Abandoned Tailings Deposits, Acid Drainage and Alluvial Sediments Geochemistry, in the Arid Elqui River Basin, North-Central Chile

Journal of Geochemical Exploration 115 (2012) 47–58

Contents lists available at SciVerse ScienceDirect

Journal of Geochemical Exploration

journal homepage: www.elsevier.com/locate/jgeoexp

Abandoned tailings deposits, acid drainage and alluvial sediments geochemistry, in the arid Elqui River Basin, North-Central Chile

Jorge Oyarzún a, Daniela Castillo a, Hugo Maturana a, Nicole Kretschmer b, Guido Soto c, Jaime M. Amezaga d, Tobias S. Rötting d,1, Paul L. Younger e, Ricardo Oyarzún a,b,⁎ a Departamento Ingeniería de Minas, Universidad de La Serena, Benavente 980, La Serena, Chile b Centro de Estudios Avanzados en Zona Aridas (CEAZA), Universidad de La Serena, Av. Raúl Bitrán s/n, La Serena, Chile c Centro del Agua para Zonas Aridas y Semiáridas de América Latina y del Caribe (CAZALAC), Benavente 980, La Serena, Chile d School of Civil Engineering and Geosciences, Newcastle University, Newcastle upon Tyne, NE1 7RU, UK e Newcastle Institute for Research on Sustainability, Devonshire Building, Newcastle University, Newcastle upon Tyne, NE1 7RU, UK article info abstract

Article history: Two major pollutant sources related to hydrothermal ore deposits and mining operations exist in the Elqui river Received 7 November 2011 basin, Chile: (a) acid drainage from Andean epithermal El Indio Au–Ag–Cu–As district and nearby hydrothermal Accepted 23 February 2012 alteration zones, and (b) diffuse sediment dispersion from abandoned tailings deposits in usually dry creeks in Available online 2 March 2012 the western belt of the basin. This work analyses the contribution of both sources to the current metal contents of the ﬁne grained sediments of the rivers and creeks of the Elqui basin, including a group of chemical elements Keywords: and data analysis techniques not considered in previous works carried out in the area. Analysis of “active sedi- Acid drainage ” Active sediments ments (i.e., sediments in permanent contact with surface water) in the main channel and tributaries of the Hydrothermal alteration Elqui river reveals that both pollutant sources contribute to their exceptionally high Cu contents (between 0.1 Tailings and 0.2% in the minus 60 mesh fraction). However, As pollution (0.03%) is mainly derived from the El Indio CAMINAR district. Potentially toxic heavy metals (notably Cd, Pb, Hg and Mo) are present in low concentrations and do not represent major threats to ecology or human health. Nevertheless, ongoing erosion of abandoned tailings deposits may result in soil contamination and thus be detrimental to the export-oriented agriculture of the Elqui basin. Consequently, remediation of that source should be prioritized. © 2012 Elsevier B.V. All rights reserved.

1. Introduction amalgamation of gold while the sulfide minerals were dressed for the flotation stage. As mining expanded and the number of mines multiplied, Metallic mining in Chile, mainly for Ag, Cu and Au, thrived in the a large number of tailing deposits were left behind in many places on the 1830s, after the establishment of independence from the Spanish alluvial plains of the rivers and creeks throughout the whole basin. This Crown in 1818. In the Elqui river basin, located in the Coquimbo region, occurred in practice with almost no regulation at all, and started to north-central Chile, the remains of 19th century mining include scat- change only after the Environmental law was enacted in Chile in 1994 tered slags (from the process of smelting sulfide minerals) and piles of (De la Maza, 2001; Newbold, 2006). low-grade or barren rocks near old Cu and Ag mines, with normally Much of this material has been already eroded during the episodic low contents of toxic metals like As and Cd and barely affected by leach- winter floods affecting rivers and normally dry creeks. However, over ing processes due to the dominant arid climate (Oyarzún, 2001). The a hundred deposits still remain and are a potential source of pollution introduction of the flotation process around 1908 (Valenzuela, 2005)to- in the Elqui basin. Moreover, this diffuse contamination coexists with gether with the uptake of the “trapiche” or Chilean mill (originally used Cu, Zn and As rich acid drainage from the El Indio Au–Cu–As district for wheat grinding in Europe), allowed a significant expansion of Cu and located at the NE heads of the basin and other minor Andean sources. Cu-Au mining to exploit lower grade sulfide minerals. The trapiche was a In the early 1970s El Indio, an extremely high grade Au–Cu–As district relatively cheap and easy to install and operate device, which permitted (Jannas et al., 1999) was discovered. The main deposit was mined for some 25 years and began its closure activities in 2000. A preliminary geochemical sampling performed by Oyarzún et al. (2003) revealed ex- ⁎ Corresponding author at: Departamento Ingeniería de Minas, Universidad de La tremely high contents of Cu (over 0.1%), Zn (around 0.05%) and As Serena, Benavente 980, La Serena, Chile. Tel.: +56 51 204503; fax: +56 51 223350. (0.02%), in fine grained sediments contaminated by acid drainage. This E-mail address: [email protected] (R. Oyarzún). was followed by a second study that confirmed the previous figures. 1 Present address: Technical University of Catalonia (UPC), Department of Geotech- nical Engineering and Geosciences, Campus Nord, Modulo D2, c/ Jordi Girona 1-3, Also a gypsum-rich bed, dated to ca. 9640±40 years and containing 08034 Barcelona, Spain. abundant Cu, Zn and As (up to 1.6, 14.7 and 2.3%) was discovered in

Fig. 1. Elqui river basin and active sediments sampling locations (*only sampled in 2007; **only sampled in 2008). T: Toro and Turbio rivers; C: Claro river; E: Elqui river; N: Los Negritos creek; A: El Arrayan creek. the Turbio river valley, the Andean tributary connecting the Toro and metal contents of the fine grained sediments of the rivers and creeks Elqui rivers. This probably represents the remnant of a temporary lake of the Elqui basin. These two sources contribute under very distinct bed, and records old, natural acid drainage generation from El Indio conditions of relief and climate, but are spatially closed and act within (Oyarzun et al., 2004). the same basin, where mining and modern agriculture coexist and However, in addition to this natural polluting source, Oyarzun et al. that is representative of the arid belt of the Central Andes. Further- (2006) also documented a significant increase in the Cu and As total con- more, the current research included a group of chemical elements tents of the Toro river water (i.e., 5.6 and 0.8 mg/L), the direct receptor of and data analysis techniques not considered in previous works. the district drainage, since the very beginning of full scale mining operations at the district. The well designed and carefully executed closure plan of El Indio was effective in reducing (at least temporarily) the As 2. Material and methods content of the Toro river drainage. However, neither water acidity nor its Cu content was reduced. Moreover, they have even shown some 2.1. Study area slight increasing trends lately, to judge from monitoring data of the Chilean water authority (Espejo et al., in press; Galleguillos et al., 2.1.1. Geography, climate, hydrology and main water users of the basin 2008). This fact is explained by the unfavorable environmental context The Elqui basin is located between latitudes 29° 20′ and 30° 27′ S, of the district, i.e., highly altered and fractured rocks, hosting high sulfur at the narrowest segment of the Chilean territory, limited by the high- minerals in an underground mine with some 100 km of tunnels and a est Andean peaks (east) and the Pacific ocean (west), and subjected high hydraulic gradient (Oyarzun et al., 2007). Nevertheless, the Puclaro to arid conditions produced by the Pacific anticyclone. Just 130 km irrigation dam, built in year 2000 on the Elqui river course, has per- separate the head of the basin at the Andean peaks (attaining over formed as an effective sink for As, Cu, and Fe total contents of the river 6000 m) from the river mouth on the Pacific coast (Fig. 1). Therefore, water (Galleguillos et al., 2008). Thus, although no major mining- flow regimes in the basin's rivers are turbulent, due to their steep bed agricultural conflicts for water resources availability exist in the Elqui slopes. In addition, they have a torrential regime, controlled by snow- river basin, water pollution is a source of concern for the farmers, in fall on the Andes occurring mainly during the winter months, from particular the contamination of agricultural lands by tailings materials May to August (about 200–300 mm) followed by snow melting in transported via the irrigation channels (Oyarzún and Oyarzún, 2011). spring-summer (mainly December–January). In consequence, the Within this framework, this work analyses the contribution of Elqui river flow is at a minimum during winter (some 5.5 m3 s− 1) both sources (i.e., diffuse contamination from Cu, Zn and As rich and peaks by the end of December (about 9.6 m3 s− 1)(DGA, Cade- acid drainage from El Indio and scattered tailings deposits) to the Idepe, 2004). J. Oyarzún et al. / Journal of Geochemical Exploration 115 (2012) 47–58 49

