Tracing Soil Contamination from Pre-Roman Slags at the Monte Romero Archaeological Site, Southwest Spain

: Soil serves as a repository of human history, preserving artifacts within its horizons. However, the presence of chemically reactive remnants, such as ancient slags, can significantly impact the surrounding soil environment. This paper addresses this scarcely explored issue by focusing on soil contamination arising from pre-Roman slag deposits at the Monte Romero archaeological site in southwest Spain, dating back to the Tartessian period (c. 7th century BC). Through the high-resolution microscopy examination of slag wastes and the trace element analysis of soil samples by ICP-OES, this study evaluated current contamination status using a multi-index approach. The results revealed markedly high levels of Pb (>5000 mg kg − 1 ), Cu (up to 2730 mg kg − 1 ), and As (up to 445 mg kg − 1 ) in the soil compared to a control sample. The identification of secondary complex compounds like Cu arsenates and Pb arsenates/antimonates within slag cavities suggests post-depositional weathering processes, leading to the dispersion of potentially toxic elements into the surrounding soil. Assessments through indices of contamination and potential ecological risk highlighted severe contamination, particularly concerning Ag, Pb, Sb, Cu, and As. This study underscores the importance of addressing potential environmental hazards associated with archaeological sites hosting remnants of metal production.


Introduction
Soil plays a critical role in biogeochemical cycling, sustaining life and providing ecosystem services.It also serves as a repository of our archaeological and geological heritage, hosting anthropogenic artifacts of cultural value that offer insights into ancient human activities [1].However, these artifacts can become sources of environmental contamination, depending on their composition and reactivity within the soil matrix.Archeometallurgical materials often contain elevated concentrations of potentially harmful elements [2,3], posing significant threats to soil quality.Abandoned settlements with long histories of mining and metallurgical activities provide valuable opportunities to study the long-term fate and behavior of pollutants within the soil system [4].Therefore, assessing the concentration of trace elements in soils near archaeological settings displaying indications of potential contamination is a pertinent matter of study.
Soil and sediment contamination by heavy metals derived from metallurgical waste dumps have been documented in the literature for a number of historical smelting sites across various regions, including Britain [5], Central Europe [3,6], Australia [7], South America [8], and prominent ore districts, such as the Colline Metallifere district in southern Tuscany [9] and the Iberian Pyrite Belt [10][11][12].
In metalliferous regions like the Iberian Pyrite Belt, situated in the southwest of the Iberian Peninsula, mining and extracting copper and silver significantly contributed to the civilization and development of ancient societies [13,14], but has also left adverse environmental legacies.Over centuries of intensive smelting activities during the Roman Soil Syst.2024, 8, 78 2 of 12 era, vast quantities of metallurgical waste, estimated at around 30 million tons [15], were generated.These wastes were disposed of near mining sites, forming extensive slag heaps.Evidence of pre-Roman smelting activities is also apparent, with an estimated 4 million tons of slag inferred from surface deposits [16].Remnants of ancient metallurgical processes are scattered across southwest Iberia, with notable concentrations found in Rio Tinto, where 6 million tons of copper and silver slags are piled up in the soil [17].Over time, fractions of arsenic and heavy metals, typically present in metallurgical wastes [18,19], might have migrated into the surrounding ecosystems, potentially causing significant contamination.
While the environmental impact of modern mining practices in the Iberian Pyrite Belt is well documented [20][21][22], and references therein, research concerning soil contamination arising from ancient smelting activities remains notably scarce.This study aimed to address this knowledge gap by examining the current level of soil contamination around the Monte Romero protohistoric smelter and assessing the potential ecological risks associated with the identified trace elements of concern.

The Archeological Site of Monte Romero
Monte Romero (37 • 46 37 N-6 • 47 33 W) is situated in the province of Huelva (Spain), within the municipality of Almonaster la Real, approximately 1 km southwest of the Cueva de la Mora mining village (Figure 1).Late Bronze Age, and marked the establishment of trading links between southwester Iberia and the Eastern Mediterranean.

