Water, Air, & Soil Pollution

, 225:1993

Heavy Metal and Arsenic Resistance of the Halophyte Atriplex halimus L. Along a Gradient of Contamination in a French Mediterranean Spray Zone

  • Jacques Rabier
  • Isabelle Laffont-Schwob
  • Anca Pricop
  • Ahlem Ellili
  • Gabriel D’Enjoy-Weinkammerer
  • Marie-Dominique Salducci
  • Pascale Prudent
  • Brahim Lotmani
  • Alain Tonetto
  • Véronique Masotti

DOI: 10.1007/s11270-014-1993-y

Cite this article as:
Rabier, J., Laffont-Schwob, I., Pricop, A. et al. Water Air Soil Pollut (2014) 225: 1993. doi:10.1007/s11270-014-1993-y


Elements uptake, histological distributions as well as mycorrhizal and physiological statuses of Atriplex halimus were determined on trace metal and metalloid polluted soils from the surrounding spray zones of a former lead smelter in the South-East coast of Marseille (France). Analyses of heavy metal and arsenic distribution in soil and plant organs showed that A. halimus tolerance is largely due to exclusion mechanisms. No specific heavy metal concentration in leaf or root tissues was observed. However, accumulation of salts (NaCl, KCl, Mg and Ca salts) on leaf bladders and peripheral tissues of roots was observed and may compete with metal element absorption. Occurrence of endomycorrhizal structures was detected in roots and may contribute to lower element transfer from root into the aerial parts of plants. The non-destructive measurements of leaf epidermal chlorophylls, flavonols and phenols showed a healthy state of the A. halimus population on the metal and metalloid polluted sites. Considering the low metal bioaccumulation and translocation factors along with a reduced metal stress diagnosis, A. halimus appeared as a good candidate for phytostabilization of trace metals and metalloids and notably arsenic in contaminated soils of the Mediterranean spray zone. However, its invasive potential has to be determined before an intensive in situ use.


Mediterranean saltbush Root symbioses Inorganic contamination Phytostabilization Salt-affected soils Non-invasive sensors 

1 Introduction

At the beginning of the Calanques hills (Marseille, S-E France), Mediterranean coastal ecosystems interface pollution from abandoned industrial sites and polluted sea sprays from urban effluents. Occurrence of heavy metals and metalloids from industrial activities, such as lead (Pb) and arsenic (As), was detected in sea sediments and soils along the site. This pollution is considered by the French Institute of Human Health (INVS) to be unsafe for the population living near this site (Lasalle 2007). There is a need of global understanding of the pollutant fluxes at the level of the different compartments (soil, water, plants) and the biocoenose relationships. Preliminary studies on this area were previously done: main physico-chemical characteristics, trace metal and metalloid (MM) contamination levels in soil and in two species, i.e. Rosmarinus officinalis and Globularia alypum growing within the former industrial site and not submitted to sea spray, were previously characterized (Testiati et al. 2013; Affholder et al. 2013).

Along the polluted coastline, an halophytic vegetation developed amongst with the species Atriplex halimus L. The latter is a widespread Mediterranean shrub species (Osmond et al. 1980; McArthur and Sanderson 1984; Ortíz-Dorda et al. 2005) with high resistance to various abiotic stresses such as drought (Le Houérou 1992), salinity (Bajji et al. 1998, 2002) and heavy metals (Lutts et al. 2004; Lefèvre et al. 2009; Manousaki and Kalogerakis 2009; Mateos-Naranjo et al. 2013). In addition to its ability to grow on degraded soils and in very harsh conditions such as soil salinity (Pourrat and Dutuit 1994; Martinez et al. 2004), A. halimus has the property to produce an abundant foliar biomass even during unfavourable periods of the year (Kessler 1990). These characteristics enable this plant to be used in the phytoremediation of contaminated soils (see review by Walker et al. 2014), for example, for phytoextraction of cadmium and zinc in South-Eastern Spain (Lutts et al. 2004). However, since this species shows a high genetic variability (Abbad et al. 2004; Ortíz-Dorda et al. 2005), phytoextraction capacities may greatly vary from one population to another. Since biotic and abiotic parameters may also impact the translocation abilities of this species, an alternative strategy is to use it for phytostabilization. However, the presence as a key factor in horizontal transfer of nutrients, namely, root symbionts, remains under-documented. The arbuscular mycorrhizal (AM) status of the genus Atriplex is controversial, and the Chenopodiaceae family is normally defined as non-mycorrhizal. However, more than 10 species of the Atriplex genus were found to form endomycorrhizal associations (Allen 1983; Johnson-Green et al. 1995; Aguilera et al. 1998; Barrow and Aaltonen 2001; Asghari et al. 2005; Sonjak et al. 2009). Furthermore, dark septate endophytes that can form mutualistic, mycorrhiza-like associations with their host plants (Jumpponen 2001) have been observed in roots of some species of the genus Atriplex (Sonjak et al. 2009). To the best of our knowledge, there is no report about the AM status of A. halimus in France. This study, first, aimed to determine if spontaneous A. halimus were naturally endomycorrhized.

