagrass meadows may that would otherwise enter the food chain and, potentially, cause harm. act as significant sinks for metal pollution. If that is the case, they may The only available estimates of metal emissions from the smelter are for prevent those metals from entering the marine food web, emphasising the period 1999–2014, and indicate the smelter emitted a total of the need to preserve the seagrass meadows to avoid the potential for re- 900 tonnes of Pb, 1300 tonnes of Zn and 40 tonnes of Cd (Department lease of those metals into coastal food webs. More studies are required of the Environment and Energy, 2016) during those 15 years. We can to assess the ecological service of seagrasses as pollutant sinks and to provide a preliminary estimate of the proportion of those smelter emis- determine the fate of seagrass metal pools following disturbance. sions accumulated in the seagrass meadows surrounding Port Pirie by comparing the inventory of those elements accumulated over the last Data availability 15 years in the seagrass soil (the mean of A15 and A12 metal stocks in the top 1–2 cm of soil) with the reported smelter emissions over the All the data used in this publication are available from the authors same period. We scaled Pb, Cd and Zn concentration in the seagrass upon request. cores to match the gradient detected by Ward and Young (1981) and Tiller et al. (1989) in the area where they reported the highest metals Acknowledgements concentration in the sediment. Within this area we applied those con- centrations over 70 km2 of seagrass (Miller et al., 2009) to estimate This research was funded through by Edith Cowan University (ECU), the total stocks of each metal in the area. To correct for any accumula- the Environmental Protection Authority of South Australia and South tion that might have occurred independent of the smelter operations, Australia Water Corporation, through an ECU-Industry grant we removed the metals stocks which have been observed to accumulate (G1001729). For part of the project OS was supported by an ARC over the past 15 years in a meadow that has experienced the effects of DECRA (DE170101524), AL by Holsworth Wildlife Research Endow- development following European settlement but which is removed ment – Equity Trustees Charitable Foundation and PM by the from metal smelting activities. The Oyster Harbour Posidonia meadow Australian Research Council (LE170100219). The authors gratefully ac- (Fig. 7) in SW Australia is located N2000 km from any major lead pro- knowledge the assistance of Dr S Murray-Jones in initiating this re- cessing facility but has Pb, Cd and Zn accumulation rates of 5, 6 and 10 search. We thank Dr Warwick Noble, Mr Matt Nelson and Mr Tim × 10−4 g m−2 over the same 15 year period. We took this accumulation Kildea for their assistance with field sampling, Ms Ariane Artias-Ortiz rate as a baseline and removed it from the stocks accumulating at Port for assistance with the metal and lead isotope analyses, Prof Antonio Pirie. Martinez-Cortizas for assistance in the interpretation of the PC analyses We estimated that over the last 15 years, 7% (≈60 tonnes) of Pb, 11% and Prof. Mark Taylor for providing helpful insight about Port Pirie con- (≈5 tonnes) of Cd and 15% (≈200 tonnes) of Zn emissions from the tamination history. smelter may have been accumulated in the soils underneath 70 km2 of seagrass meadows surrounding the smelter. While this suggests a po- Authors contribution tential ecosystem service provided by seagrass meadows in the region, it also raises the possibility of enhanced metal loading to the coastal wa- P.L., O.S., S.G. and M.F. conceived the project and the initial study de- ters should the meadows be disturbed. The loss of seagrass meadows sign. P.L., O.S., S.G. and M.M. collected the samples. A.L., P.L., O.S. and P.M. can result in erosion of Corg