Evaluating capability of two halophyte plants for phytoextraction of cadmium from contaminated soils

Abstract

Selecting an appropriate plant is a key factor for the phytoextraction of heavy metals from contaminated soils. The possibility of using two halophytes for the remediation of cadmium (Cd) from contaminated soils was investigated. Consequently, an experiment with Chenopodium album L. and Chenopodium botrys L. including Cd treatments of 0, 5,10, 20, 40, 60, 80 and 100 ppm was conducted. Designated analyses including fresh weight, dry weight, and Cd concentrations in shoots and roots were performed. Results indicated that for both plants, a logarithmic model performed best to demonstrate the relationship between total soil Cd with shoot Cd and dry matter. Although both halophytes had considerable capability to remediate Cd from soils, C. botrys demonstrated greater potential to accumulate Cd in shoots. The range of bioconcentration factor (BCF) in C. botrys was obtained to be 0.8 to 8.8 and in C. album was 0.4 to 1.73. For C album, BCF in most treatments was lower than 1.0, but for C. botrys, it was mostly larger than 1.0. Although both plants can be used as hyperaccumulators to remediate Cd-polluted soils, the clean-up time of C. album under higher Cd concentrations was shorter than C. botrys L.

KEYWORDS:

• Clean-up time
• heavy metal
• salt tolerant plants
• soil decontamination
• translocation factor

Introduction

Environmental pollution has been considered to be a major threat to plants as well as animals in the past few decades. Several industries have introduced enormous amounts of effluents into the environment that act as the main source of pollution (Jaiswal et al. 2018). Pollutants in the soil are generally classified into organic and inorganic contaminants. Organic pollutants mainly include solvents, paints, pesticides, dioxins and petroleum hydrocarbons. Trace elements and heavy metals are placed in the inorganic contaminants group (Fabietti et al. 2010). The major sources of heavy metal emissions into the soil are industrial activities, including mining, ore refining, discharge of sewage and waste disposal, as well as pesticides and fertilizers used in agriculture (Saldarriaga et al. 2023). In general, the availability of heavy metals is dependent on the quantity and source of metals, organic matter content, pH, quantity and type of clay, cation exchange capacity, competitiveness of other elements, and plant uptake mechanisms (Das et al. 1997).

Cadmium (Cd) is chemically similar to zinc (Genchi et al. 2020). Additional Cd absorption in a plant causes growth to stop, interference in the absorption and transfer of elements, decreased amount of chlorophyll and photosynthesis, and disorder in glucose oxidase activities (Balestrasse et al. 2001). Its consumption limit for humans and feeding livestock has been reported to be 70 µg/day and 10–20 ppm, respectively (Kabata-Pendias 2000). Heavy metals such as copper, iron, zinc, cobalt, and nickel in low concentrations are essential for plant metabolism. But lead, mercury, arsenic, Cd, chromium and silver are toxic for plants and have negative impacts on human health (Koller and Saleh 2018; Li et al. 2022; Yang et al. 2022).

Phytoremediation is a relatively new remediation strategy in which in-situ plants are used for soil decontamination (Sharma et al. 2023). With this technology, plants can remediate contaminated soils by means of absorbing and concentrating pollutants in their cells (Henry 2000). The selected plants must be compatible with the environment and resistant to pollutants, and should have lower water and nutrient requirements (Adam and Duncan 2002; Eapen et al. 2007; Arabi et al. 2017). Using glycophyte species such as sunflower (Helianthus annuus), maize (Zea mays L.), chickpea (Pisum sativum L.) and mustard (Brassica juncea L.) is most common for phytoextracting purposes (Salt et al. 1995; Cutright et al. 2010; Wuana and Okieimen 2010; Majeed et al. 2019; Fu et al. 2021). However, halophytic plants are of special interest because these plants are naturally present in environments characterized by excess toxic ions, mainly sodium, chloride and boron. Some recent studies (e.g. Garci-Caparros et al. 2022) have revealed that these plants may also tolerate other stresses including heavy metals, based on the findings that tolerance to salts and to heavy metals may, at least partly, rely on common physiological mechanisms (Manousaki and Kalogerakis 2011). These plants are also able to cope with heavy metal stress due to their developed morphological and physiological traits including synthesis and storage of osmolytes such as proline and intracellular complexation, chelation and compartmentalization of metal ions (Garci-Caparros et al. 2022). It has been speculated that many salt-resistant plants might be hyperaccumulators of heavy metals (Munir et al. 2022) and, thus, offer greater potential for phytoremediation purposes (Garci-Caparros et al. 2022).