Moreover, heavy metals may alter the physiological status of plants growing on contaminated soils (Clijsters and Assche 1985). We also aimed to determine the heavy metal and metalloid contents of roots and leaves of A. halimus and of soils where the plants were spontaneously growing, that are calcareous soils and subjected to marine aerosols deposits, and to establish if chlorophyll and some flavonoids contents of leaves were altered. All these pieces of information were gathered to define whether or not A. halimus is a good candidate for phytostabilization in the context of heavy metal contaminated soils of the Mediterranean spray zone.

2 Material and Methods

2.1 The Study Area

Spontaneous populations of A. halimus from three sites named Calanque de Saména, Calanque des Trous and Cap Croisette in the National Park of Calanques were chosen to study the potential use of this species for a phytostabilization process under the mixed influence of the sea spray and metal and metalloid (MM) pollution. In 1925, the former silver-lead smelter factory of l’Escalette (Fig. 1) ceased its activities, and approximately 20,000 m3 of slag was deposited on the old factory site, but slag was also scattered along the coast as several main deposits and as road fill. The first two sites nearby visible slag deposits were chosen as potentially polluted and the latter as reference site since it is far from slag deposits and from the former smelter factory, so potentially the less contaminated site (Testiati et al. 2013; Affholder et al. 2013). Prior localization of the biggest slag deposit was checked from aerials photography archives particularly for the reference site Cap Croisette (IGN 1926, 1961). In order to check correlations between the Na and MM concentrations in plant and soils to the sources of contamination, the distance between each plant from the sea or the road was evaluated from the multiplex GPS data and with GPS visualizer (GPS visualizer 2013).
Fig. 1

Location of the three A. halimus stations nearby the former industrial site of l’Escalette, South-East of Marseille, France (filled circle indicates A. halimus sampling sites on the main picture and individual localization of the five individuals per site on the small pictures)

2.2 Sample Collection

Sample collection was done on five healthy-looking A. halimus individuals in each population and on their mycorhizospheric soil.

Five soil samples were collected per site from the rooting zone of each selected plant on the top 15 cm of soils (after removal of the litter) and sieved at 2 mm for trace and major elements analyses (Na, K, As, Cu, Fe, Mn, Pb, Zn).

Roots and leaves were collected for MM analyses. The root samples (ca. 5 g) were collected in the top 15 cm of soils of the five-plant individuals on each site then pooled in three samples for MM analysis. More roots were randomly taken on each individual for assessment of arbuscular mycorrhizae (AM) and/or dark septate endophyte (DSE) colonization and kept separately. A total of 10 g of leaves were collected at different levels (high-medium-low) on the five plants of each site and then pooled for MM analyses. More leaves per individual were taken for the ex situ plant physiological index calibration.

2.3 Soil Salinity Estimation

Measurements of pH and salinity (calculated from conductivity) were determined by potentiometry in a 1:5 soil:water suspension using a Multi 3420 SET B-WTW pHmeter and ECmeter (Baize 1988) on the five soil samples per site.

2.4 Trace and Major Element Analysis

The five soils samples per site were dried at room temperature and ground to 0.2 mm (tungsten grinder with titanium sieve). Total metal and major element concentrations were determined after soil digestion with aqua regia (1HNO3/2HCl) in a microwave oven (milestone Start D) as described by Affholder et al. (2014).

Roots (three pooled samples per site) and leaves (one pooled sample per site) were washed apart with Milli-Q water then dried at 40 °C during 1 week. Afterwards, dry weights were determined, and the samples were ground separately to 0.2 mm. Total metal concentrations in roots and leaves were determined by acid digestion (HNO3:H2O2:ultra-pure H2O, volume proportion ratio 2:1:1) with a microwave oven (milestone Start D).

Solutions obtained for soils and plant organ mineralization were filtrated at 0.45 μm. Cu, Fe, Pb, Mn and Zn were analyzed by ICP-AES (Jobin Yvon, Spectra 2000), As was analyzed by graphite furnace AAS (Thermo Scientific ICE 3000 series AA spectrometer), while flame AES (Thermo Scientific ICE 3000) was used for Na and K measurements.

