1 Introduction

Tailings in many abandoned mines are a well-known environmental problem causing pollution of soils, surface water, and groundwater (Li et al., 2014; Pinto et al., 2015) and serious health problems (Gebeyehu & Bayissa, 2020; Singh et al., 2014). Mining activity leaves high levels of heavy metals in soils (Giri et al., 2017; Nguyen et al., 2020; Wei et al., 2018), which can contaminate large areas of neighboring lands due to the dispersion of fine-grained mine wastes (Boussen et al., 2010; Cheng et al., 2018; Othmani et al., 2013). The environmental risks associated with soil contamination by heavy metals are even greater in semi-arid areas, where extended periods of severe drought, seasonal rainfall variability, and the scarcity of vegetation cover amplify soil degradation (Mendez & Maier, 2008). Most of heavy metals are non-essential elements and, over time, are toxic to plants because of their non-biodegradable nature (Giri et al., 2017; Ma et al., 2015).

Globally, the extent of heavy metal contamination is steadily increasing in agricultural soils through eolian and run-off processes (Ghorbel et al., 2010; Gutiérrez et al., 2016). This leads to their accumulation in soils at toxic levels, which reduces soil fertility and crop yields (Cheng et al., 2018; Kicińska & Wikar, 2021; Nagajyoti et al., 2010). Heavy metals can also be absorbed and accumulated in cultivated plants and eventually transferred into the food chain (Bhatti et al., 2018; Boussen et al., 2013; Gebeyehu & Bayissa, 2020). Moreover, heavy metals affect plant development processes including seed germination, metabolism, and plant growth (Khan et al., 2015). They induce oxidative stress through the accumulation of reactive oxygen species (Hasanuzzaman et al., 2020), resulting in high production of malondialdehyde and proline and increased electrolyte leakage (Filippou et al., 2014; Gill & Tuteja, 2010; Hameed et al., 2016).

Most mining tailings are deposited in inadequate facilities with the absence of proper mine land reclamation (Doumas et al., 2018; Karaca et al., 2018). Across the globe, more than 5 million sites are potentially contaminated by heavy metals (He et al., 2015). In northern Tunisia, about 50 mining districts were listed with high concentrations of heavy metals (Elouear et al., 2016). The level of heavy metals in the soils surrounding these mines indicated that the permissible limits have been significantly exceeded (Boussen et al., 2010; Chaabani et al., 2017; Ghorbel et al., 2010; Othmani et al., 2013). In Jebel Ressas, mining activity (1869–1956) generated about 2.106 tons of toxic wastes (Elouear et al., 2016). In addition, lead (Pb), cadmium (Cd), and zinc (Zn) are among the most widespread contaminants (Mlayah et al., 2009). The average concentrations of Pb, Cd, and Zn in the natural soil were 32 mg.kg−1, 0.58 mg.kg−1, and 50 mg.kg−1, respectively (Pais & Jones, 1997). Our recent study showed high levels of heavy metals Pb (15 587 mg.kg−1), Zn (37 766 mg.kg−1), and Cd (181 mg.kg−1) in soils surrounding Jebel Ressas mine, which are mostly devoted to agricultural crops (Hachani et al., 2020). Pb and Cd are considered among the most phytotoxic heavy metals (Amari et al., 2017), while Zn is regarded as one of the most mobile elements (Gutiérrez et al., 2016).

Numerous studies have suggested methods for removing heavy metals from contaminated lands (Rascio & Navari-Izzo, 2011; Khalid et al., 2017; Karaca et al., 2018). However, the availability of heavy metals is controlled by certain factors such as pH, redox potential, cation exchange capacity, soil texture, and organic matter (Khan et al., 2015). Soil organic matter is an important element that improves the physicochemical properties and fertility of soil (Feller et al., 2012; Wang et al., 2020). A positive relationship was found between the amount of organic amendment and the heavy metal detoxification efficiency (Bacchetta et al., 2015; Liu et al., 2018; Medynska-Juraszek et al. 2020). Besides, the addition of organic amendments increases the availability of metals and improves plant growth (Wang et al., 2020). However, the selection of appropriate plant species to use as amendments is essential to ensure effective removal of heavy metals. In this context, the use of Fabaceae species as green manure can be considered a promising path for research and a suitable tool to increase soil fertility and promote the immobilization of heavy metals. Fabaceae species are able to increase the organic matter content of soils (Alcántara et al., 2017), improve biological nitrogen fixation (Cooper & Scherer, 2012), immobilize toxic elements in the roots (Mahmud et al., 2020), and grow on contaminated soils (Alexander et al., 2006; Ramos et al., 2020).