Halophytes employ effective salt-tolerance mechanisms to avoid salt damage and stay relatively ‘calm’ (Tester and Davenport 2003), whereas glycophytes ‘panic’ under salt stress due to limited salt-tolerance mechanisms (Munns and Tester 2008). Bohnert et al. (1995) showed that glycophyte species have mechanisms very similar to halophytes. Thus, it supports the hypothesis that stress tolerance mechanisms are ubiquitous. Salts accumulating glands are most common in Poaceae, Tamaricaceae, Chenopodiaceae, and Frankenaciaceae families. Furthermore, many species from these families are known to have glandular structures (Manousaki and Kalogerakis 2011).

Researchers have counted several advantages for phytoremediation including: (i) cost-effectiveness; (ii) simplicity of accomplishment and maintenance; (iii) usability for decontamination of various organic and inorganic compounds; and (iv) large-scale applicability. However, there are some limitations to this technology including: (i) slower clean-up process compared to traditional methods; (ii) climate dependency; (iii) leaching contaminants to groundwater; and (iv) bounding the remediation to the root zone (Salt et al. 1998).

The main hypothesis of this study was to find out the capability of some halophyte plants for the remediation of heavy metals from low to moderate Cd-contaminated soils. This study was then focused on evaluating the capability of two selected halophyte species to clean up Cd and to compare them with each other in order to select the most appropriate one for Cd phytoextraction purposes. Other objectives of this paper have focused on the transfer indices and clean-up time required for phytoextraction of Cd from contaminated soils by the selected halophyte plants.

Materials and methods

This research was carried out in a greenhouse located in Karaj (latitude and longitudes of 35° 47′ North and 50° 52′ East), Iran. One of the most important families with salt-accumulating glands (minerals and ions) is Chenopodiaceae and many species of this family have a glandular structure (Manousaki and Kalogerakis 2011). Consequently, two species of this family including Chenopodium album L. and Chenopodium botrys L., which are known as native halophytes in arid and semi-arid regions, were selected and the designated pot experiments were performed. The experiments were conducted in a completely randomized design with eight treatments each with four replicates.

Chemical and physical analyses of soil

The experimental soil samples (0–30 cm depth) were taken from a research farm and sieved with a 2-mm sieve. Soil pH was determined with a glass electrode in an aqueous solution of 0.01 M CaCl₂ (1:5 ratio of soil:solution). The electrical conductivity (EC) of the saturation extract was determined by using a conductivity meter apparatus. The soil texture was determined with a hydrometer method (Gee and Bauder 1986). The cation exchange capacity (CEC) was measured by the ammonium acetate saturation method (Rowell 1994). Soil organic matter was determined by using the Walkley-Black method (Walkley and Black 1934). Calcium carbonate (CaCO₃) content was obtained with the titration method (Rayment and Higginson 1992). The Cd concentration was determined by atomic absorbance spectrophotometer (Spectra AA-200), following acid digestion (4:1 concentrated HNO₃ and HClO₄ v/v) (Gupta 2016). Table 1 represents some measured physical and chemical properties of the experimental soils.

Table 1. Selected physical and chemical properties of the experimental soil.

Soil texture	pH	EC (dS/m)	%O.C	%N	CEC (meq/100 g)	%CaCO3	ρb(g/cm^3)	Cdinit(ppm)
Sandy clay loam	7.58	6.71	0.7	0.07	14	7.5	1.33	1.22

