Dissolved versus particulate forms of trace elements in the Athabasca River, upstream and downstream of bitumen mines and upgraders

Highlights

• Ag, Bi, Cd, Cu, Mo, Ni, Pb, Re, Sb, Tl, V, Zn were studied in the Athabasca River.

• Except for Re, total conc. were not significantly more abundant downstream of industry.

• Except for Mo, Re and Sb, the elements were found mainly in the particulate fraction.

• Total concentrations were below the Guidelines for the Protection of Aquatic Life.

• Trace element concentrations are similar to those reported for other major rivers.

Abstract

Employing trace metal-free analytical protocols developed for polar snow and ice, water samples were collected upstream, midstream and downstream of open pit bitumen mines and upgraders along the Lower Athabasca River (AR, ~125 km). The purpose of this study was to determine whether dissolved (< 0.45 µm) and total (i.e. acid-extractable) trace element (TE) concentrations in the AR have been significantly impacted by industrial operations. Of the TEs known to be enriched in bitumen such as V, Ni, Mo and Re, only total Re was significantly more abundant downstream of industry. Total concentrations of Ag, Bi, Cd, Co, Cu, Sb, Tl and Zn were not significantly more abundant downstream of industry, compared to upstream. Moreover, these elements occur predominantly (63–96%) in the particulate fraction. In contrast to the cationic TEs, elements which occur predominantly as anions in aqueous solution (Mo, Re and Sb) are mainly in the dissolved fraction (89%, 56%, and 73%, respectively). Vanadium, Ni and Pb in the suspended solids revealed similar concentrations downstream and upstream of industry, with concentrations proportional to those of conservative, lithophile elements (Al, La, and Th), but well below their corresponding abundance in crustal rocks. Taken together, these findings suggest that physical weathering and erosion are the primary sources of metals in the AR, both upstream as well as downstream of industry, with silicates the dominant source of most of the TEs of concern.

Introduction

The three most important mechanisms which govern the major ion composition of global rivers are atmospheric precipitation, mineral-water interactions, and evaporation-crystallization processes (Gibbs, 1970). On the other hand, the geochemical behavior of trace elements (TEs) is more complicated because they may be present in one of several forms: ‘truly dissolved’ ionic species and simple compounds that are < 1 kDa in size, ‘colloidal’ species in the range of 1–1000 nm such as TEs complexed with or adsorbed to organic matter or iron oxyhydroxides, and particulates, consisting of larger complexes, mineral particles and other sedimentary materials that are suspended within the water column (Gaillardet et al., 2014). For many TEs, total concentrations are dominated by the particulate fraction (Gibbs, 1973).

In practice, TEs are operationally separated into the dissolved (<450 nm) and particulate (>450 nm) fractions by the process of filtration. This process is useful for separating the smaller, more reactive TE species in the dissolved fraction from the more inert TEs in the particulate fraction, which are largely contained inside the crystalline structure of minerals. Due to the convenience of filtration and the greater representativeness of the dissolved fraction in terms of bioaccessibility (i.e. the proportion present in a form that is accessible for uptake into aquatic organisms; McGeer et al., 2004), the dissolved fraction is often used in regulatory criteria (e.g. US EPA, 2018). However, the effectiveness of this separation is limited because it ignores the dynamic nature of colloidal interactions and causes numerous artefacts (Buffle and Leppard, 1995; Horowitz et al., 1996; Lead and Wilkinson, 2006; Filella, 2007). For example, the TEs with sizes <450 nm that are sorbed to the surface of minerals in the >450 nm particulate fraction can be released when physicochemical conditions such as pH, redox potential, ionic strength or temperature are altered, or when the local concentration of TEs is changed (Smith, 1999; Warren and Haack, 2001; Van Leeuwen and Buffle, 2009; Cuss et al., 2020). Similarly, organic matter and small inorganic colloids can partially clog filter pores and decrease the effective pore size, resulting in the exclusion of material that is < 450 nm and the consequent underestimation of dissolved TE concentrations (Morrison and Benoit, 2001). Biologically-mediated reactions also complicate the relationship between size and bioaccessibility, such as the desorption of ions from the surface of particles >450 nm that are retained in mucus layers, and digestive processes that may liberate TEs from complexes (Tao et al., 2000; Glover and Hogstrand, 2002; Zhao et al., 2016). Indeed, while the concentration of the free metal ion is most associated with bioaccessibility/bioavailability and subsequent toxic effects, the dynamic exchanges of TEs between ionic species, the dissolved fraction, and smaller particulates that exemplify ‘colloidal behaviour’ make it challenging to identify the most appropriate target for regulation (Paquin et al., 2002; Buffle et al., 2009; Tercier-Waeber et al., 2012). Additionally, the particulate fraction of TEs can be highly relevant for some species such as bivalve molluscs, which process suspended particulate matter and sediment during their feeding processes (Tessier et al., 1984; Griscom and Fisher, 2004; Rzymski et al., 2014). For these reasons, some water quality guidelines apply a precautionary approach by regulating TEs in terms of total concentrations, such as the Canadian Council of Ministers of the Environment (CCME, 2018). While such conservative practices may lead to less effective use of sparse regulatory resources, particularly in sedimentary terrains where rivers with high flow may carry considerable suspended sediment, a quantitative understanding of both the dissolved and particulate fractions is needed to fully understand the ecological significance of TEs (Paquin et al., 2002; Lumb et al., 2006). Furthermore, natural processes such as the transport and deposition of mineral particles in rivers via the erosion of soil or the resuspension of sediments are major sources of particulate TEs to rivers during periods of high flow (Gaillardet et al., 2014). On the other hand, dissolved concentrations are controlled by other variables such as their solubility and concentrations of colloidal carriers < 450 nm (e.g. dissolved organic matter and iron oxyhydroxides; Mason, 2013). Hence, understanding the geochemical origins and processes that contribute to the concentration and distribution of TEs in rivers also requires quantitative understanding of their size-based partitioning.