Quality assurance-quality controls and accuracy were checked using standard plant reference materials (DC 73349, NCS-China) and standard soil reference materials (CRM 049-050, RTC-USA) with accuracies within 100 ± 10 %.

Two factors were calculated as an indicator of the phytoremediation potential of A. halimus according to Yoon et al. (2006): (1) bioconcentration factor (BC = metal concentration in root/metal concentration in soil) and (2) translocation factor (TF = metal concentration in leaf/metal concentration in root).

2.5 Root Symbiont Observations

The root samples were rinsed first under tap water then deionized water and stored in alcohol (60 %, v/v) at room temperature until proceeding. The percentage of symbiont colonization was estimated by visual observation of fungal colonization after clearing roots in 10 % KOH and staining with lactophenol blue solution, according to Phillips and Hayman (1970). A minimum of 50 root segments (1-cm long) per plant was counted. AM frequency was expressed as a percentage of colonization per root sample. Occurrence of DSE was also recorded at the same time, and DSE frequency was determined as previously described for AM frequency.

2.6 In Situ Non-destructive Plant Physiological Index Measurements

Plant physiological indices were optically estimated using a non-destructive measurement of constitutive and induced epidermal phenols, flavonols, anthocyanins chlorophylls and the chlorophyll-to-flavonoids ratio named nitrogen balance index (NBI) of in situ plants, using a Multiplex® 3 (FORCE-A, Orsay, France). This portable fluorimetric device uses fluorescence technology with multiple excitations. Different combinations of the blue-green, red and far-red fluorescence signals at the various excitation bands could be used as indices of different compounds (Cerovic et al. 2008; Agati et al. 2011). For each individual, 25 measurements were done on five plants for each site. The surface measured by multiplex on each plant is hence equivalent to about 10 % of the surface of an average A. halimus bush of 1 m in diameter.

2.7 Ex Situ Plant Physiological Index Calibration

To elaborate scale values of multiplex index in a range from healthy to chlorotic or necrotic tissues, 25 excised samples were chosen from apical buds, healthy leaves, yellowed leaves and dead shoots. They were laid on the 10-cm black surface accessory of the Multiplex® 3 on their adaxial and abaxial sides and measured with the multiplex sensor for five random layouts. Moreover, to estimate reference values of multiplex indices for senescing process in leaf, five sets of eight excised healthy leaves were measured as described above at time zero then after a period during which they were floated on water above three filter paper layers in 9-cm diameter Petri dishes. The Petri dishes covered with aluminium foils were incubated in the dark at 24 °C for 15 days to induce leaf senescence. The measurements done at the beginning of the treatment were repeated at the end of the incubation period.

2.8 Scanning Electron Microscopy (SEM) and Elemental Analysis

Root and leaf samples were prepared as previously described by Rabier et al. (2008). Transversal sections 30-μm thick were cut at −25 °C using a cryomicrotome (Cryo-cut II microtome Reichert-Jung), then immediately placed on SEM specimen holders and carbon metalized (10–15 nm) for observation under an ESEM Philips XL 30 microscope with detector EDAX sdd apollo 10. The energy dispersive X-ray spectra (EDXS) for the compositional analyses were performed by point analysis with 3,000–4,000 cps (count per second) and a collection time of 2 min. X-ray mapping was performed for 20 min to give the elemental distribution for each selected element (Ca, K, Mg, Cl, Mn, S, P, Ni, Al, Mg, Fe and Si). In all cases, the voltage was 20 kV. SEM images were obtained with backscattered electron or secondary electron imaging.

2.9 Statistical Analysis

Statistical analyses and control charts were performed for all data using JMP 10 statistical software (SAS Institute, Cary, NC, USA). Differences between MM concentrations in plant parts and soils, plant stress indices and fungal colonization percentages in the three populations of A. halimus were compared using the parametric Tukey’s test after log transformation of the values for data following a log normal distribution. The non-parametric Wilcoxon rank sum test (Kruskal-Wallis test) and Wilcoxon each pair test were used for data with non-normal distributions.