Experimental design and treatment

According to the maximum allowable limits of Cd in soil, which is considered to be 3 µg/g (Chiroma et al. 2014), the Cd concentrations of 0 (T1), 5, 10, 20, 40, 60, 80 and 100 mg/kg (CdCl₂) were applied to contaminate the experimental soil samples. To contaminate the experimental soils, the soil samples were first thoroughly mixed with CdCl_2: as such, CdCl₂ was dissolved in distilled water and then the soil was placed on plastic and sprayed by the designated Cd solution. The contaminated soils were then poured into the PVC pots of 20 cm height and 22 cm diameter and carefully compacted in order to reach the bulk density of 1.33 g/cm³. Irrigation was applied with tap water for 45 days to achieve equilibrium between the applied Cd and experimental soils. In this way, the conditions of pots became similar to that of the natural environment. After 45 days, the seeds were planted in the experimental pots. After germination and when four leaves had appeared, two healthier and stronger plants were held in each pot and the rest were thinned. The soil water content was kept approximately constant around the field capacity by drip irrigation during the entire experimental period.

Plant analysis

At the end of the growth period (120 days), experimental plants were carefully harvested and separated into roots and shoots, and were dried at 80°C until constant weight was obtained. Then, the dry weights of roots and shoots were determined. To assess the effect of different amounts of soil Cd on the phytoextraction of this metal, the concentrations of Cd were measured in mg/kg of dry weight for each plant part. For this purpose, Cd contents of dry shoots, dry roots and soil Cd concentrations were measured. The Cd concentrations of shoots and roots were determined by using the digestion method with nitric acid, perchloric acid, and sulfuric acid (in a volume ratio of 40:4:1) (Greman et al. 2001; Gupta 2016).

Quantitative evaluation

Usually, natural processes follow exponential or logarithmic schemes. In this study, the following linear, exponential, logarithmic and quadratic mathematical functions were, respectively, examined against the experimental data: (1) $\begin{array}{rl}y& =\mathrm{ax}+b\end{array}$(1) (2) $\begin{array}{rl}y& =a{e}^x+c\end{array}$(2) (3) $\begin{array}{rl}y& =\mathrm{bLnx}+c\end{array}$(3) (4) $\begin{array}{rl}y& =a{x}^2+\mathrm{bx}+c\end{array}$(4)

in which a, b, and c are model parameters, and y and x are dependent and independent variables, respectively.

In order to analyze the results, normality tests as well as the Pearson correlation analysis were carried out, using SPSS software. The effects of Cd on factors such as dry matter and extracted Cd in the harvestable part of plants were analyzed by using the univariate regression method.

To evaluate the phytoextraction potential of selected halophytes and to obtain the most appropriate treatment for phytoextraction, both translocation factor (TF) and bioconcentration factor (BCF) were, respectively, calculated based on the following equations (Eisazadeh et al. 2019; Testa et al. 2023): (5) $\begin{array}{rl}\mathrm{TF}& =\frac{C_{\mathrm{Shoot}}}{C_{\mathrm{Root}}}\end{array}$(5) (6) $\begin{array}{rl}\mathrm{BCF}& =\frac{C_{\mathrm{Shoot}}}{C_{\mathrm{Soil}}}\end{array}$(6) where C_Shoot, C_Root and C_Soil are the heavy metal concentration in shoot, root and soil, respectively.

The clean-up time required for phytoextraction of 15, 30 and 60 percent of Cd from contaminated soils was calculated based on that proposed by Schnoor (1997) as follows: (7) $\begin{array}{rl}K& =U/M_0\end{array}$(7) (8) $\begin{array}{rl}M& =M_0{e}^{-\mathrm{kt}}\end{array}$(8) (9) $\begin{array}{rl}t& =-(\ln M/M_0)/K\end{array}$(9) where k is the constant for uptake (year), U is the rate of contaminant uptake (kg/year), M_o is the contaminant initial mass (kg), M is the remaining contaminant mass (kg) and t is the required clean-up time. Finally, the Cd uptake rate was calculated.

