1 Introduction

Geochemical mapping represents an application of conventional chemical methods to the knowledge of the natural environment (Plant et al., 1989). This is the most helpful way to determine the current levels of geochemical amounts—geochemical baselines—to comprehend the spatial distribution and variability of the different elements, and to document variations in their classes in natural materials at the Earth's surface (Demetriades et al., 2014). Those changes, usually shown in geochemical distribution maps, can often indicate areas affected by possible anthropic pollution activities or anomalies related to known or unsuspected mineralization.

Geochemical mapping is one of the main activities performed in national geological surveys and organizations. Following the recommendations in the Final Report of Project 259 for the International and National Geochemical Programs (Darlney et al., 1995), the Geological Survey of Spain achieved the Geochemical Mapping of Spain (Locutura et al., 2012). The main results accomplished with the data of the regional soil survey in the Iberian Pyrite Belt are presented below.

The IPB has been the scene of ancient mining activity since the Roman domination or even much earlier (Phoenicians or Tartessians). This proves several great mines that are still operating today. Further, they also show remnants of Roman extractive works (Pinedo-Vara, 1963). But the great leap forward for the mining activity in the IPB took place in the second half of the nineteenth century. Since then, there has been intense extractive and exploratory activity even when operations decreased, primarily because of the low price of the metals. Nowadays, the IPB continues to arouse unusual experimental interest, which indicates that the investment in this activity is still fruitful and profitable. One reason for this is the recent commissioning of five new mines or the enlargement of older ones that have started recently.

In the IPB study area, the recent past of intense exploration has been mostly based more and more on significant development and advance of the geological and metallogenetic processes, besides ground and airborne geophysical methods like gravimetric, electric, or electromagnetic methods. In contrast, geochemistry has yet to be applied to exploration projects. Yet, we can cite the geochemical survey carried out to assess the abundance and distribution of toxic trace elements in the soils of Andalusia, as relevant research has shown (Aguilar et al., 1999). This is because the suspected pollution associated with old mining has inhibited the application of geochemical techniques based on water, sediment, or rock samples, which are considered susceptible to easy contamination in those scenarios. Just a few local lithogeochemical surveys have been implemented (Díez-Montes et al., 2017), as well as applications of gas geochemistry (Bel-lán et al. unpublished). The only exception worth mentioning corresponds to the multi-elemental soil geochemistry study performed by SEIEMSA (Bonnemaison et al., 1993). This study used a multicriteria treatment with other geological and geophysical variables to define anomalies between the transect of the Herrerías mine and Puebla de Guzmán village, finding the small Vallejín pyrite deposit.

Finally, the recent literature shows different geochemical scientific studies related to the IPB, like Galán et al. (2008), that investigate the influence of geological setting on geochemical baselines of trace elements in soils in the South-West Spain. Further, Batista et al. (2012, 2020), also evaluated the mobility of trace elements from soils to sediments and geochemical exploration in the Portuguese part of the IPB, and Luz et al. (2014), used Cu, Zn, and Pb soil geochemistry for exploration. Likewise, Gonçalves and Mateus (2019) used data from stream sediments to delimitate several geochemical anomalies, and Fernández-Caliani et al. (2020) defined geochemical anomalies for some critical elements in soils of western Andalusia (Spain).

This study aims to create a new detailed and good-quality regional knowledge infrastructure of geochemical data from residual soils in the IPB. The main targets of this research are: (1) develop a geochemical database of residual soils from the IPB which increases the knowledge of the abundance and the spatial distribution of chemical elements; (2) determine the geochemical associations that govern the general variability, behavior and distribution of the elements by multivariate analysis; and (3) distinguish the main geochemical anomalies.

2 Geological setting

The study zone corresponds to the Spanish part of IPB that covers an area of about 3.000 km2 in the Huelva and Seville provinces.

2.1 Regional geology

The Iberian Pyrite Belt (IPB) is a well-known area for its exceptional relevance and abundance of mineral deposits, extending around 230 km long and 40 km wide from Seville (Spain) to Lisbon (Portugal) (Quesada, 1996; Tornos, 2006) (Fig. 1A). It is located in the South Portuguese Zone, the southernmost of the six tectonic–magmatic zones defined by Julivert et al. (1974). This zone was sutured to the Hercynian massif during Dinantian-Namourian times, using the dismembered metamorphosed Acebuches ophiolite (Quesada, 1991).