This work is part of a phytoremediation project for abandoned mining sites combining woody species and agricultural crops, in agroforestry systems. Thus, two experiments are carried out simultaneously: one involving woody species (Hachani et al., 2020, 2022) and the other using Fabaceae species to improve phytostabilization of heavy metals. The realization of this project will contribute to improve scientific knowledge and operational applications in terms of restoration of mining sites using agroforestry in semi-arid areas in the Mediterranean area. The purpose of the present study is to identify the most suitable species to be used. This was carried out through (i) the evaluation of the effects of excess Pb, Zn, and Cd on emergence and seedling growth of four Fabaceae species during germination and early seedling stage, and (ii) the assessment of the effects of multi-metal (Pb, Zn, and Cd) contaminated soil on some traits of oxidative stress in order to study the relationships between metal toxicity, oxidative stress, and tolerance responses.

2 Material and Methods

2.1 Site Description, Soil Sampling, and Laboratory Analyses

The study area is the abandoned Pb–Zn mining site of “Jebel Ressas” located 30 km south of the Tunisian capital (36°36′21.4″N, 10°19′04.0″E). The region is characterized by a semi-arid climate. The mine is located near the former miners’ village and surrounded by agricultural lands (Ghorbel, 2012). Near the mine and the village, there are three mine disposal dumps (tailing I, II, and III) (Fig. 1). Contaminated soil samples were collected from the slope of the tailing (II) and control soil was collected from a non-contaminated area as described by Hachani et al. (2020). Analyses of contaminated and control soils were performed using three soil composite samples (six soil samples per composite sample) as described by Hachani et al. (2020). To determine heavy metal concentrations, soil samples were subjected to acid digestion then metal concentrations were determined by ICP-OES (Perkin-Elmer Optima 4300) (Hachani et al., 2020). The analysis revealed concentrations of 15.58 ± 0.79 mg.g−1 for Pb, 37.76 ± 3.21 mg.g−1 for Zn, and 0.18 ± 0.03 mg.g−1 for Cd in contaminated soil. However, in control soil, concentrations were 0.009 ± 0.0004 mg.g−1 for Pb and 0.021 ± 0.002 mg.g−1 for Zn while Cd was below detection limits (Hachani et al., 2020). A physicochemical characterization of both soils (pHwater, pHCaCl2, electrical conductivity, mineral elements, etc.) is thoroughly presented in our previous publication (Hachani et al., 2020).

Fig. 1
figure 1

Localisation of Jebel Ressas abandoned mining site in North Tunisia

Full size image

2.2 Plant Material, Experimental Design, and Germination Conditions

Seeds of Vicia faba L (variety Badi), Cicer arietinum L (variety Beja 1), Lens culinaris Medik (variety El Kef), and Medicago arborea L were selected for their healthy aspects and their homogeneity. Seeds were provided by the laboratory of field crops at the National Institute of Agronomic Research of Tunisia (INRAT) with the exception of Medicago arborea, which was furnished by the seeds department at the National Institue of Research in Rural Engineering, Water and Forests (INRGREF, Tunisia).

The experiment was conducted at the National Institue of Research in Rural Engineering, Water and Forests (INRGREF) in Tunis. Germination tests were performed according to the recommendations of the International Seed Testing Association (ISTA) (2012). For each species (V. faba, C. arietinum, L. culinaris, and M. arborea), two treatments were applied: NCS using non-contaminated (control) soil and CS using contaminated soil. The seedlings were distributed according to an experimental design in four complete random blocks. The tests were conducted in plastic trays (length: 30 cm \(\times\) width: 20 cm \(\times\) depth: 7 cm) filled with 1 kg of soil (contaminated or control). Seeds were surface sterilized in a 3% sodium hypochlorite solution (v/v) for 10 min, rinsed twice with distilled water and placed on the topsoil. Then the soil was slightly turned to bury the seeds at a depth of 1 cm. Four repetitions of 100 seeds were made for each species according to the rules of ISTA (2012). Trays were irrigated to field capacity with tap water. The trays were incubated at 25 °C and 85% relative humidity for 15 days and the emergence was monitored every 24 h. Seeds were considered germinated when the cotyledon emerge on the surface of the soil.

2.3 Growth Parameters

The number of the emerged seeds was counted every 24 h for 15 days. The emergence rate was determined by the ratio of the number of emerged seeds to the total number of seeds according to the formula described by Ashraf and Abu-Shakra (1978).