stocks and fine-grained soils (Marbà et al., interpreted the data. A.L. performed data analyses and wrote the manu- 2015), which are typically enriched in metals relative to other compo- script. All authors contributed to the final manuscript and subsequent nents of the bulk sediment (Zhang et al., 2014). Sediment disturbance revision. may also affect metal partitioning between pore water and sediment and, hence, metal bioavailability (Simpson and Batley, 2007). This ar- Competing financial interests gues strongly for the preservation of seagrass meadows in the region. While the seagrasses at Port Pirie appear to tolerate the current levels The authors declare no competing financial interests. of contamination (Gaylard, 2014) it is not clear whether ongoing con- tamination may increase the vulnerability of seagrasses to other Appendix A. Supplementary data stressors (Prange and Dennison, 2000), such as global climate change. Posidonia species have shown sensitivity to marine heatwave events Supplementary data to this article can be found online at https://doi. (Thomson et al., 2015) and extreme heat conditions in 1993 led to the org/10.1016/j.scitotenv.2018.08.400. loss of 130 km2 of seagrasses along the north eastern coast of Spencer Gulf (Seddon et al., 2000). It is possible that metal contamination References could exacerbate the effects of predicted future sea temperature in- creases, placing the pollutant stocks within the Port Pirie seagrasses at ANZECC/ARMCANZ, 2000. Australian and New Zealand guidelines for fresh and marine water quality. Aust. New Zeal. Environ. Conserv. Counc. Agric. Resour. Manag. risk of dispersion in the marine environment. Given the potentially sig- Counc. Aust. New Zeal. 1, 1–103. nificant role of seagrasses as metal sinks, further studies based on a Bell, A., 1983. How marine life copes with a lead smelter. Ecosystems 36, 17–22. A. Lafratta et al. / Science of the Total Environment 649 (2019) 1381–1392 1391 BHP, 1985. Australians in company (The Broken Hill Proprietary Company Limited in its Krishnaswamy, S., Lal, D., Martin, J.M., Meybeck, M., 1971. Geochronology of lake sedi- one hundredth year). URL. https://www.bhp.com/-/media/bhp/documents/aboutus/ ments. Earth Planet. Sci. Lett. 11, 407–414. https://doi.org/10.1016/0012-821X(71) ourcompany/our-history/150813_ourhistory_ 90202-0. australiansincompanybhpsonehundredthyear.pdf, Accessed date: 9 October 2016 Kristensen, L.J., 2015. Quantification of atmospheric lead emissions from 70 years of (WWW Document). leaded petrol consumption in Australia. Atmos. Environ. 111, 195–201. https://doi. Bollhöfer, A., Rosman, K.J.R., 2000. Isotopic source signatures for atmospheric lead: the org/10.1016/j.atmosenv.2015.04.012. Southern Hemisphere. Geochim. Cosmochim. Acta 64, 3251–3262. https://doi.org/ Kristensen, L.J., Taylor, M.P., Evans, A.J., 2016. Tracing changes in atmospheric sources of 10.1016/S0016-7037(00)00630-X. lead contamination using lead isotopic compositions in Australian red wine. Bowman, G.M., 1985. Oceanic reservoir correction for marine radiocarbon dates from Chemosphere 154, 40–47. https://doi.org/10.1016/j.chemosphere.2016.03.023. Northwestern Australia. Aust. Archaeol. 20, 58–67. Kristensen, L.J., Taylor, M.P., Flegal, A.R., 2017. An odyssey of environmental pollution: the Burdige, D.J., 2007. Preservation of organic matter in marine sediments: controls, mecha- rise, fall and remobilisation of industrial lead in Australia. Appl. Geochem. 83, 3–13. nisms, and an imbalance in sediment organic carbon budgets? Chem. Rev. 107, https://doi.org/10.1016/j.apgeochem.2017.02.007. 467–485. https://doi.org/10.1021/cr050347q. Lent, R.M., Herczeg, A.L., Welch, S., Lyons, W.B., 1992. The history of metal pollution near a Cartwright, B., Merry, R.H., Tiller, K.G., 1976. Heavy metal contamination of soils around a lead smelter in Spencer Gulf, South-Australia. Toxicol. Environ. Chem. 36, 139–153. lead smelter at Port Pirie, South Australia. Aust. J. Soil Res. 15, 69–81. https://doi.org/ https://doi.org/10.1080/02772249209357837. 