Results and discussion

Heavy metal uptake, mobility, and translocation within plant tissues are greatly dependent on plant species, heavy metal type, and their concentration (Ghori et al. 2019). Analysis of the control samples indicated that parameters including soil texture, pH, and EC did not make any limitations for the growth of the examined halophytes (Table 1). To evaluate the performance of Cd phytoextraction by C. album and C. botrys the selected models were compared in order to select the most appropriate model. By applying a univariate regression, the impact of total soil Cd (as independent variable) on several factors including dry matter, and Cd concentrations in roots and shoots (as dependent variables) was analyzed. The results of these analyses indicated that by increasing soil Cd, the Cd concentration in both roots and shoots increased. However, this increase alone cannot imply the capability of the examined plants for phytoextraction, because by increasing soil Cd, the dry biomass in both roots and shoots has decreased. Thus, the dry biomass must be multiplied by Cd concentration in the shoot in order to obtain the extracted Cd by harvestable parts of a plant. Figure 1 demonstrates the fitted models on soil Cd data and plant factors of C. botrys. As shown in this figure, by increasing soil Cd up to 30 mg/kg, the extracted Cd by shoots first increased and decreased afterwards. This decline can be attributed to the reduced plant biomass when soil Cd exceeds a concentration of 30 mg/kg. Tables 2 and 3 demonstrate the best fit of the selected models on experimentally obtained data. The related regression analyses of the effects of total soil Cd on dry biomass and the obtained Cd concentration in shoots are also given in these tables.

Figure 1. Different mathematical models were applied to the relationship among total soil Cd and factors including Cd in roots and shoots, Cd removed by shoots and dry matter of C. botrys. L (4 replicates).

Table 2. Models’ parameters and goodness-of-fit statistics for C. botrys L.

Plant factor	Model	b2	b1	Constant	Sig.	df2	df1	R^2
Cdroot	Quadratic	−0.01	2.49	8.42	0.00	30	1	0.96
		$y=-0.01x^2+2.49x+8.42$
Dry matter	Logarithmic	-	−1.537	11.620	0.00	30	1	0.834
		$\text{y}=-1.537\text{ lnx}+11.620$
Cdshoot removed	Quadratic	−0.027	2.685	83.242	0.005	29	2	0.30
		$y=-0.027x^2$+2.685$x$ +83.242
Cdshoot	Logarithmic	-	15.550	11.57	0.00	30	1	0.90
		$y=15.550\mathrm{lnx}+11.574$

Table 3. Models’ parameters and goodness-of-fit statistics for C. album.

Plant factor	Model	b2	b1	Constant	Sig.	df2	df1	R^2
Dry matter	Logarithmic	–	−1.96	18.78	0.00	38	1	0.76
		$\text{y}=-1.96\text{lnx}+18.78$
Cdshoot	Logarithmic	-	10.088	−8.07	0.00	38	1	0.95
		$\text{y}=10.088\text{lnx}-8.07$
Cdshoot removed	Quadratic	−0.024	3.58	1.18	0.00	37	2	0.93
		$\text{y}=-0.024{\text{x}}^2+3.58\text{x}+1.18$

Results of the regression analyses show that in the case of C. album, by increasing soil Cd the shoot Cd concentration increased, while the dry matter followed a declining trend. The best-fitted regression models are given in Figure 2. It is worth noticing that for finding out which plant species can clean up more Cd from the contaminated soil, both factors of high biomass and high Cd accumulation in shoots must be concurrently considered. The obtained results further indicate that although the examined regression functions (dry matter, shoot Cd concentration and extracted Cd by shoots) for C. botrys and C. album are alike, they have different interpretations that reflect the different responses of these two plants to Cd contamination. In other words, although the decreased rate of dry biomass of C. album is higher (having a larger negative slope) than C. botrys, the biomass of C. album is higher than that of C. botrys. Thus, the dry matter of C. album is eventually higher than C. botrys. The applied regression functions further suggest that the slope of Cd accumulation in shoots of C. botrys is larger than that of C. album. So that, if x = 0, C. botrys will have more Cd content (larger intercept). Considering the obtained intercept values, C. botrys has a larger Cd accumulation capability. By utilizing the multivariable regression, the linear relationship between a set of independent variables and a dependent variable can be examined in ways that the relationships between the independent variables are also taken into consideration. Regression helps to determine the variance of the dependent variable. This is accomplished somewhat through the estimation of contributing variables (two or more independent variables) in variance. In multivariate regression, the values of one variable (dependent variable, y) are estimated from other variables (two or more independent variables, x₁, x₂, …) that can be done through a linear equation as: (10) $y^{\mathrm{\prime }}=b_0+b_1(x_1)+b_2(x_2)+\dots +b_k(x_k)$(10)

Figure 2. Different mathematical models were applied to the relationship among total soil Cd and factors including Cd in shoots, dry matter and Cd removed by shoots of C. album (5 replicates).