Measurements of root system length, shoot height, and total dry mass of the seedlings were performed at the end of the experiment. For each parameter, 20 seedlings (5 seedlings per bloc) per treatment and per species were randomly selected. Total dry mass was determined after drying the seedlings from each treatment and each species at 60 °C for 48 h. The vigor index (VI) was calculated following the formula of Abdul-Baki and Anderson (1973):

$$VI\;\left(\%\right)=(\mathrm{Roots}\;\mathrm{length}+\mathrm{Shoots}\;\mathrm{height})\times\mathrm{Final}\;\mathrm{germination}\;\mathrm{rate}$$

Loss of vigor index (LVI) was expressed using the following formula:

$$LVI\;\left(\%\right)=\left(\frac{\mathrm{VI}\;\mathrm{of}\;\mathrm{the}\;\mathrm{control}-\mathrm{VI}\;\mathrm{of}\;\mathrm{the}\;\mathrm{treatment}}{\mathrm{VI}\;\mathrm{of}\;\mathrm{the}\;\mathrm{control}}\right)\times100$$

The tolerance index (TI) was calculated using the formula described by Wilkins (1978):

$$TI\;\left(\%\right)=\left(\frac{\mathrm{Fresh}\;\mathrm{mass}\;\mathrm{of}\;\mathrm{stressed}\;\mathrm{seedlings}}{\mathrm{Fresh}\;\mathrm{mass}\;\mathrm{of}\;\mathrm{control}\;\mathrm{seedlings}}\right)\times100$$

2.4 Determination of Electrolytes Leakage, Lipid Peroxidation, and Proline and Hydrogen Peroxide (H 2 O 2 ) Concentrations

For each treatment and each species, the integrity of the membrane structures was evaluated on 20 randomly selected seedlings (5 shoots/bloc/treatment) according to the method described by Blum and Ebercon (1981). Thus, 20 fragments were taken from leaves located in the median part (1 cm lenght). The fragments were washed twice with 10 ml of distilled water. Then, they were soaked in sterile test tubes containing 15 ml of distilled water in the dark at 40 °C for 1 h. Free conductivity was measured using a conductivity meter (Cellox 325 model, Multiline P3 PH/LF-SET, WTW Gmbh, Weilheim, Germany) and expressed in µS.cm−1. The test tubes were put in a water bath at 100 °C for 1 h to determine the maximum leakage of all electrolytes. After cooling, the total conductivity was measured. The rate of electrolytes leakage is defined as follows:

$$\mathrm{Electrolyte}\;\mathrm{leakage}\;\left(\%\right)=\left(\frac{\mathrm{Free}\;\mathrm{conductivity}}{\mathrm{Total}\;\mathrm{conductivity}}\right)\times100$$

Lipid peroxidation was estimated by the determination of the concentration of thiobarbituric acid reactive susbtances (TBARS) according to the method described by Heath and Packer (1968) using 20 randomly selected seedlings (5 shoots/bloc) for each treatment and each species (shoots). One gram of fresh plant material was milled in the presence of 10 ml of extraction buffer consisting of 0.5% thiobarbituric acid (TBA) (w/v), 10% trichloroacetic acid (TCA) (w/v), and 0.2 mM EDTA. The homogenate was incubated in a water bath at 95 °C for 30 min then to stop the reaction, it was immediately cooled down in an ice bath. The mixture was centrifuged at 1000 rpm for 10 min. The absorbance of the supernatant was measured at 532 nm and at 600 nm against a blank using a UV spectrophotometer (UV-1200 Model, Tomos Life Science Group Pte, Ltd, China). TBARS concentrations were calculated using the extinction coefficient (ε = 155 mM−1 cm−1).

The determination of the concentration of proline was performed according to the method described by Monneveux and Nemmar (1986) using 20 randomly selected seedlings (5 shoots/bloc) for each treatment and each species (shoots). One hundred milligrams of fresh leaves was homogenized with 2 ml of 40% methanol (v/v). The mixture was incubated in a water bath at 85 °C for 1 h. After cooling, 1 ml of the mixture was added to 1 ml of acetic acid, 25 mg of ninhydrin, and 1 ml of a solution composed of 300 ml acetic acid, 30 ml orthophosphoric acid, and 120 ml distilled water. The new mixture was incubated in water bath at 100 °C for 30 min then after cooling down, 5 ml of toluene was added. The absorbance of the supernatant was measured at 528 nm using a UV spectrophotometer (UV-1200 Model, Tomos Life Science Group Pte, Ltd, China). The concentrations of proline were determined using the standard curve.

The concentration of hydrogen peroxide (H2O2) was determined according to the method described by Sergiev et al. (1997) using 20 randomly selected seedlings (5 seedlings/bloc) for each treatment and each species. Fresh leaves (500 mg) were cold homogenized in 5 ml of 0.1% trichloroacetic acid (TCA) solution (w/v). The homogenate was then centrifuged at 1000 rpm for 15 min. Five hundred microliters of the supernatant was added to 500 µl of 10 mM phosphate buffer (pH = 7) and 1 ml of a potassium iodate solution (1 M). The absorbance was measured at 390 nm using a UV spectrophotometer (UV-1200 Model, Tomos Life Science Group Pte, Ltd, China) then H2O2 concentrations were obtained from the standard curve.

