Science of the Total Environment 778 (2021) 146070 Contents lists available at ScienceDirect Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv Natural acid rock drainage in alpine catchments: A side effect of climate warming Mario Zarroca a,⁎, Carles Roqué b, Rogelio Linares a, José G. Salminci a,c, Francisco Gutiérrez d a Geology Department, Universitat Autònoma de Barcelona, E-08193-Bellaterra, Barcelona, Spain b Àrea de Geodinàmica Externa i Geomorfologia, Universitat de Girona, E-17071 Girona, Spain c Geology and Environment Department, Instituto Nacional de Tecnología Industrial (INTI), Avenida General Paz 5445, Buenos Aires, Argentina d Earth Sciences Department, Universidad de Zaragoza, C/. Pedro Cerbuna 12, E-50009 Zaragoza, Spain 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 • Historical mapping reveals strong inten- sification of high-mountain acidifica- tion. • Natural acid rock drainages release potentially toxic metals into mountain habitats. • Both climatic and geomorphological dy- namics control the severity of the phe- nomenon. • Natural acid rock drainages likely will intensify in a future due to climate warming. a r t i c l e i n f o a b s t r a c t Article history: A historical series of aerial photographs spanning more than 70 years (1945–2018) revealed that natural acid Received 23 December 2020 rock drainage (ARD) has experienced an intensification in the Noguera de Vallferrera alpine catchment (Central Received in revised form 13 February 2021 Pyrenees) due to climate change during the last decade. ARD manifests by the precipitation of whitish aluminum- Accepted 20 February 2021 compounds that strikingly cover the beds of some gullies and streams in high-mountain catchments. The total Available online 26 February 2021 length of affected streams has increased from ca. 5 km (1945) to more than 35 km (2018). Up to 68 water sam- Editor: Christian Herrera ples were collected in three main areas to determine the spatial variation in acidity and concentration of dis- solved metals, representative of surface and subsurface waters. Concentration of aluminum clearly correlates with acidity of waters. Aluminum precipitation occurs where acidic waters, enriched in metals due ARD related Keywords: to the oxidation of sulfides, mix with non-acidic waters. In addition to aluminum, other potentially toxic trace Acidification metals are present at concentrations well above the quality standards for natural waters. Here, we show that cli- Alt Pirineu Natural Park (Central Pyrenees) mate warming and the severe droughts recorded in the last decade are the most plausible causes for the observed Climate change ARD intensification. This result is supported by a good correlation between the regional ascending rate of the High-mountain waters periglacial limits (ca. 46 m-height/decade) and the rising rate of the maximum elevations at which ARD occurs Potentially toxic trace metals (ca. 45 to 55 m-height/decade). In addition to climatic control, we also show that the local geomorphology is playing a major role. The distribution of periglacial deposits (rock glaciers, protalus ramparts, cones and talus slopes) and deep-seated gravitational slope deformations exert a strong control on the spatial patterns and hy- drodynamics of ARD. A better understanding of the phenomenon and the monitoring of its evolution can provide clues on these side effects of climate warming, here and in many other alpine catchments worldwide. © 2021 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/). ⁎ Corresponding author. E-mail address: mario.zarroca.hernandez@uab.cat (M. Zarroca). https://doi.org/10.1016/j.scitotenv.2021.146070 0048-9697/© 2021 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/). M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 1. Introduction Gault et al., 2015). The climate is alleged to exert a strong control on ARD processes (Furniss et al., 1999; Plumlee et al., 1999). Some observa- Acid drainage is considered by the United Nations one of the most tions made during the last decades suggest that climate change is pro- challenging environmental problems that society will face in the moting an increase in some water solutes in mountain areas with coming years (e.g. Tuffnell, 2017; Qian and Li, 2019). Acid drainage is metal-sulfide mineralization (Manning et al., 2013), where ARD is the the result of bio-chemical weathering of certain minerals contained in main natural mechanism for metal enrichment in waters. Studies con- sulfide-bearing rocks under water and oxygen-rich environments (e.g. ducted over 30 years in the southern sector of the Rocky Mountains Nordstrom, 2011a). The interaction of sulfides as Pyrite (FeS2) with (Colorado, USA) reveal that metals and SO2− 4 concentrations in water oxygen and water results in the formation of sulfuric acid that attacks have increased up to 100–400% in recent years (e.g. Todd et al., 2012; the rocks, dissolving Fe, Al and other minor elements (Nordstrom and Crouch et al., 2013; Manning et al., 2013). The warmer and drier condi- Alpers, 1999; Bigham and Nordstrom, 2000; Blowes et al., 2004). tions induced by the climate change are claimed to be the possible Some of these elements, in non-elevated concentrations, are considered causes for the significant quality deterioration of rivers and groundwa- essential for life (e.g. Co, Cu, Zn, and Mn) (Mertz, 1981; Nielsen, 1990), ter in the region. This problem has been also documented in other whereas others (e.g. Cd, Hg, Pb, or As) either do not have any known mountain areas worldwide, such as the Yukon Territory (Canada) physiological function and/or are very toxic, even at trace concentra- (Lacelle et al., 2007; Kwong et al., 2009; Gault et al., 2015); the Himalayas tions (Godt et al., 2006; Sarmiento et al., 2011; Olmedo et al., 2013). (Salerno et al., 2016); or the European Alps (Thies et al., 2007; Ilyashuk Therefore, acidic metal-enriched waters may cause severe degradation et al., 2014). of water masses and damage to ecosystems (e.g. Nordstrom, 2009; There is strong evidence supporting that the rate of warming is am- Nordstrom, 2011a; Nordstrom, 2011b; Talukdar et al., 2016; Ilyashuk plified by elevation (Pepin et al., 2015). The effects of elevation- et al., 2018). In addition to solutes, metals also form precipitates. For dependent warming are especially significant in alpine environments. example, Al, which is a major component of a number of common Warming rates have reached up to +0.8 °C/decade in the Tibetan silicate minerals, generally precipitates at pH > 4.5˗5.6 in the form of Plateau since the last 50 yrs. (Pepin et al., 2015), and + 0.5 °C/decade whitish-colored aluminum-hydroxides and -hydroxisulphates (AHPs) in the European Alps since the 1980s (EEA, 2009). In the Pyrenees (Bigham and Nordstrom, 2000; Takeno, 2005; Sánchez-España et al., (Fig. 1), the average temperature has raised ca. +0.2 to +0.4 °C/decade 2011, 2016; Gault et al., 2015). since the 1950s, which exceeds in 30% the world average (OPCC-CTP, Acid drainages are commonly linked to the mining industry (acid 2018). Here, positive anomalies have been recorded systematically mine drainage, AMD) (Nordstrom, 2009). However, they may also since 1980s, being the last decade the warmest since the start of the in- occur in unmined catchments rich in sulfides under specific natural strumental records (1910). This trend is even more pronounced in the conditions (acid rock drainage, ARD) (Nordstrom and Alpers, 1999). central sector of the Pyrenees, where significant positive trends for Although the study of AMD has been extensively addressed in the liter- temperature (T = +0.23 to 0.57 °C/decade) have been observed during ature (e.g. Plaza et al., 2017; Qian and Li, 2019; Rezaie and Anderson, the period 1970–2013, in contrast with the more subtle long-term 2020; Giddings et al., 2020; Galván et al., 2021), research dealing with (1910–2013) trends (T = +0.11 to +0.06 °C/decade) (Pérez-Zanón the ARD phenomenon and the associated impacts is much more limited et al., 2017). General trends regarding precipitation are not as clear, (e.g. Kwong et al., 2009; Todd et al., 2012; Ilyashuk et al., 2014, 2018; since positive and negative anomalies alternate over short time spans Fig. 1. Geological setting of the study area. (A) Geological sketch of the Pyrenees showing the main structural units (after Rosell and Linares, 2001). (B) Distribution of areas with evidence of natural acid rock drainage (ARD) in the Pyrenean axial zone, occurring in high-mountain catchments mainly underlain by sulphide-bearing metasedimentary rocks. (C) Simplified geological map of the eastern sector of the Alt Pirineu Natural Park. The Noguera de Vallfererra catchment (NVC) is associated with the Mérens shear zone, where a thick succession of metasedimentary Silurian to early Ordovician formations crop out, including the pyrite-rich slates of the Lladorre Unit in the Monteixo-Norís Range. 2 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 for the same periods (OPCC-CTP, 2018). Precipitation negative trends The contact with the atmosphere causes ferrous iron (Fe2+) to fur- have been reported for the eastern sector of the Pyrenees (−3 ther oxidize to ferric ion (Fe3+), producing additional acidity according to −8%/decade), more pronounced since 1950 (BAIC, 2019). However, to the reaction (2) (Akcil and Koldas, 2006). although this pattern is also observable in the central Pyrenees, the ampli- tude of the negative anomalies is rather small (−0.64% during 1910–1970; Fe2þ þ 1=4O2 þ Hþ →Fe3þ þ 1=2H2 O ð2Þ −2.70% for 1970–2013) (Pérez-Zanón et al., 2017; OPCC-CTP, 2018). The Pyrenean region hosts the southernmost glaciers in Europe Depending on the specific composition of the sulfide, other minor (Fig. 1A) and the warming has already caused a 50% reduction of the metals (e.g. Zn, Cu, Cd, Co) and metalloids (e.g. As, Sb) may be also re- total glaciers reported in 1984 (OPCC-CTP, 2018). Periglacial rock de- leased in the oxidation process (Nordstrom and Alpers, 1999). The ki- posits such as rock glaciers are climatically more resilient than glaciers netics of the oxidation process is very slow, but it is typically catalyzed to thaw (Jones et al., 2019), although they are also experiencing rapid at low pH by the action of acidophilic archaea, bacteria (e.g. Thiobacillus degradation due to climate warming rapid degradation due to climate ferrooxidans sp.) and photosynthetic algae, which increases the rate of warming (e.g. Thies et al., 2007; Salerno et al., 2016: Ilyashuk et al., reaction by several orders of magnitude over abiotic rates (Nordstrom 2018). Rocky periglacial deposits may constitute high-permeability and Southam, 1997; Baker and Bandfield, 2003; Jennings et al., 2008; aquifers (Jones et al., 2019), and their thawing is reported to cause Nordstrom, 2011b; Cruz-Hernández et al., 2012). ARD in many cold mountains with sulfide mineralization (Williams Al is a major component of a number of common silicate min- et al., 2006; Todd et al., 2012; Ilyashuk et al., 2014). Climatic predictions erals and thus is a frequent element in ARD-precipitates. Al is dis- do not foresee a change in the climate warming trend for the coming solved at pH < 4.5˗5.6 and precipitates above this pH in the form years (e.g. Radić et al., 2014; Huss and Hock, 2018), thus ARD could be- of whitish-colored aluminum-hydroxides and -hydroxisulfates come more common and severe in many mountain regions worldwide. (AHPs), which composition is dependent on the bedrock mineral- The consequences on the quality of water resources and ecosystems re- ogy. Common