Mining districts

Along the Sonora river upper basin there are several zones characterized by hydrothermal alteration or mineralization, which form part of the geochemical background. Hence, some of them include currently active mines while many others deposits are unexploited or underwent production. Figure 1 shows, from north to south, the main mining developments, which correspond to the Cu-Mo Buenavista del Cobre, the Pb-Zn El Gachi, the Ag (Au) Santa Elena, the Pb-Zn (Au) San Felipe, the tungsten El Jaralito and the Cu-W-Mo Washington mining sites.

Study area

The study area corresponds to the Sonora river upper basin, which is located between the Cananea town to the north, and the El Molinito dam to the south (Figure 1), with a total surface of 9292 km². This region has several mining developments, most of them with intermittent activity, and some of them under production, including the Buenavista del Cobre porphyry copper mine, which is the largest mine in Mexico, and one of the largest copper deposit at world-scale (^{Titley, 1993}).

Lithological categorization

The geochemical background depends on several factors, among which the geology and lithology are key factors (^{Hao et al., 2014}). The geology differs from lithology because it takes in account the spatial-temporal relationships and genesis of rocks, as well as structural features. The concentration of elements is influenced by many factors; however, assuming that landscape characteristics remain relatively unchanged, the surficial lithology is a key factor influencing the geochemical background. Therefore, the background values of metallic elements considerably vary from one area to another, depending on the petrographic and geochemical composition of outcropping rocks. In order to establish the representative rock types and to select sampling sites, a geological basemap was synthetized from 1:50,000 and 1:250000 geological maps published by the Mexican Geological Survey (Servicio Geológico Mexicano, SGM). The lithology was categorized into 11 units, according to their nature and regardless the age of formation (Table 1; Figure 2). For instance, considering their similar compositions, Cretaceous and Cenozoic felsic volcanic rocks were categorized into a single unit named “felsic volcanic rocks”.

Figure 2 Map showing the sub-basins of the Sonora river upper basin, defined as sectors A, B and C, with the distribution of the 11 units categorized by lithological attributes, which are: Unit 1: non-consolidated detritic deposits; Unit 2: consolidated detritic deposits; Unit 3: marine sedimentary rocks; Unit 4: quartzite; Unit 5: granitic rocks; Unit 6: volcanic felsic rocks; Unit 7: intermediate volcanic rocks; Unit 8: basalt; Unit 9: metamorphic rocks; Unit 10: contact metamorphic rocks; Unit 11: altered rocks. White circles represent samples sites. Inset shows the division of the Sonora river upper basin in sectors A, B and C. 

Sampling and Geographic Information System

The Sonora river upper basin was divided into three sectors, which correspond to the three main sub-basins (inset of Figure 2): (i) Sector A (1479 km²) corresponds to the watershed drained by the Tinajas creek, which received the spill in first instance, and the Bacanuchi river; (ii) Sector B (2278 km²) corresponds to the Bacoachi river sub-basin, and (iii) Sector C, which is the largest and southernmost sub-basin (5535 km²), and receives discharges from sectors A and B. Considering the lithological categorization, a total of 73 rock samples were selected to perform the geochemical studies (Figure 2).

Furthermore, sampling was constrained considering the area of exposure of each lithological unit, which was calculated for sectors A, B and C (Table 2), and recorded as an attribute to the sample. A geographic information system (GIS) was created using ArcGIS 10.3, in the UTM projection for the zone 12N, with WGS84 datum.

Preparation and analytical techniques

Rock samples were hammered in the collecting site to about a fist size, in order to avoid further contamination. The collected samples were crushed to fragments of about one cm in diameter, using a steel jaw-crusher at the Estación Regional del Noroeste, Universidad Nacional Autónoma de México (UNAM). A fraction of the crushed samples was sent to ActLabs laboratories in Canada, where they were powdered using an agate mill. The concentrations of major and trace elements were measured by geochemical packages Code 4B (lithium metaborate-tetraborate fusion ICP whole rock) and Code 4B2 (trace element ICP-MS), respectively.

Calculation of the geochemical background

The concentration average of each measured element was obtained considering the number of samples, and weighted taking into account the exposure area for each lithological unit from sectors A, B and C. Then, weighted average values of major and trace element for the three sectors were used to obtain the natural geochemical background of rocks from the entire Sonora river upper basin. Because of the comparatively small area covered by quartzite and regional metamorphic rocks of units 9 and 10, respectively, they were not considered in the calculation.

Major and trace elements

The average concentrations of major and trace elements for samples from the lithological units of sectors A, B and C are shown in Table 2. The weighted average of elemental values of Sectors A, B and C, as well as the entire basin, are shown in Table 3. These values are considered to represent the geochemical background of the study area, and do not have any petrogenetic significance since the lithological units are constituted by geological formations of different nature and age. Figure 3 shows that the lithological units from the three sectors have similar geochemical patterns when normalized to the Upper Continental Crust (UCC) composition proposed by ^{Rudnick and Gao (2003)}, defining a relatively narrow strip close to the reference line, for most elements, particularly for the rare earth elements group (REE).