Structurally, the Elqui basin is underlain by two main tectonic and middle reach of the Elqui river, located on the western tectonic morphological blocks of similar width, separated by a major N–S, east block of the basin, some 50 km from the coast,. In addition to the ag- dipping thrust fault near longitude 70°40′W. The western block at- ricultural, agro-industrial and mining sectors, the Elqui river system tains some 2000 m altitude at its Eastern limit and progressively de- provides drinking water to some 300,000 residents, 250,000 of creases to some 500 m at 20 km from the coast. In this part of the them in the coastal cities of La Serena and Coquimbo, 24,000 inhabi- basin, precipitation falls as rain during the winter season. However, tants in the inland city of Vicuña (next to the river, at the eastern the mean annual precipitation value is meager (about 80 mm) except limit of the western block), and about 4200 persons living in 13 during El Niño (ENSO Cycle) episodes, separated by 5 to 10 years small villages (INE, 2002). The latter are supplied with tap water by (DGA, Cade-Idepe, 2004). The eastern block corresponds to the local committees, which mainly withdraw water from shallow wells mountain heights of the Andes that attain over 6000 m. The distribu- placed close to the streams. tion of atmospheric precipitations (mainly as snow) is highly variable in this block, and few meteorological stations exist. However, it is 2.1.2. Geology, ore deposits, and mining estimated between 200 and 300 mm year− 1 in average. The Elqui river basin exhibits geological traits and metallic belts The major part of the Elqui basin surface is made up of hard rock that are characteristic of the 27° S–33° S tectonic segment of the mountain massifs, and only 0.24% of its 9645 km2 is used for agricul- Andean belt (Sillitoe, 1974). They include: a) dominant presence of tural purposes (that require both suitable land and water resources). intrusive, extrusive and volcaniclastic rocks of calc-alkalic afﬁnities, Two dams exist, both for irrigation purposes: La Laguna, on the a consequence of the oceanic tectonic plate subduction under the mountain river of the same name, with a capacity of 40 million South American plate since Upper Paleozoic times (Charrier et al., cubic meters (M m3) and the larger Puclaro dam (200 M m3) in the 2007; Oyarzún, 2000); b) a series of Mesozoic–Cenozoic intrusive

Fig. 2. Non-active sediments and tailings materials sampling locations (*denote tailings deposits where several individual samples were taken and their concentrations averaged). M: Marquesa creek; N: Los Negritos creek; A: El Arrayan creek; S: Santa Gracia creek; L: Altovalsol town. 50 J. Oyarzún et al. / Journal of Geochemical Exploration 115 (2012) 47–58 A Cu Mo As 6

4 4

3 2 2

0 0 0 0 800 1600 2400 3200 0.5 2.5 4.5 6.5 8.5 0 150 300 450 Zn Cd Pb 8 4 6

2 4 3 Frequency

0 0 0 0 600 1200 1800 2400 0.0 1.6 3.2 4.8 6.4 8.0 9.6 16 20 24 28 32 36 40 Hg S Sb 6 6 6

3 3 3

0 0 0 0.00 0.08 0.16 0.24 0.32 0.0 0.2 0.4 0.6 0.8 0.5 1.5 2.5 3.5 4.5 5.5 6.5 Concentrations B Cu Mo As 6 6 4

3 3 2

0 0 0

0 00 00 00 00 00 .0 .8 .6 .4 .2 .0 .8 .6 .4 0 50 00 50 0 0 1 2 3 4 4 5 6 10 20 30 40 50 1 3 4 Zn Cd Pb 6 4 6

3 2 3 Frequency 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 2 3 4 5 6 0 0 0 0 0 20 40 60 80 00 20 40 60 80 3 6 9 12 1 1 1 1 1 Hg S Sb 4 6 9

6 2 3 3

0 0 0

0 4 8 2 6 0 4 8 .0 .2 .4 . 6 .8 .5 .5 .5 .5 .5 .5 .0 . 0 .0 .1 .1 .2 .2 .2 0 0 0 0 0 0 1 2 3 4 5 0 0 0 0 0 0 0 0 Concentration

Fig. 3. Frequency distributions for active sediments in 2007 (Panel A) and 2008 (Panel B).

and volcanic rocks in N–S belts that become progressively younger Quaternary volcanic activity, both traits being present north and eastward; c) strong faulting, with normal faults dominant in the south of this tectonic segment; and e) the presence of several N–S western block and thrust faults in the eastern one (e.g., the Vicuña metallic belts, approximately coincident with the magmatic belts. Be- E- dipping thrust fault, that lifts up the eastern, Andean block of the tween 29°20′S and 30°27′S they include: Kiruna-type Fe, Cu–Fe–Au Elqui basin); d) the absence of a N–S tectonic valley and of and Cu–Au deposits in the coastal belt; vein, “manto” type and J. Oyarzún et al. / Journal of Geochemical Exploration 115 (2012) 47–58 51 C Cu Mo As 10 10 10

5 5 5

0 0 0 0 0 0 0 0 0 0 0 0 0 2 4 6 8 0 2 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 2 4 6 8 0 0 6 2 8 6 12 18 24 3 3 4 4 1 Zn Cd Pb 16 20 20 8 10 10

Frequency 0 0 0 0 0 0 0 0 0 0 0 2 4 6 8 0 2 4 6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ...... 6 2 8 4 0 6 2 8 4 4 8 2 0 4 8 0 0 0 0 1 1 1 1 1 1 2 3 4 4 5 1 16 2 2 2 3 Hg S Sb 40 6 10

20 3 5

0 0 0

0 5 0 5 0 5 0 5 0 .0 .2 .4 . 6 .8 .5 .5 .5 .5 .5 .5 .5 .5 . 5 .5 . 5 . 0 .2 .5 .7 .0 .2 .5 .7 . 0 0 1 2 3 4 0 1 2 3 7 8 9 0 0 0 0 0 1 1 1 1 2 4 5 6 1 Concentrations