2.5 Enzyme Assays

The determination of enzymatic activities for catalase (CAT), ascorbate peroxidase (APX), and guaïacol peroxidase (GPX) was performed at the end of the experiment using 20 randomly selected seedlings (shoots) for each treatment and each species. Four hundred milligrams of fresh shoots was homogenized at 4 °C using sterile sand and 2 ml of extraction buffer composed of 50 mM K phosphate buffer (pH = 7), 5 mM sodium ascorbate, and 0.2 mM EDTA. The homogenate was filtrated then centrifuged at 6000 rpm for 15 min at 4 °C. Enzyme activities were determined in the supernatant using a UV spectrophotometer (UV-1200 Model, Tomos Life Science Group Pte, Ltd, China).

CAT (EC1.11.1.6) activity was determined by monitoring the decrease in absorbance at 240 nm (molar extinction coefficient: ε = 0.036 mM−1 cm−1) according to the method described by Aebi (1984). The reaction mixture consisted of 2 ml of 25 mM K phosphate buffer (pH = 7), 30 mM H2O2, and 50 µl of enzyme extract.

APX (EC1.11.1.11) activity was assayed as described by Nakano and Asada (1981) by monitoring the decrease of the absorbance at 290 nm (molar extinction coefficient: ε = 2.8 mM−1 cm−1). The reaction mixture consisted of 2 ml of 50 mM K phosphate buffer (pH = 7), 5 mM H2O2, 0.5 mM sodium ascorbate, 0.1 mM EDTA, and 40 µl of enzyme extract.

GPX (EC1.11.1.7) activity was determined according to the method of Fielding and Hall (1978) by monitoring the increase of the absorbance at 470 nm (molar extinction coefficient: ε = 26.6 mM−1 cm−1). The reaction mixture consisted of 2 ml of 25 mM K phosphate buffer (pH = 7), 10 mM H2O2, 9 mM guaïacol, and 10 µl of enzyme extract.

The absorbances were recorded after mixing the enzyme extracts with the substrates. Control assays were done in the absence of the enzyme extract. Enzyme activities were expressed per g fresh mass (FM). In the case of CAT and APX, 1 unit of enzyme activity was defined as the amount of enzyme necessary to decompose 1 µmol of substrate per min at 25 °C. As for GPX, 1 unit was defined as the amount of enzyme that produces 1 µmol of guaïacol per min at 25 °C.

2.6 Metal Concentrations

Twenty seedlings (5 seedlings per bloc) from each treatment and each species (shoots and roots) were dried and reduced to a fine powder. Thirty milligrams of sample was acid digested using a mixture of HNO3: H2SO4: HClO4 (10:1:0.5; v/v/v) during 2 h at 110 °C. After cooling, the digestion products were filtered and diluted with 0.5% nitric acid (v/v) to 50 ml. Total concentrations of Pb, Zn, and Cd were determined by atomic absorption spectrometry (Perkin Elmer PinAAcle 900 T, USA). For each species, heavy metal composition was expressed as content (concentration × dry mass) per seedling to accurately represent heavy metal accumulation (Lamhamedi et al., 2013; Timmer & Miller, 1991). The effect of metal contamination (Pb, Zn, and Cd) in the four species was assessed using bioaccumulation factor (BAF) and translocation factor (TF) (Mackay & Fraser, 2000) and calculated as follows:

$$BAF=\frac{\mathrm{Concentration}\;\mathrm{of}\;\mathrm{heavy}\;\mathrm{metal}\;\mathrm{in}\;\mathrm{plant}}{\mathrm{Concentration}\;\mathrm{of}\;\mathrm{heavy}\;\mathrm{metal}\;\mathrm{in}\;\mathrm{soil}}$$
$$FT=\frac{\mathrm{Concentration}\;\mathrm{of}\;\mathrm{heavy}\;\mathrm{metal}\;\mathrm{in}\;\mathrm{shoots}}{\mathrm{Concentration}\;\mathrm{of}\;\mathrm{heavy}\;\mathrm{metal}\;\mathrm{in}\;\mathrm{roots}}$$

2.7 Statistical Analyses

Data was analyzed with ANOVA using XLSTAT software (Addinsoft Inc., NY, USA) after testing the assumptions of homoscedasticity and normality of residuals Steel et al. (1997). The differences among treatments and species regarding all measured parameters were determined by Tukey’s test at a 5% significance level.