3 Results and Discussion

3.1 Soil Main Features

Soils were alkaline with typical pH (ca 8.6) from calcareous zone (Table 1) as already reported by Testiati et al. (2013) around the former industrial site of l’Escalette. Soil salinity measurements showed that soils were from non-saline to slightly saline with no significant difference between the three sites. No significant variations of K were observed; however, Na average soil concentration was higher in Calanque de Saména than that in Cap Croisette. This may be explained by the difference of sea spray distribution. The plant population of Cap Croisette is 100–140 m away from the seashore, while those of Calanque des Trous and Calanque de Saména are both 10–40 m away (Fig. 1 and Table 1).
Table 1

Soil characteristics of the three selected sites

Determined parameters

Selected sites

Calanque de Saména

Calanque des Trous

Cap Croisette


8.6 ± 0.2 a

8.6 ± 0.2 a

8.5 ± 0.2 a

Salinity (mg kg−1)

0.5 ± 0.4 a

0.9 ± 0.8 a

0.2 ± 0.1 a

Na (mg kg−1)

2,486 ± 1,798 a

2,084 ± 1,265 ab

867 ± 242 b

K (mg kg−1)

5,440 ± 1,573 a

4,917 ± 2,969 a

4,978 ± 1,314 a

Road distance from the sample sites (m)

17 ± 13 b

8 ± 3 b

88 ± 47 a

Sea distance (m)

18 ± 7 b

26 ± 15 b

110 ± 30 a

Means (±standard deviation) followed by different letters in the same line are significantly different (Wilcoxon test, p ≤ 0.05, N = 5)

3.2 MM Concentrations in Soil and A. halimus Samples

Various metals and metalloids were detected in the soil samples in link with the historical anthropic activities (Table 2). Cd data were not shown since they were under the detection level (0.1 mg kg−1). Sb which was present at concentrations under 4 mg kg−1 in roots and far lower toxic than As (Dietl et al. 1997; Foster et al. 2005) was not shown. The main element detected was Fe. The presence of As and Pb on the sites is in accordance with contamination from the past activity of the former smelter of l’Escalette but less concentrated than the contamination measured in the slag deposit (Laffont-Schwob et al. 2011a; Testiati et al. 2013). Concentrations of As, Fe and Mn were significantly higher at Calanque de Saména than that in the two other sites. This high level of As in Calanque de Saména is probably related to remains originating from the former creeping chimneys (Affholder et al. 2013; Testiati et al. 2013). Concentrations of Pb and Zn were significantly lower in Cap Croisette than that in the other sites. Obviously, the soil concentration variations of As between 43 and 670 mg kg−1 for Calanque de Saména and Pb between 37 and 1,293 mg kg−1 for Calanque des Trous indicate a point source pollution deposit. Differently, the Pb soil concentrations variations from 17 to 129 mg kg−1 for the less polluted Cap Croisette could be mistaken with traffic-related pollution (Triboit et al. 2010). Most of the variability of concentrations within one site can be explained by the distance from the contaminated fill material located under the road (Figs. 1 and 2). This is especially the case for the main contaminants, i.e. As, Pb and Zn on the three sites, but the differences may be masked at lower concentrations as for Cu because the values are close to another non-contaminated reference site (Affholder et al. 2013). In Fig. 2, the highest element concentrations in soils were detected nearby the road in Calanques de Saména, and these concentrations proportionally decreased when more distant from the road. In Calanque des Trous, all samples were along the road, and no pattern of element distribution could be observed. In Cap Croisette, element concentrations were low along the transect from 25 to 140 m from the road. This shows clearly that the dispersion of pollution comes from the solid waste deposits used as building materials of the road. The mechanism of dispersion on the sea spray area may be different than the ridges zone uphill submitted to the plume of smoke from the former chimney of the smelter (Affholder et al. 2013; Testiati et al. 2013). Moreover, the sea sprays might have washed down most of the atmospheric pollution. Nevertheless, an observable contamination was highlighted specifically for As in soils of Saména and for Pb and Zn in soils of Calanque des Trous. Indeed, contamination factors (concentration ratio between soil and background values) calculated from an estimation of the local contamination background levels in soils (5, 7.5, 43 and 66 mg kg−1 for As, Cu, Pb and Zn, respectively as reported by Affholder et al. 2013), presented value around 60 for As in Calanque of Saména and values of 8 and 14 for Zn and Pb, respectively, in Calanque des Trous, while values of contamination factors were only of 1 and 2 for soils of Cap Croisette whatever the element.
Table 2

Average metal and As concentrations (mg kg−1) in soil samples from each site


Concentration (mg kg−1)