D Cu Mo As 6 8 20 3 4 10

0 0 0 0 0 0 0 0 0 0 0 0 2 4 6 8 0 2 4 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 5 0 5 0 5 0 5 2 4 6 8 1 1 2 2 3 3 Zn Cd Pb 20 20 10 10 10 5

Frequency 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 4 6 8 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 0 5 0 5 0 5 0 5 1 2 3 5 1 1 2 2 3 3 4 4 4 Hg S Sb 10 10 30

5 5 15

0 0 0

.0 .5 .0 .5 .0 .5 .0 .5 .0 .0 .5 .0 .5 . 0 .5 .0 .5 0 10 20 0 0 0 0 0 0 0 1 1 2 2 3 3 4 0 0 1 1 2 2 3 3 3 4 5 6 7 Concentration

Fig. 4. Frequency distributions for non-active sediments (Panel C) and tailings deposits (Panel D). porphyry type Cu deposits, vein Ag and Au deposits and stratiform minerals and minor Pb and As) creeks, but only the latter attains eco- Mn ores in the central part of the basin, and Au–Ag–Cu–As deposits nomic importance, due to the presence of the Talcuna Cu–Mn district, in the Andean tectonic block, close to the border with Argentina. that has been mined since the 1880s (Boric, 1985). Ore deposits in the Current extractive activities include iron mining in the El Romeral district crop out both at the Marquesa creek, as well as at its afﬂuent Kiruna type deposit (magnetite, minor apatite and pyrite), some the Las Cañas creek, and include vein type Cu and stratiform Cu and 21 km northeast of the city of La Serena and the Elqui river mouth, Mn deposits. The Cu ores contain unusual Pb as well as As, and are but outside the Elqui watershed boundaries (Oyarzun et al., 2003; transitional to the Ag ores of the nearby Arqueros district (Oyarzun Squeo et al., 2006). Copper is mined at the Santa Gracia (chalcopyrite et al., 1998; Reyes, 1991). Current mining in the district is performed and bornite) and Marquesa (chalcopyrite and bornite plus Mn by three small to medium size companies that have a total production 52 J. Oyarzún et al. / Journal of Geochemical Exploration 115 (2012) 47–58

Table 1 high sulﬁdation enargite-pyrite and gold-quartz mineralization with − 1 Statistics summary (all elements in mg kg except S, in %) for sediments and tailings alunite and a number of minor sulfophile metals like Zn, Sb, and Bi deposits (n: number of samples). Worldwide average contents are taken from Sparks (Jannas et al., 1999). Also, Cu grades (about 5%) and As (contained (1995) (p. 24–25). in enargite) were high. The latter became the problematic side of Cu Mo As Sb Pb Zn Cd Hg S the district, for the serious risk involved in term of occupational AS-07 (n=14) health and environment, as the rich ores needed to be roasted in Minimum 30 1 12 1 23 118 0.3 0.05 0.005 order to eliminate the arsenic content, a process which gave rise to Maximum 3030 8 341 5 39 2280 6 0.32 0.62 high air pollution. In addition to As pollution, drainage water in the Median 1520 2 153 2 26 382 1.3 0.12 0.255 Average 1420 2.6 133 2.2 28 492 1.6 0.13 0.284 district is contaminated by acid drainage rich in Cu, Zn, Fe and SO4, SD 977 1.9 95.9 1.3 6 543 1.6 0.07 0.17 a consequence of oxidation of the highly sulﬁdic mineral paragenesis, but also enhanced by the extreme grade of advanced argillic alter- AS-08 (n=14) ation and fracturing that affect the igneous rocks of the district Minimum 32 0.5 6 1 15 125 0.3 0.04 0.13 (Jannas et al., 1999). As previously stated, this contamination pre- Maximum 4230 6 347 4 101 1630 4.4 0.23 0.75 Median 1450 2 129 3 29 438 1.6 0.09 0.29 dated human activity, but was increased by the underground mining Average 1900 2.2 141 2.8 40 522 1.6 0.1 0.3 activities, and the closure operations performed in the district have SD 1720 1.4 112 1 25 403 1.1 0.05 0.16 not been entirely effective in stopping the acid drainage generation process (Galleguillos et al., 2008). Currently, the Vacas Heladas pros- NAS (n=19) Minimum 127 0.5 5 1 15 64 0.3 0.05 0.04 pect at El Indio belt is assessed in order to start a hep-leaching oper- Maximum 4710 13 84 10 516 2460 1.6 1.94 4.55 ation by a joint venture international group. Median 700 2 16 3 32 169 0.3 0.29 0.16 Average 1090 4.2 23.3 3.1 71 320 0.3 0.52 0.86 2.2. Materials, sampling and analysis SD 1140 4.1 20.9 2.1 116 541 0.3 0.59 1.47