aluminum-hydroxysulfates that precipitate in such pH main uncertain. range are aluminite (Al2(SO4)(OH)4-7H2O), basaluminite (Al4(SO4)(OH) Here we aim at investigating the ARD processes occurring in the 10·5H2O) and hydrobasaluminite (Al4(SO4)(OH)10·12-36H2O), which is Noguera of Vallferrera catchment (NVC) (Alt Pirineu Natural Park, the most ubiquitous in ARD. Aluminum-hydroxides such as gibbsite (γ- Central Pyrenees) (Fig. 1). In this regard, this research attempts to Al(OH)3), its polymorph bayerite (α-Al(OH)3) and other amorphous to provide a scientific basis on ARD that could be useful for designing nano-crystalline phases are also frequent (Nordstrom and Alpers, 1999; effective adaptation strategies to the impacts of climate warming on Bigham and Nordstrom, 2000; Takeno, 2005; Sánchez-España, 2007; water resources and ecological value of mountain habitats. The main Sánchez-España et al., 2011; Sánchez-España et al., 2016; Gault et al., specific objectives of the research are (1) to identify the scope of the 2015). ARD processes in the NVC, (2) to study their evolution during the last Geochemical composition of the whitish precipitates in the NVC was decades; and (3) to assess the main environmental factors that control determined by Igc (2012). Scanning electron microscopy with energy the phenomenon. NVC is a pristine alpine catchment in which human dispersive X-ray analysis (SEM-EDX) revealed a secondary mineral activity has been scarce to date. However, ARD produces striking whit- roughly cracked by dehydration, composed mainly of Al, O, and S in ish precipitates of aluminum along km-long stretches of some stream less quantity. The precipitates are nearly amorphous to X-ray diffraction beds (Fig. 2). In recent times, signs of ARD are expanding downstream (XRD), with just some possible nanocrystals embedded into the amor- and upstream to areas where the phenomenon had not been observed phous matrix. However, some broad reflections near 7–10° and 20–25° during the last decades (IGC, 2012). ARD manifestations had been previ- 2θ were recognized. This XRD pattern is compatible with aluminum sec- ously reported in the Pyrenees, but they had just seasonal development ondary minerals formed at pH ~ 4.5–5.0 (Sánchez-España, 2007; Du et al., and their effects were limited to some restricted areas (e.g. Zaharescu 2009; Sánchez-España et al., 2011). This geochemical signature may be et al., 2009; Zaharescu et al., 2016). ARD is actually becoming more characteristic of aluminum-hydroxides as gibbsite or bayerite, with a widespread in several watersheds and show conspicuous manifesta- minor fraction of some aluminum-hydroxysulfate as hydrobasaluminite. tions regardless of the season (Figs. 1B and 2), as is the case of the It should be noted that hydrobasaluminite is a metastable compound that NVC. This general trend leads to the hypothesis that climate (i.e. cli- tends to transform to a more stable phases as gibbsite or bayerite and, mate warming) is exerting strong control on ARD. Furthermore, therefore, it is common for them to coexist (Sánchez-España et al., 2016). some lakes and streams in the NVC headwater are acidic and show anomalous concentrations of trace metals that lead them to be qual- 3. Geological setting ified as polluted (IGC, 2012). In contrast, some other nearby lakes and streams are circumneutral or even markedly basic and of low The Pyrenees is a ca. 500-km long, E-W-trending mountain chain metal concentrations, despite having the same elevation, orientation that extends between France and Spain, in SW Europe (mean lati- and bedrock mineralogy. This suggests us that, in addition to climate, tude 42.4 o N) (Fig. 1). From the geotectonic perspective, it is a other local factors may play a key role in ARD. double-verging thrust-and-fold belt resulting from the collision be- tween the Iberian and Eurasian plates during the Mesozoic-Cenozoic 2. Acid rock drainage and precipitates Alpine orogeny (e.g. Choukroune and ECORS, 1989; Muñoz, 1992; García-Sansegundo, 1996). Three main structural domains are dif- The weathering of sulfides is a bio-chemical process that promote the ferentiated: (1) the North-Pyrenean Zone, between the Aquitanian production of sulfuric acid under oxygen-rich environments (e.g. foreland basin and the North-Pyrenean Fault, with northward verg- Nordstrom, 2011b). Pyrite (FeS2) and its polymorph marcasite are com- ing thrust-and-fold structures involving both the Variscan base- monly involved in this process, but other sulfides such as arsenopyrite ment, composed of Paleozoic rocks, and post-Variscan sedimentary (FeAsS), chalcopyrite (CuFeS2), sphalerite (ZnS) and pyrrhotite (Fe1-xS) successions; (2) the Axial Zone, that constitutes the core of the can also take part. When exposed to air and water, the oxidation and de- orogen, in which the Variscan basement forms a south-verging composition of pyrite may produce a solution of ferrous sulfate and sul- antiformal stack; and (3) the South-Pyrenean Zone, comprising a se- furic acid that causes the acidification of water (reaction 1) (Bradley and ries of thrust sheets of Mesozoic and Paleogene formations that have Worland, 2015): been transported tectonically towards Ebro foreland basin (Fig. 1A) (e.g. Barnolas and Pujalte, 2004; Teixell and Muñoz, 2000). The Pyr-   enean Axial Zone is the topographically higher structural unit with a 12FeS2 þ 45O2 þ 34H2 O→4 H3 OFeðSO4 Þ2 2FeðOHÞ3 þ 16H2 SO4 ð1Þ number of peaks above 3000–3400 m a.s.l.. The identified ARD 3 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 Fig. 2. Evidence of whitish aluminum-hydroxysulfate precipitates (AHPs) on stream beds in the Central and Eastern Pyrenees, at elevations ranging from 2200 to 2500 m a.s.l. (A and B) Views of two stretches of the Coma de l'Embut Creek affected by AHPs, at the Capçaleres del Ter i del Freser Natural Park (October 2017 (A) and September 2018 (B)). Note that the stream bed is white color regardless of the water flow condition. (C) Coma River, Valls del Comapedrosa Natural Park (2018). (D and E) Port Vell creek, Alt Pirineu Natural Park (September 2018). processes occur in the Pyrenean Axial Zone, including those of the to a sequence of cm-thick beds of pelites and arenites, and sporadic m- NVC analyzed in this work (Figs. 1B and 3). The exposed rocks of thick quartzite layers (Alins Formation) with dominant cleavage dipping the Variscan basement include an intensely deformed succession 30–60°N (Laumonier et al., 1996; Capellà and Carreras, 1996; Cochelin ca. 2500 m thick of Paleozoic metasedimentary formations. These et al., 2018). The lithology in the northern domain is more diverse, rocks were intruded during the late stages of the Variscan orogeny consisting of a thick rhythmic alternation of sandstones and pelites with by batholiths dominated by granitoids, which developed aureoles ferruginous beds (Sotllo Formation), intercalated microconglomerate len- of thermal metamorphism, in which most of the ore bodies occur. ticular beds, sills and felsic metavulcanites (ICGC, 2007). The dominant The NVC is located within the Alt Pirineu Natural Park (Spain) cleavage dips here around 60–90°N (Cochelin et al., 2018). The Al- (Fig. 1C). It is adjacent to the Comapedrosa Valley Natural Park (Principal- content in the pelitic beds of the Sotllo Formation is within the 13–23% ity of Andorra) and the Regional Natural Park of the Ariège Pyrenees range, which is in agreement with other Cambro-Ordovician Pyrenean (France). Here, the main structural unit is the Pallaresa anticline rocks (IGC, 2012). The N-dipping black slates and phyllites of the Lladorre (Choukroune and Séguret, 1973), which affects a Paleozoic low-grade Unit crop out at high elevation along a WNW-ESE strip between the metasedimentary sequence ca. 2500 m thick (Zandvliet, 1960; Capellà Lladorre (W) and Comapedrosa (E) peaks (Fig. 1C). This 250 m thick and Carreras, 1996). This asymmetric anticline is offset on its northern ridge-forming formation is the most relevant in terms of ARD since it in- limb by a north-dipping reverse fault system and the associated shear cludes significant volumes of sulfide-bearing rocks (IGC, 2012). zone (Mérens shear zone: MSZ), which separates two different domains The MSZ is associated with a 70 km-long reverse fault system with (Cochelin et al., 2018). Exposed rocks in the southern domain correspond an estimated vertical displacement of ca. 2000 m. It is one of the 4 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 Fig. 3. Geomorphological setting of the Noguera de Vallferrera catchment (NVC). (A) Geomorphological map showing abundant periglacial and paraglacial features developed on a relict glacial landscape. Periglacial processes and gravitational slope deformation dominate in the headwaters zone. (1) Baiau Lakes, (2) Escorbes Lakes, (3) Port Vell Lakes and (4) Aixeus Lake. (B) Cross-section across the Monteixo-Noris Range showing the interpretation of a deep-seated gravitational slope deformation (DSGSD) associated with the Mérens shear zone. Note bend in the Noguera de Vallferrera stream consistent with the northward displacement of the DSGSD. major orogen-parallel structures affecting the Variscan basement in the expression between the Baiau plain (Pla de Baiau) and the Boet Axial Zone (Cochelin et al., 2018). The MSZ runs along the axial zone of mountain pass (Port de Boet) (Figs. 1C and 3). Here, the intensely the Pallaresa anticline and the northern flank of the Monteixo-Norís sheared rocks are prone to the development of large gravitational range. This fault-weakened zone has a conspicuous geomorphic deformations. The slopes show fresh-looking uphill- and downhill- 5 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 facing scarps up to 25 m high attributable to recent or active deep- and boulders lose their whitish appearance when wetted, but it is seated gravitational slope deformation (i.e., sackung). Similar fea- recovered moments after becoming dry. This causes the stream beds tures related to gravitational faulting have been reported to the east af- to remain whitish most of the time. It should be noted that the photo- fecting rock glaciers (Àspres de Banyell, Valira Valley, Andorra) (Turú graphic flights are not carried out during rainy conditions. Once the pre- and Planas, 2005). Both, tectonic fracturing and the recent gravitational liminary mapping was validated, the AHP signs and its evolution were faulting have clear implications regarding the hydrology and ARD, since tracked along the complete historical series, including the lower res- they control the arrangement of the drainage network and contribute to olution and uncolored photographs earlier than 2000 (Fig. S1 in increase the permeability of the rock massif. Supplementary Material). Regarding the NVC landforms, glacial and periglacial features pre- Sampling stations were selected according to the results of the dominate above 2000 m a.s.l. (Fig. 2). Mass-wasting under periglacial AHPs mapping and other hydro-geomorphological criteria, in order conditions (e.g., frost shattering) has produced numerous deposits to obtain a representative model of the entire area. A total of 68 consisting of angular rock fragments (i.e., talus slopes and cones, rock water samples were collected from lakes, streams and seepages dur- glaciers, protalus ramparts), many of which are active in the present ing three different campaigns conducted in August (Baiau and Port time. Of importance for the present study are also deep-seated gravita- Vell-Pica Roja) and September (Baiau) 2018, and May 2019 (Aixeus) tional slope deformations