Fig. 1
figure 1

A Geological schematic map of the Iberian Pyrite Belt from Gumiel et al. (2010)

Full size image

The geology of the IPB has been studied by numerous authors (Inverno et al., 2015; Leistel et al, 1997; Quesada, 1996; Routhier et al., 1978; Schermerhorn, 1971; Strauss and Madel 1974; Tornos, 2006; among other). There is an agreement among most of the authors on the stratigraphic succession of the IPB, which consists of Upper Palaeozoic (Middle Devonian–Mississippian) sedimentary and igneous rocks that have been classically subdivided into three central lithostratigraphic units. From bottom (older) to top (younger) these are the Devonian Phyllite–Quartzite Group (PQ Group), the Late Devonian–Mississippian Volcano Sedimentary Complex (VSC) and the Late Visean–Moscovian post-volcanic succession (Culm Group).

One of the most significant features of the IPB is the great abundance of submarine volcanic processes giving place to continuous accumulations of volcanic and sedimentary materials with certain lateral continuity, forming a volcano-sedimentary pile or complex of hypabyssal, effusive, pyroclastic, and epiclastic-sedimentary rocks. Those materials of volcanic sources (Volcanic Sedimentary Complex, VSC), from late Devonian to Visean age, evolved along some volcanic phases, showing evidence of essential interactions with hydrothermal fluids from the volcanic or mixed volcanic-sedimentary origin, mobilized and activated by the most important thermal sources.

The PQ Group forms the base of the succession. It consists of a detrital sequence of shale and quartz-sandstone representative of shallow-water deposition, probably in a shallow marine platform (Mantero et al., 2011 and Moreno, 1993). There are two main outcrops: the eastern part of the Valverde del Camino antiform and the western part of la Puebla de Guzman antiform.

The VSC mainly comprises volcanic and subvolcanic rocks interbedded in a detrital sedimentary sequence of shales and volcano-derived sandstones. It is characterized by bimodal volcanism, with the presence of dacitic-rhyolitic dome complexes and sills, basaltic lava flows and sills, and thick pumice and crystal-rich felsic volcaniclastic units interbedded with detrital sedimentary rocks, mostly mudstone with some greywacke and sandstone (Leistel et al., 1997 and Tornos, 2006). In the’70s, Schermerhorn (1971) defined a synthetic column for the whole IPB with three different volcanic episodes, the first being the most relevant for the ore deposits.

The volcanic rocks mostly show a felsic composition ranging from dacite to rhyolite. Abundant albite, quartz phenocrysts, and minor biotite, partially or replaced by chlorite, appear in the rhyolitic rocks, which show perlitic and spherulitic textures. However, dacites show porphyritic textures (Rosa et al., 2010).

This is relevant to the study area, as outcrops with volcanic rocks of intermediate composition -andesites- only predominate in local parts of the northern IPB. These andesites are formed mainly by plagioclase, occasionally albitized, Fe-oxides, clinopyroxenes and rare quartz and biotite. Plagioclase are often altered to calcite, muscovite and epidote (Mitjavila et al., 1997).

In the upper part of the volcano-sedimentary sequence, there is a locally developed horizon, up to 30 m thick, of hematitic radiolaria-rich, purple-colored shale with Mn-bearing jasper lenses (Silva et al., 1990), called Manganesiferous Formation, which is an excellent stratigraphic reference level in the area.

Finally, towards the top, the VSC gradually evolves into the Culm group sequence, consisting of a synorogenic thick turbidite sequence of shales and sandstones that filled the foreland basin of the orogen (Leistel et al., 1997).

Plutonic rocks are confined to the northeastern part of the belt, forming the Sierra Norte de Sevilla Batholith (SNSB). This mass has diorites, tonalites and leucotonalites with a lesser proportion of gabbros, and granites (IGME, 1982; Schitz et al., 1987). The geochemical similarities between SNSB plutonic rocks and the materials of the volcanic successions (Díez-Montes and Bellido, 2008; Díez-Montes et al., 2011; Schütz et al., 1987; Thiéblemont et al., 1997), as well as the available U/Pb evidence (Dunning et al., 2002; Barrie et al., 2002) suggest that SNSB plutonic unit is genetically related with the VSC volcanic rocks.

The Pulo do Lobo sequence, in the northern part, consists of phyllites and quartzites that represents an accretionary complex developed in Variscan times and marks the SW Iberian suture, separating SPZ from the Ossa Morena Zone (OMZ) (Munhá et al., 1989; Pereira et al., 2008).