TM (n=11) 2.2.1. Samples collection Minimum 66 1 5 1 2 45 0.3 0.17 0.12 The study included the sampling of fluvial sediments interacting Maximum 3320 13 678 63 426 4420 7.4 3.1 3.48 with the water flow of the Elqui river and its tributaries (hereafter Median 1360 3 39 4 59 160 0.3 0.39 0.43 generally termed “active sediments”, AS) at 21 locations (Fig. 1), Average 1480 3.6 94 7.4 104 550 1.1 0.84 0.85 SD 1030 3.1 170 15.1 122 1080 1.8 0.96 0.9 both in October 2007 and May 2008 (hereafter termed the Spring Worldwide average 33 2 7.7 1.2 19 95 0.17 0.19 0.22 and Autumn campaigns respectively). However, due to the turbulent content of river and variable river flow conditions, just 15 of the 21 locations pre- sediments sented adequate fine active sediments in both periods. Therefore, 4 AS: active sediments; NAS: non-active sediments; TM: tailings material. (E12b*, N15a*, N15b*, and A15c*) and 2 (E12a**, E15d**) locations were sampled only in 2007 and 2008, respectively. Sampling of these sediments was performed according to the European proce- of some 40 t day− 1 of copper concentrates, grading about 30% Cu and dures described in Oyarzún et al. (2003), and simultaneously water 800 g t− 1 Ag. These operations generate some 1200 t day− 1 of tail- pH measurements were performed using a Hanna HI-9828 multipara- ings materials. Although much of the tailings deposits accumulated meter probe. Also, 21 samples of sediments from the alluvial plains of by the mining operations of the district have been already eroded normally dry creeks of the western block of the basin (here termed and partly incorporated to the Elqui river fluvial sediments (e.g., ”non active sediments”, NAS), were taken in October 2007, as well as 2 M t of tailings materials were eroded by a flood in 1997), several 13 composite samples of tailings materials (TM) deposited in the flood- M t still remain. Also, operational accidents have resulted in several plains of the same creeks. The geographic distribution of samples is pre- spills, including the pollution of irrigation channels with 12,000 m3 sented on Figs. 1 and 2, labeled according to the river or creek to which of tailings materials in 2002 (Galleguillos, 2004). they belong. The NAS samples were taken from several points along a In the Andacollo area (Oyarzun et al., 1996), Cu mining started in line parallel to the creek axes, in order to attain a better representation the 1950s. It was followed by a medium scale heap leaching opera- of the metallic contents of the sediments. Also, TM samples were taken tion, Carmen, that mined about 60 M t of supergene ore (0.8% from the top and lateral slopes of the deposits and their chemical vari- grade) of a porphyry copper deposit, during the 1996–2011 period, ations evaluated. After drying at room temperature, the AS samples, as and was recently expanded to mine about 400 M t hypogene ore re- well as those of NAS and TM, were sieved under 64 μm for in this frac- serves (0.4% Cu) of the same deposit, at a rate of 60,000 t day− 1. Its tion (silt and clays) concentrates heavy metals (CENMA, DGA, 2010), mineralogy includes chalcopyrite-pyrite plus supergene chalcocite and sub-sampled by systematic splitting. with molibdenite and minor gold contents. Currently, Carmen is producing some 900 t day− 1 of Cu concentrates (30% Cu) from its new 2.2.2. Sample chemical analysis and data processing operation plus 50 t of copper cathodes from the remaining of the su- Chemical analysis (Atomic Absorption) for As, Cd , Cu, Hg, Mo, Pb, pergene ores. The whole operation involves the accumulation of some Sb, and Zn, were first performed at the Geoanalítica Laboratory in 60,000 t day−1 of tailings materials. Besides, about 90 t of gold had al- Coquimbo, a laboratory that had already been involved in the previ- ready been mined from surrounding Au-deposits in the Andacollo ous studies referred to above. The sediments and tailings deposits district by 1996, when a medium scale operation began, which was samples were digested in hot aqua regia (3:1 HCl:HNO3), followed active until 2006 at a rate of 16,000 t day− 1, grading 0.7 g t− 1 Au, by dissolution with HCl (25%) which leaves behind a silica-only resi- and producing around 4 t year− 1 of gold. Although mining has almost due. The method was satisfactory for Cu, Zn, As and Pb (100% of the ceased, gold recovery from the heap leaching piles continues. These data over the detection limit, i.e., 10 ppm for Cu, Pb, and Zn, and piles contain some 60 M t of processed materials. In addition, the 5 ppm for As). However, only about 55% of Cd, 40% of Mo, 22% of Andacollo district is full of scattered tailings deposits generated by Hg, and 3% of Sb data were above detection limit. In consequence, the small and artisanal mining activities. the atomic absorption method was complemented by ICP-AES deter- Finally the El Indio Au–Cu–As district (Au-minerals, Cu-sulfosalts), minations. This complementary analytical method was applied to 14 discovered in the early 1970s and mined for some 25 years until its samples of active sediments of the 2007 and 2008 campaigns respec- closure in 2000, was famous for its exceptional Au grades (attaining tively, in addition to 19 samples of non active sediments and to 17 over 200 g t-1 in the “direct shipping ores”). It contains complex composite ones from 11 different tailings deposits (it was not TM NAS AS, 08 AS, 07 1000 2000 3000 4000 5000 Concentration (ppm) 1000 2000 3000 4000 5000 1000 2000 3000 4000 5000 1000 2000 3000 4000 5000 10000 10000 0 0 0 0 1000 1000 100 100 10 10 1 1 uM sZ dP gS S Sb Hg Pb Cd Zn As Mo Cu uZ bAs Pb Zn Cu uZ bAs Pb Zn Cu 10 12 14 10 12 14 10 12 14 10 12 14 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 i.6. Fig. 200 400 600 800 200 400 600 800 200 400 600 800 200 400 600 800 nr-alnsdpstcmoiinlvraiiy(:M5 :N6 :N7 :N8 :Q-9). E: N-8; D: N-7; C: N-6; B: M-5; (A: variability compositional deposit Intra-tailings 0 0 0 0 .Oazne l ora fGohmclEpoain15(02 47 (2012) 115 Exploration Geochemical of Journal / al. et Oyarzún J. D A i.5. Fig. 1000 2000 3000 4000 5000 1000 2000 3000 4000 5000 1000 2000 3000 4000 5000 1000 2000 3000 4000 5000 10000 10000 o lt igasfrtenn tde elements. studied nine the for diagrams plots Box 1000 1000 0 0 0 0 100 100 10 10 1 1 uZ bAs Pb Zn Cu uZ bAs Pb Zn Cu 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 100 200 300 400 500 600 100 200 300 400 500 600 100 200 300 400 500 600 100 200 300 400 500 600 0 0 0 0 B E 10000 – 1000 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 58 100 10 1 uZ bAs Pb Zn Cu 10 20 30 40 50 60 10 20 30 40 50 60 10 20 30 40 50 60 10 20 30 40 50 60 0 0 0 0 0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5 C 53 54 J. Oyarzún et al. / Journal of Geochemical Exploration 115 (2012) 47–58

Table 2 Table 4 Metal content ratio for different sample types. Correlation coefficients for active sediments in 2008 (bold font indicates 0.05 significance). Cu Mo As Sb Pb Zn Cd Hg S Cu Mo As Sb Pb Zn Cd Hg AS/NAS (average) 1.53 0.56 5.71 0.83 0.41 1.58 4.43 0.22 0.34 AS/NAS (median) 2.12 1.00 8.81 0.83 0.88 2.43 5.80 0.34 1.69 Mo 0.49 TM/NAS (average) 1.36 0.84 4.09 2.47 1.46 1.72 3.14 1.62 0.99 As 0.61 0.64 TM/NAS (median) 1.94 1.50 2.44 1.33 1.84 0.95 1.00 1.34 2.69 Sb 0.51 0.44 0.77 AS/TM (average) 1.12 0.67 1.46 0.34 0.28 0.92 1.41 0.14 0.34 Pb −0.27 0.41 −0.12 0.04 AS/TM (median) 1.09 0.56 3.62 0.63 0.47 2.56 5.80 0.12 0.32 Zn 0.78 0.56 0.48 0.07 −0.19 Cd 0.69 0.52 0.39 0.05 −0.02 0.87 AS: active sediments; NAS: non-active sediments; TM: tailings material. Hg −0.04 −0.11 −0.20 −0.04 −0.03 −0.26 −0.03 S −0.23 0.41 0.44 0.65 0.56 −0.10 −0.01 0.16 possible to analyze Sb, Cd, Mo, and Hg in the whole set of original samples due to the lack of enough material conserved after the first analytical procedure). The ICP-AES determinations of these four ele- aspect. This is also the case for all the elements in the non-active sed- ments and sulfur were performed at the ALS Chemex Mineral Division iments and tailings samples. Laboratory in Coquimbo, also using aqua regia sample digestion. De- Table 1 presents a general summary of the basic statistical param- tection limits were 0.5 ppm for Cd; 0.01 ppm for Hg; 1 ppm for Mo; eters for the group of nine elements analyzed in sediments and tail- 2 ppm for Sb; 100 ppm for S. Standard QA/QC procedures (i.e., blanks ings materials. Worldwide average figures for river sediments have and duplicates) were performed by the laboratory. been included for comparison purposes. This information is comple- Data processing started with rather simple, standard statistical mented with a graphic box-plot representation (Fig. 5). Of particular analysis methods (central tendency and dispersion parameters). In note are the unusually high Cu, Zn, and As contents, when compared addition to the determination of ordinary statistical parameters, mul- to the worldwide averages, as well as to other rivers in the country tivariate statistical methods were used. Given that some of the statis- (Segura et al., 2006). These results are also consistent with those of tical techniques require normally distributed data (e.g., Thyne et al., previous studies (Oyarzun et al., 2004, 2003). 2004; Yidana et al., 2008) our analysis started with the Anderson- For AS, they are mainly explained by the contributions of acid Darling and Ryan-Jones normality tests (MINITAB, 2008). Although drainage from El Indio district and from neighbouring hydrothermal the number of samples is rather low to establish definitive conclu- alteration zones, through the Toro-Turbio rivers, although the Inca- sions, the majority of the distributions obtained may be assimilated guaz river also contributes important Cu and Zn loads. A comparison to the lognormal type. Spearman's rho correlation coefficient, a non of the average contents of the spring and autumn samples of active parametric test suitable for not normal distributed data, was calculat- sediments shows that Cu, Zn and As, the main elements contained ed (Kottegoda and Rosso, 2008) and a Q-mode hierarchical cluster in acid drainage from El Indio district (Oyarzun et al., 2004, 2003), analysis (HCA) was performed, on log-transformed and standardized, are higher in autumn, whereas the rest of the elements exhibit little i.e., z-scale transformed variables, in order to avoid misclassification difference. This fact may be explained considering that the autumn due to differences in data dimensionality (Chandra et al., 2006; peak concentrations are a consequence of the oxidation and dissolu-

