The Ríotinto mining district, located within the Iberian Pyrite Belt, is one of the world’s largest polymetallic mining areas, with a rich history dating back to Roman times. Extensive mining activities have significantly impacted the environment, including landscape alteration, deforestation, and pollution of soil, rivers, and air. This is largely due to open-pit mining, timber use in smelting, acid rain from sulfur emissions, and contamination from mine waste dumps exposed to erosion. The surrounding soils, within a 5–7 km radius, contain elevated concentrations of potentially toxic elements (Table 1) (Vázquez‐Arias et al., 2023).

Currently, the mine is operated as an open-pit copper mine by Atalaya Mining, processing 15.5 million tons of ore annually with a copper grade of 0.37% and a cut-off grade of 0.14%. The site covers 35 km² and is divided into three main areas: the southern sector (with benches and waste dumps), the central sector (housing processing facilities), and the northern sector (tailings ponds and water management). The tailings ponds—Gossan, Cobre and Aguzadera—are situated within the Odiel river basin (Diaz et al., 2024).

Soil was collected from the site designated for a future field experiment (37°41′26.4″N, 6°34′01.8″W), located within the Riotinto mining area. To obtain a representative sample of the site, systematic sampling was conducted across an area of approximately 600 m² in collaboration with Atalaya Minning. Fifteen evenly distributed sampling points were selected across the site. At each point, around 10 kg of soil were extracted from the top 0–20 cm of the soil profile using a shovel. The collected soil was placed in individual sacks, transported to the greenhouse, and air-dried. To minimize spatial variability, the contents of the sacks were thoroughly mixed and homogenized before being used in the experiment.

The experiment took place at the experimental farm “La Hampa” (Coria del Río), part of the IRNAS-CSIC research facilities. A total of 65 pots (2 L capacity each) were prepared: 50 were filled with soil from Riotinto (RT) and 15 with control soil (C) sourced from agricultural land on the same farm. The RT and C soils differ markedly in their physicochemical characteristics. The RT soil is highly acidic, with a pH of 3.07 (in KCl) and 3.28 (in water), whereas the C soil is alkaline with a pH of 7.91 (in water), indicating contrasting acidity–alkalinity regimes. Electrical conductivity (EC) is substantially higher in RT (0.896 dS m^-1) compared to C (0.231 dS m⁻¹), reflecting a greater concentration of soluble salts in the contaminated soil. Organic matter (OM) content is notably depleted in RT (0.30%) relative to C (2.54%), suggesting poor nutrient availability and limited biological activity. In terms of texture, both soils are classified as sandy loam.

Given the strongly acidic nature of the contaminated RT soils, a sugar lime amendment (SL) was incorporated at a rate of 30 t ha⁻¹ (RT + A) in the first 10 cm of the soil of the pots. Sugar lime, which contains 70%–80% calcium carbonate (CaCO₃), effectively neutralizes soil acidity through chemical buffering reactions, gradually raising the soil pH from highly acidic to near-neutral levels (Madejón et al., 2010). This amendment was necessary to improve the soil chemical properties and create favorable conditions for plant establishment. Due to the high complexity, cost, and extended duration of the hybrid field experiment, applying the amendment prior to planting was essential to ensure plot viability and plant survival. A non-amended control soil was not included, as the extreme acidity of the native RT soil would have precluded plant growth and compromised the feasibility of the experimental design.

Five poplar varieties, consisting of one parental species and four hybrids, were evaluated in this study. The hybrids were supplied by Phytowelt (Germany), an industrial partner in the European EDAPHOS project. Hybrid P10 was derived from a cross between Populus nigra—recognized for its high wood quality—and Populus maximowiczii, known for its rapid growth rate. The additional hybrids, P148, P149, P152, and P153, were generated from backcrossing P10 with Populus alba. The latter species is characterized by drought and pathogen resistance and its ability to thrive in nutrient-poor soils, although the hybrids used in this study have not been previously assessed for growth in contaminated environments.

In early November 2023, uniform cuttings (25–30 cm in length) containing only buds (no leaves) were planted for each hybrid. A total of 13 cuttings per hybrid were established, with 10 planted in contaminated Riotinto soil and 3 in uncontaminated control soil. After planting, sprinkler irrigation was scheduled for 3 min, three times per week, to maintain adequate moisture.

