The Monte Carlo simulations for sampling points within the Imetjoki catchment resulted in an overall interquartile range between 10 and 37% MSR for both the spring and summer, with the catchment-scale median MSR being 19% (mean: 27%; see full output details in Supporting Information Table S3‒4). A clear difference can be seen between the seasons (Figure 3), where the Monte Carlo MSR-output in the spring (yellow probability density distributions) lie closer to the 0%-line compared to the MSR-distributions in the summer (purple density distributions). For locations sampled in both seasons, the difference in catchment-scale MSR was 9 percentage units over the summer (i.e. 14% median MSR in the spring and 23% in the summer; Fig. 3). The highest increase in MSR was detected at the Northern Lake outlet (from a median of 5% in spring to 24% in summer) and the Imetjärvi inlet (from 4% to 20%, respectively), which is also shown in Figure 2b where these two sampling points (IDs 3 and 4) had the highest increase in measured isotopic value from spring to summer. The smallest seasonal change was found in samples representing the Industrial area (ID 9) where high median MSR was found both in the spring (32%) and in the summer (34%).
The samples from the Khibiny catchments show an interquartile range of MSR-values between 0 and 15%, with a catchment-scale median MSR value of 7% for Yuksporiok (mean: 16%), and as low as 2% for Vuonnemiok (mean: 7%; see full output details in Supporting Information Table S5). The highest MSR (14%) in Khibiny was found in the Belaya River (ID 22), which may be impacted by potential MSR in the Vudyavr Lake and the other two (unmonitored) tributaries (Vudyavriok and Saamsky; Fig. 1). However, the estimation for this location is uncertain as it may be biased from a potential release of additional sulfur from Kirovsk City, e.g. as municipal wastewater having another and here unaccounted for isotopic signal.
Discussion
Catchment conditions affecting natural MSR
The Imetjoki and Khibiny catchments share similar climatic features (e.g. temperature and number of days with snow coverage) and the sulfur concentrations of their water systems show overlapping ranges. However, we found that catchment-scale median MSR-values in Imetjoki (summer season) were about 18 percentage units higher than the catchment-scale median MSR-values in Khibiny during the same summer season (23% MSR vs. 5%). This supports our working hypothesis that some fundamental conditions are more favorable for MSR in the Imetjoki catchment compared to the Khibiny catchments. For instance, organic material was more abundant in the Imetjoki stream water most likely due to its forest cover, yielding higher DOC concentrations from decomposing vegetation. Direct effects of an increase in organic matter concentration on natural MSR have previously been observed in southern Finland (Bomberg et al., 2015), where organic material was added to acidic water in flooded mine shafts to induce MSR, which over the course of 15 years successfully increased the pH and lowered the metal concentrations. Furthermore, the anoxic environments required for MSR (Pester et al., 2012) as well as relatively slow residence times allowing for longer exposure to SRM (Nelson et al., 2009) are likely present at multiple “hot-spot” locations within the Imetjoki catchment, which contains 11 small lakes and has about 16% of its total area covered by peatlands (SGU, 2020). High δ34S-values in wetland-dominated sub-catchments (relative to other sub-catchments) has previously also been attributed to MSR (Björkvald et al., 2009). Notably, although most surface waters were only slightly acidic (pH 5.7-7.1), a few locations within the Imetjoki catchment displayed high metal concentrations and/or strongly acidic water (pH 3.3-4.8), which is potentially harmful to SRM (Xu and Chen, 2020). These strongly acidic waters were nonetheless associated with non-negligible MSR-levels. For example, the Industrial area (ID 8) had in the spring a pH of 3.3 and a Cu concentration of 5000 µg/L and still showed a median MSR of 17%. An explanation could be that the local stream network have developed acid-tolerant SRM, considering that mine drainage from the Nautanen mines has been developing during a long time over the past 110 years, reaching a condition that can be described as nearly steady-state with regard to the considerable downstream pollution transport (Fischer et al., 2020).
The water from Belaya and Vuonnemiok catchments of Khibiny on the other hand showed low median MSR-values (5%), which is consistent with the prevailing high alpine environments characterized by bare rocks and snow fields remaining over the summer, and the fact that they have been increasingly industrialized over the last century (Moiseenko et al., 2009). Both these factors contribute to the sparse forest cover of the Khibiny catchments, occurring only in the lower valleys outside of built areas, yielding generally low DOC concentrations (1-3 mg/L), as measured within the catchments in 2017. Furthermore, the high alpine environment spanning elevations between 100 to 1200 meters above sea level is characterized by steep slopes and a thin soil layer, which contributes to relatively fast through-flow of precipitation (i.e., shorter hydrological residence times). For instance, the Gakman sub-catchment (6.2 km2; IDs 16‒18) has a slope of 0.07 m/m, while the Imetjoki catchment (6.6 km2) has a slope of 0.03 m/m. Whereas the Gakman sub-catchment in Khibiny on average had relatively similar pH (6.8-7.4) as Imetjoki, the Yuksporiok mining area in Khibiny (ID 20‒21) had highly basic pH (10.2-10.6), which potentially is another factor that can limit the MSR-values. Conclusive evidence regarding how much (high) pH may limit natural MSR seem however still to be lacking, since some studies have shown presence of considerable SRM in alkaline waters, e.g. up to pH 9.8 in mine shafts in southern Finland (Bomberg et al., 2015) and in mine tailings with pH up to 9.3 in New Zealand (Chappell and Craw, 2003).