The north of the study area also crops out rocks that belong to the OMZ, and materials from the Guadalquivir Basin in the southeastern part (Fig. 2).

Fig. 2
figure 2

Geological detailed map (modified from Matas et al., 2015) with mineral occurrences (García-Cortes, 2011)

Full size image

The current structure of the IPB represents the culmination of a complex process of cortical extension and associated magmatism that was produced by the oblique collision of the continental block of the SPZ with the active margin of the Iberian block located to the north and which ended with the obduction of the latter on that of the SPZ (Almodóvar et al., 2019; Leistel et al., 1997).

2.2 Mineral deposits

The Iberian Pyrite Belt is one of the most significant accumulations of metals contained in volcanogenic massive sulfide (VMS) deposits worldwide (Almodóvar et al., 2019; Leistel et al., 1997) with more than 80 known deposits containing > 1700 Mt of sulfide ore (mined and reserves), and some tens of Mn occurrences (Fig. 2).

Different types of mineral deposits in the IPB include mineralization in VMS, stockworks, gossans and Cu-rich shales. Mn is also extracted from small open pits linked to hydrothermal mineralizations and volcanism (Locutura, 2011).

More than 80 ore deposits and 100 mineral occurrences are located within the studied area in the IPB. We can highlight some of the VMS ore deposits over 50 Mt (e.g. Rio Tinto, La Zarza, Tharsis, Aznalcollar-Los Frailes, Sotiel-Migollas, Masa Valverde and Aguas Teñidas) and the most important Mn ore, Soloviejo.

Nowadays, mining companies are working in Aguas Teñidas, La Magdalena, Río Tinto and Sotiel-Migollas deposits. This illustrates the great mining importance of the area with tonnages greater than 50 Mt. Another active mine in the area is Las Cruces, which is working a blind orebody of Cu-rich minerals under 120 m of Tertiary marls covering, in the alluvial plain of the Guadalquivir River, near Seville. The initial reserves of this deposit were 17.2 Mt grading 6.2% Cu in the secondary enrichment zone and some additional resources in the primary mineralization and the gossan (Yesares et al., 2015).

3 Material and methods

3.1 Sampling and sample preparation

The type of sample selected for the geochemical mapping of the study zone was residual soil, which followed the recommendations of the Geochemical Baseline Program (Salminen, et al. 1998).

During the field campaign from spring 2016 to spring 2017 were collected up to 6197 soil samples, at an average sampling density of two samples per km2 (a grid of 500 × 1000 m) in N-S profiles, and an average weight of 3 kg. They were taken from 2 to 20 cm depth when possible due to the poor development of soils in the Pyrite Belt. Most of the sampled soils are classified Leptosols (WRB, 2014). These soils rarely reach more than 10–20 cm above bedrock. This is why they show bedrock fragments and low contents of new formation minerals, clays and Mn and Fe hydroxides (Fernández-Caliani et al., 2009) (Fig. 3C).

Fig. 3
figure 3

A Distribution of 6197soil samples in the study area. B Diagram of composite sampling method and sampling soil profile. C Soil example of the study area. D Grid of sampling 500 m × 1000 m under satellite view

Full size image

Samples were collected as composite samples from five pits within 100 m2 (Fig. 3B) (Locutura et al., 2012). Soil samples were firstly sieved to < 3 mm during the field campaign. For quality control purposes, a field duplicate was taken at every 30 sample site, with an offset distance of ca. 10–20 m from the original sample site.

All samples were shipped to a central sample preparation facility at the Geological Survey of Spain (IGME). All samples were processed by the protocol described by Salminen et al. (2005). Each sample was oven dried at 40 °C, followed by break up, homogenization and, finally, split into sub-samples. Soil samples passed through a 100 mesh nylon screen (150 µm). Before sending the samples to laboratory, all of them were randomized, and analytical replicates prepared from the field duplicates, following the procedure of Locutura et al. (2012).