Sampling and Sample Preparation
Three composite soil samples (MR-1, MR-2, and MR-3) were collected from a barre Geologically, it lies in the northernmost part of the Iberian Pyrite Belt, renowned as one of the world's largest volcanogenic massive sulfide provinces.The Monte Romero orebody is a polymetallic (Pb-Zn-Cu-Ag) massive sulfide deposit hosted in a volcanic-sedimentary sequence dominated by felsic volcanic rocks [23].The deposit was primarily mined for copper, yielding 600,000 tons of ore during the first third of the 20th century [24].Past mining and ore processing activities have resulted in significant deforestation, leading to the destruction of natural vegetation and the conversion of soils into marginal lands, currently supporting sparse vegetation or remaining barren.Eutric Cambisol is the dominant soil group developed on the felsic volcanic rocks [25], although large parts of natural soils have been transformed into mine Technosols.The region has a Mediterranean continental climate with some Atlantic influences, and is characterized by hot, dry summers, and mild, wet winters.
The protohistoric smelter was first discovered in 1975 near the modern mining area of Monte Romero, during the Huelva Archaeo-Metallurgical Exploration Project [26].The archaeological survey at Monte Romero unveiled extensive traces of silver and copper smelting activities.Two distinct types of slag were recovered: (1) free-silica (quart-bearing), plano-convex slags, characterized by a spherical shape with a flattened underside; and (2) flat slags displaying a casting structure and significant matte content.Based on field observations, the volume of the slag deposit is estimated to be roughly 300 m 3 , with an average density of 3.6 g cm −3 , resulting in a total mass of about 1080 tons.Other archaeological finds included grooved stone mining tools, tuyeres bearing litharge impregnations, and fragments of copper-silver matte ( [13], p. 123).The archaeological excavation conducted in 1986 [27] led to the conclusion that Monte Romero served as an important mining center and metallurgical workshop linked to the production of silver and copper from complex polymetallic sulfide ores [28].
The Monte Romero archeological site has been listed in the General Catalog of Andalusian Historical Heritage since 2005 [29].At this site, all stages of mineral processing, from polymetallic ore extraction to cupellation, were conducted through a sophisticated pyrometallurgical process ( [13], p. 127) [30,31].Typological analyses of broken pieces of ceramic materials found at the site, including handmade decorated vessels and Phoenician amphorae, date this protohistoric settlement to the mid-7th to early 6th centuries BC [32].This period, locally termed the Tartessian period, corresponds to the Orientalizing Late Bronze Age, and marked the establishment of trading links between southwestern Iberia and the Eastern Mediterranean.

Sampling and Sample Preparation
Three composite soil samples (MR-1, MR-2, and MR-3) were collected from a barren area (20 × 30 m) within the archaeological site boundaries (Figure 1).This area corresponds to the location of a former smelting workshop, as indicated by the occurrence of scattered slags within the collected soil.Concurrently, representative samples of the slags were also gathered from the same sampling points.In addition, a control soil sample (BJ-0) was obtained from a location (37 • 47 07 N-6 • 48 51 W) positioned outside the modern mining area of Monte Romero, roughly 2 km northwest of the archaeological site.This control sample showed no discernible signs of contamination and served as a benchmark for comparison.
At each sampling site, three discrete samples were collected from the topsoil layer (0-10 cm depth) using a polypropylene hand shovel.The samples were combined in the field to create composite samples weighing approximately 3 kg, and were stored in zip-lock polyethylene bags.Upon reaching the laboratory, the composite samples were spread out on absorbent paper and left to air-dry at room temperature.After drying, mechanical disaggregation was performed with a wooden roller, followed by homogenization and sieving through a 2 mm mesh sieve to obtain a consistent particle size fraction.Representative subsamples were formed by coning and quartering.An aliquot was then finely ground in an agate mortar and passed through a 63 µm sieve to achieve a uniform powder for chemical analysis.

Soil Texture and Electrochemical Properties
Soil texture was determined using particle-size distribution data obtained from laser diffraction analysis performed on a Mastersizer 3000 analyzer (Malvern Instruments Ltd., Worcestershire, UK).Potentiometric measurements of pH and electrical conductivity were conducted by stirring 10 g of soil in 25 mL of deionized water for 5 min, followed by a standing period of 30 min.pH measurements were carried out using a Crison Basic 20 pH meter with a glass electrode calibrated with pH 4 and 7 buffer solutions at 25 • C. Electrical conductivity measurements were conducted using a Crison GLP 31 conductivity meter equipped with a glass electrode calibrated with two standard KCl solutions.