10.1071/SR9770069. Macreadie, P.I., Allen, K., Kelaher, B.P., Ralph, P.J., Skilbeck, C.G., 2012. Paleoreconstruction Corbin, T., Wade, S., 2004. Heavy Metal Concentrations in Razorfish (Pinna bicolor) and of estuarine sediments reveal human-induced weakening of coastal carbon sinks. Sediments Across the Northern Spencer Gulf. Environment Protection Authority, Ad- Glob. Chang. Biol. 18, 891–901. https://doi.org/10.1111/j.1365-2486.2011.02582.x. elaide, South Australia. Marbà, N., Arias-Ortiz, A., Masqué, P., Kendrick, G.A., Mazarrasa, I., Bastyan, G.R., Garcia- Department of Mines, 1966. Mining review no. 125. URL. https://sarigbasis.pir.sa.gov.au/ Orellana, J., Duarte, C.M., 2015. Impact of seagrass loss and subsequent revegetation WebtopEw/ws/samref/sarig1/wcir/Record?r=0&m=1&w=catno=2022453, on carbon sequestration and stocks. J. Ecol. 103, 296–302. https://doi.org/10.1111/ Accessed date: 15 August 2017 (WWW Document). 1365-2745.12370. Department of Mines and Energy South Australia, 1988. Mineral resources review no. Martinez-Cortizas, A., López-Merino, L., Bindler, R., Mighall, T., Kylander, M., 2013. Atmo- 156. URL. https://sarigbasis.pir.sa.gov.au/WebtopEw/ws/samref/sarig1/image/DDD/ spheric Pb pollution in N Iberia during the late Iron Age/Roman times reconstructed MRR156.pdf, Accessed date: 15 August 2017 (WWW Document). using the high-resolution record of La Molina mire (Asturias, Spain). J. Paleolimnol. Department of the Environment and Energy, 2016. National Pollutant Inventory. URL. 50, 71–86. https://doi.org/10.1007/s10933-013-9705-y. http://www.npi.gov.au/npidata/action/load/emission-by-individual-facility-result/ Masque, P., Sanchez-Cabeza, J.A., Bruach, J.M., Palacios, E., Canals, M., 2002. Balance and criteria/state/SA/year/2011/jurisdiction-facility/SA0018, Accessed date: 23 Novem- residence times of Pb-210 and Po-210 in surface waters of the northwestern Mediter- ber 2016 (WWW Document). ranean Sea. Cont. Shelf Res. 22, 2127–2146. https://doi.org/10.1016/S0278-4343(02) Edwards, J.W., Edyvane, K.S., Boxall, V.a., Hamann, M., Soole, K.L., 2001. Metal levels in 00074-2. seston and marine fish flesh near industrial and metropolitan centres in South Mateo, M.a., Renom, P., Michener, R.H., 2010. Long-term stability in the production of a Australia. Mar. Pollut. Bull. 42, 389–396. https://doi.org/10.1016/S0025-326X(00) NW Mediterranean Posidonia oceanica (L.) Delile meadow. Palaeogeogr. 00168-5. Palaeoclimatol. Palaeoecol. 291, 286–296. https://doi.org/10.1016/j. Environmental Protection Authority, 2003. Annual report 2002–2003. URL. https://www. palaeo.2010.03.001. epa.sa.gov.au/files/8273_annualreport0203.pdf, Accessed date: 12 January 2016 Maynard, E.J., Calder, I.C., Phipps, C.V., 1993. The Port Pirie Lead Implementation Program: (WWW Document). Review of Progress and Consideration of Future Directions. Public and Environmental Ettler, V., 2015. Soil contamination near non-ferrous metal smelters: a review. Appl. Health Servive, South Australia Health Commission, Adelaide, South Australia. Geochem. 64, 56–74. https://doi.org/10.1016/j.apgeochem.2015.09.020. Maynard, E.J., Frangs, L.J., Malcolm, M.S., 2005. The Port Pirie Lead Implementation Pro- Ferguson, J., 1983. Concentrations and speciation of lead, zinc and cadmium in seawater- gram: Future Focus and Directions. Department of Health, Adelaide. like smelter effluent and adjacent marine environments, Port Pirie, South Australia. Miller, B.D., Westphalen, G., Jolley, A.M., Eglinton, Y., 2009. Final Report to the Northern Mar. Freshw. Res. 34, 375–385. https://doi.org/10.1071/MF9830375. and Yorke Natural Resources Management Board for the Project: Sustaining Marine Gacia, E., Duarte, C.M., 2001. Sediment retention by a Mediterranean Posidonia oceanica Biological Health. Prepared by Department for Environment and Heritage, Coast meadow: the balance between deposition and resuspension. Estuar. Coast. Shelf Sci. and Marine Conservation Branch. Adelaide, South Australia. 