Seasonal conditions affecting natural MSR
The presented results regarding impacts of seasonality show that the catchment-scale MSR-values under spring conditions with on-going snow-melt was lower (14%) compared to MSR-values under summer conditions (23%). This difference suggests that there is a summer “boost” in the activity of SRM. For instance, between early spring (May) and summer (August), the average catchment-wide water temperatures rose from 3.8 to 11.5 °C, which is more favorable even for cold-tolerating bacteria (Virpiranta et al., 2019). However, the isolated effect of temperature (and the seasonal changes of it) on field-scale MSR is difficult to determine since multiple factors are likely interacting (Khan et al., 2019; Praharaj and Fortin, 2004). Two such factors that showed pronounced seasonality in Imetjoki are DOC, which increased on average with 30% over the summer (Fischer et al., 2020), and sulfur concentrations, which increased with 26% (Figure 2b; excluding ID 8 which decreased by almost 3 times). This also means that sites in warmer climates with usually increasing summer bioproductivity and DOC turnover would likely enable even higher such “boosting”.
Another factor to consider is that the sampled spring stream water represents a mixture of groundwater dominated base flow and melt water. The latter may remain on top of, or near the soil surface, since the ground would still be frozen during spring, hindering infiltration and groundwater recharge. Hence, it is likely that many of the considered water samples from the spring have a relatively high proportion of sulfur from atmospheric deposition stored in the melting snow pack, which has not been in contact with groundwater nor been exposed to prior MSR, hence diluting the MSR-values form the (mine-drainage impacted) base flow (e.g. as similarly observed by Mörth et al., 2008). Accounting for such possible dilution, the actual MSR of the mine-drainage impacted waters alone must then be equal to or higher than the (mixed) sample value of 14%. A possible explanation for the locally high MSR-value of 30% during spring at ID 9 may then be that it represents a groundwater flow dominated brook, thus being less impacted by meltwater dilution. Observations of relatively high MSR (before onset of dilution of melt water) was observed at the Krycklan catchment in northern Sweden (e.g. Björkvald et al., 2009), and at Lake Mjösjön in central Sweden where winter ice-coverage gave a gradual decrease in lake water oxygen levels and by that favored MSR (e.g., Andersson et al., 1992).
Implications - MSR as a nature-based solution to mine drainage under ambient changes
The present study supports previous cold region studies (e.g. Fischer et al., 2022) in detecting high MSR-values locally (30-40% under natural conditions). We additionally found novel evidence of a more general presence of MSR over different seasons, even under the considered cold (Arctic) conditions, which indicates that there is large potential for using MSR as part of nature-based solutions to mitigate adverse impacts of (acid) mine drainage in the Arctic and elsewhere. This can for instance be obtained through managing and taking advantage of favorable MSR-conditions in (constructed) wetland and lake systems. Enhancement measures could be carried out by adding organic material in sedimentation ponds or mining shafts (e.g., Bomberg et al., 2015) utilizing already existing structures to provide prolonged residence times and possible anaerobic bottom layers.
The Arctic is currently subject to considerable climate-driven shifts in the carbon cycle that are projected to continue in the foreseeable future. There is notably an observed trend of increasing DOC-values in streams (Guo et al., 2007) which is favorable for MSR (Bomberg et al., 2015), and will therefore probably act to enhance MSR over vast regions. Conversely, the increasing number of snow- and ice-free days (e.g., Box et al., 2019) will increase vertical fluxes of water and oxygen in upper soil layers, which in turn may decrease MSR in some locations by removing associated pockets of anaerobic conditions (e.g., see similar discussion in Palomo et al., 2013). Unless such MSR-driven changes and their impact on the retention of metals and other substances along their transport pathways are recognized and understood, they may – if detected through downstream monitoring – be misinterpreted as changes in metal mobilization. For instance, increased retention by (neglected) MSR may be at risk of being misinterpreted as decreased source zone mobilization of metals. The metal mobilization (e.g., at mining waste heaps) are notably also expected to be impacted by on-going hydro-climatic changes, although through quite different processes (e.g., Hotton et al., 2020; Jarsjö et al., 2020; Shrestha et al., 2020). A key challenge in understanding the role of MSR in mitigation solutions to (acid) mine drainage is therefore to consider how it may impact the balance between mobilization and retention, under conditions of future ambient change.
Conclusions
We here interpret and compare results from multiple stream water measurements in two major Arctic mining regions, allowing us for the first time to quantify net impacts of catchment conditions and seasonality on large-scale MSR. This reflects a combination of multiple drivers and large-scale processes that e.g. are difficult to reproduce in laboratory experiments. Specifically, we conclude that:
A likely range of catchment-scale MSR-values in the Arctic is ~5-20% during the summer season, with catchments located in vegetated terrain containing wetlands, lakes or voluminous groundwater systems that allow longer reaction times for SRM being more likely to show median MSR-values of around ∼20%.
Local values of MSR can be as high as ~40% at hot spot conditions e.g. near lakes, indicating large potential for using MSR as part of nature-based solutions to mitigate adverse impacts of (acid) mine drainage.
Evidence of persistent field-scale MSR over different seasons in the Arctic indicate that microbial processes and their interactions with the environment may be more persistent than previously anticipated.
The present results showing wide-spread and persistent MSR even under cold conditions call for more detailed investigations regarding potential field-scale correlations between MSR and individual landscape and hydro-climatic characteristics (e.g., DOC, water temperature, ice-cover, vegetation, slope), which e.g. can be supported by the here utilized isotopic fractionation and mixing scheme.