Bulk Chemical Analysis
The soil samples underwent a multi-acid digestion process (HClO 4 + HNO 3 + HF + HCl) and the resulting extract solutions were analyzed by inductively coupled plasma optical emission spectrometry (ICP-OES) at Activation Laboratories (Ancaster, ON, Canada).This laboratory adheres to ISO/IEC 17,025:2017 standard [33], ensuring the reliability and quality of the analytical results.In this study, we analyzed the total concentrations of 19 trace elements, including Ag and several environmentally significant elements, such as As, Bi, Cd, Co, Cr, Cu, Ni, Pb, Sb, and Zn.
The quality of the analytical procedure was controlled by certified reference materials from the OREAS series (13b, 45d, 70b, 72b, 77b, 96, 98, 101b, 247, 521, 620, 681, 753, and 904), reagent blanks, and replicate samples to verify the accuracy and reproducibility of the data (Table S1).The relative standard deviations from duplicates and the certified values for the majority of the reported elements were found to be less than 10%, indicating high data quality.

High-Resolution Microscopy and Microchemical Analysis
To analyze the composition and microstructure of the slag samples, they were carefully fragmented to expose fresh surfaces.Selected fragments were then coated with a thin layer of carbon to enhance their electrical conductivity and examined using a fieldemission scanning electron microscope (FESEM) equipped with an energy-dispersive X-ray spectroscopy (EDS) microanalysis system and detectors for secondary and backscattered electrons.The FESEM (JSM-IT500HR instrument, JEOL, Tokyo, Japan) operated under standard conditions with an accelerating voltage of 20 kV, a working distance of 10 mm, and a variable beam diameter tailored to the size of the analyzed particle.

Indices of Contamination and Potential Ecological Risk
Soil contamination by trace elements was assessed using various indices, including the degree of contamination proposed by Hakanson [34], the pollution load index introduced by Tomlinson et al. [35], and the ecological risk index, also suggested by Hakanson [34].Based on the contamination degree index, Abrahim and Parker [36] classified the contamination level into seven grades, ranging from nil to very low contamination (Cd < 1.5) to ultra-high degree of contamination (Cd ≥ 32).
These indexes rely on the contamination factor (Cf ), computed by dividing the concentration of each element in the soil (C s ) by its concentration in the control sample, taken as the local background value (C b ), as shown in Equation (1): In this study, the control sample BJ-0 was used as a reference.
Soil Syst.2024, 8, 78 The degree of contamination (Cd) was obtained by summing the individual contamination factors (Cfi), as follows (Equation ( 2)): The pollution load index (PLI) was derived from the geometric mean of the contamination factors (Cf ) for the elements of concern, as described in Equation ( 3): The potential ecological risk index (ERI), primarily driven by contaminants such as As, Cd, Pb, and Cu [34], was computed using Equation ( 4): where ERI is the potential ecological risk, Cfi denotes the contamination factor of element i, and Tfi indicates the corresponding toxicological factor according to Häkanson [34] (As = 10, Cd = 30, Cu = Pb = 5).Based on the ERI values, the following grades were used to describe the potential ecological risk: low risk (ERI < 150); moderate risk (150 ≤ ERI < 300); considerable risk (300 ≤ ERI < 600); and very high risk (ERI ≥ 600).