Shrestha and Kazama, 2007), using MINITAB software. Ward linking tion of sulfide minerals like enargite (Cu3AsS4) during the summer, method and Euclidean distance were used as measures of similarity, when temperature is higher and liquid water is available. for they produce an efficient samples group classification (Güler et The “Andean” metallic source is complemented by the contribution al., 2002; Thyne et al., 2004; Yidana et al., 2008). The clustering (i.e., of the metallic districts of the western tectonic block through the dry linkage distance) was conducted following the Sneath's index of creeks that tribute to the Elqui river and is intermediated by the tailings

Dlink/Dmax b2/3 (Astel et al., 2007; Shrestha and Kazama, 2007).This materials deposited on these creeks. In fact, due to the narrow alluvial analysis was performed for the major four elements related to ore plains of the dry creeks and to the abundant tailings material deposited deposits in the basin which are also those presenting the higher con- on their surfaces, in particular on the Marquesa creek, their sediments tents in the samples analyzed: Cu, Zn, Pb, and As. are polluted to variable degrees and their metal contents reflect that of the tailings deposits. Except for the case of Andacollo, where different types of deposits coexist, the tailings materials are mineralogically and 3. Results and discussion chemically homogeneous; likewise the ore bodies from which they come. This homogeneity is shown in Fig. 6, which represents the inter- 3.1. General analysis of the data nal chemical variability in five tailings deposits (two or three composite samples were taken in each deposit). Figs. 3 and 4 present the frequency distribution histograms for the A comparison of the relative enrichment of metals in both sources concentrations of the nine elements in both types of sediments and in according to the data is presented in Table 2. Considering the average tailings samples. For active sediments all the elements, except Pb in figures, AS compared to NAS are very enriched in As and Cd (by a 4 to the spring samples and Sb in the autumn ones, present a log-normal 9 fold factor), and richer in Cu and Zn (a factor of 1 to 3), but have less

Table 3 Table 5 Correlation coefficients for active sediments in 2007 (bold font indicates 0.05 Correlation coefficients for non-active sediments (bold font indicates 0.05 significance). significance).

Cu Mo As Sb Pb Zn Cd Hg Cu Mo As Sb Pb Zn Cd Hg

Mo 0.19 Mo 0.67 As 0.36 −0.07 As 0.31 −0.13 Sb 0.02 −0.18 0.48 Sb 0.68 0.34 0.42 Pb −0.21 0.47 −0.07 0.04 Pb 0.14 −0.33 0.77 Zn 0.97 0.18 0.40 −0.02 −0.15 Zn 0.63 0.28 0.79 0.61 0.56 Cd 0.55 0.00 −0.33 −0.13 0.07 0.55 Cd 0.39 −0.11 0.64 0.58 0.64 0.64 Hg −0.09 −0.37 0.02 −0.07 −0.04 −0.12 −0.19 Hg 0.78 0.71 0.18 0.53 −0.16 0.55 0.14 S −0.07 0.30 0.18 0.35 0.37 −0.06 0.20 −0.07 S 0.81 0.63 0.35 0.69 0.01 0.62 0.19 0.85 J. Oyarzún et al. / Journal of Geochemical Exploration 115 (2012) 47–58 55