Physiological monitoring began 8 weeks post-planting, coinciding with leaf emergence. Every 2 weeks, leaf chlorophyll content was assessed on all five poplar varieties using a SPAD-502 chlorophyll meter (Konica Minolta, Japan). For each plant, three measurements were taken from comparable leaves and averaged to obtain individual chlorophyll values (Chang and Robinson, 2003). The plants were maintained under these conditions for a total of 3 months.

The experiment was carried out in a controlled-environment growth chamber at IRNAS-CSIC research facilities. Environmental conditions were maintained with a 16-h photoperiod (light from 07:00 to 23:00) and a constant temperature of 24 °C. A total of 24 pots (1 L each) were prepared for each plant species, with eight pots assigned to each soil type (RT, RT + A, and C), consistent with the hybrid poplar experiment setup.

To evaluate the performance of the companion herbaceous species under highly acidic conditions, seeds were also sown in untreated RT soil without amendment. For B. juncea, 1 g of seeds was evenly distributed per pot, whereas for L. purpureus, four seeds were planted per pot. Plants were irrigated every 2 days to maintain optimal soil moisture and grown for 1 month.

Physiological measurements were performed during the third and fourth weeks of growth. Leaf chlorophyll content was assessed in L. purpureus using a SPAD-502 chlorophyll meter (Konica Minolta, Japan), following the same protocol used in the poplar hybrid experiment. Chlorophyll measurements could not be obtained for B. juncea due to its limited leaf size at this stage.

After 1 month of growth, the entire aerial biomass from each pot was harvested. Fresh biomass was recorded immediately, followed by oven-drying at 70 °C for 48 h to determine dry biomass. The dried plant material was subsequently analyzed for nutrient and trace element concentrations.

Soil samples were collected both at the start and end of the experiment. Samples were air-dried at 40 °C, sieved to <2 mm, and analyzed for chemical properties. Soil pH was measured in a 1 M KCl extract (1:2.5, m/v) using a CRISON micro pH 2002 m, and electrical conductivity (EC) was determined in a 1:5 soil-to-water extract. Total organic carbon (TOC) content was quantified using a ThermoFlash 2000 NC analyzer. For trace element determination, soil was finely ground and sieved to <60 μm. Pseudototal trace element concentrations were determined via aqua regia digestion (HNO₃:HCl, 1:3) using a Digiprep MS block digester (SPS Science), while bioavailable fractions were estimated through extraction with 0.01 M CaCl₂ (1:10, m/v) following Houba et al. (2000). Trace elements in all the extracts were determined by ICP-MS (Inductively Coupled Plasma Mass Spectrometry (Agilent 7800). For plant analyses, only leaves of the poplars were sampled to determine nutrient and trace element concentrations. Due to the limited vegetative growth observed, all leaves from each pot were harvested at the end of the 3-month period to ensure adequate biomass for chemical analysis. For each hybrid and soil type, leaves from 3 to 5 pots (depending on the available biomass) were combined to create a composite sample, resulting in three replicates per hybrid and soil. In the case of B. juncea and L., purpureus for the RT + A and RT soils (RT only for Lablab), two to three pots per soil type were selected—depending on biomass availability—to prepare composite samples, also resulting in three replicates per treatment.

Vegetal samples were rinsed in a 0.1 N HCl solution for 15 s, followed by a rinse in distilled water for 10 s, and then oven-dried at 70 °C. Dried plant material was ground and sieved through a 500-μm stainless-steel sieve before analysis preparation.

Trace elements in plant tissues were determined using ICP-MS after wet digestion with concentrated HNO₃ in a Digiprep MS block digester. Analytical quality was ensured using reference materials, with ERM CC141 for soil and INCT-OBTL-5 (tobacco leaves) for plants, achieving recovery rates between 90% and 105%. All trace elements and nutrients concentrations were expressed on dry weight.