Slag Description
The macroscopic examination of the slag samples revealed distinctive globular and oval morphologies (Figure 2).Upon fracturing, the freshly exposed surfaces displayed a dark gray or black glassy matrix containing numerous, visible (naked eye) irregularly shaped grains of quartz.These features, including the presence of ball-shaped slags with flattened bases, are strongly indicative of free-silica (unreacted quartz) silver slag, a type commonly produced in the Iberian Peninsula during the Tartessian period [37,38].Similar slags have been documented in the archaeological site of Monte Romero [26,27] and other settlements and production centers across southwest Iberia ( [5], pp.79-154) [38].Another common feature observed in the slags is the occurrence of numerous vesicles.These are small, spherical or ellipsoidal cavities formed by the entrapment of gas bubbles during slag solidification, similar to those found in volcanic glass.This vesicularity effectively increases the overall surface area of the slag, thereby enhancing its chemical reactivity.Under binocular microscopy, some of these cavities and fissures appear filled or coated with crystalline aggregates of secondary minerals showing greenish and bluish hues (Figure 3).This suggests the precipitation of newly formed minerals within the open spaces.Chemical analyses of three free-silica type slags from Monte Romero (samples PH-146, PH-148, and PH-160), as reported by Rothenberg and Blanco [26], yielded an average composition of 22.4% silica (SiO 2 ); 19.93% Fe; 3.80% Pb; 0.97% S; 0.67% Cu; 0.50% Zn; 0.17% As; 0.16% Sb; and 87 mg kg −1 Ag.
Based on the above metal contents, the slag heaps at Monte Romero contain approximately 41 t of Pb, 7.2 t of Cu, 5.4 t of Zn, 1.8 t of As, and 90 kg of Ag.To contextualize these figures, the total mass of slag (~1080 t) and contained metals at this archaeological site is significantly smaller than modern slag dumps in the Iberian Pyrite Belt.Contemporary copper smelting operations typically produce slag materials that are several orders of magnitude greater.For instance, the slag dumps adjacent to the derelict pyritic smelter complex of Rio Tinto (10 million tons) contain 43 t of Ag, 1100 t of As, 44,100 t of Cu, and 20,900 of Pb [10].Similarly, the slags generated at São Domingos (4.7 million tons) contain approximately 1000 t of As, 16,000 t of Cu, 14,000 t of Pb, and 40,000 t of Zn [39].The Monte Romero site represents historical pollution from smaller-scale operations with lower technological capabilities, leading to less efficient metal extraction and higher residual metal content in the slag, particularly Ag, Cu, and Pb.
Another common feature observed in the slags is the occurrence of numerous vesicles.These are small, spherical or ellipsoidal cavities formed by the entrapment of gas bubbles during slag solidification, similar to those found in volcanic glass.This vesicularity effectively increases the overall surface area of the slag, thereby enhancing its chemical reactivity.Under binocular microscopy, some of these cavities and fissures appear filled or coated with crystalline aggregates of secondary minerals showing greenish and bluish hues (Figure 3).This suggests the precipitation of newly formed minerals within the open spaces.Another common feature observed in the slags is the occurrence of numerous vesicles.These are small, spherical or ellipsoidal cavities formed by the entrapment of gas bubbles during slag solidification, similar to those found in volcanic glass.This vesicularity effectively increases the overall surface area of the slag, thereby enhancing its chemical reactivity.Under binocular microscopy, some of these cavities and fissures appear filled or coated with crystalline aggregates of secondary minerals showing greenish and bluish hues (Figure 3).This suggests the precipitation of newly formed minerals within the open spaces.Backscattered electron (BSE) images acquired with the FESEM (Figure 4) provided a more detail view of the vesicular surface of the slags and their porous structure, as well as the appearance of minerals precipitated within some vesicles.These minerals exhibited diverse crystalline habits, such as botryoidal (grape-like clusters) and coralloid forms, often forming radial groups of acicular (needle-like) crystals.Backscattered electron (BSE) images acquired with the FESEM (Figure 4) provided a more detail view of the vesicular surface of the slags and their porous structure, as well as the appearance of minerals precipitated within some vesicles.These minerals exhibited diverse crystalline habits, such as botryoidal (grape-like clusters) and coralloid forms, often forming radial groups of acicular (needle-like) crystals.
These secondary minerals may also result from alteration of metallurgical phases within complex slags exposed to weathering for centuries [40].Slag is inherently porous and chemically reactive, which leads to the formation of secondary minerals through reactions with atmospheric oxygen and surface water, typically under conditions with a pH above 4 [41].The precipitation of these secondary phases provides evidence of slag weathering and the subsequent release of trace elements into the surrounding soils, as reported in other ancient smelting wastes [3].At the Monte Romero site, the arsenates filling the slag vesicles can be interpreted as precipitates from meteoric solutions enriched in As and heavy metals.This enrichment likely occurred due the breakdown or destabilization of the primary metallurgical phases of the slags within the soil environment.Dissolution and recrystallization reactions significantly influence the mobility of metal(loid)s in dumps [7].The presence of relatively soluble complex arsenates suggests that historical slag wastes pose a potential environmental threat, acting as sources of As and heavy metals that could be released over time.However, it is noteworthy that repeated cycles of dissolution and reprecipitation commonly lead to the formation of the least soluble phases, thereby potentially reducing long-term environmental risks.