To study the effects of three independent variables including total Cd, soluble Cd, and Cd in the shoot on produced dry matter (dependent variable), the normality of the data was examined using the Kolmogorov–Smirnov test at a 5% confidence level (Table 4). By conducting this test, the normality of the data was confirmed. Performing the Pearson correlation between the shoot Cd, total Cd and dissolved Cd with produced biomass in both species showed a negative correlation between the first three parameters and biomass at the 1% confidence level (Table 5). For C. album and C. botrys species, the regression models were fitted, respectively, with one and two independent variables (Tables 6 and 7).

Table 4. The output of conducted Kolmogorov-Smirnov test for C. album and C. botrys species.

	Dissolved Cd (ppm)	Dry matter (g/pot)	Cdshoot (ppm)	Cdtotal (ppm)
	C. album (C. botrys)	C. album (C. botrys)	C. album (C. botrys)	C. album (C. botrys)
Mean	0.093 (0.1)	7.36 (13.33)	54.65 (22.68)	30.62 (35.971)
Std. deviation	0.0436 (0.047)	2.26 (3.19)	21.95 (14.7)	30.96 (28.75)
Absolute	0.151 (0.137)	0.071 (0.168)	0.194 (0.17)	0.221 (0.127)
Positive	0.130 (0.137)	0.070 (0.168)	0.1 (0.128)	0.221 (0.127)
Negative	−0.151(−0.136)	−0.071 (−0.101)	−0.194 (−0.17)	−0.165 (−0.119)
Kolmogorov-Smirnov	0.854(0.773)	0.399 (0.897)	1.08 (0.894)	1.249 (0.828)
Sig.	0.459 (0.589)	0.997 (0.397)	0.187 (0.401)	0.088 (0.499)

Table 5. The coefficient of Pearson correlation between the independent and dependent variables for C. album and C. botrys species.

	Dissolved Cd	Biomass	Cdshoot	Cdtotal
	C. album
Cdtotal	0.889**	−0.745**	0.925**	1
Cdshoot	0.949**	−0.839**	1	0.827**
Biomass	−0.816**	1	−0.838**	−0.745**
Dissolved Cd	1	−0.820**	0.867**	0.838**
	C. botrys
Cdtotal	0.838**	−0.728**	0.827**	1
Cdshoot	0.867**	−0.830**	1	0.827**
Biomass	−0.820**	1	−0.830**	−0.728**
Dissolved Cd	1	−0.820**	0.867**	0.838**

**P<0.01 significant.

Table 6. The obtained coefficients of determination and correlation models fitted to C. album and C. botrys species.

	Coefficient of determination (R^2), Coefficient of correlation (R)
Model	C. botrys	C. album
1	(R^2  = 0.751) (R = 0.867)	(R^2 = 0.715) (R = 0.839)
2	(R^2 = 0.815) (R = 0.891)	–

Table 7. Coefficients of C. album and C. botrys plant.

	Unstandardized Coefficients		Standardized Coefficients
Model	Std. Error	B	Beta	T	Sig.
C. album
Constant	0.522	16.822	–	32.199	0.00
Dissolved Cd	0.021	−0.175	−0.839	−8.431	0.00
C. botrys
Constant	0.510	12.176		23.894	0.00
Dissolved Cd	7.621	−29.529	−0.570	−3.875	0.001
Cdshoot	0.015	−0.037	−0.363	−2.467	0.02

In order to assess the validity and accuracy of the examined models, the coefficient of determination (R²) and root mean square error (RMSE) were calculated by: (11) $\mathrm{RMSE}=\sqrt{\frac{1}{n}\sum _1^n{(O-P)}^2}$(11) where O and P are, respectively, the measured and estimated values, and n is the number of samples.