Table 6 maintains the ratio for Cu, increases the enrichment in As, Sb, and Correlation coefficients for the tailings deposits materials (bold font indicates 0.05 Cd, and incorporates Zn into the group. On the other hand, the significance). lower contents of Sb, Hg, Mo, and S in active sediments are main- Cu Mo As Sb Pb Zn Cd Hg tained, indicating that these elements are mainly contributed by the Mo −0.39 tailings materials. As 0.21 −0.32 Among the Spearman's rho coefficients presented on Tables 3 to 6, Sb 0.05 −0.29 0.65 highlight the following results (over 0.7): 1. — active sediments, (a) − Pb 0.13 0.66 0.73 0.45 spring campaign: Cu–Zn; autumn campaign: Cu–Zn, As–Sb, Zn–Cd; − Zn 0.02 0.53 0.57 0.31 0.69 — – – – – – – Cd −0.06 −0.30 0.60 0.21 0.77 0.71 2. non active sediments: Cu Hg, Cu S, Zn As, Pb As, Hg S, Mo Hg −0.56 0.61 0.13 0.19 −0.19 −0.14 −0.05 Hg; and 3. — tailing materials: Pb–As, Zn–Cd, and Pb–Cd. These results S −0.63 0.66 −0.33 −0.45 −0.42 −0.26 −0.09 0.50 are also consistent with the participation of two principal metallic sources. The principal one is the El Indio enargitic Cu–Au–As mineralization district (Jannas et al., 1999), which contributes Cu, Zn, As, Sb, Pb, Hg, Mo, and S. If the medians are considered, the observed enrich- and Cd. This source is dominant in the active sediments, in particular ment in As, Cd, Cu, and Zn is enhanced. The enrichment in Pb and Mo in those of the autumn campaign, for the reasons stated before (i.e. is only moderate, while S is higher in the AS. Comparing the average active ore mineral weathering during summer). The second source metal contents of TM and NAS samples, the former are highly is the Talcuna Cu (Mn) district in Marquesa creek (Oyarzun et al., enriched in As and Cd (3 to 5 fold), fairly enriched in Cu, Zn, Pb, Sb, 1998), where the Cu ore minerals present a polymetallic affinity, and Hg (1 to 2 fold), but contain less Mo. However, taking into ac- expressed in higher Pb contents, a metal also related to As in this dis- count the fact that the data approach the log normal distributions, trict. The possibility of metal dispersion in the Talcuna district due to the comparison of the medians is more illustrative. Thus, when they acid drainage is hampered both by the dry conditions prevailing in are considered, the relations for Cu, Pb, As, Sb, and Hg stand. In ex- the western block and the low sulfur/metal ratios of this district. change, Zn and Pb exhibit similar contents in both types of samples, Therefore, the metallic contents exhibited by the non active sedi- and Mo and S are higher in TM samples. Finally, the comparison of ments of the dry creeks is mainly due to the polluting effect of mobi- AS and TM samples is particularly important, as they represent the lization of the tailings deposits located on its normally dry alluvial two principal sources of metallic pollution: the “Andean” El Indio dis- plains, either by wind erosion or by episodic rains (related to the trict and the tailings materials of the western tectonic block, respec- ENSO cycles). The contribution of these creeks is high in Cu, and in tively. Comparison of averages indicates that active sediments are the case of Marquesa, also high in Pb and Zn, but moderate to low in more enriched in Cu, As, and Cd, but less enriched in Pb, Sb, Hg, Mo, As. Regarding Sb, Cd, Hg and Mo contents, they are moderate in the and S, but have a similar Zn content. Comparison of medians basin, although some high values for Hg and Mo appear in dry creeks sediments and tailings samples. A third minor source, expressed by the correlation coefficient, corresponds to the tailing deposits generated A by low tonnage gold mining operations in the Arrayán and Negritos 14.18 creeks, that produce higher Mo and Hg contents (the latter due to Hg used in gold amalgamation). The negative Cu–Hg and Cu–S correlations are particularly interesting, and could be explained by the Cu-poor characteristic of these gold deposits, that are, in contrast, rich in pyrite 9.45 (FeS2). As a consequence, its tailings deposits exhibit a positive correlation of Hg with S (the latter, contained in pyrite) but a negative one of G1 G2 both elements with Cu. Distance 4.73 The cluster analysis performed for the active sediments of both campaigns revealed the presence of two principal groups (Fig. 7). One of them gathers samples with low to moderate metals contents, mainly from the Claro river and from the lower 0.00 part of the Elqui river, downstream the Puclaro dam. The other 1 3 2 * 8 4 6 7 * * 3 5 9 6 7 0 1 * 4 T 1 T c C 1 1 a b T T T T T 1 1 T E 5 E E E1 5 5 E E 2b 1 1 1 1 group mainly includes samples of the Turbio river and Elqui A N N E Samples river(forthelatter,upstreamofthePuclarodam).Fornon- active sediments and tailings samples, the cluster analysis also B defines two major groups (Fig. 8). A first one mainly comprises 13.91 sediments from the Marquesa creek, with a component of tailings from the Marquesa, Talca and Arrayán creeks. The second group includes non active sediments and tailings samples of the Santa Gracia and Arrayan creeks and by samples presenting lower me- 9.27 tallic contents of Marquesa creek. G1 G2 3.2. Geochemical traits of the rivers and creeks Distance 4.64 The average and median metallic and sulfur contents for AS, NAS and TM samples in each of the rivers and creeks sampled in the current study are presented in Table 7. Also, the average pH of water in contact with AS 0.00 is indicated. Special mention is merited by the following issues: 1 3 6 4 2 * 8 4 * 3 5 6 7 9 0 1 T 1 1 T T * C 1 17 * T T T T T 1 1 E E E E d E E 2a 5 The Toro river AS samples, which are the most directly affected by 1 1 E E acid drainage from El Indio district, exhibit relatively low Cu, Zn and Samples Cd contents, a fact that contrasts with the high Cu and Zn water con- Fig. 7. Clustering for active sediments in 2007 (Panel A) and 2008 (Panel B); * denotes tent of this river water (Oyarzun et al., 2006, 2003). This apparent samples only taken in 2007, ** denotes samples taken only in 2008. contradiction is explained by the fact that the low pH of water favors 56 J. Oyarzún et al. / Journal of Geochemical Exploration 115 (2012) 47–58

24.83

16.55

G1 G2 Distance

8.28

0.00 1 2 3 1 2 4 0 4 6 9 0 1 1 5 7 0 9 1 8 7 5 8 6 2 3 4 3 3 2 7* 5* 9* 1 1 2 1 2 1 1 1 1 1 1 1 1 1 1 6* 8* 1 -M -M -M -M -M -M L -M -M S S S S -M -N A -A A -A A A A A A A A -M S S -N M Q - - - - - S - S - S ------N -N - S S S M M M - - M S S S S M S S S S S S S S S S S M M M A A A T T M T M M T A A A A T A A A A A A A A A A A A A T T M M T N N N T T T N N N N N N N N N N N N N N NA N N N T T Samples

Fig. 8. Clustering for non active sediments and tailings materials samples; * denote tailings deposits where several individual samples were taken and their concentrations averaged. the presence of these elements under its ionic, soluble forms, pre- in terms of mining exploration opportunities. In contrast, the Claro venting their transference to the fine grained sediments. In contrast, river sediments present low contents of all the elements tested. as most As is mobilized under a non ionic form (like As2O3), which The Puclaro dam plays an important role as a sink for Cu, Zn and is incorporated to the alluvial sediments at low pH, it therefore, at- As, which decrease to about 1/4, 1/2 and 1/5 of those in the upper tains its highest contents in the Toro river sediments. course of the Elqui river. This role had already been established by The La Laguna river sediments present low Cu contents and mod- Galleguillos et al. (2008) for water quality, but not until now in rela- erate concentrations of the rest of the elements. In this case, low Cu is tion to sediments. Other elements are less affected or even increase in not a consequence of pH (7.6), but of the ore mineralogy of this sub the lower course. This is the case for Pb and Hg, as a consequence of basin. the inputs of the dry creeks (Marquesa and Los Negritos-El Arrayan, There are extremely high Cu and Zn contents of the Turbio river respectively) to the Elqui river sediments. sediments, as a consequence of the mixing of the Toro and La Laguna There is a striking similarity of the average Cu, Zn, Pb, As and Sb flows, resulting in neutral to slightly basic water pH, producing hy- contents of NAS and TM of the Marquesa creek. It is not possible to drolysis, precipitation and transference of both metals to the fine sed- explain this similarity in terms of lithogeochemical reasons: this sub iments. In contrast, As and Pb attain an average content between basin is dominated by andesitic rock outcrops which have only mod- those of both confluent rivers. erate (Cu, Zn) to low (Pb, As, Sb) concentrations of these elements. In The extremely high Zn and high Cd contents of the Incaguaz river addition, as stated before, acid drainage is not generated at the Tal- sediments, together with high Cu and moderate As and Mo, are not cuna district, due to mineralogical and hydrological conditions. explained by the present geological literature and could be of interest Therefore, the only reasonable explanation is the incorporation of

Table 7 Average (and medians) metallic content of sediments and tailings materials in rivers and creeks. n: number of samples; *pH of stream water in contact with sediments (Carvajal, 2009); UP: Upstream of Puclaro Dam; DP: Downstream of Puclaro Dam.

n pH* Cu Mo As Sb Pb Zn Cd Hg S

mg kg− 1 mg kg− 1 mg kg− 1 mg kg− 1 mg kg− 1 mg kg− 1 mg kg− 1 mg kg− 1 %

Active sediments Toro River 2 3.7 380 3 340 3.5 56 185 b0.5 0.09 0.69 La Laguna River 2 7.6 31 1 90 2.0 19 245 0.9 0.10 0.12 Turbio River 9 7.4 (7.3) 2920 (2900) 3.0 (3.0) 216 (207) 3.0 (3.0) 28 (26) 600 (600) 2.0 (2.0) 0.13 (0.10) 0.29 (0.30) Incaguaz River 2 7.3 2240 5.5 49 1.0 49 1950 5.2 0.05 0.17 Claro River 2 7.7 45 2.5 9 1.5 34 130 b0.5 0.18 0.17 Elqui River (UP) 5 7.8 (7.9) 2465 (2315) 1.6 (2.0) 151 (142) 2.4 (3.0) 23 (23) 515 (527) 1.4 (1.8) 0.08 (0.06) 0.2 (0.2) Elqui River (DP) 6 7.9 (8.1) 657 (429) 1.4 (2.0) 31 (24) 2.2 (3.0) 37 (29) 218 (160) 1.6 (1.3) 0.10 (0.11) 0.40 (0.42)