Soil Properties and Total Trace Element Concentrations
Soil samples collected from the archaeological site exhibited a textural class (silt loam) and physicochemical characteristics (slightly acidic reaction and low electrical conductivity) consistent with the control soil.On the other hand, they revealed remarkably higher total concentrations of trace elements, as detailed in Table 2.In fact, all soil samples hosting the Monte Romero slags exhibited exceptionally high levels of Pb, exceeding the upper limit of quantification (5000 mg kg −1 ); hence, the Pb contents reported in Table 2 represent minimum estimates.Copper concentrations were also markedly elevated compared to the control sample, with an average value of 2480 mg kg −1 .Other trace elements of potential concern, such as Zn reaching up to 705 mg kg −1 , and As up to 445 mg kg −1 , were also found at relatively high levels.These findings indicate significant contamination with these potentially toxic elements, with total concentrations comparable to those reported in soils around abandoned mine sites in the Iberian Pyrite Belt [22].
The concentrations of the remaining analyzed elements were all below 100 mg kg −1 .However, the content of Ag in the soil samples (29.4-40.9mg kg −1 ) was notably elevated compared to both the control sample and the average concentration of Ag (0.13 mg kg −1 ) reported in soils worldwide [42].This enrichment suggests a potential anthropogenic influence from past metallurgical activities, likely linked to inefficient metal recovery processes.

Trace Elements of Concern and Potential Ecological Risk
The primary contaminant elements introduced into the soil by the slag wastes (Ag, As, Bi, Cu, Pb, and Sb) are clearly distinguishable when comparing their mean concentrations in the contaminated soil with those in the control soil (Figure 5).This distinct geochemical association is consistent with the chemical composition of the polymetallic ores historically mined at the now-abandoned Monte Romero mine [24].

Trace Elements of Concern and Potential Ecological Risk
The primary contaminant elements introduced into the soil by the slag wastes (Ag, As, Bi, Cu, Pb, and Sb) are clearly distinguishable when comparing their mean concentrations in the contaminated soil with those in the control soil (Figure 5).This distinct geochemical association is consistent with the chemical composition of the polymetallic ores historically mined at the now-abandoned Monte Romero mine [24].Table 3 summarizes the calculated values of contamination factor (Cf ), contamination degree (Cd), pollution load index (PLI), and potential ecological risk index (ERI).
The results revealed that Ag is the most enriched metal in the soil, with an average Cf of 85.75, followed by Sb with an average of 28.93, and Pb with a Cf exceeding 18.80 in all samples.The values are indicative of very high contamination.Other trace elements, including Cu, As, and Bi, also showed considerable contamination (Cf > 4), though to a lesser extent than in most mine soils of the Iberian Pyrite Belt [22].The remaining analyzed trace elements generally displayed Cf values around unity, suggesting a minimal contribution to overall contamination.
The degree of contamination (Cd), derived from the sum of individual Cf values for the elements of concern (Ag, As, Bi, Cd, Cu, Pb, and Sb), exceeded 141 in all samples, indicating severe anthropogenic contamination at the site.Similarly, the PLI, calculated from the geometric mean of individual Cf values, presented relatively high values (PLI > 9), further supporting the conclusion of significant anthropogenic influence.
The right-censored ERI value of 242 surpassed the low ecological risk threshold, indicating moderate-to-high ecological risk [33], although the near-neutral pH level observed in the soil suggests a relatively low potential for metal mobility.This finding is important as it suggests that the contaminants are less likely to migrate through the soil.This multi-elemental approach highlights significant soil contamination at the Monte Romero site, particularly with Ag, Sb, and Pb.The slag mass (~1080 tons) confined in the small smelting workshop area (~600 m 2 ) suggests a localized contamination footprint.However, erosion processes pose the potential threat of dispersing the contaminants over a broader area.This raises concerns about the potential spread of contamination beyond the immediate vicinity, affecting nearby soils and water sources.While the current ecological risk assessment categorizes the situation as moderate, the potential negative impact on the surrounding environment warrants further investigation.