For data analyses, the IBM SPSS Statistics 29.0 software was used. Analysis of variance (one-way ANOVA) and Tukey HDS post-hoc tests were conducted to evaluate the effect of different soil on the pH, TOC, psuedototal and available trace elements in soils and plants. Previously, the data were subjected to the Kolmogorov-Smirnov test to check for normal homogeneity of variances. If the distribution was not normal, a logarithmic transformation was applied, and the normality test was performed again. If the Kolmogorov-Smirnov test was still not satisfied, a non-parametric test was used (Mann-Whitney U for mean comparisons and Kruskal-Wallis for variance analysis). In all cases, a significance level of p < 0.05 was employed.

The addition of the amendment (SL) raised the soil pH by 5 units, from 3.10 to 8.15, in the top 10 cm of soil. An increase in EC was also observed after the amendment addition (from 0.896 to 1.422 dS m⁻¹). The RT soil had low soil TOC, which was improved from 0.30% to 0.46% after the SL addition. This increase in OM may also enhance the low fertility of this soil.

At the beginning of the experiment, the pseudototal concentrations of trace elements were measured in both the C and RT + A soils. The analysis revealed significantly higher concentrations of all examined trace elements in the RT + A soils. Among these, Pb exhibited the greatest increase relative to the control, followed by As, Cd, and Cu, while Zn showed the smallest difference (Table 2). Notably, the addition of the amendment did not affect the total metal content in the soil (data not shown).

The area of the contaminated soils corresponds to the Iberian Pyrite Belt, specifically the South-Portuguese zone (Galán et al., 2008). The geochemical background values on this zone are 47.5, 46.4, 143 and 89 mg kg⁻¹ for As, Cu, Pb and Zn. The RT soils far exceed these values, with Pb concentrations 77 times higher than background levels, Cu 45 times higher, and As 20 times higher.

Trace element availability was assessed in C, amended RT + A and unamended RT soils before the experiments (Table 3). As expected for mine-impacted soils, CaCl₂-extractable concentrations were much lower than the pseudototal values. Notably, Pb and especially As exhibited extremely low availability, with concentrations below 0.01% of their pseudototal content. In contrast, Cd and Zn displayed higher availability, reaching approximately 10% of their pseudototal concentration.

A key difference between RT and RT + A soils, was the previously mentioned variation in pH values. This resulted in a drastic and significant reduction in the availability of cationic trace elements (Cd, Cu, Pb, and Zn) in the RT + A soil. However, in the case of As, which exists predominantly as an anion, the pH increase resulted in a substantial rise in its availability, with concentrations increasing from 0.004 mg kg⁻¹ at pH 3 in RT soils to 0.30 mg kg⁻¹ in RT + A soils. It should be noted that the CaCl₂ extraction method has limited reliability in accurately estimating As availability. Therefore, the results obtained using this method should be interpreted with caution and considered as indicative rather than definitive.

After 3 months of growth, poplar development was evaluated based on the proportion of cuttings producing green leaves in each soil type (Table 4). In the control soil, the parental hybrid P10 demonstrated the highest growth performance, with 100% of cuttings producing green leaves. In the RT + A soils, hybrid P152 showed the highest proportion of cuttings producing green leaves (80%), followed by P10 and P153. Conversely, hybrids P148 and P149 exhibited poor performance in RT + A, with only 10% and 30% of cuttings producing green leaves, respectively, indicating limited suitability for these conditions. Moreover, the low survival rates observed under no-contaminated soil suggest that additional factors may be influencing the growth limitations of these two hybrids.

Brassica juncea failed to establish in the untreated RT soil, and its biomass in the RT + A soils was significantly lower than in the control soil (1.00 ± 0.13 g vs. 2.86 ± 0.10 g dry weight; U = 0.00, p < 0.001). In contrast, L. purpureus successfully germinated and grew in all soil types. Due to its larger plant size compared to B. juncea, biomass was expressed on a per-plant basis, preventing direct species-to-species comparisons. Mean biomass values were 0.81 ± 0.01 g in C soil, 0.44 ± 0.20 g in RT + A, and 0.18 ± 0.03 g in RT. Biomass differences across the three soils were statistically significant (F = 28.397, p < 0.001).