3.2 Chemical analyses and quality control

Chemical analyses of 64 elements were carried out at Actlabs Laboratories Ltd. (Ontario, Canada), after a tetra-acid digestion, by a combination of inductively coupled plasma atomic emission spectroscopy (ICP-AES) (Al, Ca, K, Mg, Mn, Mo, Na, P, S, Ti and V), inductively coupled plasma mass spectrometry (ICP-MS) (Ag, Ba, Be, Bi, Cd, Co, Cr, Cs, Cu, Ga, Ge, Hf, Hg, In, Li, Nb, Ni, Pb, Rb, Re, Se, Sn, Sr, Ta, Te, Th, Tl, U, Y, Zn, Zr, La, Ce, Pr, Nd, Sm, Eu, Gd, Dy, Tb, Ho, Er, Tm, Yb and Lu) and instrumental neutron activation analysis (INAA) (As, Au, Br, Fe, Hf, Ir, Na, Sb, Sc, W, La, Ce, Nd, Sm, Eu, Tb, Yb and Lu).

Chemical element concentrations were determined using a tight internal and external quality control procedure. Quality control of the analytical approach included using a method reagent blank, duplicates, and certified reference materials to ensure the accuracy and precision of results. The relative deviations of the standard analyses to the reference values were typically below 10%. The internal quality control (IQC) in Actlabs Laboratories Ltd. was based on introducing into the batches of an international standard material each 10–15 samples (Fig. 4). The results showed no systematic errors or instrumental deviations (less than 100% ± 2 standard deviation) of these quality control supplies by the laboratory.

Fig. 4
figure 4

Quality control report from some elements from Actlabs Laboratories (international standard materials)

Full size image

External quality control (EQC) was based on (a) a field duplicate taken at a rate of 1 in 20 samples, which can determine the sampling variance, sampling preparation variance, and analytical variance. The variation in the EQC is larger than in the IQC due to the sum of the different variances. We have defined a confidence band of 25%, which almost all elements fit. Nevertheless, elements with a variation range close to the lower detection limit (Ag, Au, Bi, Cd, Ge, Hf, Hg, In, Mo, Re, S, Se, Sn, Ta, Te and W) and elements that appear in refractory minerals (Ti, V and Zr) show higher but acceptable errors. The quality control follows the descriptions in Demetriades et al. (2014).

3.3 Statistical analysis

A complete database with all analytical results, data from field campaigns and other outstanding characteristics were prepared for statistical computation, based on Reiman et al. (2008). The element concentrations below the instrumental detection limit (IDL) have been assigned a value corresponding to half their detection limit. The results of the chemical analyses were processed using STATISTICA software to generate the statistical parameters used to guide the interpretation of the results.

3.3.1 Univariate statistical analysis and spatial distribution

Univariate statistical analyses were performed to show the single-element geochemical distribution. The spread of element concentration values, distributions and identification of multiple populations and outliers, were examined through calculations of different parameters summarized in Table 1. Also, deviations from normal or log-normal behavior were investigated for each element (Reiman et al., 2005). The representation of the populations of the different elements in the form of histograms and box plots has made it possible to observe that some trace elements deviate from the normality patterns. Spatial distribution maps of the chemical elements have been made by interpolating the inverse of the weighted distance (ID2). The main features of element geochemical distributions through point maps and contour maps can display possible relationships with geologic features (Reiman, 2005). The interpolation grid was defined by a cell of 800 m (x axis) and 400 m (y axis), in accordance with the sampling grid. The radio search (anisotrope) was 3000 × 2000 m with a direction of N110E. Contours maps were generated using ArcGIS software and statistical percentiles (10th, 25th, 50th, 75th, 90th and 97.5th).

Table 1 Statistical parameters for major, minor and trace elements in the IPB soils (6194 samples, 0–20 cm horizon, analyzed by ICP-MS, ICP-AES or INAA, Det. Lim: laboratory detection limit, P: Percentil, < BDL: Below detection limit
Full size table

3.3.2 Multivariate treatment

Multivariate statistics is a valuable tool for treating geochemical data in geochemical surveys, as it allows investigating several variables and their interactions simultaneously. There are different options for multivariate treatment in geochemical exploration including principal component analysis (PCA), factor analysis (FA), independent component analysis, cluster analysis and multi-dimensional scaling (Zuo et al., 2021).

According to Grunsky (2010), PCA and FA are the most popular multivariate techniques for managing geochemical exploration data because these techniques reduce many elements into several factors based on the correlation between variables. We have used FA between these two options because PCA accounts for a maximum variance of all variables, so all variables have to enter into the factor and show the total structure in the data (all variables are “forced” into the result and in contrast, FA is based on the correlation structure of the variables and factors do not have to explain the total variation of the data, there are unusual variables will not enter the common factors (Reiman et al., 2002).

