Relationships between Potentially Toxic Elements in intertidal sediments and their bioaccumulation by benthic invertebrates

Summary

The bioaccumulation of Potentially Toxic Elements (PTEs) by benthic invertebrates in estuarine sediments is poorly understood. We sampled and analysed PTEs in sediments and benthic invertebrates from five sites in the Skeena Estuary (British Columbia, Canada), including sites adjacent to an abandoned cannery and a decommissioned papermill. Our aim was to elucidate baseline levels of PTE concentrations at sites that may be recovering from disturbance associated with prior industrial development and identify organisms that could be used to biomonitor the impact of future industrial developments. There was no indication that sediments of the salmon cannery were polluted, but acidic sediments adjacent to the papermill contained elevated concentrations of Cd, Cr, Hg and Pb. Benthic invertebrate community assemblages confirm that sediments have mostly recovered from prior industrial development associated with discharge of papermill sludge. Overall, we did not observe any relationship between PTE concentrations in the sediment and PTE concentrations in invertebrate tissues. However, we did observe a negative relationship between sediment pH and the Biota-Sediment Accumulation Factor (BSAF) of most PTEs for Oregon pill bugs (Gnorimosphaeroma oregonensis). G. oregonensis, observed at all sites, feeds on the fibers associated with the papermill discharge. Thus, G. oregonensis is a useful biomonitors for quantifying the impact of the decommissioned papermill, and are candidate biomonitors for assessing the impact of similar industrial development projects on intertidal ecosystems.

Keywords:

Biology and life sciences – Organisms – Eukaryota – Physical sciences – Chemistry – Engineering and technology – Animals – Invertebrates – Mathematics – Earth sciences – Ecology and environmental sciences – Pollution – Ecology – Ecosystems – Marine and aquatic sciences – Bodies of water – Discrete mathematics – Combinatorics – Permutation – Physical chemistry – Manufacturing processes – Geology – Petrology – Sediment – Sedimentary geology – Estuaries – Molluscs – Bivalves – Chemical deposition – Surface treatments

Introduction

Intertidal estuarine habitat is both ecologically and economically important, providing nursery habitats to several species, such as juvenile pacific salmon, that support important commercial fisheries [1]. Estuaries are often very biologically diverse, exhibiting an abundant array of flora and fauna [2–6]. One abundant taxonomic group often found in estuaries are benthic invertebrates [5, 7]. Benthic invertebrates are often used as indicators of environmental pollution since they live in sediments and are directly exposed to any contaminants located within the sediment. Further, they are prey for many commercially important fish species, often responding to pollution before commercially viable species are negatively impacted, warning of the potential threat of pollution [8–10]. Benthic invertebrates inhabiting estuaries are inherently resistant to physical and chemical change as they have adapted to living in a dynamic environment with wide spatial and temporal ranges of chemical and physical properties, such as pH, redox potential, salinity, and particle size [11–13].

Potentially Toxic Elements (PTEs) are naturally occurring elements (e.g. Arsenic Cadmium, Cobalt, Chromium, Nickel, Mercury, Lead and Zinc) found in all ecosystems, but are typically associated with contamination and potentially toxic to plants, animals and humans when present in elevated concentrations [14]. Anthropogenic activities have impacted the biogeochemical cycling of Potentially Toxic Elements (PTEs) in even the most remote ecosystems on our planet [15, 16]. Because the chemical and physical properties of estuarine sediment influence the bioavailability of PTEs to benthic invertebrates and are temporally and spatially dynamic, it is difficult to predict the impact of industrial development on the fate and impact of PTEs [17]. As such, we do not currently understand the relationship between sediment properties, food web structure and PTE bioaccumulation by sediment-dwelling invertebrates well enough to predict the impact of pollution from industrial developments upon the estuarine environment. Quantifying the disruption of natural biogeochemical cycles of PTEs by anthropogenic activities is difficult if concentrations prior to industrial development are unknown. Therefore, it is important to conduct baseline surveys to assess the bioavailability and bioaccumulation of PTEs by organisms to predict the likely impact of further pollution on the ecosystem.