Values of SPAD chlorophyll readings were recorded at several time points. For poplar hybrids, the figure presents data collected at 2 and 3 months after planting, while for L. purpureus, measurements were taken at 3 and 4 weeks of growth (Figure 1). SPAD measurements were not possible for B. juncea due to its very small leaves. No significant differences in SPAD values were detected among poplar hybrids or between soil types (control and RT + A); however, hybrids P148 and P149 could not be fully assessed because most cuttings failed to produce green leaves. In contrast, L. purpureus displayed significant variation in SPAD values among soil types both after 3 weeks (F = 3.973, p < 0.05) and 4 weeks (F = 13.805, p < 0.001) of growth. After 4 weeks, SPAD values indicated that control (C) and RT plants reduced their chlorophyll levels (Figure 1).

Consistent with the soil chemical analyses, the most pronounced differences between control and RT + A soils were observed for Pb and As (Figure 2). Statistical analysis was conducted only for hybrids P152, P153, and P10 due to limited sample availability for P148 and P149 (single data point per condition). Significant differences in trace element accumulation were detected between C and RT + A soils for all analyzed elements in poplars. Lead and As exhibited the largest differences, mirroring their high soil concentrations. Variation among hybrids was significant only for Cd (F = 6.102, p = 0.012) and Fe (F = 3.857, p = 0.045). Overall, hybrids P152 and P153 presented the highest concentrations of most trace elements (As, Cd, Cu, Fe, Hg, Mn, Pb, and Zn). Concentrations of As, Cd, and Pb exceeded normal leaf thresholds for all hybrids, while Mn and Zn occasionally surpassed reference levels (Kabata-Pendias, 2011). No established reference values were available for Fe.

For B. juncea, trace element accumulation was significantly higher in RT + A compared to C soils for all elements except Ni, which was not elevated in the soil (Figure 3). Arsenic levels were particularly high, and Cu, Pb, and Zn concentrations exceeded those found in poplar leaves after 4 weeks of growth.

Lablab purpureus successfully grew even in RT soils. Overall, trace element concentrations in aerial biomass were lower than those measured in B. juncea and poplar hybrids, despite the higher trace element availability in RT soils (Figure 4; Table 3). Except for Pb in RT, concentrations did not exceed normal plant thresholds.

Macronutrient concentrations in poplar hybrid leaves are displayed in Figure 5. No significant differences were detected among hybrids within the same soil. Potassium showed the lowest values in RT + A leaves, with significantly higher levels in C soil for hybrids P152, P153, and P10 (F = 49.069, p < 0.01). Similar trends were observed for P148 and P149. Phosphorus concentrations were also significantly lower in RT + A soils (F = 5.009, p < 0.001), whereas Mg and S levels were significantly higher in RT + A compared to C soils (F = 15.207, p < 0.001 and F = 87.852, p < 0.001, respectively). No significant differences were found for Ca or Na.

In B. juncea, nutrient patterns largely mirrored those of the poplar hybrids (Figure 6). Ca and K levels were significantly lower in RT + A plants (F = 107.229, p < 0.001 and F = 201.593, p < 0.001, respectively), while Mg, Na, and S concentrations were significantly higher compared to C plants (F = 202.127, p < 0.001; F = 267.457, p < 0.001; F = 508.134, p < 0.001). Phosphorus contents remained similar between both soils.

For L. purpureus, nutrients distribution followed a pattern similar to that observed in B. juncea (Figure 7). Growth at pH 3 (RT soil) notably influenced the results. Calcium concentration was lowest in the RT soil (H (2) = 0.027), while K levels differed significantly across all three soils (F = 157.314, p < 0.001). Magnesium and S concentrations were significantly higher in plants grown in contaminated soils (RT and RT + A) compared to controls (F = 29.972, p = 0.001 and F = 78.215, p < 0.001, respectively). Additionally, P content was significantly greater in RT plants than in those from the RT + A and C (F = 66.825, p < 0.001).

The soil at the Ríotinto mine is highly acidic, which severely limits plant establishment and growth (Che et al., 2023). Previous research has demonstrated the effectiveness of soil amendments, such as SL, in ameliorating acidic conditions and promoting vegetation development. For example, Pérez de Mora et al. (2011) reported that SL application facilitated spontaneous vegetation colonization, while Ciadamidaro et al. (2013) showed enhanced growth and survival of P. alba saplings following SL treatment.