4 Results and discussion

4.1 Univariate spatial distribution

The spatial distribution of a chemical element in a given area can be of various types or forms. The high values, those of the intermediate range and the very low ones can be adjusted approximately well, to the contours of geological or lithological units. These guidelines reflect a special relationship between the unit above and the element that can define the unit, for example, Na in the IPB. Another form of distribution is one in which the values of the element bear no relation whatsoever to the contours of the units or the geological structure, but rather appear with an uneven distribution, in small (or intermediate) spots of color (range of generally high or very high) scattered, cutting contours. However, this distribution may be more accentuated in some geological enclosures than others. This distribution modality can result from existing mineralization elements mobilized by secondary or primary geochemical dispersion. Depending on the nature of the element, they can also be associated with industrial or agricultural pollution, for example, Cu in the IPB. Finally, some elements present mixed distribution patterns with characteristics of the two previous distribution models. They usually occur in the case of trace elements uniformly distributed in a lithology or unit that can be mobilized to a certain degree in the regional context, producing a distribution pattern adjusted to a unit and, at the same time, an erratic dispersion outside it (Reiman, 2005).

Sodium is a metal with a litophile affinity. It forms its own minerals: silicates like albite (Na2AlSi3O8), chlorines like halite (NaCl), carbonates like sosa (Na2CO3), and fluorides like criolite (Na3AlF4). However, this element is more common in minerals that form rocks, and it appears included in its crystallographic structure, both in form of the trace element and important components. For example, silicates, like feldspars, sodium micas, amphiboles and pyroxenes. Its natural associations are Cl, Na, Mg, Br, S, its relationship with marine breezes and inputs, and with brines and evaporitic environment. Other natural associations are Si, Al, Na and K, related to acid igneous rocks. Sodium increases its concentration in magmatic differentiation processes from basic to acid rocks, so, it presents a higher concentration in this type of igneous rocks (Locutura et al., 2012).

Sodium Clarke’s is 2.27% (Taylor, 1964) and in soils Na2O median amount is estimated 1.3%. In the European Union, the average in soils is 0.8% (Salminen et al., 2005), while in Spain is 0.31% (Locutura et al., 2012), all of these results were measured on the fraction < 2 mm.

In our studied zone for the fraction < 150 µm, Fig. 5A sodium median contents in soils are 0,53% and Na2O median contents are 0,71%, higher than the average contents of Spain, 0.87%. This is due to the higher values of the VS complex (up to 1%), with a high quantity of acid rocks that increase the baseline of the entire zone. Sodium clearly distinguishes the northwest igneous rocks, with higher values (up to 2%) in the acid units (Variscan granites) and the VS complex; the volcanism shows compositions from basalt to rhyolite. However, the most felsic terms dominate, as domes and sills are associated to volcanoclastic deposits with similar chemical composition.

Fig. 5
figure 5

A Monoelemental contour map of sodium (Na). B Monoelemental contour map of copper (Cu)

Full size image

Copper is a metal with a high chalcophile affinity. It forms a broad group of its minerals, like sulfides, sulfur-antimonides, sulfo-arsenides, carbonates, and oxides, the most frequent being chalcopyrite (CuFeS2), chalcocite (Cu2S), bornite (Cu5FeS4), covelline (CuS), cuprite (Cu2O), malaquite and azurite, but also native copper. Therefore, it appears in the crystal lattice of silicates (biotite, pyroxenes and amphiboles and of sulfides (sphalerite and oxides (magnetite) in substitution of Fe or Mg, so, it is more abundant in basic igneous rocks than acid igneous rocks (Locutura et al., 2012).

Copper Clarke’s is 68 mg/kg (Mielke, 1979) and in soils the median content of Cu is between 13 and 24 mg/kg (Kabata-Pendias, 2001). For the EU the median content in soils is 13 mg/kg (Salminen et al., 2005); in Spain, the median content in soils is 18.5 ppm and finally, the median content in Andalucía soils is 24 mg/kg (Locutura et al., 2012), all of these results was measured on the fraction < 2 mm. Copper can be relatively immobile in soils but its fixation processes (adsorption, occlusion and co-precipitation, organic chelation and complexing, microbial fixation), among other soil parameters, can be controlled by pH (Kabata-Pendias, 2001).