The Skeena is British Columbia’s second largest river and the estuary provides important nursery habitats for juvenile salmon [3, 18]. Coastal areas to the north of the estuary surrounding the small port cities of Prince Rupert and Port Edward have been extensively developed [19–22]. Industrial developments included an international port, a papermill, and several canneries. Findings of previous surveys of the benthic invertebrates inhabiting the intertidal sediment in the Skeena estuary reveal an infaunal community that is relatively undisturbed at the estuary scale, but which still shows the scars of historic industrial developments at specific locations [5, 11, 23, 24]. High densities of amphipods throughout the Skeena estuary [5, 11] suggests that current disturbances to intertidal areas are relatively limited. Similarly, 40 intertidal invertebrate species are observed in this area, including multiple species at all trophic levels within the food web [5, 11], and such a complex community is often associated with non-disturbed habitats [21, 25–27]. Conversely, disturbed sites are often more easily invaded by invasive species [28, 29], and Capitellidae (Capitella capitata species complex) polychaetes are often observed in disturbed habitats, particularly areas that have been organically enriched [21, 25, 30, 31]. Capitellidae polychaetas have a clumped distribution within the Skeena estuary and can be locally abundant at the scale of a 1m² plot [5, 11]. Similarly, abundant but localised populations of the invasive cumacean, Nippoleucon hinumensis, have also been observed in the Skeena estuary [5, 11], and on other historically impacted mudflats along BC’s north coast [31]. While the universal presence and high abundances of amphipods, coupled with the complexity of the intertidal community at the estuary scale strongly suggest that these intertidal communities are relatively undisturbed, the locally abundant populations of invasive species and Capitellidae polychaetes offer contradictory evidence. These indicators of disturbance suggest that either localised disturbances are currently occurring, or this biological signal is a remnant of past industrial developments. To the best of our knowledge, no current industrial developments are occurring in the Skeena estuary that can explain the localised indicators of disturbance observed in the biological community [5, 11, 24, 31]. Therefore, we hypothesise that localised disturbance in the Skeena estuary is a legacy of past industrial developments [5, 11]. It is likely that these disturbances are at least partially related to discharge from the papermill on Watson Island, which was released into the immediate near shore area (Porpoise Bay), strongly depressing the invertebrate communities in this area during the 1970s [20, 32, 33]. As such, we postulate that these intertidal areas have been passively recovering from this disturbance for ~50 years. However, it is not clear the extent to which the papermill discharge has been buried by fresh ‘clean’ sediment and is no longer bioavailable to benthic invertebrates.

Considerable developments have been proposed in the Skeena estuary, including oil and gas pipelines, super-tanker routes, potash loading facilities, and a liquid natural gas (LNG) terminal. The Skeena salmon run contributes an estimated $110 million dollars annually to the local economy [34], so pollution of the intertidal nursery habitat of the Skeena Estuary could have devastating consequences on both the economy and ecosystem [35, 36]. It is therefore critical to predict the impact that future developments will have on the bioaccumulation of PTEs by benthic invertebrates inhabiting intertidal sediments of the Skeena Estuary. Our aim was to investigate the extent to which intertidal sediments in the Skeena Estuary are contaminated by historic industrial developments and identify organisms that could be used to biomonitor the impact of future industrial developments. As such, the objectives of this study were to (i) explore the relationship between sediment depth, sediment properties and the fate of PTEs from historic industrial development in the Skeena Estuary, and (ii) quantify the impact of PTEs on the community assemblages of benthic invertebrates inhabiting intertidal sediments and the bioaccumulation of PTEs into their tissues. To achieve these objectives we conducted a survey of sediments and invertebrates at five intertidal sites north of the Skeena river mouth to examine baseline levels of PTE concentrations in sediments and benthic invertebrates at sites that may be recovering from previous disturbances and subject to future industrial development.