developed on oversteepened slopes carved by (Fig. 3). Samples were collected in 15 mL PP conical tubes, acidified the glaciers and debuttressed during the deglaciation. These paraglacial to pH < 2 by addition of 5 M HNO 3 (ultrapure grade, Merck). The features have been also reported in other sectors of the Axial Pyrenees samples were preserved at T < 5–10 °C until their analysis in the lab- (e.g. Gutiérrez et al., 2005, 2008, 2012; Guerrero et al., 2013; Jarman oratory. Physicochemical parameters (i.e. temperature, pH and elec- et al., 2014; McCalpin and Corominas, 2019). In the study area, these trical conductivity-EC) were measured at the water sampling points landforms display a remarkable development in the northern by a HI9810 pH-meter and HI98192 EC-meter (by HANNA), both slope of the Monteixo-Norís range (Fig. 2). Here, the upper- equipped with temperature probes for enable compensation of the middle sector of the slopes is riddled by a dense system of fresh- pH and EC values to a reference temperature of 25 °C. The meters looking downhill- and uphill-facing scarps up to 2 km long. Some were multiple-point recalibrated prior to each sampling campaign of such uphill-facing scarps affect talus slopes, suggesting that by certified pH-standard buffers (4.01, 7.01, 9.18, 10.01 and 12.45; they may be currently active. These features have clear implication ±0.01 pH at 25 °C) and EC-standard solutions (0.00, 84.01413 and on the local hydrology and thus on ARD. 5000 μScm−1 at 25 °C). The samples were controlled and conserved in the laboratory at a 4. Methods T < 5 °C. The pH of the acidized samples was checked to remain pH < 2, ultrapure HNO3 were added if necessary, and dilution was The area was investigated by integrating geological and hydro- checked prior to the lab analysis. The total concentrations of dissolved physico-chemical data. Historical aerial photographs and orthoimages trace-metals and metalloids (Ti, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Cd, Au, from dates spanning from 1945 to 2018 with spatial resolutions up to Hg and Pb) were determined by inductively coupled plasma–mass 25 cm/px (source: Spanish National Geographic Institute-IGN and Car- spectrometry (ICP-MS) (e.g. Longerich et al., 1990; Jarvis and Jarvis, tographic and Geological Institute of Catalunya-ICGC) were used for 1992) by an Agilent 7500c spectrometer. Al, major elements (Na, K geomorphological mapping and the identification of AHPs coating and Ca), and some high-concentration trace-elements were determined stream and gully beds. Particularly valuable were the aerial photographs by inductively coupled plasma-optical emission spectrometry (ICP- taken in 1945–46 and 1956–57 (American flights A and B; Cartographic OES) by an Agilent 5100. Standard QA/QC lab protocols were observed and Photographic Air Force Center-CECAF) that, despite their lower spa- for the analysis, including blanks and initial calibration verification by tial resolution (~5 m/px and ~ 2 m/px, respectively), allowed us to ex- 6 to 9 different certified standards for each trace metal, continuing pand the observation period to almost 75 years. The comparison calibration verification every 10 samples, and dilution check. This between the photographs used for the analysis was possible since all approach enabled achieving an accuracy in the determination of the they were captured during the snow-free season, ranging from the trace concentrations with limits of detection (LoD) of (Al <5 μg/L), (Ti end of June to the ends of September, with the only exemption of <0.5 μg/L), (Cr <0.5 μg/L), (Mn <1 μg/L), (Fe <6 μg/L), (Co <0.02 μg/L), 2009's series collected in October–November. We excluded the 2017's (Ni <0.5 μg/L), (Cu <0.01 μg/L), (Zn <0.35 μg/L), (As <0.2 μg/L), (Cd series from the analysis because, despite the photographs were col- <0.05 μg/L), (Au <0.1 μg/L), (Hg <2 μg/L), (Pb <0.5 μg/L). lected at the ends of September, the land surface was partially covered Multivariate statistical analysis was performed to better un- by snow above 2200 m a.s.l.. derstand the correlations between pH, EC and trace metals con- AHPs signs were identified in orthoimages as stream sections with centrations. Covariance and correlation coefficient matrices were anomalous whitish colors, in contrast with the almost black appearance calculated for the sampling campaigns conducted in Baiau (sum- of the unaffected stretches (Fig. 4). The blackish tone of the unaffected mer and autumn 2018) and Port Vell-Pica Roja (summer 2018) sections is related to the dark color of the gravels and boulders derived sectors. The calculated correlation coefficients were analyzed to from Paleozoic rocks. Infrared colored imagery captured since 1996 assess spatial and seasonal correlation patterns. The sampling were also analyzed, at resolutions ranging from 2.5 m/px (1996) to conducted in Aixeus (spring 2019) was excluded from the analy- 0.5 m/px (since 2008) (spectral band centered at 729.45 nm (red chan- sis to avoid possible over interpretations. nel with infrared zone information), 623.82 nm (green channel with red Meteorological data was analyzed to determine temperature and zone information) and (543.27 nm blue channel with green zone infor- precipitation anomalies. The data were registered by the autono- mation), source ICGC). The stretches affected by AHPs are shown in in- mous meteorological stations of Vielha e Mijaran (1002 m a.s.l.) frared imagery in a color band that ranges from white or turquoise to (UTM N31T 319310E 4,729,700 N), Esterri d'Àneu (950 m a.s.l.) bright blues, while the unaffected stretches are shown dark green (UTM N31T 345637E 4,721,441N), and Certascan (2400 m a.s.l.) (Fig. 4). According to these criterions, a preliminary mapping of the spa- (UTM 31T 358470E 4,728,981N) (source Meteorological Service of tial distribution of AHPs was conducted analyzing the imagery younger Catalunya, Spain). The Standardized Precipitation Index for cumula- than 2012. The areas were visited during 2017–2019 to corroborate the tive precipitation values of 12 months (SPI12) (McKee et al., 1993) reliability of our mapping. Direct observations confirmed that APHs are was computed by the SPI Generator software (National Drought Mit- more striking during summer baseflow, but they are still visible during igation Center - UNL) for the historical period 1955–2017, according spring and autumn runoff, moment when the waters acquire a charac- to the meteorological data recorded by the Esterri d'Àneu meteoro- teristic turquoise color (Fig. 4). In this regard, the AHPs coating gravels logical station. 6 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 Fig. 4. Historical true-color and infrared aerial orthoimages of a stretch of the Baiau River at the Pla de Baiau plain (2170–2175 m a.s.l., centered at WGS84 UTM N31T 370220E 4,718,950 N), captured during wet (October-2009) and dry (August-2011) seasons (source Cartographic and Geological Institute of Catalunya-ICGC, Spain). Stream waters depict different color pattern depending on whether they are affected or unaffected by aluminum-precipitates (AHPs), regardless of the flow regimen. Affected waters appear turquoise in true-color and bright blue in infrared imagery, while unaffected waters look black (true-color) or dark green (infrared). Bottom photography, panoramic view of a section of the Baiau River upstream the Pla de Baiau (UTM N31T 370630E 4,718,325N; see location in Fig. 6). Note the turquoise color of the stream waters, characteristic of the rivers affected by AHPs. 5. Results seepages and lake outlets and maintain their condition until they con- verge with non-acidic ones. Some acidic springs and seepages are at ele- 5.1. Hydrochemistry and geochemistry vations as low as 1860 m a.s.l. (Mossen Batlle; pH ~ 4.8), 1650 m a.s.l. (seepage at the foot of Monteixo peak; pH ~ 5.4) and 1220 m a.s.l. (Areu The water samples were classified into three groups according to their village spring; pH ~ 5.6). Non-acidic waters are the most abundant in pH: acidic (pH < 5.6), circumneutral to basic (hereinafter so called “non- the NVC. For instance, Escorbes lakes at 2374 m a.s.l. (8.2 < pH < 10.1) acidic”) (pH > 6.5) and water mixes (5.6 < pH < 6.5) (Fig. 5 and Table S1 and Port Vell lake at 2465 m a.s.l. (pH ~ 9.5) (labeled as 2 and 3 in in supplementary material). The acidic waters (i.e., lakes, streams and Fig. 3), and the majority of the streams and creeks, with pH values typi- seeps) are ascribed to locally recharged groundwater that has flowed cally within the 6.5˗9.5 range. These are essentially surface waters that through sulfide-rich rocks or deposits and are altered by ARD. Some had none or limited subsurface circulation. Water mixes result from the lakes in the NVC headwaters as e.g. Baiau Lakes at 2480 m a.s.l. mixing at the surface of acidic and non-acidic waters. The mixing mostly (pH ~ 4.7) and Aixeus Lake (pH ~ 4.9) at 2365 m a.s.l. (labeled as 1 and initiates below 2500 m a.s.l. (e.g. Baiau cirque-Portvell, 2470 m a.s.l.; Port 4 in Fig. 3) are among the most acidic ones in the Pyrenees (Catalan Vell-Pica Roja sector, 2480 m a.s.l.) (Figs. 6 and 7). et al., 1993; Pla, 2001; Jean-Christophe Auguet and Casamayor, 2013). Water samples have low electrical conductivity (EC ~ 9 to 326 Such lakes are fed by acidic and non-acidic seepages that supply inflows μScm−1) as expected for low-mineralized mountain waters (Fig. 5 and throughout the year, in addition to surface runoff, including snow melt Table S1 in supplementary material). EC is higher in late summer and direct rainfall. The acidic gullies and streams are sourced by acidic (post-thaw baseflow) than in autumn (rainy season) and acidic waters 7 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 Fig. 5. Physico-chemical features of the water samples collected during the summer of 2018 in the Baiau (labeled as BM-) and Port Vell-Pica Roja sectors (labeled as APM-) (see additional data in Table 1 in supplementary material). Three water types are distinguished: acidic (pH < 5.6, enriched in metals), non-acidic (pH > 6.5, non-enriched in metals) and water mixes (pH ~ 5.6–6.5). Note the marked differences in the water quality of the nearby Escorbes (2370 m a.s.l.) and Baiau (2480 m a.s.l.) lakes (Fig. 7B), despite their similar elevation and orientation on the north-facing side of the ridge, and bedrock lithology. generally have higher EC than the non-acidic ones, although with a Al mainly occurs as Al3+ (aq) in acidic waters (pH <5.0˗5.6) (e.g. difference below one order of magnitude. EC is also above the average Takeno, 2005), but where they mixes with non-acidic waters, the acid- (X EC + σ) in the non-acidic Escorbes Lakes, and in the lower part of ity of the former is buffered, and the conditions for Al precipitation are the catchment (Pla de Boet). More significant are differences in terms reached, typically in the form of whitish AHPs. The precipitation of the of metal concentrations, exceeding two orders of magnitude (×500) in AHPs in the NVC roughly initiate around 2400–2500 m a.s.l. and extents creeks located just a few meters away, despite the mineralogy of the down to ca. 1800–1700 m a.s.l.. As an example, ephemeral AHPs start to substrate in its headwaters is the same. The most relevant differences occur at ca. 200 m (2420 m a.s.l.) downstream the small Baiau Lake out- correspond to the aluminum-ion, ranging from <10 to 4890 μg/L-Al. let (2480 m a.s.l.), (Figs. 6 and 7). AHPs form where the acidic waters Overall, the concentration of minor- and trace-metals are higher in converge and start mixing with creeks of non-acidic