3 Results

3.1 Growth Parameters

The presence of heavy metals in the soil significantly affected seed emergence kinetics (Fig. 2). After 15 days, the effect of heavy metals on the emergence rates of the seedlings varied depending on the species (Fig. 2). In fact, a latency period preceding seedlings emergence was longer for V. faba seedlings under both treatments (CS and NCS) compared to the rest of the species. A significant stimulation was observed for seedling emergence rate in stressed V. faba (CS) kinetics early in the experiment. However, after 15 days, no significant difference was recorded between the stressed V. faba seedlings (CS) and the control (NCS) (Fig. 2a). The kinetics of emergence of C. arietinum, L. culinaris, and M. arborea seedlings were significantly slowed compared to the controls (NCS). For C. arietinum and L. culinaris, the final emergence rates were significantly (p < 0.05) reduced by 35% and 22%, respectively compared to the controls (Fig. 2b and c). However, by the 10th day, the kinetics of M. arborea increased significantly exceeding the control (NCS). The final emergence rate significantly increased by 23% (Fig. 2d).

Fig. 2
figure 2

Emergence rates of Vicia faba (a), Cicer arietinum (b), Lens culinaris (c), and Medicago arborea (d) seeds exposed to control soil (NCS) and Pb–Zn-Cd contaminated soil (CS) for 15 days. The values represent mean ± SD of 4 replicates of 100 seeds

Full size image

At the end of the experiment, seedlings subjected to contaminated soil (CS) exhibited no noticeable effect on their morphological aspect (Fig. 3). Furthermore, data revealed no significant (p > 0.05) difference in terms of the total dry mass regardless to the species (Fig. 4). Exposure to contaminated soil (CS) had no significant effect (p > 0.05) on root and shoot growth of V. faba, C. arietinum, and L. culinaris compared to the control soil (NCS), with the exception of M. arborea, which showed a significant decrease by 20% for shoots height (p = 0.035) and by 44% for root length (p = 0.022) compared to the control (NCS) (Table 1).

Fig. 3
figure 3

Morphological aspects of Vicia faba, Cicer arietinum, Lens culinaris, and Medicago arborea seedlings after 15 days of exposure to control soil (NCS) and Pb–Zn-Cd contaminated soil (CS)

Full size image
Fig. 4
figure 4

Total dry masses of Vicia faba, Cicer arietinum, Lens culinaris, and Medicago arborea seedlings exposed to control soil (NCS) and Pb–Zn-Cd contaminated soil (CS). For each species, means (± SD, n = 20) with different letters significantly differ from each other based on Tukey’s tests at p < 0.05

Full size image
Table 1 Shoot height, root length, loss of vigor index, and tolerance index of Vicia faba, Cicer arietinum, Lens culinaris, and Medicago arborea seedlings after 15 days of growing under control soil (NCS) and Pb–Zn-Cd contaminated soil (CS) treatments
Full size table

M. arborea exhibited the highest loss of vigor index reaching 51% while V. faba and C. arietinum showed the lowest loss of vigor index by 21% and 20%, respectively. According to the tolerance index, M. arborea was classified as the least tolerant species followed by L. culinaris then V. faba while C. arietinum is the most tolerant to contaminated soil (Table 1).

3.2 Electrolyte Leakage, Lipid Peroxidation, and Proline and Hydrogen Peroxide Concentrations

In contaminated soil (CS), electrolyte leakage increased significantly for L. culinaris (p = 0.0001) and M. arborea (p = 0.041) reaching 97% and 71%, respectively compared to the controls (CT). No statistical differences were recorded for V. faba (p = 0.911) and C. arietinum (p = 0.997) (Fig. 5a).

Fig. 5
figure 5

Electrolytes leakage (a) and thiobarbituric acid reactive susbtances (TBARS) (b), proline (c) and hydrogen peroxide (H2O2) (d) concentrations in Vicia faba, Cicer arietinum, Lens culinaris, and Medicago arborea shoots exposed to control soil (NCS) and Pb–Zn-Cd contaminated soil (CS). For each species, means (± SD, n = 3, composite sample) with different letters significantly differ from each other based on Tukey’s tests at p < 0.05

Full size image

Oxidative damage to cell membranes was assessed by lipid peroxidation using an estimate of thiobarbituric acid reactive susbtances (TBARS) concentration. In respect to the controls (NCS), TBARS concentrations increased significantly for L. culinaris (p = 0.003) and M. arborea (p = 0.0001) reaching 86% and 84%, respectively. No significant differences were recorded for V. faba (p = 0.998) and C. arietinum (p = 0.999) (Fig. 5b).