Methods

Field sampling

Sediment cores and benthic invertebrates were collected from five intertidal sites north of the Skeena river mouth (Fig 1) during July 2017: Cassiar Cannery (CC), Inverness Passage (IP), Papermill Bay (PB), Wolfe Cove (WC), and Tyee Banks (TB). Cassiar Cannery (N54° 10’ 40.4, W130° 10’ 40.4) is a former salmon cannery that closed in 1983 and is now an ecotourism lodge. Inverness passage (N54° 10’ 05.9, W130° 09’ 40.4), a mudflat ~3 km closer to the Skeena mouth than Cassiar Cannery, has intertidal habitat similar to Cassiar Cannery, but was never directly impacted by a cannery or other anthropogenic development. Inverness passage was thus selected as a reference site proximal to Cassiar Cannery. Papermill Bay (N54° 13’ 59.3, W130° 17’ 07.5) is a small bay located directly adjacent to a large decommissioned papermill. The papermill was closed in 2001, ceasing all operations and discharge [37–39]. Wolfe Cove (N54° 14’ 33.0, W130° 17’ 34.5) is a mudflat located approximately 1 km from the papermill, selected as a reference site proximal to Papermill Bay. Finally, Tyee Banks (N54° 11’ 59.1, W129° 57’ 36.7) is located 20km upstream from the mouth of the Skeena River on a large intertidal mudflat. At some point in the past, this area had a small-scale sawmill operating and accumulations of sawdust and woodchips are still present in the upper intertidal sediment. Tyee Banks was selected as a reference site in a remote location, much further away from Cassiar Cannery and Papermill Bay than the proximal reference sites. As this work was carried out on nearshore habitat in Canada, focusing upon common intertidal invertebrates, no permits or specific permissions were required to conduct this work.

Sediments were sampled by pushing polycarbonate cores (5 cm diameter, 20 cm length) into the mud with a rubber mallet, digging out with a spade, and wrapping with cling film, before extruding and dividing into strata; 0–5 cm, 5–10 cm, 10–15 cm and 15-20cm. Sediment samples from each strata were dried at 40°C for 48 hours and shipped to the UK for analysis. A total of 15 cores were collected from each site along five transects comprising three samples taken from the upper, mid and lower shore in a random stratified sampling design [40]. Sediments from the upper, mid and lower shore (but the same strata) were pooled and homogenized prior to analysis. At each location that a sediment core was collected, two more cores (7 cm diameter) were taken to a depth of at least 5 cm (well into the anoxic layer [41]), transferred to a plastic bag in the field and then passed through a 250μm stainless steel sieve to retain sediment-dwelling benthic infauna (animals living in the sediment). This infauna sampling strategy was supplemented by opportunistic digging in areas of the site in which infauna were clearly present (e.g. burrow openings or surface casting) to collect macrofauna (animals retained on 1mm mesh). All invertebrates found in sufficient quantities for chemical analysis (0.5 g dry weight) were identified to the species level [5, 11], pooled into a single sample for each site, rinsed in deionised water, frozen, shipped to the UK, and freeze-dried prior to analysis.

Laboratory analysis

The particle size distribution of sediments was determined using laser granulometry (Malvern Mastersizer 3000). Sub-samples were then ground to a fine powder using an agate ball mill and analysed for total organic carbon and nitrogen content using a Thermo Scientific Flash 2000 Organic Elemental Analyser. Sediment pH was determined in a soil-water suspension after shaking with water for 15 min at a 1:10 w/v ratio based on BS7755-3.2 [42]. The total concentration of PTEs in sediments was determined by ICP-MS (Inductively Coupled Plasma Mass Spectrometry) analysis of 0.5 g sediment samples digested in reverse aqua regia (9 ml of nitric acid and 3 ml of hydrochloric acid) using a MARS 6 microwave digestion system, based on EPA [43]. After preliminary analysis of a large range of PTE concentrations, the elements selected for further investigation were Cadmium (Cd), Cobalt (Co), Chromium (Cr), Nickel (Ni), Lead (Pb) and Zinc (Zn). The total concentration of Mercury (Hg) in sediments was determined using thermal degradation–gold amalgamation atomic absorbance spectroscopy as outlined in EPA [44] using a Nippon MA-3000 analyser. The bioavailability of PTEs to biota was estimated by extracting 2.5 g of sediments with 25 ml of 0.05M EDTA (Ethylenediaminetetraacetic acid) at 20°C for one hour, centrifuging, filtering and analysing PTEs (As, Cd, Co, Cr, Cu, Ni, Pb and Zn) in the extract using ICP-OES (Inductively Coupled Plasma Optical Emission Spectrometry). The total concentrations of PTEs in invertebrates were determined by ICP-MS analysis of 0.5 g of sample digested in slightly diluted nitric acid (2 ml of ultra-pure water and 8 ml of nitric acid) using a MARS 6 microwave digestion system. Details of quality control undertaken during laboratory analysis can be found in the supporting information file.