Non active sediments Marquesa creek 6 1330 (710) 1.3 (1.0) 45 (42) 3.8 (3.0) 167 (105) 710 (425) 0.6 (0.4) 0.25 (0.20) 0.16 (0.17) Santa Gracia creek 3 285 (271) b1.0 (b1.0) 16 (16) 2.3 (2.0) 28 (29) 105 (111) b0.5 (b0.5) 0.11 (0.11) 0.08 (0.09) Arrayán creek 9 1090 (1100) 6.4 (8.0) 12 (13) 2.6 (3.0) 27 (26) 152 (157) b0.5 (b0.5) 0.81 (0.53) 1.28 (0.26) Los Negritos creek 1 2454 13 11 5.0 21 160 b0.5 0.76 3.71

Tailings material Marquesa creek tailings deposits 6 1727 (1381) 1.5 (1.0) 57 (50) 4.3 (3.5) 198 (158) 520 (349) b0.5 (b0.5) 0.28 (0.25) 0.32 (0.31) Santa Gracia creek tailings deposits 1 2260 0.7 23 2.3 18 110 b0.5 0.10 0.08 Talca creek tailings deposits 4 2700 (2560) 2.8 (3.0) 236 (128) 4.8 (3.7) 46 (32) 104 (107) b0.5 (b0.5) 0.30 (0.30) 0.25 (0.20) Andacollo tailings deposits 6 840 (800) 6.2 (4.5) 41 (29) 2.8 (2.5) 27 (20) 675 (95) 1.4 (b0.5) 1.74 (1.80) 1.76 (1.49) J. Oyarzún et al. / Journal of Geochemical Exploration 115 (2012) 47–58 57 tailings materials to the alluvial sediments of the creek, a well docu- nearby agricultural area of La Cantera basin in terms of the future mented process, as stated previously. wind erosion of these deposits. For the Santa Gracia creek, no significant incorporation of tailings materials to the alluvial sediments could be established on the basis of geochemical data. This fact may be explained both by the wider al- Acknowledgments luvial plain of this creek and by the comparatively minor mining activity developed in the area, in comparison with that of the This contribution has been prepared in the context of the Project Marquesa creek. The different geochemical patterns of the Marquesa Catchment Management and Mining Impacts in Arid and Semiarid South- and Santa Gracia creeks also include the high positive correlation co- America (CAMINAR), partially funded by the European Commission, efficient between Cu–Zn (0.94), Cu–Pb (0.94) and Pb–Zn (0.83) of the 6th Framework Program, contract number INCO-CT2006-032539 (this Marquesa creeks sediments (that points to a single source, i.e., the article does not represent the official opinion of the European Commis- tailings materials). In contrast, those of Santa Gracia exhibit a good sion). R. Oyarzún was also supported by the Research Office of the Uni- Pb-Zn correlation (0.5), but a negative one of both elements with versidad de La Serena, Project DIULS CD093401. This work has been Cu (−1.0 and −0.5, respectively), suggesting the possible influence made as part of the Sustainable Mining Program (PROMIS) of the Depar- of two types of ore minerals. tamento Ingeniería de Minas, Universidad de La Serena. The paper The Arrayán creek connects the Elqui basin with the Andacollo sub benefited from the comments of two anonymous reviewers. basin, some 20 km southward, through the Los Negritos creek, offer- ing the opportunity to sediments and tailings materials from Anda- collo to be transported under the exceptional floods, occurring a References few times each century. Abandoned tailings deposits are abundant Astel, A., Tsakovski, S., Barbieri, P., Simeonov, V., 2007. Comparison of self-organizing in the Los Negritos creek, and sediments of El Arrayán creek are maps classification approach with cluster and principal component analysis for high in Cu and have moderate enrichment in Mo, Hg and S. large environmental data sets. Water Research 41, 4566–4578. Boric, R., 1985. Geología y yacimientos metálicos del Distrito Talcuna, IV Región de Finally, alluvial sediments of Los Negritos creek have in average Coquimbo. Revista Geológica de Chile 25–26, 57–75. three times the Cu content of the tailings materials sampled in the Carvajal, M.J., 2009. Caracterización química e isotópica del sistema hidrológico de la area (Andacollo). However, the latter have two times the Cd and cuenca de Elqui. Tesis de Licenciatura, Ingeniero Civil Ambiental. La Serena, Chile. fl four times the Zn and As contents of the sediments (an illustration CENMA, DGA, 2010. Análisis de la composición físico-química de los sedimentos u- viales y su relación con la disponibilidad de metales en agua. Tomo II: Introducción of the mineralogical diversity of the Andacollo district). y Metodología. Centro Nacional del Medio Ambiente y Dirección General de Aguas, Santiago, Chile. Chandra, U., Kumar, S., Rath, P., Bihari, B., Bhatta, D., 2006. Application of factor and cluster analysis for characterization of river and estuarine water systems—a case 4. Conclusions study: Mahanadi River (India). Journal of Hydrology 331, 434–445. Charrier, R., Pinto, L., Rodríguez, M.P., 2007. Tectonostratigraphic evolution of the An- Acid drainage from El Indio district and nearby hydrothermal al- dean Orogen in Chile. In: Moreno, T., Gibbons, W. (Eds.), The Geology of Chile. – fi The Geological Society of London, pp. 21 114. teration zones, conveyed via the Toro river, is con rmed as the prin- De la Maza, C.L., 2001. NEPA's influence in developing countries. Environmental Impact cipal source for the exceptionally high Cu and As contents in active Assessment Review 21, 169–171. sediments of the Turbio and Elqui rivers. Secondarily, the Incaguaz DGA, Cade-Idepe, 2004. Diagnóstico y clasificación de los cursos y cuerpos de agua según objetivos de calidad, cuenca del río Elqui. Dirección General de Aguas y river delivers sediments rich in Cu and Zn to the Turbio river. The Cade-Idepe Consultores en Ingeniería, Santiago, Chile. Puclaro dam, located in the middle reach of the watershed, performs Espejo, L., Kretschmer, N., Oyarzún, J., Meza, F., Núñez, J., Maturana, H., Soto, G., Oyarzo, as a sink for Cu-, Zn- and As-rich alluvial sediments. This (in addition P., Garrido, M., Suckel, F., Amézaga, J., Oyarzún, R., in press. Application of water quality indices and analysis of the surface water quality monitoring network in to its similar effect over the Cu, As and Fe content of the river water) semi-arid North Central Chile. Environmental Monitoring and Assessment doi:10. confirms findings of a previous study. 1007/s10661-011-2363-5. The normally dry creeks, confluent to the Elqui river downstream Galleguillos, G., 2004. Efectos de la actividad minera y de las obras hidráulicas en la fl the Puclaro dam, in particular the Marquesa creek, have the potential calidad de las aguas del río Elqui y de sus a uentes. Tesis de Licenciatura, Ingeniero Civil Ambiental. La Serena, Chile. to contribute Cu, Zn, As, Pb, Sb, Cd, Hg, and Mo to the Elqui river sys- Galleguillos, G., Oyarzún, J., Maturana, H., Oyarzún, R., 2008. Retención de arsénico en tem. This potential is based on the tailings deposits located on its al- embalses: el caso del Rio Elqui, Chile. Ingeniería Hidráulica en México 23, 29–36. luvial plains, which have already contaminated the creeks Güler, C., Thyne, G., Mc Cray, J., Turner, A., 2002. Evaluation of graphical and multivariate statistical methods for classification of water chemistry data. Hydrogeology sediments. Indeed, pollution events have occurred during episodic Journal 10, 455–474. floods and as a consequence of operational accidents in the Marquesa INE, 2002. XVII Censo nacional de población y VI de vivienda. Instituto Nacional de Estadísti- creek. cas. Available at: http://www.ine.cl/canales/chile_estadistico/censos_poblacion_vivienda/ fi censo2002/mapa_interactivo/mapa_interactivo.htm. (Last accessed 2011-06-20). In terms of eco-toxicological risk, the high metals contents of ne Jannas, R.R., Bowers, T.S., Petersen, U., Beane, R.E., 1999. High-sulfidation deposit types grained sediment could represent both an objective and subjective in the El Indio district, Chile. Society of Economic Geologists 7, 219–266 (Special threat in terms of the pollution of agricultural lands via the irrigation Publication). Kottegoda, N.T., Rosso, R., 2008. Applied Statistics for Civil and Environmental Engi- channels. However, of the potentially toxic elements analyzed, such neers. Blackwell Publishing. (718 pp.). as As, Pb, Hg, Sb, Cd, and Mo, only As was detected in high concentra- MINITAB, 2008 . Normality Test, ID 1167 Available at: http://www.minitab.com/es-ES/ tions. Cu and Zn, although present in high concentrations, are benefi- support/answers/answer.aspx?log=0&id=1167 (Last accessed 2010–04–15). Newbold, J., 2006. Chile's environmental momentum: ISO 14001 and the large-scale cial in moderate content and not harmful, except for persons having mining industry — case studies from the state and private sector. Journal of Cleaner metabolic handicaps. Nevertheless, considering that an important Production 14, 248–261. part of these agricultural produces is exported, the eradication of Oyarzún, J., 2000. Andean metallogenesis: a synoptical review and interpretation. In: abandoned tailings deposits should be addressed given its likely neg- Cordani,V.G.,Milani,E.J.,ThomasFhilo,A.,Campos,D.A.(Eds.),TectonicEvolu- tion of South America: 31st International Geological Congress, Rio de Janeiro, ative perception effects on foreign markets. pp. 725–753. Although no significant agricultural activity exist at the Andacollo Oyarzún, J., 2001. Algunos metales y metaloides con propiedades tóxicas o carcinogén- district, the proliferation of abandoned tailings deposits is harmful for icas: distribución natural y riesgos por contaminación en Chile. Revista Chilena de Salud Pública 5, 96–101. the air quality of the town and presents a highly negative aspect to Oyarzún, J., Oyarzún, R., 2011. Sustainable development threats, inter-sector conflicts those visitors that are sensible to environmental impacts. Besides, and environmental policy requirements in the arid, mining rich, northern Chile the high tonnage mining that is currently carried out at the district territory. Sustainable Development 19, 263–274. Oyarzun, R., Ortega, L., Sierra, J., Lunar, R., Oyarzún, J., 1996. The manto-type gold de- will accumulate several hundred millions of tailings materials over a posits of Andacollo (Chile) revisited: a model based on fluid inclusion and geologic relatively high, arid and windy area. They represent a threat for the evidence. Economic Geology 91, 1298–1309. 58 J. Oyarzún et al. / Journal of Geochemical Exploration 115 (2012) 47–58