In the study area for the fraction < 150 µm, (Fig. 5B) Cu median content in soils, 43.7 mg/kg, are higher than the medium contents worldwide, Europe, Spain and Andalusia. This is due to the vast quantity of Cu-rich sulfur deposits. The area's central part, mainly made up of materials from the Volcano-Sedimentary Complex, contains a dispersion of minor anomalies, related to the numerous existing deposits on a yellow background (40–60 mg/kg). However, in most peripheral areas, with limited deposits and more sterile lithologies (PQ Group and Culm), the Cu backgrounds are lower (less than 35 mg/kg) and with a completely uneven distribution. Also, we can observe a high and wide anomaly in the IPB with values of 56.5 mg/kg or even higher. More significant abnormalities are related to the more enormous ore deposits in this zone, like La Zarza, Sotiel, Aguas Teñidas, Cueva de la Mora, San Telmo, Tharsis, and Rio Tinto, there are other anomalies unrelated to already known deposits and so, with a prospective interest.

4.2 Multivariate spatial distribution

Factor analysis was used to investigate the complex multivariate relationships among variables. The treatment aims to explain the variation in an extensive database by as few “factors” as possible. Factor analysis served to identify different groups of chemical elements with approximately the same geochemical pattern. However, FA is sensitive to outliers in the data, so it needs a normal distribution of the population before carrying out a classical FA (Reiman et al., 2002).

We have made several attempts to change the number of factors and elements. We have finally chosen the normal population instead of the log transformation population because we have deleted the extreme values and outliers with the help of the representation of the populations in box plots (Fig. 6A, B examples), which has allowed us to differentiate them which have been eliminated for this multivariate treatment. We have noted a higher consistency of the results with this treatment in the normal population. This solution accounts for 69.22% of the total variance.

Fig. 6
figure 6

A As a Box plot of the total population of samples with outliers (blue) and extremes (red). B Zn Box plot of the total population of samples with outliers (blue) and extremes (red)

Full size image

The different factors obtained were studied and interpreted by their hypothetical origin (natural, anthropogenic, or mixed). ArcGis software has been used for producing color surface maps showing element and association factor score distributions. Maps of factor score distributions are handy to correlate lithological units with physicochemical and pollution processes of the surface environment at each sampled site (Reiman et al., 2002). Inverse distance weighted (MIDW) is the interpolation method with the same parameters as contours maps.

We have used factorial analysis by main component extractions with “varimax rotation”. Elements with loadings over 0.25 are considered representative members of each association resulting from the chosen factor model (Table 2).

Table 2 Varimax rotated factors (six-factor model) for 6194 soils samples from IPB
Full size table

The element association F1: Be-Rb-K-Th-Al-U-Li-(Ba-Sm-Tl-Eu-Gd-V) in contrast to Ca-Na-(Mg-Sr) accounts for 15.48% of data variability, and the F2 association: Er-Ho-Tm-Y-Yb-Dy-Lu-Tb-Gd-(Na) accounts for 15.11% of data variability; F3: Bi-Au-Sb-Ag-As-Sn-Pb-Ba-(S-Cu-Hg) and F4: Fe-S-Co-Cr-Ni-V-Al-Sr-(P-Ga-Mn-Li) account for 11.04% and 11,09% of data variability, while F5: Cd-Zn-S-Hg-(Cu-Pb-Tl-Ag-Sb-In-As) and F6: Sm-Nd-Pr-La-Eu-Ce-Gd-Tb-(Th-U) in contrast to (Na) account for 6.19% and 10.39% of data variability, respectively. In each association, elements are listed with decreasing loadings, and elements in brackets have loadings between 0.5 and 0.25.

The element association F1: Be-Rb-K-Th-Al-U-Li-(Ba-Sm-Tl-Eu-Gd-V) in contrast to Ca-Na-(Mg-Sr) show a spatial distribution with a litoestratigrafic control (Fig. 7). It’s mostly controlled by the presence of soil developed on PQ group, but is important to differentiate between the east zone, in Valverde del Camino antiform with values up to 0.47, in contrast with La Puebla de Guzman antiform in the west part with values lower than 0.47, that could suggest differences in its origin and composition. Leistel et al. (1997) highlight the importance of good information in areas of PQ rocks in antiform core zones for mineral exploration. The PQ core of Puebla de Guzmán antiform can be considered as a nappe thrust over a substratum of VSC. Moreover, it shows high values related to Gil-Márquez granites and facies from Pulo do Lobo formation and Santa Barbara formation due to the presence of feldspars in shales, lithoarenites and feldespatic sandstones in these formations. The Culm Group shows an erratic distribution for this factor, while both VSC and SNSB present a negative correlation.