Statistical analysis

A Geoaccumulation index was calculated following [45] to determine if the potentially polluted sites (Cassiar Cannery and Papermill Bay) are polluted with respect to proximal (Inverness Passage and Wolfe Cove, respectively) and remote (Tyee Banks) reference sites using the following equation:

Sites were classified as unpolluted if Igeo ≤ 0, unpolluted to moderately polluted if 0 ≤ Igeo ≤ 1, moderately polluted if 1 ≤ Igeo ≤ 2, moderate to strongly polluted if 2 ≤ Igeo ≤ 3, strongly polluted if 3 ≤ Igeo ≤ 4, strongly to extremely polluted if 4 ≤ Igeo ≤ 5, and extremely polluted if Igeo ≥ 5.

Biota-Sediment Accumulation Factors (BSAFs) were calculated by dividing the concentrations of PTEs in the invertebrate tissues by the average concentration in the sediments from the site from which the invertebrates were collected (average across all transects and depths).

The influence of site and sediment depth on sediment PTE concentrations and properties was quantified using analysis of variance (ANOVA) and permutational multivariate analysis of covariance (PERMANCOVA) [46, 47]. The relationship between PTE concentrations and sediment properties was first tested using PRIMER’s RELATE function [48]. This function compares two resemblance matrices looking for any relationships. Relationships were further explored using principal component analyses (PCA) on the variance-covariance matrix of all sediment PTE and sediment property data. Relationships between the PTE concentrations observed in the sediments and in the collected invertebrates were examined in several configurations using RELATE and plotted using non-metric multidimensional (nMDS) scaling plots. A full description of the statistical analysis undertaken is provided in the supporting information file and outputs provided in S1, S2 and S3 Tables.

Results and Discussion

Inverness Passage, Wolfe Cove and Tyee Banks are suitable reference sites

The median particle diameter of sediments was used to determine whether sediment deposited at different sites were similar to each other so that the suitability of reference sites could be assessed. The PERMANCOVA analysis indicates that properties of sediments (pH, median particle diameter, C and N) are significantly influenced by both site and sediment depth, with site explaining more than 50% of the observed variance in the data (S1 Table). There was no significant (p > 0.05) difference in the median sediment particle diameter between Inverness Passage and Cassiar Cannery, or between Wolfe Cove and Papermill Bay (Fig 2 and S3 Table). This observation supports our assumption that the sediment deposited in the sites potentially contaminated by industrial development (Cassiar Cannery and Papermill Bay) and their respective proximal reference sites (Inverness Passage and Wolfe Cove) have the same geogenic origin and are subject to a similar depositional environment. Thus, any differences in sediment chemical properties between potentially contaminated sites and their proximal reference sites we infer is due to anthropogenic influences.

Tyee Banks had a significantly (p < 0.05) larger median particle size than Inverness Passage and Cassiar Cannery and a significantly (p < 0.05) smaller median particle size than Wolfe Cove and Papermill Bay (Fig 2 and S3 Table). The location of Tyee Banks is closer to the mouth of the Skeena River and, therefore, only receives sediment from the Skeena, whereas the other four sites may also receive sediment from the Nass River, to the north of the Skeena Estuary [49]. However, because Tyee Banks is much further from potential pollution sources, the site can be used to determine whether proximal reference (Inverness Passage and Wolfe Cove) sites are contaminated with PTEs from the adjacent historic industrial developments.

Papermill Bay sediments are contaminated with Cd, Cr, Hg and Pb

Concentrations of PTEs at potentially polluted sites were compared with reference sites to assess whether the potentially polluted sites are contaminated. The PERMANCOVA analysis (S2 Table) indicated significant differences in sediment total and EDTA extractable PTE concentrations between different sites, but not between different depths. We observed significantly (p < 0.05) greater concentrations of As, Cr, Hg and Pb in the Papermill Bay sediments than the proximal reference sediments at Wolfe Cove (Fig 3 and S3 Table). However, Co, Cu and Ni concentrations were significantly (p < 0.05) lower in the Papermill Bay sediments (Fig 3 and S3 Table). The Geoaccumulation index, using Wolfe Cove as an uncontaminated reference site, indicated that Papermill Bay is ‘unpolluted’ with As, Cd, Co, Cr, Cu, Ni and Zn, but ‘unpolluted to moderately polluted’ with Pb and ‘moderately polluted’ with Hg (Table 1). When Tyee Banks was used as the reference site, Papermill Bay is classified as ‘unpolluted to moderately polluted’ with Cd, Cr and Hg. Thus, there is evidence to suggest that the sediments in Papermill Bay are contaminated with Cd, Cr, Hg and Pb, most likely emanating from the sludge [50] discharged by the decommissioned papermill on Watson Island (Fig 1).