The pseudototal concentrations of trace elements in the studied soil were significantly higher than the commonly accepted tolerance levels for plants, especially in the case of As (15 mg kg⁻¹), Cu (200 mg kg⁻¹), and Pb (500 mg kg⁻¹) (Mendez and Maier, 2008). The ecological and toxicological effects of contaminated soils are mainly determined by the bioavailable fraction of the contaminants. This fraction dictates the actual exposure levels for soil organisms and the potential for these elements to move through the food chain (Chen et al., 2023). Specifically, the transfer of trace elements from soil to plants—and the related health risks—depends largely on this soluble, bioavailable fraction (Antoniadis et al., 2019).

Trace element mobility can change substantially over time as soil properties evolve. An increase in soil pH is particularly relevant, as it typically reduces the solubility of cationic metals and thereby decreases their bioavailability for plant uptake (Antoniadis et al., 2017). In the present study, SL application increased soil pH by approximately five units, substantially reducing the availability of most cationic trace elements. Notably, the bioavailable fractions of Pb and As in untreated RT soils were already low relative to their total concentrations. This reduced mobility aligns with the natural aging of anthropogenic contaminants, which typically results in decreased solubility and plant transfer potential over time (Antoniadis et al., 2017).

It is important to emphasize that As behaves differently from most trace metals. While cationic metals typically exhibit reduced bioavailability with increasing pH, As exists predominantly as an anion in soils. Consequently, As adsorption capacity tends to decrease under higher pH conditions, leading to increased mobility and potential bioavailability (Beesley et al., 2014). Consequently, pH amendments such as SL may have contrasting effects on arsenic mobility compared to other metals, influencing its environmental fate and potential uptake by plants.

Trace elements can severely impair plant growth by disrupting physiological and biochemical functions (Bharti and Sharma, 2022). This detrimental effect was evident in the poplar hybrids, where a reduced proportion of cuttings produced green leaves in the contaminated soil.

Soil acidity, nutrient deficiencies, and elevated levels of potentially toxic elements in mine soils strongly limited germination. Brassica juncea failed to germinate in untreated RT soil, consistent with findings from Fernández-Landero et al. (2024), who also reported poor germination of this species in acidic mine soils. Clemente et al. (2005) similarly demonstrated that soil pH strongly constrains plant growth and biomass production. In contrast, L. purpureus successfully germinated and survived in untreated contaminated soil, albeit with delayed emergence and reduced biomass. Its resilience is likely linked to larger seed size, which provides additional nutritional reserves during early growth stages. However, high trace element availability and low fertility, particularly potassium deficiency (threefold lower in RT than in control soils), contributed to reduced biomass, consistent with reports linking K availability to Lablab productivity (Ruthrof et al., 2018).

Trace elements are known to disrupt photosynthesis by impairing chlorophyll synthesis and electron transport (Bharti and Sharma, 2022). Surprisingly, chlorophyll content (SPAD values) did not significantly differ between poplar hybrids grown in control and RT + A soils, suggesting a potential tolerance to contaminated conditions. Similarly, L. purpureus showed no significant SPAD differences between RT + A and control soils, with only untreated RT soil causing reduced chlorophyll levels. This indicates that the amendment mitigated stress effects and that L. purpureus possesses physiological adaptations—such as efficient nutrient uptake or tolerance mechanisms—that support photosynthetic activity under contamination (Naeem et al., 2020).

Poplars, as members of the Salicaceae family, are well adapted to nutrient-poor soils and high trace element concentrations (Punshon, 2001). Their capacity for Cd and Zn accumulation was evident, particularly in hybrids P152 and P153, despite low soil bioavailability for these metals. Similar leaf-to-soil concentration patterns have been reported by Madejón et al. (2004), supporting the role of poplars as biomonitors. Arsenic and Pb were the dominant contaminants in the study site. Soil As concentrations exceeded those reported in similar studies (Madejón et al., 2004), whereas Pb levels were exceptionally high. Although Pb is typically immobile in soils, a meta-analysis by Tőzsér et al. (2023) showed substantial leaf accumulation in poplars, suggesting the need for further investigation into Pb uptake dynamics and remediation potential.