Conclusions
Ancient metallurgical practices left a distinct geochemical signature on the local soil environment of Monte Romero.Trace element analysis unveiled significant concentrations of Ag, Cu, As, Pb, Sb, and Bi in soil samples adjacent to pre-Roman slag sites compared to a control area.This enrichment can be ascribed to post-depositional weathering processes involving the free-silica, ball-shaped slags scattered throughout the site, remnants of silver and copper smelting activities dating back to the Tartessian period (c.7th century BC).
The presence of secondary Cu and Pb arsenates within the slag cavities suggests the breakdown of primary metallurgical phases, leading to the subsequent release and redistribution of potentially harmful elements.Consequently, the surrounding soil acts as a reservoir for these contaminants, exhibiting severe contamination primarily driven by Pb, Sb, and Ag, along with a moderate ecological risk associated with As, Pb, and Cu.
These findings underscore the long-lasting environmental repercussions of historical smelting operations on soil quality and the need to implement proper management strategies for archaeological sites hosting remnants of metal production.

Figure 1 .
Figure 1.Geographic setting of Monte Romero, showing (a) the archaeological site and (b) the sam pling locations within the site.

Figure 1 .
Figure 1.Geographic setting of Monte Romero, showing (a) the archaeological site and (b) the sampling locations within the site.

Figure 2 .
Figure 2. General appearance of the free-silica silver slags recovered from Monte Romero.

Figure 2 .
Figure 2. General appearance of the free-silica silver slags recovered from Monte Romero.

Figure 2 .
Figure 2. General appearance of the free-silica silver slags recovered from Monte Romero.

Figure 3 .
Figure 3. Close-up view of cavities and fissures within the slag, lined with secondary minerals exhibiting (a) greenish and (b) bluish hues.

Figure 3 .
Figure 3. Close-up view of cavities and fissures within the slag, lined with secondary minerals exhibiting (a) greenish and (b) bluish hues.

Figure 4 .
Figure 4. High-resolution FESEM-BSE images depicting (a) the vesicular surface of the slag; (b) botryoidal habit of secondary minerals; (c) coralloid habit of secondary minerals; and (d) radial aggregates of acicular olivenite crystals filling a vesicle.

Figure 4 .
Figure 4. High-resolution FESEM-BSE images depicting (a) the vesicular surface of the slag; (b) botryoidal habit of secondary minerals; (c) coralloid habit of secondary minerals; and (d) radial aggregates of acicular olivenite crystals filling a vesicle.

Figure 5 .
Figure 5. Isocon diagram illustrating the enrichment of trace elements, notably Ag, As, Bi, Cu, Pb, and Sb, in soil hosting the slags.Contaminant elements (red dots) falling below the diagonal line (isoline) indicate their significant enrichment compared to background levels (control sample).The reported Pb content in the affected soil is considered a minimum estimate.Table3summarizes the calculated values of contamination factor (Cf), contamination degree (Cd), pollution load index (PLI), and potential ecological risk index (ERI).

Figure 5 .
Figure 5. Isocon diagram illustrating the enrichment of trace elements, notably Ag, As, Bi, Cu, Pb, and Sb, in soil hosting the slags.Contaminant elements (red dots) falling below the diagonal line (isoline) indicate their significant enrichment compared to background levels (control sample).The reported Pb content in the affected soil is considered a minimum estimate.

Table 1 .
Chemical composition (wt%) of representative secondary minerals identified within slag cavities, as determined by energy-dispersive X-ray spectroscopy (EDS).

Table 1 .
Chemical composition (wt%) of representative secondary minerals identified within slag cavities, as determined by energy-dispersive X-ray spectroscopy (EDS).

Table 2 .
Soil properties and total concentrations of trace elements analyzed by ICP-OES in the <2 mm fraction of potentially contaminated soil samples and the control sample.Right-censored Pb values were replaced with the upper limit of quantification (5000 mg kg −1 ). * Right-censored Pb values were replaced with the upper limit of quantification (5000 mg kg −1 ). *

Table 3 .
Contamination factor (Cf for each trace element per sample and their respective averages, alongside multi-elemental contamination indices: degree of contamination (Cd), pollution load index (PLI), and ecological risk index (ERI).