Proline concentrations showed substantial increases for M. arborea (p = 0.036) and L. culinaris (p = 0.037) reaching 284% and 149%, respectively compared to the controls (NCS). V. faba and C. arietinum exhibited less pronounced increases reaching 101% (p = 0.0001) and 54% (p = 0.045), respectively in respect to the control (Fig. 5c).

In contaminated soil (CS), H2O2 concentrations significantly increased by 63% for C. arietinum (p = 0.048) and 62% for M. arborea (p = 0.009) compared to the controls (NCS). Greater increases were recorded for L. culinaris (p = 0.0001) and V. faba (p = 0.042) reaching 227% and 168%, respectively compared to the controls (Fig. 5d).

3.3 Enzyme Activities

In contaminated soil (CS), enzyme activities generally increased significantly but varied depending on the species (Fig. 6). M. arborea showed a significant rise in APX and GPX activities reaching 114% (p = 0.008) and 18% (p = 0.026) respectively. No significant difference was recorded for CAT (p = 0.094). Conversely, V. faba revealed a significant overproduction of CAT by 131% (p = 0.033) while no significant change was noted for either APX (p = 0.919) nor GPX (p = 0.099). L. culinaris and C. arietinum both exhibited a great overproduction of CAT up to 628% (p = 0.0001) and 43% (p = 0.0001), respectively compared to the controls (NCS), and substantial increases of GPX by 57% (p = 0.001) and 27% (p = 0.015), respectively compared to controls (NCS) (Fig. 6).

Fig. 6
figure 6

Enzymatic activities of catalase (CAT) (a), ascorbate peroxidase (APX) (b), and guaïacol peroxidase (GPX) (c) in Vicia faba, Cicer arietinum, Lens culinaris, and Medicago arborea seedlings exposed to control soil (NCS) and Pb–Zn-Cd contaminated soil (CS). For each species, means (± SD, n = 3, composite sample) with different letters significantly differ from each other based on Tukey’s tests at p < 0.05

Full size image

3.4 Heavy Metal Content

At the end of the experiment, significant increases (p < 0.05) appeared in Pb and Zn contents, especially in the roots of all species. Results showed that Zn is the most accumulated element followed by Pb, whereas only traces of Cd were detected in the roots of all species (Zn > Pb > Cd) (Table 2). Significant variations in heavy metal accumulation were noted between species. C. arietinum and V. faba had the highest Zn and Pb contents while L. culinaris and M. arborea exhibited the lowest contents (Table 2).

Table 2 Heavy metal (Pb, Zn, and Cd) contents (mg.seedling.−1) in shoots and roots of Vicia faba, Cicer arietinum, Lens culinaris, and Medicago arborea seedlings after 15 days of growing under control soil (NCS) and Pb–Zn-Cd contaminated soil (CS)
Full size table

At the end of the experiment, C. arietinum showed significant increases in Pb shoot content by 48% (p = 0.001) compared to the control (NCS). No significant difference was recorded for V. faba (p = 0.682), L. culinaris (p = 0.594), and M. arborea (p = 0.991). However, Pb root contents were 24-fold higher for V. faba (p = 0.0001), 21-fold higher for C. arietinum (p = 0.0001), and 19-fold higher for L. culinaris (p = 0.032) compared to the controls (NCS). M. arborea exhibited lower increase in Pb root content by only eightfold (p = 0.061) compared to the control (Table 2).

In addition, C. arietinum and V. faba both showed significant increases in Zn shoot contents reaching 562% (p = 0.0001) and 470% (p = 0.0001), respectively compared to the controls (NCS). No statistical differences were revealed for L. culinaris (p = 0.933) and M. arborea (p = 0.987) in terms of Zn shoot contents. In contrast, Zn content in roots was substantialy increased by 89-fold for V. faba (p = 0.0001), 70-fold for C. arietinum (p = 0.0001), 43-fold for L. culinaris (p = 0.043), and 41-fold for M. arborea (p = 0.028) compared to the controls (NCS) (Table 2).

Regardless to the species, translocation factor (TF) values showed that the accumulation of Pb and Zn occurred in the roots (TF < 1). V. faba exhibited the lowest translocation of Pb and Zn among all species. M. arborea and L. culinaris showed the highest TF of Pb and C. arietinum had the highest TF of Zn. (Table 3). The bioaccumulation factor (BAF) values indicate that the accumulation of Pb, Zn, and Cd tends to take place in the soil rather than the seedlings (BAF < 1). C. arietinum had the highest BAF for all elements. The lowest BAF of Pb was recorded in L. culinaris and M. arborea, of Zn in V. faba, L. culinaris, and M. arborea and of Cd in V. faba and L. culinaris (Table 3).