Brassica juncea accumulated the highest metal concentrations in aerial tissues among all studied species, especially As, Cu, Hg, and Pb in amended contaminated soil, despite the low availability caused by the addition of lime to this acidic soil (Rani et al., 2023). L. purpureus, while less efficient in trace element accumulation, demonstrated exceptional tolerance, growing in untreated contaminated soil without exceeding normal tissue concentrations (except for Pb in RT soils). Its leguminous nature, stress tolerance, and adaptability to challenging environments (Aguilar-Garrido et al., 2023) highlight its suitability for use as a pioneer species in phytoremediation, particularly in regions like the Iberian Pyrite Belt.

Trace metals often disrupt nutrient uptake by competing with essential macronutrients such as Ca and Mg, leading to nutrient imbalances that impair plant health (Angon et al., 2024). In this study, the low K and Ca concentrations observed in poplar hybrids under the RT + A likely resulted from the reduced fertility of contaminated soil compared to agricultural control soil, as well as competitive interactions between essential nutrients and trace metals. Nonetheless, K concentrations remained slightly below the adequate threshold (1.5%) (Reuter and Robinson, 1997), suggesting only mild deficiency. Magnesium concentrations were elevated in contaminated soils, exceeding adequate thresholds (0.15%) and values reported in similar environments (Madejón, 2003), possibly due to enhanced uptake under stress conditions. Sulfur concentrations were also higher, reflecting naturally elevated soil S levels. Conversely, phosphorus was deficient in RT + A soils (below the 0.3% adequacy threshold), which may limit biomass production. No clear Ca uptake trend emerged, suggesting additional soil or physiological factors influence its absorption.

For B. juncea, nutrient concentrations were similar between soil types, although Ca and K deficiencies were evident. L. purpureus generally exhibited low nutrient levels, except for K, indicating potential reliance on efficient nutrient use or tolerance mechanisms for survival in poor soils. The higher concentrations of K, P, and S in RT compared to RT + A soils may stem from lower biomass that can produce a concentration effect, as well as pH-driven differences in nutrient availability within the surface soil layer.

The optimal amendment dosage should be tailored to the specific remediation objective. When aiming to enhance phytoextraction, the amendment rate should be carefully calculated to raise soil pH only to the extent necessary to improve soil productivity, without creating overly alkaline conditions that decrease the availability of cationic trace elements. In contrast, for phytostabilization strategies, applying a higher amendment dosage to maintain a more alkaline pH is preferable, as this reduces cationic metal mobility and limits their uptake by plants. In the case of As, the increase in pH did not result in concentrations that are environmentally significant. Among the poplar hybrids evaluated, P152 and P153 exhibited the highest proportions of cuttings producing green leaves identifying them as the most promising candidates for phytoremediation applications, irrespective of the specific remediation strategy. In this study, the low availability of Cd and Zn in the soil limited their potential for phytoextraction; thus, these hybrids primarily served as bioindicators for these elements. Under conditions with greater Cd and Zn availability, however, they may be more effective for phytoextraction. For most other trace elements, with the exception of Pb, P152 and P153 appear better suited for phytostabilization approaches. Given the extremely high Pb concentrations in the contaminated site, additional research is required to elucidate Pb uptake mechanisms in poplars and to assess their potential contribution to Pb remediation.

Both companion herbaceous species, B. juncea and L. purpureus, demonstrated notable potential for phytoremediation in contaminated soils. Interestingly, L. purpureus exhibited strong tolerance to untreated contaminated soils, sustaining growth despite adverse physicochemical conditions. This resilience underscores its ecological value and suitability for future research. Furthermore, as a nitrogen-fixing legume, L. purpureus could function as a pioneer species in phytoremediation strategies, contributing to soil recovery by enhancing nitrogen availability and improving soil structure.

These results highlight the necessity of the selection of most appropriate plant species and amendments according to site-specific conditions and remediation objectives. Integrating both phytoextraction and phytostabilization approaches may provide a more efficient and sustainable solution for restoring contaminated soils.