Fig. 7
figure 7

A Factor scores distribution of element association F1: Be-Rb-K-Th-Al-U-Li-(Ba-Sm-Tl-Eu-Gd-V) in contrast to Ca-Na-(Mg-Sr). B Distribution contour map of Be. C Distribution contour map of Rb

Full size image

This factor is similar to F2 in the total extraction of topsoils in Atlas Geoquímico de España (Locutura et al., 2012), that also shows higher values in the PQ Group to the eastern part of the IPB. This factor has a geochemical signature of more evolved granites, differentiating between Gil-Marquez granites and the rest of the granites of the SNSB due to their different origin (Castro et al., 1995). The higher values of the eastern zone of Valverde del Camino antiform suggest a possible sub-outcrop granite similar to Gil-Marquez granite. We can also observe these higher values in monoelemental contour maps like beryllium and rubidium (Fig. 7B, C.)

The element association F2: Er-Ho-Tm-Y-Yb-Dy-Lu-Tb-Gd-(Na) is formed mainly by HREE and Na. It is mostly related to igneous units (Fig. 8). This factor shows an essential relationship between the igneous rocks from the SNSB and the volcanic rocks from the VSC. This geochemical association confirms the idea that the Variscan rocks from the BSNS are the parental rocks of the volcanic materials from the VSC (Thiéblemont et al., 1997).

Fig. 8
figure 8

Factor scores distribution of element association F2: Er-Ho-Tm-Y-Yb-Dy-Lu-Tb-Gd-(Na)

Full size image

There are some differences between different the VSC. Higher values are related to E-W band of VSC (rhyolites and dacites) at the northeast part of the project. This factor is more related to acid terms that are also influenced by the erosion of the BSNS granites to the north. In contrast, VSC andesites (intermediate terms) at the northwest part of the project show lower values due to geochemical characteristics and because this zone does not influence the erosion of BSNS granites.

We can also see high values (up to 0.27) in the eastern part of PQ outcrops in Valverde del Camino antiform, possibly related to a sub-outcrop granite observed in F1.

The F3 factor: Bi-Au-Sb-Ag-As-Sn-Pb-Ba-(S-Cu-Hg) is a geochemical association with several elements related to the metallogenic process so associated with ore deposits (Fig. 9). It is an “ore factor.” There are VMS ore deposits, gossans, and stockworks in the IPB, but some of the main elements in this group, like Bi, Au, and As are typical from stockwork associations (Leistel et al., 1997; Marcoux et al., 1996; Velasco et al., 2013).

Fig. 9
figure 9

Factor scores distribution of element association F3: Bi-Au-Sb-Ag-As-Sn-Pb-Ba-(S-Cu-Hg)

Full size image

The factor scores represent an irregular and erratic distribution controlled by the ore deposits. In detail, we can observe the higher values (from 0.09 to 18.83), related to Rio Tinto, not only related to VMS and stockwork but also with dumps and mining wastes in the surrounding areas of several open pits (Atalaya, Salomón, Dehesa, Lago) due to the contamination of the soil affected by sulfides extraction (López et al., 2008).

Similar to higher values around other ore deposits such as La Zarza and Tharsis that are also accompanied by stockworks (Tornos, 2006) and different high concentrations with maximum values which are not so high around ores with a description of stockworks; Sotiel-Migollas (Santos et al., 1996), Aguas Teñidas (Tornos, 2006), Aznalcollar (Pascual et al., 1996), Angostura and San Miguel (Tornos, 2006). Other “anomalies” unrelated to known ore deposits need follow-up studies to consider their importance in future mineral exploration.

The main geological units show no relevant correlation with this factor, so we can only remark on the geochemical difference between the two outcrops of the PQ Group in the different antiforms again.

The association F4: Fe-S-Co-Cr-Ni-V-Al-Sr-(P-Ga-Mn-Li) is explained by elements reflecting a lithological factor. This factor contains several ferromagnesian elements related to the geochemical signature of basic igneous rocks (Fig. 10). We can observe this association in the distribution of higher factor score values in soils developed on mafic rocks outcrops of Variscan terrains from the SNSB and primary facies from the VSC.