Table 3 Translocation (TF) and bioaccumulation (BAF) factors of Vicia faba, Cicer arietinum, Lens culinaris, and Medicago arborea exposed to Pb–Zn-Cd contaminated soil
Full size table

4 Discussion

Heavy metals are one of the major causes of environmental pollution that restrict biodiversity and plant development. However, some species have the ability to grow under restrictive conditions without a significant decline in growth and could be recommended for mining sites rehabilitation. In the present study, we evaluated the growth parameters, stress indicators, and heavy metal accumulation of four Fabaceae species (V. faba, C. arietinum, L. culinaris, and M. arborea) in response to heavy metal contaminated soil (Pb, Zn, and Cd).

Our findings showed that V. faba and C. arietinum were the most tolerant species; while L. culinaris and M. arborea appeared to be less tolerant (Figs. 2, 3, 4, 5, and 6; Table 1). Thus, a relative sensitivity was noted at the germination stage and was clearly manifested in C. arietinum and L. culinaris (Fig. 2b and c). Seed germination is sensitive to environmental conditions and variable among species due to disturbances in carbohydrate metabolism and reduced water and nutrient availability (Debouza et al., 2021; Liu et al., 2012). For M. arborea, the inhibition of seedling emergence was lifted after the 10th day exceeding the control (NCS) but without reaching total seedling emergence (Fig. 2d). Similar results were reported by Peralta et al. (2001) showing an increase of seed germination of Medicago sativa L in the presence of high amounts of Cd and Cr. Our results revealed that V. faba was notable with an emergence rate reaching 100% at the end of the experiment (Fig. 2a). This is attributed to the intrinsic capacity of the seedlings to tolerate high levels of heavy metals (Júnior et al., 2016). Furthermore, the emergence rate of V. faba in contaminated soil (CS) showed a slight stimulation compared to the control (Fig. 2a), which agrees with the findings of Islam et al. (2007) showing an increased germination rate of Elsholtzia argyi H Lév at high Pb concentration.

It was reported that early seedling growth was more sensitive to the toxicity of heavy metals in comparison to germination stage (Baruah et al., 2019; Bouslimi et al., 2021; Li et al., 2005). On the contrary, the present study showed that metallic stress had no significant effect on total dry mass of the four Fabaceae species (Fig. 4). Similar results showed no significant reduction in mass production of Atriplex halimus L under Pb-contaminated soil (Bankaji et al., 2019). Our results did not reveal any significant effect of heavy metals on the primary shoot and root growth in V. faba, C. arietinum, and L. culinaris (Table 1). V. faba stands out in terms of not only mass production but also yield (Fig. 4). It is important to note, however, that heavy metals did not affect all species with the same severity, as evidenced by the growth reduction noted in M. arborea (Table 1). Growth retardation could be considered a potential adaptation mechanism, which may positively influence the seedling survival by reducing their growth rate and biomass allocation (Feng et al., 2014; Grotkopp et al., 2002). In the same context, other studies showed growth inhibition of Medicago sativa seedlings exposed to Pb (Hattab et al., 2016) and of Zea mays L exposed to Zn and Cd (Xu et al., 2014).

Under heavy metals contaminated soil, V. faba and C. arietinum exhibited minimal loss of vigor index, high values of tolerance index (Table 1), and low levels of oxidative injury (Fig. 5), which shows a protection against harmful oxidative damage. On the contrary, in L. culinaris and M. arborea, electrolyte leakage and TBARS concentartions were increased compared to the control (Fig. 5). Several studies showed that exposure to high concentrations of heavy metals increase the production of reactive oxygen species (ROS), leading to membrane damage and lipid peroxidation (Rucinska-Sobkowiak, 2010; Rai et al., 2016; Hasanuzzaman et al., 2020). The damaging effect of heavy metals may vary among plant species. Heavy metal-tolerant species have adaptation mechanisms such as the accumulation of compatible solutes, like proline which acts as an antioxidant and a stabilizing osmolyte of the membrane (Dar et al., 2016; Filippou et al., 2014; Gill & Tuteja, 2010). V. faba showed a substantial increase of proline compared to the control and the highest production of proline among all the species (Fig. 5c). Similar results have been reported by Nadgórska-Socha et al. (2013) and El-Amier et al. (2019) in V. faba and Pisum sativum L.