waters, flowing acidic waters, especially aluminum that shows the broadest concentra- from the slopes on both margins of the Baiau river. Progressive mixing tion range (Fig. 5 and Table S1 in supplementary material). Aluminum, causes AHPs to grade downstream from ephemeral to discontinuous manganese, cobalt, zinc, cooper, and cadmium concentrations show and become clearly continuous just downstream of the junction with good correlation with acidity, regardless the location or the season the Escorbes creek (2325 m a.s.l.). AHPs also become discontinuous, (Fig. 8). For example, the maximum aluminum-concentrations but still evident from the Pla de Baiau (2170 m a.s.l.) down to the Boet (>2000 μg/L) are reached at pH < 4.9 (with the sole exception of one plain (Pla de Boet) (1800 m a.s.l.) (Fig. 7B). Analogous patterns have sample at pH ~ 5.2). More disparity is on the concentrations of the been observed in the Port Vell – Areste creeks (Fig. 7A), where AHPs ini- other trace metals lead, titanium, arsenic, gold, and mercury regarding tiate at 2530 m a.s.l., Sotllo creek (at 2450 m a.s.l.), Medacorba and the pH. In the case of nickel, high concentrations (>30 μg/L-Ni) are in Barèitas rivers (at 2300 m a.s.l.), Vallpeguera creek (at 2510 m), and acidic waters, but also in the non-acidic waters of the Escorbes Lakes Gardelha river (at 2580 m a.s.l.) (Fig. 9). The total length of affected (33˗37 μg/L-Ni at pH = 7.9˗8.3). A particular pattern is observed for the stream stretches exceeded 35 km in 2018. iron-concentration, suggesting that aluminum and iron may have different ARD behavior –iron is found at high concentrations in a higher pH range, 5.2. Spatial-temporal patters of ARD during the period 1945–2018 with the majority of concentrations >100 μg/L-Fe reached at pH > 6. Regarding the aluminum precipitation, aluminum comes from the A detailed analysis of the spatial-temporal distribution of the AHPs chemical weathering of feldspars and phyllosilicates present in the using the 1945 to 2018 historical orthoimages reveals that the extent siliciclastic metasedimentary bedrock (e.g. micas, albite, orthoclase). of the affected areas has changed considerably throughout the past 8 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 Fig. 6. Water mixing downstream of the Baiau lakes (pH 4.9; 4890 μgL−1-AL). Aluminum-hydroxysulfates precipitation (AHPs) occurs in areas were the acidic Baiau River (Aluminum- enriched) converge with the meteoric Escorbes creek. Yellow arrow indicates the position of the panoramic view in Fig. 4. White arrow points to N (oblique view from Google Earth, centered in YTM 31 T 370320E 4,718,200 N). decades (Figs. 9 and S1 on supplementary material). Signs of AHPs were 5.3. Temperature and precipitation anomalies already observable in the American flights from 1945 to 46 and 1956–57, respectively. AHPs affected the Areste and Port Vell creeks Temperature anomalies have been observed in the local meteorolog- from 2380 m a.s.l., and the Arcalís creek down to 1890 m a.s.l.. The im- ical stations of Vielha e Mijaran (1002 m a.s.l.), Esterri d'Àneu (950 m a.s. ages suggest that AHPs were continuous in the Port Vell-Areste creeks, l.) and Certascan (2400 m a.s.l.), located respectively 50 km-W, 30 km- while they were dim and intermittent in the Arcalís creek. Evidence of W and 15 km-NNW from the NVC (source Meteorological Service of AHPs were also observable in the Noguera de Vallferrera river, ceasing Catalunya) (Figs. 10 and 11). A rough positive mean annual temperature at ca. 1 km downstream the confluence of the Port Vell-Areste and (Tmean) trend of ca. +0.36 °C/decade was recorded in Vielha e Mijaran Arcalís creeks (ca. 1700 m a.s.l.). AHPs were most likely already present station since 1980, while for 1955–1979 it was negative (−0.04 °C/de- before 1945, at least on the northern edge of the NVC. The local names of cade) (Fig. 10). The warmest years were 1997, 2006 and 2011, with some affected areas in the French eastern slope such as Étang Rouge (Tmean) exceeding +1.8σ the average value (XT), and 2014–2016, during (red lake) and Riou Blanc (white river) (Fig. 9) indicate that the phe- which (Tmean) was constantly above (XT + 1.6σ). Regarding precipita- nomenon must have been frequent since ancient times. Noticeable is tion anomalies, the standard precipitation Index for cumulative pre- that the total length of affected streams in 1945–46 were just around cipitation values of 12 months (SPI12) computed for the historical 5.1 km, and that no AHPs were observable above 2380 m a.s.l. in the period 1955–2017 (Esterri d'Àneu station data) reveals that severe Areste-Port Vell creeks and above 2150 m a.s.l. in the Vallpeguera to extreme (−1.5 < SPI12 < −2.5) and exceptional (SPI12 > −2.5) creek, which are elevations 200 to 360 m lower than the current maxi- droughts were rare between 1955 and 1980, while they have been mum elevation of the AHPs in the same areas. frequent in the last 3 decades, being especially noticeable the one Time-lapse historical series since 1945–46 allow tracking the evolu- that preceded 2012 (Fig. 10). tion of the AHPs up to the present (Fig. 9). No significant differences are observed in the 1956 images, compared to those of the previous decade. Nevertheless, the images younger than 1956 show an overall expansion 6. Discussion trend, both in the total affected lengths and in the elevation at which the AHPs initiate. The AHPs expansion was not constant over time, but show 6.1. Trace metals in waters peaks in 1997–98, 2007, and a maximum in 2012. In 1997–98, AHPs ex- tended along the Noguera de Vallferrera river downstream to the con- Chemical analysis of waters reveals that, in some areas of the NVC, fluence with the Sotllo creek, and affected D'Aixeus and Mercat creeks concentrations for potentially toxic trace metals such as Al, Cd, Co, Cu, for the first time, from elevations of 2350 and 2180 m a.s.l., respectively. Ni or Zn largely exceed quality standards for aquatic ecosystems (e.g. In 2007, AHPs spread to the streams located on the slopes south of Pla de GBC, 2018) and drinking-water laws (e.g. World Health Organization Boet and Pla de Baiau (up to 2150 m a.s.l.) and north (2350 m a.s.l.) of (WHO); U.S. Environmental Protection Agency (EPA); Directive 98/83/ Pla de Baiau. AHPs covered drainage sections with a total length of CE). Some spatial and temporal patterns regarding pH, EH and trace 38.3 km in 2012, which is the maximum for the whole record period; metals concentrations correlations are also observed (Fig. 8). The corre- and the AHPs elevation scaled up to 2300–2580 m a.s.l. in all the catch- lations between trace metals and acidity and EC are higher in Port Vell- ments (Table S2 in supplementary material). AHPs have continued to be Pica Roja than in Baiau sectors. The correlations in Baiau are higher in widespread during the following period (2013–2018), but their extent summer than in autumn. The correlations with EC are moderate to and maximum elevation have not exceeded the maximum recorded low in Baiau, while they are high in Port Vell-Pica Roja. Differences be- in 2012. tween Port Vell-Pica Roja and Baiau likely are because the acidic streams 9 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 10 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 It should be noted that the analysis of the higher-resolution colored images, coupled with the infrared images, has offered very consistent results when compared with the infield mapping observations. Further- more, such observations are also consistent with the climatic patterns recorded instrumentally during the same time span, and a relationship between the local effects of climate warming (i.e. temperature increas- ing and variation in the precipitation regimen) and the ARD aggravation can be established. In this regard, one of the most critical periods to sup- port our analysis, such as the period 2007–2012, is covered by high- resolution colored and infrared imagery. This leads us considering the interpretations for the entire period as reasonable approximations. On the other hand, this approach attempts providing information that would not otherwise be possible, considering the gap of knowledge on historical water quality data in areas where, as in the NVC, no controls have been carried out in the past due to its remoteness. The results also indicate that the acidification and metal enrichment of the analyzed waters are essentially caused by natural ARD processes in the NVC. AMD does not seem to be a likely origin because, although extensive mining activities have been carried out in the Vallferrera area in the past, to our knowledge extraction sites were always at eleva- tions below 2400–2500 m a.s.l.. Some atmospheric contribution of airborne metals, either anthropogenic or natural, cannot be ruled out (Bacardit and Camarero, 2009; Camarero et al., 2009a; Camarero et al., 2017; Corella et al., 2017, 2018). However, the fact that different water masses in the same area show highly heterogeneous metal con- centrations (e.g. Baiau vs. Escorbes and Port Vell lakes) point to the po- Fig. 8. Multivariate correlation coefficients of the correlations between pH - electrical tential atmospheric input is very limited compared to the geogenic conductivity (EC) and concentration of trace metals, for the campaigns conducted in Baiau (BA) and Port Vell-Pica Roja (PV) sectors in summer and autumn 2019. contribution. The elevated metal concentrations recorded in the area also corroborates it. Thus, the projections that foresee a progressive acidification recovery of the surface waters due to the decrease in indus- in Port Vell-Pica Roja are recharged by acidic seepages discharging from trial activity, which peaked during the 1960s to 1980s in this region (e.g. the Pica Roja slopes (Fig. 7A); while in Baiau, acidic waters are both in Camarero et al., 2009b), are not applicable in the high-elevation NVC. lakes as in streams, which are sourced by the Baiau lakes' outlets Climate has been invoked to exert a strong control on ARD (Furniss (Fig. 7B). More uncertainty is regarding the stational pattern observed et al., 1999; Plumlee et al., 1999), and climate warming to be the cause in Baiau. Here, the correlation with trace metals is lower in summer of recent water quality deterioration in high-mountain areas worldwide than in autumn, despite acidity and EC are higher in summer, indicating (Nordstrom, 2009; Todd et al., 2012; Manning et al., 2013; Ilyashuk lower dilution of waters altered by ARD during the dry season. It should et al., 2014, 2018; Zaharescu et al., 2016). Variations in precipitation and be noted that the summer campaign was a bit more extensive than the temperature have several effects on ARD. On the one hand, a decrease autumn campaign, which hampered the direct comparison (Fig. 5 and in precipitation amount may cause (a) less dilution of solute-rich flows Table S1 in supplementary material); and only two campaigns do not that feed streams and lakes; and (b) increasing of the fresh rock surface seem sufficient to conduct a rigorous statistical analysis. Moreover, in exposed to oxidation due to water table decline. On the other hand, an in- addition to pH and EC variation, other processes can influence the corre- crease in air temperature causes (c) a reduction in snow and ice covers; lations such as biological activity and flow regimen. Thus, the observed (d) ground-ice thawing from permafrost, including detrital glacial and correlations should be taken with caution and only general trends be periglacial deposits (e.g. moraines, rock glaciers, talus