Figure 3 Multi-element diagrams normalized to the Upper Continental Crust (UCC) values of ^{Rudnick and Gao (2003)} for sectors A, B and C of the Sonora river upper basin; The thick solid line represents the weighted average value for each sector, considering the exposure area of the lithological units. See text for discussion. 

The concentrations of CaO and P₂O₅ are around 1.5 times the UCC abundance, which is relatively more obvious in Sector C (Figure 4). The CaO anomaly may be associated with limestone, intermediate and felsic volcanic rocks, as well as with altered rocks. The relatively high values of P₂O₅ may be due to the influence of volcanic rocks. In contrast, Na₂O is slightly depleted, probably indicating it was partially removed in solution. Moreover, manganese is commonly removed by hydrothermal fluids and weathering processes, and further deposited in cavities, fractures and crusts, mainly as oxides, hydroxides or carbonates (Figure 3).

Figure 4 Multi-element diagram normalized to the Upper Continental Crust (UCC) of ^{Rudnick and Gao (2003)} showing the weighted average values for sectors A (dotted line), B (dashed line), C (thin solid line) and for the entire basin (blue solid line). 

Regarding the trace element concentrations, the three sectors display similar patterns with relative enrichments in As, Ag, Sb, Pb and Bi relative to the UCC (Figure 4). This enrichment is higher in Sector C, which is attributed to the influence of mineralized zones, particularly in San Felipe, Santa Elena and Baviácora (Figure 1). The enrichment of Ag, Sb and As is about 10 times the UCC; Ag and As concentration values ​are correlated to intermediate volcanic rocks and unconsolidated detrital deposits, whereas Sb is preferentially associated with felsic volcanic rocks, unconsolidated detrital deposits, and altered rocks. The other elemental positive anomalies (Ba, Cu, Pb, Bi, Th and U) are relatively more influenced by rocks of the Báucarit Formation and unconsolidated detrital deposits of the Sonora Group and, into a lesser extent, by altered rocks. The highly positive anomaly of W in Sector C is associated with the occurrence of scheelite-rich skarn deposits located in the Aconchi batholith, southwest of Baviácora (Figure 1).

Conservative reference elements

Elements that are abundant in the Earth´s crust, and their concentrations are not substantially changed by human activities and the biogeochemical cycles, are referred to as conservative or reference elements (^{Galuszka and Migaszewski, 2011}), which include Al, Zr, Sc, Cs, Ti, Y, Ce, among others. The coefficient of variation, defined as the ratio of the standard deviation () to the mean (x), is used to determine the conservative capability of an element (^{Reimann and de Caritat, 2005}). However, some elements, traditionally considered as conservative, may display a similar or higher variability than certain elements considered as pollutants in environmental studies (^{Reimann and Garrett, 2005}).

Commonly, the determination of environmental indexes (e.g. enrichment factor, geo-accumulation index, pollution index, concentration factor) is based on comparing the concentration of an element with a globally accepted reference, for instance, the continental crust (^{Monsalve et al., 2017}; ^{Saleem et al., 2018}; ^{Hasan et al., 2018}). However, to obtain more realistic reference values, the using of elemental concentrations of either local or regional geochemical values is preferred by many authors (e.g.^{Kříbek et al., 2014}; ^{Wu et al., 2014}; ^{Zhang et al., 2014}; ^{Christou et al., 2017}; ^{Pan et al., 2017a}; ^{Pan et al., 2017b}; ^{Shang et al., 2017}; ^{Akopyan et al., 2018}). Therefore, for the calculation of environmental indexes, particularly the enrichment factor, it is essential the using of conservative reference elements.

In the case of rocks of the Sonora river upper basin, the geochemical background shows a great variability in some elements due to the lithological diversity and the presence of locally altered and mineralized zones. According to the calculated coefficient of variation values (Figure 5), eight conservative reference elements were identified on the basis of their low coefficient of variation values (Sc, Y, Zr, Ce, Yb, Fe, Ti and Al). Therefore, these elements are recommended to assess contamination in environmental media (sediments, soils, surficial or underground water, etc.) to perform further studies in the Sonora river upper basin.

Regarding the elements commonly considered as potentially toxic, only As, Sb, Pb, W, Bi, Zn and Cu exhibit positive elemental anomalies in the geochemical background with respect to the UCC (Figure 4). Moreover, these elements are also characterized by having higher coefficient of variation values, as pointed out for soil samples from two large-scale geochemical studies in northern Europe (^{Reimann and de Caritat, 2005}).

Figure 5 Probability diagram showing the most conservative reference elements for the Sonora river upper basin. Aluminum and yttrium yield the lowest coefficient of variation. 

Implications for mining activities in Mexico

In Mexico, there is a lack of environmental regulations to establish a geochemical background prior to the development, or during the operation of a mine. As far as we know, this is the first large-scale study in Mexico that was carried out with the purpose of defining the geochemical background in a region, where mineralized areas are widespread, and mining industry is a priority activity. The establishment of the geochemical background should be considered as a basic tool to avoid misinterpretations of abnormally high elemental values, particularly those considered as potentially toxic, which can be wrongly attributed to mining activity instead of natural concentrations. For the case of the Sonora river upper basin, the geochemical background serves as a reference for the studied media in the region (e.g. soil, dust, sediments).