Hydrogen peroxide (H2O2) is a toxic ROS that accelerates oxidative damage in metal-stressed seedlings (Cuypers et al., 2016). This study showed a significant accumulation of H2O2 in the four tested species compared to the controls (Fig. 5d). The most significant increase was recorded in L. culinaris (Fig. 5d). Several studies have reported an increase in hydrogen peroxide contents in various plant species subjected to metallic stress (Cuypers et al., 2016; Hasanuzzaman et al., 2020). It was demonstrated that at low levels, H2O2 acts as a signaling regulatory component of stress-response in plants (Sofo et al., 2015; Hameed et al., 2016). Moreover, the phytotoxicity generated by heavy metals leads to oxidative stress, which is revealed by the accumulation of hydrogen peroxide (Cuypers et al., 2016). The increased production of H2O2 triggers the activation of antioxidant enzymes (Smirnoff & Arnaud, 2019). This occasioned an overproduction of CAT by sevenfold and 2.3-fold in L. culinaris and V. faba respectively, compared to the controls (Fig. 6). In addition, L. culinaris exhibited the greatest overproduction of GPX, while M. arborea showed a strong increase in APX in heavy metal-stressed seedlings (Fig. 6). The patterns of enzymatic activity seem to be linked not only to the tolerance level of the seedlings but also to the enzyme itself. For example, GPX activity was increased in Vicia faba subjected to Pb, Zn, Cd, Ni, and Cu stress (Nadgórska-Socha et al., 2013). However, Bankaji et al. (2015) showed a decline of CAT, APX, and GPX activities in Suaeda fruticosa L exposed to heavy metals (Cd and Cu). Similarly, inhibition of CAT and GPX activities was observed in Pb-stressed Jatropha curcas L, while superoxide dimutase was the major antioxidant enzyme (Shu et al., 2012). It was shown that significant variations occur in antioxidant defense system among plants (Hasanuzzaman et al., 2020; Sachdev et al., 2021). The activation of antioxidant enzymes in response to heavy metals stress has been recognized as an early mechanism of plant adaptation (Alscher et al., 2002; Hasanuzzaman et al., 2020; Verma & Dubey, 2003). Our results indicate that an upregulation of antioxidant defense system was maintained by enzymatic antioxidants (Fig. 6) and through higher proline accumulation (Fig. 5c) to maintain cell membrane stability, particularly in V. faba (Fig. 5a).

The present study showed that translocation of Pb, Zn, and Cd were reduced due to their sequestration in the roots for the four species (Table 2). This effect can be simply explained by a preferential accumulation of lead, zinc, and cadmium, in the roots (Tables 2 and 3) and it was more pronounced in V. faba seedlings (Table 3). Our findings are in agreement with those of Souguir (2009) showing an excessive cadmium accumulation in the roots of V. faba. On the other hand, plants can cope with heavy metal toxicity by chelating metal ions with organic acids or amino acids and their sequestration inside the vacuoles (Pourrut et al., 2011; Rascio & Navari-Izzo, 2011). In many heavy metal-tolerant species compartmentalization is a physiological defense mechanism that restrict heavy metal movement inside roots and may also limit their transport to the shoots (Bankaji et al., 2016; Mateos-Naranjo et al., 2014; Weis & Weis, 2004).

Our results showed that the lowest TF values of lead and zinc were recorded in V. faba seedlings (Table 3), which underline its phytostabilization potential and support its possible use as green manure to accelerate the remediation. Indeed, the use of Fabaceae as green manure has been shown to increase nutrient availability, cation exchange capacity, and soil organic matter in the soil and to reduce mobility and transfer of toxic heavy metals (Aghili et al., 2014; Bai et al., 2017; Botelho and Müller 2020). These effects are associated with the ability of Fabaceae to develop symbiotic associations with nitrogen-fixing bacteria in the soil (Cooper & Scherer, 2012). Baghaie and Aghilizefreei (2020) revealed that the green manure amendment increased Fe uptake in wheat grown in Pb contaminated soil. In addition, the green manure amendment was shown to increase the abundance of bacteria and fungi in heavy metals contaminated soils (Ai et al., 2020; Bai et al., 2017), which can be effective in the ecological restoration of mining sites.

5 Conclusion and Research Needs

Overall, the present study showed that V. faba had the best performance among all the tested species, exhibiting a particular tolerance towards metallic contamination of the soil. The high phytostabilization potential, shown by the lowest TF values, together with the high production of total dry mass, makes V. faba a promising candidate to be used as a green manure amendment for aided phytoremediation of heavy metals contaminated soils. However, as the present study is part of a program to restore an abandoned mining site heavily contaminated by Pb, Zn, and Cd, using ectomycorrhizal Pinus halepensis Mill. seedlings (Hachani et al., 2020, 2022), it would be wise to combine V. faba with ectomycorrhizal P. halepensis seedlings in an agroforestry system using the intercropping model and to evaluate the efficiency of phytostabilization of heavy metals under controlled conditions. As a next step, it would be interesting to continue the long-term evaluation of this agroforestry system on a mining site to test its viability, sustainability, and economic and environmental profitability.