The RMSE values for C. album and C. botrys species were obtained to be 1.745 and 1.008, respectively. The calculated coefficient of determination for each plant is given in Table 6. The following relations were obtained for the examined halophyte species: (12) $\begin{array}{rl}\underset{\_}{C.\mathrm{album}}\text{biomass }& =16.822\text{--}0.175(\text{C}{\text{d}}_{\text{shoot}})\end{array}$(12) (13) $\begin{array}{rl}\underset{\_}{C.\mathrm{botrys}}\text{biomass }& =12.176\text{--}29.529(\text{dissolved Cd})\\ & \text{--}0.037(\text{C}{\text{d}}_{\text{shoot}})\end{array}$(13)

The conducted correlation analysis (Table 5) shows a significant (P<0.01) negative correlation between the produced biomass and three factors of total Cd, dissolved Cd and shoot Cd. It further shows that C. album is more impacted by the total and dissolved soil Cd than C. botrys. On the other hand, according to the regression model, C. album demonstrates a smaller decline in dry matter. Thus, it can be concluded that this plant is more resistant to Cd contamination.

The linear multiple regression method was also used for finding the dominant factors affecting Cd uptake, as well as for predicting Cd concentrations in both species. Based on Eqs. (12) and (13), the biomass production of C. album was impacted only by shoot Cd. But for C. botrys, two variables of shoot Cd and dissolved Cd were involved. The halophyte C. botrys has a greater capability to accumulate Cd in its shoots. Due to the low root mass of C. album the potential accumulation of Cd in roots was not evaluated. A comparison of extracted Cd from shoots indicated that although C. botrys may have high accumulation capability, it can be better used to remediate soils containing up to 30 mg Cd/kg soil. All our observations suggest that C. album has a higher capability to extract Cd to its shoots.

Results of the conducted simple regression method show that for both examined plants, the logarithmic model performs best to demonstrate the relationship between total soil Cd with shoot Cd and dry matter. Furthermore, the quadratic model performs best to demonstrate the relationship between total soil Cd and Cd removed by shoots. In a study on switchgrass (Panicum virgatum) both linear and exponential models have been reported to be suitable for demonstrating the correlation between soil Cd and dry matter (Chen et al. 2012). However, Tudoreanu and Phillips (2004) have stated that the best estimate of their experimental data was obtained by linear functions. Khodaverdiloo and Homaee (2008) suggested a deterministic model for the phytoremediation of heavy metals from contaminated soil. They reported that their proposed model with the linear adsorption isotherm can reasonably well predict the time needed for remediation of Pb and Cd-contaminated soils with Barbarea verna and Spinacia oleracea L species. Brus et al. (2009) developed a multiple regression model using 0.43M HNO₃ extractable Cd, pH, clay, and soil organic matter as predictors of Cd levels in rice grain. They reported that the model performed much better (r²_adj= 0.661) than the linear model using only 0.01M CaCl₂ extractable Cd as a predictor (r²_adj= 0.281). In another study, Azizian et al. (2011) conducted a pot experiment to investigate the response of lettuce to different Cd levels of irrigation water (0, 5, 10 and 20 mg/l) under different irrigation intervals (1, 2 and 4 days). They reported that shoot Cd has a significant positive correlation with the final accumulated Cd. They concluded that shoot Cd can be predicted by a simple linear regression model.

Our obtained results further indicated that by increasing soil Cd concentration, the dry matter of C. album becomes higher than that of C. botrys. But in both species, by increasing the Cd concentration, the dry matter decreases significantly (Figure 3). This finding is consistent with that reported by Khodaverdiloo et al. (2020). As can be seen, the lowest amount of dry matter of C. album and C. botrys in the 100 mg/kg Cd treatment (T8) were 0.858 and 0.292 kg/m²/yr, respectively. By increasing the soil Cd concentration from 5 to 100 mg/kg, the dry matter of C. botrys and of C. album was reduced to about 60.5% and 35%, respectively. According to Sharma and Dubey (2005), heavy metal stress is one of the limiting factors affecting root growth. Lack of proper development and expansion of the root system reduces the nutrient and water uptake, which affects physiological processes such as transpiration, respiration and photosynthesis, and ultimately reduces the growth of different parts of plant and biomass.

Figure 3. Dry matter accumulation in the different levels of contaminated soils. Values are means ±SD (3 replicates).