Discharges by papermills have previously been associated with pollution of the environment with PTEs, including Cr, Cd, Cu, Pb and Zn [51, 52]. This pollution can originate from a number of sources, such as metals in the wood entering the papermill, or atmospheric deposition of metals from the smoke stacks [51]. PTE-containing materials may also be used during the pulp making processes. For example, chromate bricks are used in the recovery furnace of kraft paper mills to reduce chemical attack by the spent liquors [53]. However, Cr concentrations observed in the sediments of Papermill Bay (Fig 3) are considerably lower than the 52.4 mg kg^-1 observed in a deposit of pulp covering the sediments of a Swiss lake [54]. They are also lower than the concentrations of up to 197 mg kg^-1 observed in a fibrous deposit in the Ångermanälven river estuary, on the east coast of Sweden [55].

The carbon and nitrogen content of the sediments at Papermill Bay was significantly (p < 0.05) and considerably greater than all of the other sites (Fig 2 and S3 Table). It was evident from visual inspection of the sediments themselves that they contained a high proportion of organic fibers (large and small strings of cellulose, and wood chips), presumably discharged into the bay from the papermill during the operational phase. The PTEs observed to be elevated in the Papermill Bay sediments (Cd, Cr, Hg and Pb) all bind strongly to organic matter in sediments [56–58]. Therefore, their presence could be due to: (i) the discharge and deposition of PTE contaminated organic material; (ii) the discharge of organic material alongside the discharge of PTEs in contaminated effluent that later bind with the organic material; or (iii) the discharge of organic material which binds to naturally present PTEs in the sediment and prevents their export to the overlying water.

In contrast to the observations made in the Papermill Bay sediments, none of the PTEs analysed in the Cassiar Cannery sediments were found to be present in significantly (p > 0.05) greater concentrations than the Inverness Passage sediments (Fig 3 and S3 Table). In fact, Co and Ni concentrations were significantly lower in Cassiar Cannery sediments than at Inverness Passage. When Inverness Passage is used as a reference site, all the PTEs analysed can be classified by the Geoaccumulation index as ‘unpolluted’ (Table 1). This observation is a clear indicator that any impact resulting from either the former use of the Cassiar Cannery site (up until 1983) as a salmon cannery, or its current use as an ecotourism lodge, can no longer be observed in the concentrations of PTEs in the top 20 cm of sediment. Any sediment contaminated by cannery waste at this site may have been buried deeper than 20 cm by fresh sediment, rendering the PTEs inaccessible to benthic invertebrates.

Relationship between sediment properties and PTE concentrations

Alongside greater total concentrations of Cd, Cr, Hg and Pb in the Papermill Bay sediments, compared to Wolfe Cove, we observed significantly (p < 0.05) greater EDTA extractable concentrations of Cd, Co, Cr, Ni, Pb and Zn (Fig 4 and S3 Table), indicating a greater availability of these elements to benthic invertebrates [59]. The significantly (p < 0.05) greater carbon content and the slightly finer texture of the Papermill Bay sediments, compared to Wolfe Cove (Fig 2 and S3 Table) should provide more sites for PTEs to bind to [58, 60, 61]. Thus, the greater concentrations of EDTA extractable PTEs observed was largely due to the Papermill Bay sediments being significantly (p < 0.05) more acidic than the Wolfe Cove sediments. Lower pH results in greater competition between PTEs and hydrogen ions for the binding sites on the sediment surface and leads to greater PTE lability [62, 63].