slopes), which considered. Further research is required to better understand the distri- have been related to ARD (Williams et al., 2006; Todd et al., 2012); and bution of dissolved trace metals and their seasonal patterns, which are (e) an acceleration of sulfide bio-chemical oxidation due to its positive key aspects to take into account for a correct management of this correlation with temperature (Ahonen and Tuovinen, 1991; Schoonen hydro-environmental problem. et al., 2000), especially in the case of the oxidizing action of acidophilic ar- chaea, bacteria, and photosynthetic algae (Baker and Bandfield, 2003; 6.2. Climatic control Nordstrom, 2011a; Cruz-Hernández et al., 2012). Furthermore, once trig- gered, sulfide oxidation is an extremely exothermic reaction that can The results point towards ARD processes has become more wide- elevate sufficiently the temperature on the groundwater-ore oxidation spread and severe in the NVC in recent decades. This hypothesis has interphase even at low temperatures, making oxidation processes even been verified since 2005 by direct observations (IGC, 2012), and it has more efficient (Bigham and Nordstrom, 2000; Gault et al., 2015). also been supported by the space-time analysis of the 1945–2018 his- The main AHPs peaks in the NVC occurred in 1998, 2007 and 2012, torical aerial imagery. We are aware on the limitations of this approach, being the latter the most pronounced one to date. Noteworthy correla- especially were the analysis basses on the lower-resolution and unco- tions between the temporal AHP peaks and sharp precipitation and lored orthoimages younger to 2000. However, extending the observa- temperature anomalies may be established (Figs. 10 and 11). This sup- tion beyond 2000, although it partially compromises the resolution, ports that the recent warmer and dryer conditions have enhanced has the potential to offer a more representative longer-term diagnosis. ARD processes. Systematic positive trends in temperature (Tmax = Fig. 7. Geomorphological features and signs of Aluminum-hydroxysulfates precipitation (AHP) in the (A) Port Vell-Pica Roja and (B) Baiau sectors (UTM N31T). Acid and moderately-acid waters (pH < 5.6) are derived from the Pica Roja slopes and the acidic Baiau lakes. AHPs initiate downstream of the acidic Pica Roja seepages and Baiau lakes and change from ephemeral to continuous in the Port Vell creek and downstream of the confluence between the Baiau river and the Escorbes creek. Acidic waters drain from the north-facing slopes of the Pica Roja ridge and the Baiau cirque, with extensive periglacial deposits, while waters from the Port Vell and Escorbes lakes areas, with less abundant periglacial deposits, are circumneutral to moderately basic, supporting a geomorphological control on AHPs. 11 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 Fig. 9. Map showing the temporal (1945 to 2018) and spatial evolution of the areas affected by natural Aluminum-compounds precipitation (AHP) in the Noguera de Vallferrera catchment (NVC). The temporal pattern indicates that AHPs have significantly spread since 90s, with precipitation peaks in 1997–98, 2007 and a maximum in 2012. The total length of affected streams and creeks has increased from ca. 5 km (1945–56) to 38.3 km (2012). Noticeable is that the elevations at which AHPs initiate have climbed ca. 100 to 400 m, reaching up to 2580 m a.s.l., which matches with the elevation of the potential mountain permafrost limit in the Pyrenees (Serrano et al., 2009; Oliva et al., 2016). +0.57 °C/decade; Tmin = +0.23 °C/decade) and less significant negative In addition to thermal anomalies, AHP peaks also occurred after pre- precipitation trends (ca. −3%/decade to −5%/decade), with greater dis- cipitation negative anomalies. The SPI12 calculated from the Esterri persion along the mountain range, have been identified in the central d'Àneu station data for the historical period 1955–2017 reveals that se- Pyrenees since the 1970s–1980s (Pérez-Zanón et al., 2017; OPCC-CTP, vere to extreme and exceptional droughts have been more frequent in 2018). Similar temperature anomalies have been observed in the the last 3 decades. All AHP peaks occurred within the dry period local meteorological stations of Vielha e Mijaran (1002 m a.s.l.), (Fig. 10). The more local, but shorter precipitation data from Certascan Esterri d'Àneu (950 m a.s.l.) and Certascan (2400 m a.s.l.), which station also confirm that a drought period preceded the 2007 peak may be reasonably representative of the climate in NVC. Vielha (Fig. 11), but this is not so evident for the 2012's peak, the most intense e Mijaran is the only station with a long temperature dataset event observed to date. While at regional scale the period 2010–2012 (1955–2018), covering almost the entire time span of the historical was exceptionally dry (Fig. 10), it was not as dry in the NVC (Fig. 11). orthoimages (Fig. 10). The rough positive mean annual temperature However, it highlights that the highest thermal anomaly (+0.8 °C), only (Tmean) trend of ca. +0.36 °C/decade recorded since 1980 exceeds in exceeded in 2015, and the maximum thermal amplitude (Tmax − Tmin) 30% the world average (OPCC-CTP, 2018). Noticeable is that the was in 2012 (47.4 °C). It suggests that both low precipitation and high 1998, 2007 and 2012 AHP peaks occurred after the warmest years temperature aggravate ARD, but once triggered, temperature is probably of the record period 1997, 2006 and 2011 (Figs. 10 and S1 in supple- the main catalyzer for the phenomenon. This hypothesis also supports mentary material). why chemically unstable compounds such as AHPs are still very patent 12 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 Fig. 10. Standardized Precipitation Index for a time-scale of 12 months (SPI12) for the period 1955–2017 based on the meteorological data collected at the Esterri d'Àneu station (UTM N31T 345637E 4,721,441N), and monthly average minimum temperature (Tmin) at the Vielha station (UTM N31T 319310E 4,729,700N) for the same period (source: Meteorological Service of Catalunya). A change in trend both in precipitation (negative) and temperature (positive) is observed since 1980s. The main Aluminum precipitation (AHPs) peaks detected in the Noguera de Vallferrera Catchment (NVC) occur after drought periods, and especially high-temperature episodes. despite precipitation has increased largely since 2012 (e.g. SPI24 for Sep- 6.3. Elevation-dependent warming and elevation range of ARD 2017 to Sep 2019 was > +3, source: Meteorological Service of Catalunya). A likely cause is that above-average temperatures were more common in Results also suggest that climate warming plays some role in the ex- recent years and the average temperature stands up on historical maxi- pansion of the elevation range at which ARD occurs. We postulate that mums of the last hundred years (OPCC-CTP, 2018). the rise in elevation of the edge of the periglacial and permafrost realms Fig. 11. Meteorological data for the period 2003–18 from Certascan automated station (15 km-NNW away from the Noguera de Vallferrera Catchment, UTM 31T 358470E 4,728,981N) (source Meteorological Service of Catalunya). The graph shows a good correlation between the low-precipitation period of 2003–2007 and the Aluminum- precipitation (AHPs) peak in 2007. The relationship with precipitation in the case of the 2012-peak is not as clear, while it coincides with a thermal amplitude peak. Note also that the average temperature in the 2011 is almost the maxima of the instrumental record. 13 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 is the main cause of the upward expansion of the AHPs in the NVC in re- recharge zones (i.e. Cambro-Ordovician rhythmic sequence of schists cent decades. Here, the presence and extent of the permafrost is not well and slates). An ARD model frequently invoked in mineralized catch- documented. However, based on geomorphological indicators and tem- ments is that shallow subsurface flows derived from the melting of perature data, the existence of permafrost above 2500–2700 m a.s.l. is the ground ice of rocky periglacial deposits cause the acidification of plausible in this area (Serrano et al., 2009; Oliva et al., 2016), especially high-elevation waters (e.g. Thies et al., 2007; Zaharescu et al., 2016; in the north-facing slopes of the Monteixo-Norís, Comapedrosa, Pica Santolaria et al., 2017; Ilyashuk et al., 2018). It is because their open- Roja and Areste peaks (Fig. 9). Noticeable is that the most acidic waters work structure, which may host groundwater, interstitial ground ice occur in the Baiau Cirque, Port Vell-Areste and Aixeus areas, where sur- and even solid ice lenses, favors the water-rock interaction and it con- face waters are sourced from the slopes of these peaks with a dominant trols the groundwater and oxygen flows in areas dominated by low- north aspect. permeability bedrock (Jódar et al., 2017; Jones et al., 2019). All the acidic Climate warming is accelerating changes in the hydrological and waters in the NVC drain from areas where extensive rocky periglacial cryospheric dynamics in alpine catchments (Pepin et al., 2015). One of deposits occur. This pattern is observed in areas such as the Pica Roja the most relevant effects of the elevation-dependent warming is that gla- (Port Vell) and Baiau cirque sectors (Fig. 7). Here, acidic and metal- cial and periglacial boundaries are retreating dramatically in the last few enriched seepages drain only from the extensive rock glaciers, talus decades. Generally, where mean annual air temperatures (MAAT) remain slopes and cones located along the north aspect slopes. On the contrary, below +3 °C, ground ice and permafrost are expected to persist for some the Port Vell and Escorbes lakes, where the periglacial deposits are periods throughout the year, which is considered as a periglacial environ- much scarcer, are not acidic, despite they are located nearby of the ment (French, 2011, 2018). In areas where MAAT remains below −2 °C, Pica Roja acidic seepages and Baiau acidic lakes. In the case of the Port groundwater persists as ground ice, which causes Fe2+ biochemical oxi- Vell lake, just some talus slopes of limited extent are found in its re- dation to slow down and oxygen available for the reaction to drop se- charge zone, which would explain their lower acidity. Rocky periglacial verely. Conversely, in periglacial belts where MAAT ranges between deposits are more powerful in the Escorbes recharging zone (Escorbes −2 °C and + 3 °C, the moisture and oxygenation conditions of the rock peak and Portella de Vallpeguera slopes) (Fig. 7B), but not as much as massif vary throughout the year, allowing sulfide oxidation (Ilyashuk in the Pica Roja and Baiau sectors. This could justify that, despite not et al., 2018). The MAAT recorded by the Certascan automated weather being so acidic, their waters show significant concentration in trace station ranged between +1.5 and + 3.7 °C during the period 2003 to metals such as nickel. 