The Athabasca River (AR) watershed has attracted national and international concern regarding possible contamination by TEs (Kelly et al., 2010; Schindler, 2014), especially the lower reaches of the river which are in the vicinity of the Athabasca Bituminous Sands (ABS). The AR is one of the largest rivers in Canada (~1200 km long) and has a great socioeconomic significance to the province of Alberta. However, ongoing large scale mining and upgrading activities in the region of the ABS are claimed to be a serious concern for water quality in the AR (Kelly et al., 2009, 2010). There are numerous risks of contamination to the river through either direct impacts such as potential leakage from tailings ponds (Timoney and Lee, 2011; Ross et al., 2012; Savard et al., 2012; Holden et al., 2004; Moncur et al., 2015; Roy et al., 2016) or indirect sources such as airborne emissions from open pit mines and upgraders (Kelly et al., 2009, 2010; Jautzy et al., 2013; Schindler, 2014, 2015; Chibwe, 2019; Landis et al., 2019a, 2019b; McNaughton et al., 2019; Kurek et al., 2013; Bari et al., 2014; Evans et al., 2016; Guéguen et al., 2016). Two recent reviews provide insight into some of the issues associated with the ABS and the AR (Alexander and Chambers, 2016; Huang et al., 2016). To add to the chemical complexity of the lower AR watershed, in addition to the extensive bitumen deposits, there are abundant tributaries, saline seeps and springs (Gibson et al., 2011; Jasechko et al., 2012; Gue et al., 2015, 2018; Cuss et al., 2020). Clearly, in this type of complicated geological setting, it is essential to carefully distinguish anthropogenic inputs from natural ones.

Specifically, the focus of our attention is the claim that ABS mining and upgrading activities are a significant source of TEs considered priority pollutants under the U.S. Environmental Protection Agency's Clean Water Act (Ag, Be, Cd, Cu, Ni, Sb, Tl and Zn) to the AR and its tributaries (Kelly et al., 2010). Previous work showed that the dissolved (<0.45 μm) concentrations of V, Ni, and Mo were slightly, but significantly (p < 0.05) greater downstream of industry, and Re was profoundly (approximately 3-fold) more abundant downstream (Shotyk et al., 2017). These four metals (V, Ni, Mo and Re) are all enriched in bitumen, relative to their abundance in crustal rocks, and occur primarily in the bitumen fraction of the ABS (Hitchon and Filby, 1983; Selby and Creaser, 2005; Bicalho et al., 2017). However, it is not yet clear whether the differences in abundance of dissolved V, Ni, Mo and Re upstream versus downstream of industry are due to natural or anthropogenic inputs (Shotyk et al., 2017). In respect to the potentially toxic chalcophile metals (Ag, Cd, Cu, Pb, Sb, Tl and Zn), they were not significantly more abundant downstream versus upstream of industry (Shotyk et al., 2017).

To further our understanding of the impact of ABS mining and upgrading activities on the AR, water samples were collected in 2 consecutive years (2014 and 2015) from the main stem AR and its tributary streams. Here, we report total concentrations of trace metals including V, Ni, Mo and Re, as well as Ag, Bi, Cd, Co, Cu, Sb, Tl and Zn. In particular, we compare total concentrations upstream and downstream of industry to identify possible anthropogenic contributions, and view them in light of CCME guideline values. We also compare our results with the data reported by Regional Aquatics Monitoring Program (RAMP), a dedicated environmental monitoring program to determine the potential effects of resource development on aquatic systems. We have previously considered the distribution of TEs in the dissolved fraction (Shotyk et al., 2017), and amongst major colloidal species in the dissolved fraction (Javed et al., 2017; Cuss et al., 2018). To estimate lithogenic contributions, we compare the abundance of conservative, lithophile elements (Al, Cr, Ga, La, Th) with the abundance of TEs in the particulate fraction, and the TEs that are present in the suspended solids trapped on filters (V, Ni and Pb). We also calculated the TE fluxes of the AR upstream and its major tributaries in both 2014 and 2015. The main goals of the study are to determine the potential impact of bitumen mining and upgrading activities on the AR and to discern the natural and anthropogenic trace elements inputs to the AR. To minimize the risk of contamination, we employed modified sampling, handling and analytical protocols and procedures that were developed for the determination of Pb and stable Pb isotope ratios in Arctic ice cores (Shotyk et al., 2005a; Zheng et al., 2007), and have been previously applied to surface and ground waters (Shotyk and Krachler, 2010; Shotyk et al., 2010; Javed et al., 2017; Javed and Shotyk, 2018; Cuss et al., 2018).

In this study, the “total” concentration of TEs is used in the same way as the CCME (and regulatory agencies in Canada), which is actually “strong acid leachable” and equivalent to “acid-extractable” or “acid-soluble” (or “quasi-total”) in other jurisdictions i.e., US EPA and ISO (US EPA Method 200.2, 1994; ISO 15587-2, 2002; US EPA 3005A; US EPA 3015A; US EPA Method 3051A, 1999; US EPA 3020A, 1992).