The Cd concentrations in the shoots and roots of both examined plants are presented in Figure 4. As depicted in this figure, changes in Cd in roots and shoots are affected by corresponding concentrations of Cd in the soil; as such, by increasing soil Cd, the Cd concentration in both roots and shoots was increased significantly. The highest Cd concentrations in the 100 mg/kg treatment (T8) for roots and shoots of C. botrys were obtained to be 160.58 and 79.55 mg/kg, respectively. These were 64.45 and 39.97 mg/kg, respectively, for C. album. In this regard, Dodangeh et al. (2018) reported that the accumulation of lead in the root was higher than its aerial part in gladiolus, which is consistent with our findings. In another research focusing on heavy metal phytoextraction, it has been reported that chive (Allium schoenoprasum L.) can accumulate Cd mainly in its roots (Eisazadeh et al. 2019). In another investigation, Ozturk et al. (2019) reported considerable differences in plant ability to accumulate or exclude various elements. They reported an interesting finding indicating that concentration values of trace elements vary between the species and even in the same species at different sites.

Figure 4. The Cd is taken up by shoots (a) and roots (b) of the examined plant species. Values are means ±SD (3 replicates).

Bioconcentration factor (BCF) and translocation factor (TF)

The results obtained for TF and BCF factors are presented in Figure 5. As noted earlier, the BCF indicates the transfer of Cd from the soil to the plant. The TF factor indicates the transfer of Cd from the roots to the aerial parts of the plant. The obtained results show that although, by increasing soil Cd concentration, the TF decreases for C. botrys and also this factor was less than 1 for some treatments in C. album, the values of TF in both species indicate that the transfer of Cd from roots to shoots is more than other plants. The range of BCF changes was obtained to be 0.8 to 8.8 and 0.4 to 1.73 for C. botrys and C. album, respectively. For C. album, the obtained BCFs of shoots in most treatments were less than 1.0, but for C. botrys most of the obtained BCFs were larger than 1.0. A similar conclusion was also reported for Amaranthus hybridus by Zhang et al. (2010). Eisazadeh et al. (2019) have also reported that by increasing soil Cd concentration, the TF value decreases from 0.62 to 0.06. The trend of TF changes reported by Eisazadeh et al. (2019) is consistent with our findings in this study. Dodangeh et al. (2018) have reported TF values of lead for three ornamental plants between 0.1 and 0.41. Accumulation characteristics of Cd in A. hybridus L. were studied by Zhang et al. (2010). Their results show that BCFs in soil culture and hydroponics solutions were 0.58 to 1.22 and 5.18 to 17.55, respectively. Their reported TF values were 0.64 to 1.50 and 0.33 to 0.92, respectively, concluding that A. hybridus has a reasonable capability for phytoextraction of Cd from contaminated soils. In another study the TF factor of Cd for some plants including F. dibotrys, I. cylindrical, R. patientia, O. undulatifolius, E. annuus, P. vittata, D. erythrosora, P. aquilinum and P. americana were reported to be 2, 2.19, 0.72, 0.39, 0.20, 2.06, 0.78, 1.26 and 1.61, respectively (Fu et al. 2019). The BCFs of these plant species were also reported to be 0.22, 0.77, 0.96, 2.19, 0.92, 16.27, 4.98, 28.55 and 4.18, respectively (Fu et al. 2019). Results obtained by Chaabani et al. (2017) showed that 13 species among their studied plants had TF values of larger than 1.00, while in the case of BCF, all samples indicated a rate that was less than 1.00. Due to the genetic diversity of different plant species, the ability to absorb and transfer heavy metals could also be different in different plant species (Peris et al. 2007).

Figure 5. The obtained translocation factor (TF) and bioconcentration factor (BCF) for the examined two halophyte plants (3 replicates).

Clean-up time and uptake rate of cadmium

The required clean-up times calculated for 15%, 30% and 60% levels of soil contamination in the topsoil (0–10 cm) for the examined halophytes are presented in Figure 6. The maximum clean-up time related to the 60% level of soil contamination in the 100 mg/kg treatment was obtained to be 35 and 27 years for C. botrys and C. album, respectively. However, these were about 16 and 12 years for C. botrys and C. album, respectively, when the Cd concentration was considered being 40 ppm. However, for both halophytes the minimum clean-up time was related to 15% level of soil contamination in the 5 mg/kg treatment. In general, the results presented in Figure 6 show that by increasing the soil Cd concentration, the required clean-up time also increases. This can be attributed to the reduction of plant dry matter. The trend of Cd clean-up time reported by Eisazadeh et al. (2019) is consistent with the trend obtained in this study.