2018, with an average of +2.8 °C (σ = 0.6 °C). This means that the That model that relates rocky detrital deposits with ARD fits well for periglacial limit is located within the elevation range of ca. 2400 to the headwaters of the NVC, but it does not explain why acidic ground- 2600 m a.s.l.. Note that 10.5 km2 of the study area are situated above water is also discharging via seepages and springs at the foot of the 2600 m a.s.l., which is almost the 10% of the total. Moteixo-Norís range, at lower elevations (1860–1220 m a.s.l.), where The average rising rate of the AHPs upper limit observed in the rocky periglacial landforms are no longer found. It should be noted Areste, Baiau, Vallpeguera and Sotllo creeks over the 1945–2012 period that the Moteixo-Norís range is affected by a sackung-type DSGSD is around 50–60 m/decade, and 45–55 m/decade extending the period structure (McCalpin and Irvine, 1995; Cruden and Varnes, 1996; to 2018. Although we are aware of the significant limitations of the Soldati, 2004; Pánek and Klimes, 2016), related to the Mérens shear data derived from the interpretation of the historical orthoimages, the zone (Fig. 3). Sackungs are slow-moving complexes that displace sev- observed general trends are quite consistent for the different catch- eral hm3 and reach several hundred meters in depth (Soldati, 2013; ments, supporting that they may provide a reasonable approximation Gutiérrez et al., 2012; Pánek and Klimes, 2016). These large-scale grav- to the phenomenon. In this sense, noteworthy is that the rising rates itational deformations cause the fracturing of rock massifs and deep cir- are closely in agreement with the rise of the 0 °C isotherm estimated culation of fluids (e.g. Gutiérrez et al., 2005, 2008). We propose that the for the region at 46 m/decade, considering a temperature rise of ca. Monteixo-Norís sackung favors ARD processes (Fig. 12) due to the fol- 0.23 °C/decade and a vertical thermal gradient of ca. 0.5 °C/100 m lowing factors: (1) it forms enclosed depressions and uphill-facing (Lampre, 2001). This value is comparable to those reported for the scarps that contributes to enhance surface water infiltration; (2) it cre- European Alps (up to 70 m/decade) (Brocard et al., 2013), underpinning ates fractures that experience significant dilation due to gravitational the relationship between elevation-dependent warming and ARD. spreading (horizontal separation) increasing the permeability of the Based on this relationship a conceptual model is proposed (Fig. 12). rock massif and the bedrock-groundwater interaction surface; (3) facil- The climate variation (i.e. warming) controls the periglacial limit migra- itates deeper and more rapid circulation of water and oxygen-rich air. tion upwards and, thus, the upper limit at which ARD processes are trig- The main groundwater discharge points in this sector are at the frontal gered and/or aggravated. However, this model does not explain why part of the sackung, at the Noguera de Vallferrera valley, which shows a some nearby lakes and creeks have such different acidity condition. clear northward deflection attributable to the progressive displacement This suggests that the model must consider other controlling factors of the DSGSD. In such hydro-geomorphological context, several moder- as the local geomorphology. ately acidic (pH ~4.8–6) seepages and springs discharge at elevations of 1600 to 1700 m a.s.l (pH ~4.8–5.6); and at 1220 m a.s.l. (Areu village 6.4. Geomorphological control spring). Noticeable is the low temperature (3.9 °C) of the Mossen Batlle spring (Pla de Boet) (Fig. 3), located at a slightly higher elevation Availability of sulfide-bearing rocks is a pre-requisite for the devel- (1860 m a.s.l.). The temperature was recorded during a water sampling opment of ARD, but not the unique condition. So we propose that the conducted in August 2018, while the other groundwaters were above geomorphological control, namely the nature, geometry, and arrange- 11 °C (Table S1 in supplementary material). This supports that the ment of the periglacial deposits (rock glaciers, protalus ramparts, spring is not related to a shallow local flow but to a deeper and probably cones and talus slopes), and the DSGSD structure that affects the fast-moving groundwater flow with an elevated recharge zone (proba- Moteixo-Norís range (Fig. 3B) partially causes the large variations in bly above 2400–2500 m a.s.l.). The Mossen Batlle spring (pH ~ 4.8) ex- the acidity of the waters, both in the headwaters and in the lower hibits high concentration of metals (e.g. 1280 μgL−1-Al), in the same parts of the catchment (Fig. 12). The hydrochemistry of the surface wa- order of magnitude as Baiau Lakes (2482 m a.s.l.) (1830–2140 μgL−1-Al, ters in the NVC headwaters show large differences over relatively lim- with an outlier of 4890 μgL−1-Al). Metal concentrations in the western ited areas, despite their analogous lithological and elevation contexts. sackung seepages are lower (e.g. 80–100 μgL−1-Al). This apparent For instance, the Baiau lakes show an acidity below 4–5 pH units com- disagreement could be due to different reasons: (a) Samplings were con- pared by the Escorbes lakes, located just 800 m to the NW and 120 m ducted in different seasons, summer 2018 for Mossen Batlle and spring below (Fig. 7B), despite there is no difference in the lithology at their 2019 for the western seepages, coinciding with a snow melt peak; 14 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 Fig. 12. Conceptual model for the different natural acid rock drainage (ARD) and Aluminum precipitation compounds (AHP) observed in the Noguera de Vallferrera catchment (NVC) controlled by climatic and geomorphological factors. Periglacial deposits and rockslides favor the development of ARD from shallow groundwater flows, while deep-seated gravitational slope deformations (DSGSDs) facilitate deeper water-rock interaction. AHPs form where acidic waters mixed with non-acidic ones. Climate warming and droughts enhance the phenomenon and the elevation at which it initiates, which has climbed hundreds of meters in recent times. The rising rates are comparable with the estimated ascending rates of the periglacial limits in the region, underpinning the proposed relationship between the effects of elevation-dependent warming and ARD intensification. (b) Pyrite-rich slates and phyllites of the Lladorre unit crops out at eleva- (NVC) during the last decade. ARD produces the precipitation of whitish tions above 2500 m a.s.l. in the northern flank of the Monteixo-Norís aluminum-compounds that strikingly cover some gullies and streams in Range (Fig. 1). This suggests that the metal-rich Mossen Batlle waters the catchment headwaters. The total length of affected streams has are recharged above this elevation, while western seepages are not. increased from ca. 5 km (1945–46) to more than 35 km (2018) (c) The residence time of the groundwater may affect its mineral enrich- and the maximum elevation at which ARDs initiate has climbed ment. Mossen Batlle spring is located at the NE edge of the structure, above 2500 m a.s.l., at rates of 45–55 m/decade. Precipitation of while the western seepages are in the frontal part of the sackung. Here, aluminum-compounds occurs where acidic (metal-enriched) wa- gravitational faulting is denser and flow paths are likely to be shorter. ters are neutralized by mixing with non-acidic ones. The concen- Unfortunately, we have no data on groundwater ages to support this tration of dissolved aluminum clearly correlates with acidity of hypothesis. waters, regardless of location or season. However, further research Climatic projections foresee a temperature increase of ca. 10 to 25% is needed to determine the patterns of seasonal variation of such with respect to the present values in the Pyrenean region for the next concentration. In addition to aluminum, waters also exhibit anom- 3 decades (Barrera-Escoda et al., 2014; Gonçalves et al., 2014), and sim- alous concentrations of other potentially toxic trace metals such as ilar trends are predicted for other mountain regions worldwide (Pepin cadmium, cobalt, copper, nickel or zinc, which largely exceed qual- et al., 2015). This future scenario leads to hypothesize that ARD pro- ity standards for aquatic ecosystems, and drinking water. The po- cesses could affect larger areas in many alpine catchments with suitable tential adverse effects on these pristine and vulnerable alpine forming rocks. Therefore, a better understanding of the factors that ecosystems and the public health are largely unknown. control ARD can be valuable at designing optimal adaptation strategies The data obtained point at there is strong climatic and geomorpho- for these vulnerable spaces. Furthermore, monitoring the evolution logical control on the ARD processes. Climate warming, markedly of ARD may provide useful data on climate change impacts, comple- sharp since 1980s, in combination with the severe droughts recorded mentary to effects such as deglaciation and permafrost thawing. in the last decade is the most plausible cause for the ARD intensification However, a comprehensive knowledge on the local geomorphological in the study area. The rise in the maximum elevation of the ARD and hydrogeological context is essential for a correct interpretation of (45–55 m/decade) is consistent with the retreat and ascent rate of the ARD patterns and to avoid potential simplifications and misleading di- periglacial limit in the region, estimated at 46 m/decade. The consequent agnoses. In this regard, the proposed conceptual model, which con- degradation of permafrost facilitates that ARD intensifies at higher ele- siders both climatic and geomorphological factors controlling the ARD vations. Geomorphological analysis reveals that pyrite-rich periglacial processes in the NVC, would be easily exportable to many other alpine rocky deposits are the main ARD sources. Their high permeability and catchments worldwide (Fig. 12). specific surface facilitate water and oxygen circulation and the bio- chemical oxidation of sulfides. Acidic groundwater and throughflows 7. Conclusions derived from such periglacial deposits cause the acidification of the creeks and lakes located above 2400–2500 m a.s.l. Conversely, in areas The historical series of aerial photographs spanning more than where these deposits are scarce, the phenomenon is less severe or 70 years (1945–2018) reveals that natural acid rock drainage (ARD) has does not manifest. Acidic seepages and springs are also at lower eleva- experienced an intensification in the Noguera de Vallferrera Catchment tions (ca. 1860 to 1220 m a.s.l.). In these cases, acidic flows occur 15 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 associated with a deep-seated gravitational slope deformation (DSGSD) Barnolas, A., Pujalte, V., 2004. La Cordillera Pirenaica. In: Vera, J.A. (Ed.), Definición, límites y división. Geología de España. Sociedad Geológica de España – IGME, Madrid, that affects the Monteixo-Norís range. Here, gravitational faulting and pp. 233–241. dilation of fractures enables the circulation of deeper oxidizing ground- Barrera-Escoda, A., Gonçalves, M., Guerreiro, D., Cunillera, J., Baldasano, J.M., 2014. Projec- water flows, permitting ARD to occur in the lower parts of the catch- tions of temperature and precipitation extremes in the North Western Mediterra- nean Basin by dynamical downscaling of climate scenarios at high resolution ment. On the basis of these results, a conceptual model integrating the (1971–2050). Climate Change 122, 567–582. different ARD hydrodynamics observed in the NVC is proposed. The Bigham, J.M., Nordstrom, D.K., 2000. Iron and aluminum hydroxysulfates from acid sulfate model proposes that, while climate warming and droughts control the waters. Rev. Mineral. Geochem. 