Fig. 10
figure 10

Factor scores distribution of element association F4: Fe-S-Co-Cr-Ni-V-Al-Sr-(P-Ga-Mn-Li)

Full size image

The elevated values of the factor scores in the VSC are explained by basalts that outcrops in the North and West factor of the Valverde and the Puebla de Guzman antiforms (from 1.72 to 4.58). Also, high values in Variscan igneous rocks are related to mafic compositions, mainly gabbros, diorites and cuarzodiorites at Northeast and small bodies of gabbros, micrograbros and diabases in El Berrocal, Sotiel and at South of Valverde del Camino village. Basaltic rocks in the IBP show high contents in Ni, Cr and Mg, elements present in this factor (Mitjavila et al., 1997).

We foreground the difference between the northern and the southern Culm Group. Matas et al. (2015) explain geological variations between this zone (undifferentiated Culm), in the north part with organic-rich black shales that could be also enriched in ferromagnesian elements and the other type of Culm in the south region.

The element association F5: Cd-Zn-S-Hg-(Cu-Pb-Tl-Ag-Sb-In-As) is characterized by elements that reflect a mineralization factor (Fig. 11). It shows several discrepancies with Factor 3. In this case, is a typical volcano-sedimentary geochemical association with Zn-Cd-Cu-Pb. However, some of these elements can present a specific geological control (Pb, Sb and Tl) and could appear in silicate minerals of igneous rocks.

Fig. 11
figure 11

Factor scores distribution of element association F5: Cd-Zn-S-Hg-(Cu-Pb-Tl-Ag-Sb-In-As)

Full size image

Again, we can observe the highest concentration of values (from 1.1 to 24.85) around Rio Tinto ore deposits, like in factor 3. We can also see higher matters related to the same deposits, like factor 3: Tharsis, Sotiel, and La Zarza and other not observed so clearly like Aznalcollar, Aguas Teñidas and La Magdalena, with values up to 1.1. Different interesting high concentrations are related to some other small VMS ore deposits like Sierrecilla in the eastern part, a Zn deposit, Tinto y Santa Rosa and Cerro Buitron in the central region Lagunazo and San Telmo.

Like in factor 3, some high concentrations need future studies because they do not correlate with ore deposits.

Such as factor 1, we also noted the geochemical differences, between the Variscan rocks of the SNSB with low values of factor scores (from 0.05 to 3.44) and Gil-Marquez with high values (from 0.05 to 1.1), confirming the different origin (Castro et al., 1995).

The other main geological units do not show any particular relation with this factor except the sedimentary materials from the Gualdaquivir Basin in the southeastern part of the project, probably related to crops and the anthropic influence of agricultural fertilizers.

Finally, the F6 association Sm-Nd-Pr-La-Eu-Ce-Gd-Tb-(Th-U), in contrast to (Na), is mainly formed by light REE, and some related elements (Th and U). The elevated values of the factor scores of this association are related to the Culm Group, so these elements have an important affinity with the geological materials of these sandstones and shales in turbiditic sequences (Fig. 12).

Fig. 12
figure 12

Factor scores distribution of element association F6: Sm-Nd-Pr-La-Eu-Ce-Gd-Tb-(Th-U) in contrast to (Na)

Full size image

There is also a correlation between this factor and materials from the PQ Group in the Valverde antiform. Although the lithological control is still being determined, we can again observe the difference between two PQ outcrops, as shown in factor 1.

5 Conclusion

Spatial distribution of chemical elements in the residual soils of the Iberian Pyrite Belt is determined by lithological and metalogenic factors, as well as a partial influence from mining wastes.

Firstly, although there are many studies regarding geology in the IPB, geochemical studies can help to improve the geological maps discriminating and differentiating between zones with similar geology, like both outcrops of the PQ Group that show a different geochemical affinity.

Also, these studies can help to uncover possible non-outcropped bodies like the likely granite in the southwest of Valverde antiform, detailed geophysical studies are proposed in this area to confirm this hypothesis.

Secondly, we could partially discriminate between this metallogenic province's different types of ore deposit characteristics. In addition, some high concentrations of factor scores from F3 and F5 are not related to already known ore deposits, some with high values of critical elements so that they could be new exploration targets for follow-up surveys in the future. This fact could limit the areas for future mining exploration.

Finally, this geochemical database and the factor analyses, combined with geological and geophysical information in GIS analyses, help build exploration models.