Figure 6. Clean-up time for 15%, 30%, and 60% levels of contaminated soil for the examined species. Values are means of ±SD (3 replicates).

The obtained results of uptake rate for 15%, 30%, and 60% levels of contaminated soils are presented in Figure 7. This figure clearly shows that the Cd uptake rate follows a similar trend in all three levels of soil contamination. The maximum Cd uptake rate was related to a 60% level of soil contamination in 40 and 100 mg/kg treatments for C. botrys and C. album, respectively.

Figure 7. The uptake rate of Cd for 15%, 30%, and 60% levels of contaminated soil for the examined plants. Values are means of ±SD (3 replicates).

It is worth noting that at the time of handling this research, there was no information in the literature about growing these two halophyte plants. We therefore decided to designate five replicates rather than three replicates, which is traditional for these sorts of studies. These two extra treatments, in fact, were two ‘backups’ as a precaution. During the experimental period, one replicate of C. botrys L. was lost due to a broken greenhouse desk upon which this replicate had been settled. Since we have been inclined to use all the data collected in the first section of the study (the model fitting section) for a better fit of the examined mathematical models, we used five replicates for C. album L., but the remaining four for C. botrys L. On the other hand, in the case of C. album L. (T8), the concentration of Cd in the shoot was extremely low and almost zero in two replicates, but it was measurable in the other three replicates. Therefore, no figure for the concentration of 100 mg/kg was given in Figure 2, because there were only three and not five detectable concentration replicates. Thus, for 100 mg/kg Cd treatment, the concentration of cadmium was monitored at the background level. Finally, in the second part of the article (Figures 3–7), which reflects the phytoremediation capability, the presented graphs were made for both plants using the average of three replicates.

Conclusions

We aimed to find out the possibility of using some high salt-tolerant plants for decontaminating Cd from low to moderate-contaminated soils. Consequently, two halophyte species including C. botrys and C. album were designated to find out if halophyte plants can be used to remediate Cd from contaminated soils. Our findings reveal that C. botrys has a greater capability to accumulate Cd in its shoots. However, due to its lower dry matter under very high levels of soil Cd, it can be used for phytoextraction of Cd at concentrations up to 30 mg/kg. Under high soil Cd concentrations, C. album produces more dry matter. Although high Cd concentration tended to diminish the yield of both species, the reduced yield of C. botrys was more than that of C. album. Our results further indicate that by increasing soil Cd contamination, the clean-up time also increases. Although both examined halophyte species can be used as hyperaccumulators to remediate Cd from contaminated soils, under Cd concentrations higher than 60 ppm, the clean-up time of C. album was shorter than that of C. botrys. This can be attributed to higher dry matter of C. album, resulting in reduced clean-up time, and thus increased uptake rate. These findings support our main hypothesis that halophyte species are promising plants for phytoextraction of heavy metals from contaminated soils. Future studies might focus on some other halophyte species as well as on different heavy metals to evaluate their capability for soil remediation.

Acknowledgements

Conceptualization and design: Mahboubeh Mazhari, Safoora Asadi Kapourchal, Mehdi Homaee. Analysis and interpretation of the data: Mahboubeh Mazhari, Safoora Asadi Kapourchal, Mehdi Homaee. Drafting of the paper: Mahboubeh Mazhari, Safoora Asadi Kapourchal. Funding acquisition: Mehdi Homaee. Investigation: Mahboubeh Mazhari, Safoora Asadi Kapourchal, Mehdi Homaee. Methodology: Mahboubeh Mazhari, Safoora Asadi Kapourchal, Mehdi Homaee. Supervision: Mahboubeh Mazhari, Mehdi Homaee. Writing, review and editing: Mahboubeh Mazhari, Safoora Asadi Kapourchal, Mehdi Homaee.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The data are openly available at https://doi.org/10.6084/m9.figshare.23020052.v5.

Additional information

Funding

This work was supported by the Agrohydrology Research Group , Tarbiat Modares University, grant number IG-39713.