40, 351–403. Blowes, D.W., Ptacek, C.J., Jambor, J.L., Weisener, C.G., 2004. The geochemistry of acid mine intensity of ARD and its upper limits, the local hydro-geomorphology drainage. The geochemistry of acid mine drainage. In: Lollar, B.S., Holland, H.D., governs the spatial distribution of the affected areas. Turkerian, K.K. (Eds.), Environmental Geochemistry. Treatise on Geochemistry. vol. Climatic projections foresee a sustained increase in temperature for 9. Elsevier-Pergamon, Oxford, UK, pp. 149–204. the coming decades, which could lead to the deterioration of water re- Bradley, M.W., Worland, S.C., 2015. Bibliography for acid-rock drainage and selected acid- mine drainage issues related to acid-rock drainage from transportation Activities. sources in many cold mountain regions worldwide. Better understand- Open-File Report 2015- 1016. U.S. Department of the Interior & U.S. Geological Sur- ing and monitoring of the evolution of natural ARD processes would be vey, Reston, Virginia https://doi.org/10.3133/ofr20151016 (24 pp.). useful for predicting impacts related to climate warming and designing Brocard, E., Philipona, R., Jeannet, P., Begert, M., Romanens, G., Levrat, G., Scherrer, S.C., 2013. Upper air temperature trends above Switzerland 1959–2011. J. Geophys. Res. adaptation strategies. Understanding the geomorphological contexts of Atmos. 118, 4303–4317. these alpine catchments, a factor which is frequently neglected, is an es- Camarero, L., Botev, I., Muri, G., Psenner, R., Rose, N., Stuchlik, E., 2009a. Trace elements in sential task to achieve sound diagnoses. alpine and arctic lake sediments as a record of diffuse atmospheric contamination across Europe. Freshw. Biol. 54 (2518-253). Camarero, L., Garcia-Pausas, J., Huguet, C., 2009b. A method for upscaling soil parameters CRediT authorship contribution statement for use in a dynamic modelling assessment of water quality in the Pyrenees. Sci. Total Environ. 407, 1701–1714. Camarero, L., Bacardit, M., de Diego, A., Arana, G., 2017. Decadal trends in atmospheric de- Mario Zarroca: Conceptualization, Methodology, Investigation, position in a high elevation station: effects of climate and pollution on the long-range Writing – review & editing. Carles Roqué: Conceptualization, Methodol- flux of metals and trace elements over SW Europe. Atmos. Environ. 167, 542–552. Capellà, I., Carreras, J., 1996. La zonación estructural del hercínico del Pirineo central en el ogy, Investigation, Writing – review & editing. Rogelio Linares: Concep- anticlinorio de la Pallaresa. Estud. Geol. 52, 51–56. tualization, Methodology, Investigation, Writing – review & editing. José Catalan, J., Ballesteros, E., Gacia, E., Palau, A., Camarero, L., 1993. Chemical composition of Salminci: Conceptualization, Methodology, Investigation, Writing - disturbed and undisturbed high-mountain lakes in the Pyrenees: a reference for acid- Review & Editing. Francisco Gutiérrez: Conceptualization, Methodol- ified sites. Water Res. 27, 133–141. Choukroune, P., ECORS Team, 1989. The ECORS Pyrenean deep seismic profile reflection ogy, Investigation, Writing – review & editing. data and the overall structure of an orogenic belt. Tectonics 8, 23–39. Choukroune, P., Séguret, M., 1973. Tectonics of the Pyrenees, role of the gravity and com- pression. In: de Jong, K.H., Scholten, R. (Eds.), Gravity and Tectonics. Wiley, Nueva Declaration of competing interest York, pp. 141–256. Cochelin, B., Lemirre, B., Denèle, Y., de Saint Blanquat, M., Lahfid, A., Duchêne, S.M., The authors declare that they have no known competing financial 2018. Structural inheritance in the Central Pyrenees: the Variscan to alpine tectonometamorphic evolution of the axial zone. J. Geol. Soc. Lond. 175, 336–351. interests or personal relationships that could have appeared to influ- Corella, J.P., Valero-Garcés, B.L., Wang, F., Martínez-Cortizas, A., Cuevas, C.A., Saiz-Lopez, ence the work reported in this paper. A., 2017. 700 years reconstruction of mercury and lead atmospheric deposition in the Pyrenees (NE Spain). Atmos. Environ. 155, 97–107. Corella, J.P., Saiz-Lopez, A., Sierra, M.J., Mata, M.P., Millán, R., Morellón, M., Cuevas, C.A., Acknowledgment Moreno, A., Valero-Garcés, B.L., 2018. Trace metal enrichment during the industrial period recorded across an altitudinal transect in the southern Central Pyrenees. Sci. This work was funded by the Fundación Biodiversidad, Ministerio Total Environ. 645, 761–772. Crouch, C.M., McKnight, D.M., Todd, A.S., 2013. Quantifying sources of increasing zinc para la Transición Ecológica (Gobierno de España) (grant Climate from acid rock drainage in an alpine catchment under a changing hydrologic regime. Change – 2017, PRCV00604). The work carried out by FG has been sup- Hydrol. Process. 27, 721–733. ported by project CGL2017-85045-P of the Ministerio de Ciencia e Cruden, D.M., Varnes, D.J., 1996. Landslide types and processes. In: Turner, A.K., Schuster, Innovación (Gobierno de España). M.Z. has a Serra Húnter fellowship R.L. (Eds.), Landslides: investigation and mitigation: Transportation Research Board, U.S. National Academy of Sciences, Special Report. 247, pp. 36–75. at the Autonomous University of Barcelona. We thank the support of Cruz-Hernández, P., Carrero, S., Sarmiento, A.M., Pérez-López, R., Cánovas, R.C., Nieto, Marc Garriga, Director of the Parc Natural de l'Alt Pirineu, Eric Peramos J.M., 2012. ¿Por qué mis ríos son rojos? Hidroquímica de un ecosistema único. and Blai Rosés for their technical assistance during the field work, and Comunicaciones del XVII Simposio sobre Enseñanza de la Geología, Huelva, pp. 344–348. the anonymous reviewers and editor for their insightful comments Du, X., Wang, Y., Su, S., Li, J., 2009. Influences of pH value on the microstructure and phase and suggestions, which greatly improved the manuscript. transformation of aluminum hydroxide. Powder Technol. 192, 40–46. French, H.M., 2011. Periglacial. In: Singh, V.P., Singh, P. (Eds.), Haritashya (Encyclopedia of snow, ice and glaciers). Springer Dordrecht, the Netherlands, pp. 827–841. Appendix A. Supplementary data French, H.M., 2018. The Periglacial Environment. 4th ed. John Wiley & Sons, Chichester, UK 544 pp. Supplementary data to this article can be found online at https://doi. Furniss, G., Hinman, N.W., Doyle, G.A., Runnells, D.D., 1999. Radiocarbon-dated ferricrete org/10.1016/j.scitotenv.2021.146070. provides a record of natural acid rock drainage and paleoclimatic changes. Environ. Geol. 37, 102–106. Galván, L., Olías, M., Cerón, J.C., Fernández de Villarán, R., 2021. Inputs and fate of contam- inants in a reservoir with circumneutral water affected by acid mine drainage. Sci. References Total Environ. 762, 143614. García-Sansegundo, J., 1996. Hercynian structure of the axial zones of the Pyrenees: the Ahonen, L., Tuovinen, O.H., 1991. Temperature effects on bacterial leaching of sulfide min- Aran Valley cross-section (Spain-France). J. Struct. Geol. 18, 1315–1325. erals in shake flask experiments. Appl. Environ. Microbiol. 57, 138–145. Gault, K.B.F., Gammon, P., Fortin, D., 2015. A geochemical characterization of cold-water Akcil, A., Koldas, S., 2006. Acid mine drainage (AMD): causes, treatment and case studies. natural acid rock drainage at the Zn–Pb XY deposit, Yukon, Canada. Appl. Geochem. J. Clean. Prod. 14, 1139–1145. 62, 35–47. Bacardit, M., Camarero, Ll., 2009. Fluxes of Al, Fe, Ti, Mn, Pb, Cd, Zn, Ni, Cu, and As in GBC (Government of British Columbia Ministry of Environment & Climate Change monthly bulk deposition over the Pyrenees (SW Europe): The influence of meteorol- Strategy British Columbia), 2018. British Columbia Approved Water Quality Guide- ogy on the atmospheric component of trace element cycles and its implications for lines for Aquatic Life. Summary Report, Victoria, Canada, Wildlife & Agriculture high mountain lakes. J. Geophys. Res.-Atmos. 114, G00D02. (36pp). BAIC, 2019. Butlletí annual d’Indicadors Climàtics 1950-2018. Meteorological Service of Giddings, L.A., Chlipala, G., Kunstman, K., Green, S., Morillo, K., Bhave, K., 2020. Character- Catalunya 85 pp. URL:. https://static-m.meteo.cat/wordpressweb/wp-content/up- ization of an acid rock drainage microbiome and transcriptome at the Ely Copper loads/2018/05/29153739/BAIC2017_Resum_executiu.pdf (late access November Mine Superfund site. PLoS One 15 (8), e0237599. 2019). Godt, J., Scheidig, F., Grosse-Siestrup, C., Esche, V., Brandenburg, P., Reich, A., Groneberg, Baker, B.J., Bandfield, J.F., 2003. Microbial communities in acid mine drainage. FEMS D.A., 2006. The toxicity of cadmium and resulting hazards for human health. J. Occup. Microbiol. Ecol. 44, 139–152. Med. Toxicol. 1, 22. 16 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 Gonçalves, M., Barrera-Escoda, A., Guerreiro, D., Baldasano, J.M., Cunillera, J., 2014. Sea- Muñoz, J.A., 1992. Evolution of a continental collision belt: ECORS-Pyrenees crustal bal- sonal to yearly assessment of temperature and precipitation trends in the North- anced cross-section. In: McClay, K.R. (Ed.), Thrust Tectonics. Chapman and Hall, Western Mediterranean Basin by dynamical dowscaling of climate scenarios at high London, pp. 235–246. resolution (1971–2050). Climate Change 122, 243–256. Nielsen, F.H., 1990. New essential trace elements for the live sciences. Biol. Trace Elem. Guerrero, J., Gutiérrez, F., Carbonel, D., Bonachea, J., García-Ruiz, J.M., Galve, J.P., Lucha, P., Res. 26-27, 599–611. 2013. 1:5000 landslide map of the upper Gállego Valley (central Spanish Pyrenees). Nordstrom, D.K., 2009. Acid rock drainage and climate change. J. Geochem. Explor. 100, Journal of Maps 8, 484–491. 97–104. Gutiérrez, F., Acosta, E., Rios, S., Guerrero, J., Lucha, P., 2005. Geomorphology and geochro- Nordstrom, D.K., 2011a. Hydrogeochemical processes governing the origin, transport and nology of sackung features (uphillfacing scarps) in the central Spanish Pyrenees. Geo- fate of major and trace elements from mine wastes and mineralized rock to surface morphology 69, 298–314. waters. Appl. Geochem. 26, 1777–1791. Gutiérrez, F., Ortuño, M., Lucha, P., Guerrero, J., Acosta, E., Coratza, P., Piacentini, D., Soldati, Nordstrom, D.K., 2011b. Mine waters: acidic to Circumneutral. Elements 7, M., 2008. Late quaternary episodic displacement on a sackung scarp in the central 393–398. Spanish Pyrenees. Secondary paleoseismic evidence? Geodin. Acta 21, 187–202. Nordstrom, D.K., Alpers, C.N., 1999. Geochemistry of acid mine waters. In: Plumlee, G.S., Gutiérrez, F., Linares, R., Roqué, C., Zarroca, M., Rosell, J., Galve, J.P., Carbonel, D., 2012. In- Logsdon, M.J. (Eds.), The Environmental Geochemistry of Mineral Deposits. Vol. 6A vestigating gravitational grabens related to lateral spreading and evaporite dissolu- (G.S). Soc. Econ. Geol., Littleton, Colorado, pp. 133–160 Economic Geology. tion subsidence by means of detailed mapping, trenching, and electrical resistivity Nordstrom, D.K., Southam, G., 1997. Geomicrobiology of sulfide mineral oxidation. tomography (Spanish Pyrenees). Lithosphere 4, 331–353. In: Banfield, J.F., Nealson, K.H. (Eds.), Geomicrobiology: Interactions Between EEA (European Environment Agency), 2009. Regional Climate Change and Adaptation: Microbes and Minerals. Mineralogical Society of America, Washington D.C., the Alps Facing the Challenge of Changing Water Resources. Report 8, 143 pp. URL. pp. 361–390. https://www.eea.europa.eu/publications/alps-climate-change-and-adaptation-2009 Oliva, M., Serrano, E., Gómez-Ortiz, A., González-Amuchastegui, M.J., Nieuwendam, A., (late access November 2019). Palacios, D., Pellitero-Ondicol, R., Pérez-Alberti, A., Ruiz-Fernández, J., Valcárcel, M., Huss, M., Hock, R., 2018. Global-scale hydrological response to future glacier mass loss. Vieira, G., 2016. The periglaciation of the Iberian Peninsula. Quat. Sci. Rev. 137, Nat. Clim. Chang. 8, 135–140. 