Oyarzun, R., Ortega, L., Sierra, J., Lunar, R., Oyarzún, J., 1998. Cu, Mn, and Ag mineraliza- Segura, R., Arancibia, V., Zúñiga, M.C., Pastén, P., 2006. Distribution of copper, zinc, tion in the Quebrada Marquesa quadrangle, Chile: the Talcuna and Arqueros dis- lead and cadmium concentrations in stream sediments from the Mapocho tricts. Mineralium Deposita 33, 547–559. River in Santiago, Chile. Journal of Geochemical Exploration 91, 71–80. Oyarzun, R., Oyarzún, J., Ménard, J.J., Lillo, J., 2003. The Cretaceous iron belt of northern Shrestha, S., Kazama, F., 2007. Assessment of surface water quality using multivariate Chile: role of oceanic plates, a superplume event, and a major shear zone. Minera- statistical techniques: a case study of the Fuji river basin, Japan. Environmental lium Deposita 38, 640–646. Modelling and Software 22, 464–475. Oyarzún, J., Maturana, H., Paulo, A., Pasieczna, A., 2003. Heavy metals in stream sedi- Sillitoe, R.H., 1974. Tectonic segmentation of the Andes: implications for magmatism ments from the Coquimbo region (Chile): effects of sustained mining and natural and metallogeny. Nature 250, 542–545. processes in a semi-arid Andean basin. Mine Water and Environment 22, 155–161. Sparks, D.L., 1995. Environmental Soil Chemistry. Academic Press. Oyarzun, R., Lillo, J., Higueras, P., Oyarzún, J., Maturana, H., 2004. Strong arsenic enrich- Squeo, F., Aravena, R., Aguirre, E., Pollastri, A., Jorquera, C., Ehleringer, J., 2006. Ground- ment in sediments from the Elqui watershed, Northern Chile: industrial (gold min- water dynamics in a coastal aquifer in north-central Chile: implications for ing at El Indio–Tambo district) vs. geologic processes. Journal of Geochemical groundwater recharge in an arid ecosystem. Journal of Arid Environments 67, Exploration 84, 53–64. 240–254. Oyarzun, R., Guevara, S., Oyarzún, J., Lillo, J., Maturana, H., Higueras, P., 2006. The As- Thyne, G., Güler, C., Poeter, E., 2004. Sequential analysis of hydrochemical data for wa- contaminated Elqui river basin: a long lasting perspective (1975–1995) covering tershed characterization. Groundwater 42, 1–13. the initiation and development of Au–Cu–As mining in the high Andes of northern Valenzuela, A.E., 2005 . History of mining in Chile (part 2), CIM Bulletin 98(1084), Chile. Environmental Geochemistry and Health 28, 431–443. January/February 2005 Available at: http://www.cim.org/bulletin/bulletinlive/ Oyarzun, R., Oyarzún, J., Lillo, J., Maturana, H., Higueras, P., 2007. Mineral deposits and articles.cfm?Issue_ID=124&row=1&Type=1&Segment_ID=46 (Last accessed Cu–Zn–As dispersion–contamination in stream sediments from the semiarid 2011–09–12). Coquimbo Region, Chile. Environmental Geology 53, 283–294. Yidana, S.M., Ophori, D., Banoeng-Yakubo, B., 2008. A multivariate statistical analysis of Reyes, M., 1991. The Andacollo strata-bound gold deposit, Chile, and its position in a surface water chemistry data—the Ankobra basin, Ghana. Journal of Environmental porphyry copper–gold system. Economic Geology 86, 1301–1316. Management 86, 80–87.