52, 505–514. https://doi.org/10.1006/ecss.2000.0753. Morelli, G., Gasparon, M., Fierro, D., Hu, W.-P., Zawadzki, A., 2012. Historical trends in Gacia, E., Granata, T., Duarte, C., 1999. An approach to measurement of particle flux and trace metal and sediment accumulation in intertidal sediments of Moreton Bay, sediment retention within seagrass (Posidonia oceanica) meadows. Aquat. Bot. 65, southeast Queensland, Australia. Chem. Geol. 300–301, 152–164. https://doi.org/ 255–268. https://doi.org/10.1016/S0304-3770(99)00044-3. 10.1016/j.chemgeo.2012.01.023. Gaylard, S., 2014. Chapter 27: marine pollution in Spencer Gulf, South Australia. In: Nriagu, J.O., 1990. Global metal pollution: poisoning the biosphere? Environ. Sci. Policy Shepherd, S., Madigan, S., Gillanders, B., Murray-Jones, S., Wiltshire, D. (Eds.), A Nat- Sustain. Dev. 32, 7–33. https://doi.org/10.1080/00139157.1990.9929037. ural History of Spencer Gulf. Royal Society of South Australia, pp. 378–391. Nyrstar, 2007. Annual report 2007-think zinc. URL. http://www.nyrstar.com/investors/ Gaylard, S., Thomas, S., Nelson, M., 2011. An assessment of the current status of bioavail- en/Nyr_Documents/English/ar2007.pdf, Accessed date: 20 November 2016 (WWW able metal contamination across South Australia using translocated mussels Mytilus Document). galloprovincalis. Trans. R. Soc. South Aust. 135, 39–54. https://doi.org/10.1080/ Nyrstar, 2008. Annual Report 2008-navigating the future prudently pragmatically proac- 03721426.2011.10887149. tively. URL. http://www.nyrstar.com/investors/en/Nyr_Documents/English/Nyrstar_ Geoscience Australia, 2013. Zinc-lead-silver. URL. http://www.ga.gov.au/scientific-topics/ Annual_Report-FINAL_(EN).pdf, Accessed date: 20 November 2016 (WWW minerals/mineral-resources/zinc-lead-silver#heading-6, Accessed date: 17 Novem- Document). ber 2016 (WWW Document). Nyrstar, 2010. Annual report 2010-a year of delivery. URL. http://www.nyrstar.com/in- Glew, J., Smol, J., Last, W., 2001. Sediment core collection and extrusion. Tracking Environ- vestors/en/shareholderinformation/Documents/Annual General Meeting - 27 April mental Change Using Lake Sediments. Volume 1: Basin Analysis, Coring, and Chrono- 2011/Annual Report 2010.pdf, Accessed date: 20 November 2016 (WWW logical Techniques. Kluwer Academic Publishers, Dordrecht, The Netherlands, Document). pp. 73–105. Nyrstar, 2012. Strategy into action 2012 annual report. URL. http://www.nyrstar.com/in- Gulson, B.L., Tiller, K.G., Mizon, K.J., Merry, R.H., 1981. Use of lead isotopes in soils to iden- vestors/en/Nyr_Documents/English/Nyrstar_Annual Report 2012_EN.pdf, Accessed tify the source of lead contamination near Adelaide, South Australia. Environ. Sci. date: 20 November 2016 (WWW Document). Technol. 15, 691–696. https://doi.org/10.1021/es00088a008. Nyrstar, 2014. Annual report 2014. URL. https://investor.twitterinc.com/annuals-proxies. Gulson, B.L., Mizon, K.J., Korsch, M.J., 1983. Fingerprinting the source of lead in Sydney air cfm, Accessed date: 20 November 2016 (WWW Document). using lead isotopes. In: Carras, J.N., Johnson, G.M. (Eds.), The Urban Atmosphere— Ohmsen, G.S., 2001. Characterization of fugitive material within a primary lead smelter. J. Air Sydney, a Case Study. John Wiley & Sons, Ltd, pp. 223–244 https://doi.org/10.1002/ Waste Manage. Assoc. 