The nature of the relationships between sediment properties, sediment depth and PTE concentrations were investigated. The RELATE revealed a statistically significant correlation between the concentration of PTEs and the sediment physicochemical properties resemblance matrices (Rho: 0.46; p = 0.001, Permutations = 9999). The relationship between total and EDTA extractable PTE concentrations and sediment properties was further explored using PCA (Fig 5). The principal component scores plot (Fig 5B) reveals a clear separation between the Papermill Bay sediments and the proximal and remote references sites of Wolfe Cove and Tyee Banks, but no separation between Cassiar Cannery and Inverness Passage, re-enforcing the conclusion that the sediments of Cassiar Cannery show little evidence of anthropogenic contamination. Principal component 2 (Fig 5), separates the Inverness Passage and Cassiar Cannery sediments from the other three sites (Papermill Bay, Wolfe Cove, and Tyee Banks), which is attributed to a different geochemical matrix composition at Cassiar Cannery and Inverness Passage, as also observed by [64]. Principal component 1 (Fig 5) reveals a separation with depth in the Papermill Bay sediments. Papermill Bay sediments have greater total and EDTA extractable concentrations of several PTEs, higher C and N, and lower pH, all of which also increased with depth in the Papermill Bay sediments. This observation of deeper layers of sediment with a lower pH and higher availability of PTEs indicates that contaminated sediment is overlain by less contaminated sediment, deposited since the discharge of sludge ceased. Although most benthic invertebrates live in surficial sediments, it is important to consider the likelihood that deeper sediment will be exposed and mobilised either by future industrial development [65], or by bioturbation from deep burrowing invertebrates [66, 67].

Sediments host a healthy invertebrate population with no relationship between PTE concentrations in sediments and invertebrates

Overall, 22 different taxa (listed in S4 Table, along with PTE concentrations) were observed in this study, with Wolf Cove and Cassiar Cannery the most diverse sites, with 16 and 14 species collected, respectively. These taxa represent a subset of the 40 intertidal species that are commonly observed in the Skeena Estuary [5, 11]. Findings of previous studies in the Skeena Estuary reveal an infaunal invertebrate community predominantly dominated by cumaceans, polychaetes, oligochaetes, nematodes, copepods, amphipods, and bivalves [11, 68]. The 40 intertidal invertebrate species observed in this area during these previous investigations include multiple species at all trophic levels within the food web [5, 11], and such a complex community is often associated with non-disturbed habitats [21, 25]. This complexity was also observed in our study, with both predators (Alitta brandti, Neries vexillosa, Glycinde picta, ribbon worms and crabs) and primary consumers (Macoma balthica, isopods, amphipods, and sessile polychaete worms) represented [47, 69, 70]. We therefore provide evidence to support previous research indicating that the intertidal ecosystem has been passively recovering for ~50 years from past disturbances related to discharge from the papermill.

When all invertebrates collected at all five sites are considered, we found no relationship between PTE concentrations in invertebrates and either total (Rho: 0.27; p = 0.32, Permutations = 9999), EDTA extractable (Rho: 0.24; p = 0.79, Permutations = 9999), or both total and EDTA extractable (Rho: 0.78; p = 0.08, Permutations = 9999) PTEs in the sediment. This finding is confirmed by inspecting a nMDS plot of invertebrate PTE loadings, which includes all organisms collected at all five sites (Fig 6A). Bivalves of the same species (e.g. M. balthica, Mytilus edulis, and Mya arenaria), collected at different sites, seem to cluster together in Fig 6A. There is slightly more separation in the PTE concentrations measured in polychaete worms (e.g. Abarenicola pacifica, N. vexillosa, Paranemertes peregrina, G. picta, A. brandti, Nephtys caeca, Neotrypaea californiensis, and Streblospio benedicti) which are the cause of much of the dissimilarity in the dataset.

The suitability of G. oregonensis as a biomonitor of sediment PTE bioavailability

Selecting a benthic invertebrate to biomonitor PTE pollution requires consideration of how cosmopolitan their distribution is and the way in which they are exposed to pollutants in the environment [71]. The invertebrates sampled in this study (e.g. mussels, clams isopods, polychaete worms) acquire food using a diverse range of feeding strategies, including suspension feeding, deposit feeding at the surface or at depth, and predation of other infauna [72, 73]. Polychaete worms display all these feeding strategies. This large diversity of feeding strategies may have led to the greater range of PTE concentrations observed in Fig 6A for polychaete worms, compared to the bivalves and crustacea. Furthermore, the soft tissues of polychaete worms may only represent exposure to PTEs in the recent past due to large temporal fluctuations [74–76]. The whole body (including soft tissues and shells) of bivalves may better represent a record of time-averaged bioavailability of PTEs in the water column over the lifetime of the organism, albeit not without difficulties in interpretation [77, 78].