176–199. ICGC (Institut Cartogràfic i Geològic de Catalunya), 2007. Mapa geològic comarcal de Ca- Olmedo, P., Pla, A., Hernández, A.F., Barbier, F., Ayouni, L., Gil, F., 2013. Determination of talunya 1:50.000, Pallars Sobirà. toxic elements (mercury, cadmium, lead, tin and arsenic) in fish and shellfish sam- IGC (Institut Geològic de Catalunya), 2012. Les pàtines blanques de la capçalera de la ples. Risk assessment for the consumers. Environ. Int. 59, 63–72. Noguera de Vallferrera. Composicions, processos i efectes associats. GAO-003/12 OPCC-CTP, 2018. (The Pyrenees Climate Change Observatory), 2019. El cambio climático (89 p. Unpublished report). http://parcsnaturals.gencat.cat/web/.content/home/alt_ en los Pirineos: impactos, vulnerabilidades y adaptación. Bases de conocimiento para pirineu/coneix-nos/centre_de_documentacio/fons_documental/biblioteca_digital/ la futura estrategia de adaptación al cambio climático en los Pirineos, Madrid, Spain hidrologia/les_patines_blanques_a_la_cap_alera_de_la_noguera_de_vallferrera/3les_ 150 pp. https://www.opcc-ctp.org/sites/default/files/documentacion/opcc-informe- patines_blanques_de_la_cap_alera_de_la_noguera_de_vallferrera.pdf (Late access es-print.pdf (late access April 2020). 01/26/2021). Pánek, T., Klimes, J., 2016. Temporal behavior of deep-seated gravitational slope deforma- Ilyashuk, B.P., Ilyashuk, E.A., Psenner, R., Tessadri, R., Koinig, K.A., 2014. Rock glaciers out- tions: a review. Earth Sci. Rev. 156, 14–38. flows may adversely affect lakes: lessons from the past and present of two neighbor- Pepin, N., Bradley, R.S., Diaz, H.F., Baraer, M., Caceres, E.B., Forsythe, N., Fowler, H., ing water bodies in a crystalline rock watershed. Environ. Sci. Technol. 48, Greenwood, G., Hashmi, M.Z., Liu, X.D., Miller, J.R., Ning, L., Ohmura, A., Palazzi, E., 6192–6200. Rangwala, I., Schöner, W., Severskiy, I., Shahgedanova, M., Wang, M.B., Williamson, Ilyashuk, B.P., Ilyashuk, E.A., Psenner, R., Tessadri, R., Koinig, K.A., 2018. Rock glaciers in S.N., Yang, D.Q., 2015. Elevation-dependent warming in mountain regions of the crystalline catchments: hidden permafrost-related threats to alpine headwater world. Nat. Clim. Chang. 5, 424–430. lakes. Glob. Chang. Biol. 24, 1548–1562. Pérez-Zanón, N., Sigró, J., Ashcroft, L., 2017. Temperature and precipitation regional climate Jarman, D., Calvet, M., Corominas, J., Delmas, M., Gunnell, Y., 2014. Large-scale rock slope series over the Central Pyrenees during 1910-2013. Int. J. Climatol. 37, 1922–1937. failures in the eastern Pyrenees: identifying a sparse but significant population in Pla, S., 2001. Chrysophycean cysts from the Pyrenees. Bibl. Phycol. 109, 1–179. paraglacial and parafluvial contexts. Geografiska Annaler, Series A, Physical Geogra- phy 96, 357–391. Plaza, F., Wen, Y., Perone, H., Xu, Y., Liang, X., 2017. Acid rock drainage passive remedia- tion: potential use of alkaline clay, optimal mixing ratio and long-term impacts. Sci. Jarvis, I., Jarvis, K.E., 1992. Inductively coupled plasma-atomic emission spectrometry in Total Environ. 15, 572–585. exploration geochemistry. J. Geochem. Explor. 44, 139–200. Plumlee, G.S., Smith, K.S., Montour, M.R., Ficklin, W.H., Mosier, E.L., 1999. Geologic con- Jean-Christophe Auguet, J.C., Casamayor, E.O., 2013. Partitioning of Thaumarchaeota pop- trols on the composition of natural waters and mine waters. In: Filipek, L.H., ulations along environmental gradients in high mountain lakes. FEMS Microbiol. Ecol. Plumlee, G.S. (Eds.), The Environmental Geochemistry of Mineral Deposits. Part B: 84, 154–164. Case Studies and Research Topics. v. 6B. Society of Economic Geologists, Reviews Jennings, S.R., Neuman, D.R., Blicker, P.S., 2008. Acid Mine Drainage and Effects on Fish in Economic Geology, pp. 373–432 Chapter 1. Health and Ecology: a Review. Reclamation Research Group Publication, Bozeman, MT (26 pp). Qian, G., Li, Y., 2019. Acid and metalliferous drainage–a global environmental issue. Jour- nal of Minning & Mechanical Engeneering 1, 1–4. Jódar, J., Cabrera, J.A., Martos-Rosillo, S., Ruiz-Constán, A., González-Ramón, A., Lambán, L.J., Herrera, C., Custodio, E., 2017. Groundwater discharge in high-mountain water- Radić, V., Bliss, A., Beedlow, A.C., Hock, R., Miles, E., Cogley, J.G., 2014. Regional and global sheds: a valuable resource for downstream semi-arid zones. The case of the Bérchules projections of twenty-first century glacier mass changes in response to climate sce- River in Sierra Nevada (southern Spain). Sci. Total Environ. 593–594, 760–772. narios from global climate models. Clim. Dyn. 42, 37–58. Jones, D.B., Harrison, E., Anderson, K., Whalleyc, W.B., 2019. Rock glaciers and mountain Rezaie, B., Anderson, A., 2020. Sustainable resolutions for environmental threat of the acid hydrology: a review. Earth Sci. Rev. 193, 66–90. mine drainage. Sci. Total Environ. 717, 137211. Kwong, Y.T.J., Whitley, G., Roach, P., 2009. Natural acid rock drainage associated with Rosell, J., Linares, R., 2001. Grandes deslizamientos en el frente de la lámina cabalgante del black shale in the Yukon territory. Canada. Appl. Geochem. 24, 221–231. Montsec (Sierra del Montsec, Prepirineo Central). Rev. Soc. Geol. Esp. 14, 255–268. Lacelle, D., Doucet, A., Clark, I.D., Lauriol, B., 2007. Acid drainage generation and seasonal Salerno, F., Rogora, M., Balestrini, R., Lami, A., Tartari, G.A., Thakuri, S., Tartari, G., 2016. recycling in disturbed permafrost near Eagle Plains, northern Yukon Territory, Glacier melting increases the solute concentrations of Hilalayan glacial lakes. Envirn. Canada. Chem. Geol. 243, 157–177. Sci. Technol. 50, 9150–9160. Lampre, F., 2001. Clima de alta montaña y sistemas morfoclimáticos fríos en el macizo de Sánchez-España, J., 2007. The behavior of Iron and aluminum in acid mine drainage: spe- la Maladeta (Pirineo aragonés). Treballs de la Societat Catalana de Geografia 52, ciation, mineralogy, and environmental significance. In: Letcher, T.M. (Ed.), Thermo- 195–231. dynamics, Solubility and Environmental Issues. Elsevier Science, Amsterdam, The Laumonier, B., Abad, A., Alonso, J.L., Baudelot, S., Bresiére, G., Besson, M., Bouquet, C., Netherlands, pp. 137–150. Bourrouilh, R., Brula, P., Carreras, J., Centéne, A., Courjault-Rade, P., Courtessole, D., Sánchez-España, J., Yusta, I., Diez, M., 2011. Schwertmannite and hydrobasaluminite: a re- Fauconnier, D., García-Sansegundo, J., Guitard, G., Moreno-Eiris, E., Perejón, A., evaluation of their solubility and control on the iron and aluminum concentration in Vizcaïno, D., 1996. Cambro-Ordovicien. Synthèse géologique et géophysique des Pyr- acidic pit lakes. Appl. Geochem. 26, 1752–1774. énées. 1. Édition Bureau de Recherches Géologiques et Minières (France)-Instituto Sánchez-España, J., Yusta, I., Burgos, W.D., 2016. Geochemistry of dissolved aluminum at Tecnológico y Geominero de España (BRGM-ITGE), pp. 157–209. low pH: Hydrobasaluminite formation and interaction with trace metals, silica and Longerich, H.P., Jenner, G.A., Fryer, B.J., Jackson, S.E., 1990. Inductively coupled plasma- microbial cellsunder anoxic conditions. Chem. Geol. 441, 124–137. mass spectrometric analysis of geological samples: a critical evaluation based on Santolaria, Z., Arruebo, T., Pardo, A., Rodríguez-Casals, C., Matesanz, J.M., Lanaja, F.J., Urieta, case studies. Chem. Geol. 83, 105–118. J.S., 2017. Natural and anthropic effects on hydrochemistry and major and trace ele- Manning, A.H., Verplanck, P.L., Caine, J.S., Todd, A.S., 2013. Links between climate change, ments in the water mass of a Spanish Pyrenean glacial lake set. Environ. Monit. As- water-table depth, and water chemistry in a mineralized mountain watershed. Appl. sess. 189 (7), 324. Geochem. 37, 64–78. Sarmiento, A.M., DelValls, A., Nieto, J.M., Salamanca, M.J., Caraballo, M.A., 2011. Toxicity McCalpin, J., Corominas, J., 2019. Postglacial deformation history of sackungen on the and potential risk assessment of a river polluted by acid mine drainage in the Iberian northern slope of Pic d'Encampadana, Andorra. Geomorphology 337, 134–150. Pyrite Belt (SW Spain). Sci. Total Environ. 409, 4763–4771. McCalpin, J.P., Irvine, J.R., 1995. Sackungen at the Aspen Highlands ski area, Pitkin County, Schoonen, M., Elsetinow, A., Borda, M., Strongin, D., 2000. Effect of temperature and illu- Colorado. Environ. Eng. Geosci. 1 (3), 277–290. mination on pyrite oxidation between pH 2 and 6. Geochem. Trans. 1, 1–23. McKee, T.B., Doesken, N.J., Kleist, K., 1993. The relationship of drought frequency and du- Serrano, E., Morales, C., González Turba, J.J., Martín, R., 2009. Cartografía del permafrost de ration of time scales. Eighth Conference on Applied Climatology. American Meteoro- montaña en los Pirineos españoles. Finisterra XLIV 87, 45–54. logical Society, Jan17–23, 1993, Anaheim, CA, pp. 179–186. Soldati, M., 2004. Deep-seated gravitational slope deformation. In: Goudie, A.S. (Ed.), Mertz, W., 1981. The essential trace elements. Science 213 (4514), 1332–1338. Encyclopedia of Geomorphology. vol. I. Routledge, New York, pp. 226–228. 17 M. Zarroca, C. Roqué, R. Linares et al. Science of the Total Environment 778 (2021) 146070 Soldati, M., 2013. Deep-seated gravitational slope deformation. In: Bobrowsky, P.T. (Ed.), Tuffnell, S., 2017. Acid Drainage: the Global Environmental Crisis You've Never Heard Of. Encyclopedia of Natural Hazards. Encyclopedia of Earth Sciences Series. Springer, https://theconversation.com/acid-drainage-the-global-environmental-crisis-youve- Dordrecht, pp. 151–155. never-heard-of-83515. Takeno, N., 2005. Atlas of Eh-pH Diagrams. INTERCOMPARISION Of Thermodinamic Data- Turú, V., Planas, X., 2005. Inestabilidad de vertientes en los valles del Valira. Datos y bases. National Institute of Advanced Industrial Science and Technology, Geological dataciones para el establecimiento de una cronología, posibles causas. Andorra y Alt Survey of Japan Open File Report No. 419, Japan 285 pp. Urgell (Pirineos orientales). VI Simposio Nacional sobre Taludes y Laderas Inestables, Talukdar, B., Kalita, H.K., Baishya, R.A., Basumatary, S., Sarma, D., 2016. Evaluation of ge- Valencia (Abstracts). netic toxicity caused by acid mine drainage of coal mines on fish fauna of Simsang Williams, M.W., Knauf, M., Caine, M., Liu, F., Verplanck, P.L., 2006. Geochemistry and River, Garohills, Meghalaya, India. Ecotoxicol. Environ. Saf. 131, 65–71. source water of rock glacier outflow. Colorado Front Range. Permafr. Periglac. Process. Teixell, A., Muñoz, J.A., 2000. Evolución tectono-sedimentaria del Pirineo meridional 17, 13–33. durante el Terciario: Una síntesis basada en la transversal del Río Noguera- Zaharescu, D., Hooda, P., Burghelea, C., Polyakov, V., Palanca-Soler, A., 2016. Climate Ribagorçana. Rev. Soc. Geol. Esp. 13, 251–264. change enhances the mobilization of naturally occurring metals in high altitude envi- Thies, H., Nickus, U., Mair, V., Tessadri, R., Tait, D., Thaler, B., Psenner, R., 2007. Unexpected ronments. Sci. Total Environ. 560–561, 73–81. response of high alpine lake waters to climate warming. Environ. Sci. Technol. 41, Zaharescu, D.G., Hooda, P.S., Fernandez, J., Palanca-Solera, A., Ionela-Burgheleaa, C., 2009. 7424–7429. On the arsenic source mobilisation and its natural enrichment in the sediments of a Todd, A.S., Manning, A.H., Verplanck, P.L., Crouch, C., McKnight, D.M., Dunham, R., 2012. high mountain cirque in the Pyrenees. J. Environ. Monit. 11, 1973–1981. Climate-change-driven deterioration of water quality in a mineralized watershed. Zandvliet, J., 1960. The geology of the upper Salat and Pallaresa Valleys, Central Pyrenees, Environ. Sci. Technol. 46, 9324–9332. France - Spain. Leidse. Geol. Meded. 33, 191–254. 18