51, 1443–1451. https://doi.org/10.1080/10473289.2001.10464371. qj.49711046517 (CSIRO, Melbourne). Pasminco Ltd, 1990. Annual Report 1990. Gulson, B.L., Law, A.J., Korsch, M.J., Mizon, K.J., 1994. Effect of plumbing systems on lead Pasminco Ltd, 1994. Annual Report 1994. content of drinking water and contribution to lead body burden. Sci. Total Environ. Pasminco Ltd, 1999. Pasminco Annual Report. URL. http://s3.amazonaws.com/zanran_ 144, 279–284. https://doi.org/10.1016/0048-9697(94)90447-2. storage/www.pasminco.com.au/ContentPages/888846846.pdf, Accessed date: 10 No- Harbison, P., 1986. Mangrove muds - a sink and a source for trace metals. Mar. Pollut. Bull. vember 2016 (WWW Document). 17, 246–250. https://doi.org/10.1016/0025-326X(86)90057-3. Pasminco Ltd, 2003. Annual Report 2003. Jackson, C., Abbot, P., 2008. Keeping it South Australian — Angas base metals mine sends Prange, J.A., Dennison, W.C., 2000. Physiological responses of five seagrass species to trace concentrates to Port Pirie Smelter for high-end processing. MESA J. 50, 11–13. metals. Mar. Pollut. Bull. 41, 327–336. https://doi.org/10.1016/S0025-326X(00) Järup, L., 2003. Hazards of heavy metal contamination. Br. Med. Bull. 68, 167–182. https:// 00126-0. doi.org/10.1093/bmb/ldg032. Rodionov, S.N., 2004. A sequential algorithm for testing climate regime shifts. Geophys. Kalnejais, L.H., Martin, W.R., Bothner, M.H., 2010. The release of dissolved nutrients and Res. Lett. 31, 2–5. https://doi.org/10.1029/2004GL019448. metals from coastal sediments due to resuspension. Mar. Chem. 121, 224–235. Ross, K.E., Bidwell, J.R., Williams, M., Boland, J., 2003. Trace metals associated with seston https://doi.org/10.1016/j.marchem.2010.05.002. collected from Spencer Gulf, South Australia, near a lead smelter: spatial distribution, Kämpf, J., Payne, N., Malthouse, P., Beach, W.P., Ka, J., 2010. Marine connectivity in a large temporal variability and uptake by bivalves. Trans. R. Soc. S. Aust. 127, 33–45. inverse estuary. 26, 1047–1056. https://doi.org/10.2112/JCOASTRES-D-10-00043.1. SACOME, 2016. Mining & exploration in South Australia. URL. https://www.sacome.org. Koinig, K., Shotyk, W., Lotter, A., Ohlendorf, C., 2003. 9000 years of geochemical evolution au/policy-and-advocacy.html, Accessed date: 2 December 2017 (WWW Document). of lithogenic major and trace elements in the sediment of an alpine lake – the role of Sanchez-Cabeza, J.A., Masque, P., Ani-Ragolta, I., 1998. 210Pb and 210Pb analysis in sedi- climate, vegetation, and land-use history. J. Paleolimnol. 30, 307–320. https://doi.org/ ments and soils by microwave acid digestion. 227, 19–22. https://doi.org/10.1007/ 10.1023/a:1026080712312. BF02386425. 1392 A. Lafratta et al. / Science of the Total Environment 649 (2019) 1381–1392 Seddon, S., Connolly, R.M., Edyvane, K.S., 2000. Large-scale seagrass dieback in northern smelting operations. Aeolian Res. 14, 35–52. https://doi.org/10.1016/j. Spencer Gulf, South Australia. Aquat. Bot. 66, 297–310. https://doi.org/10.1016/ aeolia.2014.03.003. S0304-3770(99)00080-7. The Register, 1921. Port Pirie early history. http://trove.nla.gov.au/newspaper/article/ Serrano, O., Mateo, M.A., Dueñas-Bohórquez, A., Renom, P., López-Sáez, J.A., Martínez 63420592. Cortizas, A., 2011. The Posidonia oceanica marine sedimentary record: a Holocene ar- Thomson, J.A., Burkholder, D.A., Heithaus, M.R., Fourqurean, J.W., Fraser, M.W., Statton, J., chive of heavy metal pollution. Sci. Total Environ. 409, 4831–4840. https://doi.org/ Kendrick, G.A., 2015. Extreme temperatures, foundation species, and abrupt ecosys- 10.1016/j.scitotenv.2011.08.001. tem change: an example from an iconic seagrass ecosystem. Glob. Chang. Biol. 21, Serrano, O., Martínez-Cortizas, A., Mateo, M.A., Biester, H., Bindler, R., 2013. Millennial 1463–1474. https://doi.org/10.1111/gcb.12694. scale impact on the marine biogeochemical cycle of mercury from early mining on Tiller, K.G., Merry, R.H., Zarcinas, B.A., Ward, T.J., 1989. Regional geochemistry of metal- the Iberian Peninsula. Glob. Biogeochem. Cycles 27, 21–30. https://doi.org/10.1029/ contaminated surficial sediments and seagrasses in upper Spencer Gulf, South 2012GB004296. Australia. Estuar. Coast. Shelf Sci. 28, 473–493. https://doi.org/10.1016/0272-7714 Serrano, O., Davis, G., Lavery, P.S., Duarte, C.M., Martinez-Cortizas, A., Mateo, M.A., (89)90024-3. Masqué, P., Arias-Ortiz, A., Rozaimi, M., Kendrick, G.A., 2016a. Reconstruction of Tkalich, E., Sundarambat, R., 2008. Coastal and marine environment. URL. https://www. centennial-scale fluxes of chemical elements in the Australian coastal environment epa.sa.gov.au/soe_2013/downloads/SOESA-WEB-06Coastal-03(29Aug13).pdf, using seagrass archives. Sci. Total Environ. 541, 883–894. https://doi.org/10.1016/j. Accessed date: 10 February 2017 (WWW Document). scitotenv.2015.09.017. Tonkin Consulting, 2009. Port Pirie Regional Council: Port Pirie Stormwater Management Serrano, O., Lavery, P., Masque, P., Inostroza, K., Bongiovanni, J., Duarte, C., 2016b. Seagrass Plan. URL. https://www.pirie.sa.gov.au/webdata/resources/files/1.6 Port Pirie sediments reveal the long-term deterioration of an estuarine ecosystem. Glob. Chang. Stormwater Management Plan 20_05_09.pdf, Accessed date: 12 May 2016 (WWW Biol. 22, 1523–1531. https://doi.org/10.1111/gcb.13195. Document). Simpson, S.L., Batley, G.E., 2007. Predicting metal toxicity in sediments: a critique of cur- USEPA, 1996. Method 3052: microwave assisted acid digestion of siliceous and organi- rent approaches. Integr. Environ. Assess. Manag. 3, 18–31. https://doi.org/10.1002/ cally based matrices. Test Methods for Evaluating Solid Waste. ieam.5630030103. Ward, T.J., Hutchings, P.A., 1996. Effects of trace metals on infaunal species composition in Sims, M.A., 1973. The Impact of Three Heavy Metals on the Marine Flora of an Inlet at Port polluted intertidal and subtidal marine sediments near a lead smelter, Spencer Gulf, Pirie. University of Adelaide, Adelaide, Australia (Honours Thesis). South Australia. Mar. Ecol. Prog. Ser. 135, 123–135. https://doi.org/10.3354/ South Australia Health, 2016. Port Pirie blood lead levels. Analysis of blood lead levels for meps135123. the nine months of 2016 (1 January–30 September 2016). URL. http://www.sahealth. Ward, T.J., Young, P.C., 1981. Trace metal contamination of shallow marine sediments sa.gov.au/wps/wcm/connect/109450804e3754db9062d8c09343dd7f/Port%2BPirie% near a lead smelter, Spencer Gulf, South Australia. Mar. Freshw. Res. 32, 45–56. 2BBlood%2BLead%2BLevels%2BAnalysis%2Bfor%2Bthe%2Bthird%2Bquarter%2B2016% https://doi.org/10.1071/MF9810045. 2B%25281%2BJanuary%2B30%2BSeptember%2B2016%2529.pdf?MOD= Weis, J.S., Weis, P., 2004. Metal uptake, transport and release by wetland plants: implica- AJPERES&CACHE=NONE&CONTENTCACHE=NON, Accessed date: 10 December tions for phytoremediation and restoration. Environ. Int. 30, 685–700. https://doi.org/ 2016 (WWW Document). 10.1016/j.envint.2003.11.002. Stuiver, M., Reimer, P.J., Bard, E., Beck, J.W., Burr, G.S., Hughen, K.A., Kromer, B., McCormac, Zhang, X., Yang, L., Li, Y., Li, H., Wang, W., Ye, B., 2012. Impacts of lead/zinc mining and G., Van Der Plicht, J., Spurk, M., 1998. INTCAL98 radiocarbon age calibration, smelting on the environment and human health in China. Environ. Monit. Assess. 24000–0 cal BP. Radiocarbon 40, 1041–1083. https://doi.org/10.2458/azu_js_ 184, 2261–2273. https://doi.org/10.1007/s10661-011-2115-6. rc.40.3781. Zhang, C., Yu, Z., Zeng, G., Jiang, M., Yang, Z., Cui, F., Zhu, M., Shen, L., Hu, L., 2014. Effects of Taylor, M.P., Davies, P.J., Kristensen, L.J., Csavina, J.L., 2014. Licenced to pollute but not to sediment geochemical properties on heavy metal bioavailability. Environ. Int. 73, poison: the ineffectiveness of regulatory authorities at protecting public health 270–281. https://doi.org/10.1016/j.envint.2014.08.010. from atmospheric arsenic, lead and other contaminants resulting from mining and