There were only two benthic invertebrate species observed at all five sites; Baltic clams (M. balthica) and Oregon pill bugs (G. oregonensis). When PTE concentrations in M. balthica, and G. oregonensis were compared to sediment concentrations, no relationship was observed with either total (Rho: 0.31; p = 0.19, Permutations = 9999), EDTA extractable (Rho: -0.0.; p = 0.48, Permutations = 9999), or both total and EDTA extractable (Rho: 0.21; p = 0.28, Permutations = 9999) sediment PTE concentrations. These findings contrast to numerous articles in the literature quantifying relationships between sediment and benthic invertebrate PTE concentrations [58, 71, 79, 80]. When we plot the PTE concentrations in M. balthica, and G. oregonensis in multidimensional space (Fig 6B), we reveal that the elemental profiles of M. balthica, and G. oregonensis contrast greatly. M. Bathica collected at Papermill Bay contained lower concentrations of Cr, Co, Ni, Zn, Cd, As and Pb than those collected at the other four sites, but higher concentrations of Cu. In contrast, G. oregonensis collected at Papermill Bay have higher concentrations of Cr, Ni, Zn, and Pb than those collected at the other four sites.

Relationships between sediment properties and BSAFs (calculated by dividing the concentrations of PTEs in the invertebrate tissues by the concentration in the sediments from the site from which the invertebrates were collected) reveal the importance of pH in explaining the difference in the bioaccumulation of PTEs by M. balthica, and G. oregonensis. For most PTEs there is a positive relationship between pH and the BSAF for M Balthica (Table 2), including a significant relationship with the Cr BSAF (Fig 7A). This is an unexpected finding that does not have an immediately obvious explanation. However, we observe a negative relationship between sediment pH and the BSAF of most PTEs for G. oregonensis (Table 2), including Cr (Fig 7B). This relationship is intuitive since metal cations dissociate from mineral surfaces at lower pH levels [62, 63]. Because G. oregonensis feeds on organic material on the surface of the sediments, the concentrations of PTEs associated with acidic sediments contaminated by papermill sludge at Papermill Bay are more likely to be assimilated by G. oregonensis and become bioaccumulated in their tissues. A survey of intertidal areas adjacent to papermills on the British Columbia coastline (including the papermill on Watson Island) identified G. oregonensis as tolerant of papermill impacted shorelines [81]. Different species of isopods have also been associated with papermill discharges on the Canadian shores of Lake Superior [82], the Swedish coast of the Gulf of Bothnia [83], and the Firth of Lome on the west coast of Scotland [84]. The greater abundance of isopods (including G. oregonensis) in close proximity to papermills, also observed by Robin, Harger [85], is attributed to the provision of pulp fibers from the mill as a food source creating a stressed environment which enables these isopods to thrive.

Conclusions

We found evidence to indicate that the discharge of papermill sludge containing PTEs from the decommissioned papermill on Watson Island has changed the sediment geochemistry at Papermill Bay by reducing the pH and increasing the total and EDTA extractable concentrations of Cd, Cr, and Pb. However, the benthic invertebrate community composition confirms that the population has mostly recovered from previous disturbance. Oregon pill bugs (G. oregonensis) were one of only two benthic invertebrate species observed at all five study sites. G. oregonensis is a cosmopolitan species and is tolerant of sites contaminated with papermill effluent because it uses the fibers discharged as a food source. Thus, we conclude here that G. oregonensis makes an excellent candidate biomonitor species to assess recovery from the environmental impact of the papermill on Watson Island and is a candidate biomonitor for monitoring the future impacts of similar industrial development projects on intertidal ecosystems.

Supporting information

S1 File [pdf]
Supporting information for: Relationships between Potentially Toxic Elements in intertidal sediments and their bioaccumulation by benthic invertebrates.

S1 Table [pdf]
PERMANCOVA showing sediment properties (pH, median particle diameter, C and N) varied by site, depth, and transect.

S2 Table [pdf]
PERMANCOVA showing that sediment total and EDTA extractable (potentially bioavailable) PTEs varied by site and transect.

S3 Table [tb]
Analysis of Variance for sediment properties. F-statistics of a two-way ANOVA with ‘site’ and ‘depth’ as the two factors.

S4 Table [pdf]
Concentrations (mg kg) of Cr, Co, Ni, Cu, Zn, As, Cd and Pb in benthic invertebrates sampled from five intertidal mudflats along the north coast of British Columbia, Canada.