Calanque de Saména

Calanque des Trous

Cap Croisette


300 ± 310 a

(60, 125, 606, 668, 43)a

13 ± 10 b

(13, 29, 14, 6, 4)a

8 ± 7 b

(5, 6, 7, 4, 20)a


27 ± 6 a

28 ± 17 a

18 ± 11 a


33,860 ± 14,612 a

11,566 ± 3,998 b

10,600 ± 3,699 b


540 ± 240 a

191 ± 99 b

271 ± 78 b


263 ± 141 a

(164, 211, 432, 393, 116)a

628 ± 633 a

(405, 1,293,1,317, 90, 37)a

45 ± 47 b

(16, 31, 19, 30, 129)a


355 ± 299 a

(258, 213, 884, 262, 156)a

528 ± 446 a

(348, 1,016, 997, 195, 86)a

88 ± 58 b

(56, 50, 67, 77, 190)a

Means (±standard deviation (sd)) followed by different letters in the same line are significantly different (Wilcoxon test, p ≤ 0.05, N = 5)

aFor major pollutants, when sd was too high, value of each replicate is given in brackets to describe pollution heterogeneity

Fig. 2

As, Fe, Pb and Zn concentrations (mg kg−1) in each soil samples versus the distance to the road. Filled circle indicates main element concentrations in soil (mg kg−1). Grey line indicates smooth line (JMP 10 statistical software). Cu and Mn values where omitted for the sake of clarity

However, only low concentrations of MM were found in roots and leaves compared with the levels of pollutants in the soils (Table 3). Similar results of low MM transfer from soil to root in the surroundings of the three sites were previously reported in the rare species Astragalus tragacantha (Laffont-Schwob et al. 2011a) and in Tamarix gallica L. (Abou Jaoudé et al. 2012), respectively. Further investigations on stem MM contents may be of interest to conclude on the potential MM translocation in A. halimus.
Table 3

Major and trace element average concentrations (mg kg−1) in root and leaf parts from A. halimus samples on each site

Plant part


Concentrations (mg kg−1)

Calanque de Saména

Calanque des Trous

Cap Croisette



























40.5 × 103

50.6 × 103

46.1 × 103


51.8 × 103

43.3 × 103

44.2 × 103



7.1 ± 0.5

26 ± 1

7 ± 1


12.0 ± 0.2 c

15.9 ± 0.5 a

13.7 ± 0.3 b


1,144 ± 9 c

3,988 ± 142 a

2,344 ± 7 b


94 ± 2 c

467 ± 48 a

134 ± 6 b


58 ± 7 a

200 ± 24 b

25 ± 5 c


72.8 ± 0.5 ab

145 ± 4 a

68 ± 1b


(5.16 ± 0.57) × 103 ab

(7.11 ± 0.58) × 103 a

(2.45 ± 0.55) × 103 b


(12.48 ± 0.45) × 103 a

(9.24 ± 0.13) × 103 b

(10.87 ± 0.24) × 103 ab

Means for element concentrations in leaves are from only one pooled sample for each site. For comparison of means (±standard deviation, N = 3) for element concentrations in roots, the data in a same row followed by different letters were significantly different using

udl under detection level (<1.4 mg kg−1 for As)

*p ≤ 0.05, Tukey’s parametric multiple comparison test

**p ≤ 0.05, nonparametric comparisons for all pairs using Dunn method for joint ranking

From a point of view of potentially using this plant species as a phytoextraction candidate, no individual exhibited concentrations in leaves higher than 1.0 g kg−1 for Cu, Pb or As, and higher than 10.0 g kg−1 for Zn. Therefore, A. halimus cannot be considered as a hyperaccumulator of any of these elements according to the criteria set by Baker and Brooks (1989). No accumulation of MM was shown as BC and TF were inferior to one except for Zn-translocation factors on all sites (Table 4). These on field results are in agreement with previous studies under controlled conditions on A. halimus seedlings for Pb, Zn and Cu (Lotmani et al. 2011) and A. halimus cuttings for Cd and Pb (Manousaki and Kalogerakis 2009).
Table 4

Average transfer factors of elements in A. halimus from the three sites


Selected sites

Calanque de Saména

Calanque des Trous

Cap Croisette

Bioconcentration factor (element concentration in root/element concentration in soil)


































Translocation factor (element concentration in leaf/element concentration in root)

































udl under detection level

Moreover, MM accumulation was below the domestic animal toxicity (Mendez and Maier 2008). Whatever the detected element, all bioconcentration factors were low. For Na and K, apparent BC were two- to threefold higher, and especially, TF were high in accordance with the excretion capacity of salt bladders of A. halimus (Grigore and Toma 2010). However, contribution from sea sprays is not negligible, and TF may be overestimated.

As hypothesized by Manousaki and Kalogerakis (2011), salt excretion organs of halophytes such as salt bladders of A. halimus are not always specific to sodium and chloride ions, and other potentially toxic ions such as zinc, copper or lead may be excreted through the salt bladders from leaf tissues onto the leaf upper layer. As shown in Tables 3 and 4, the MM transfer is negligible in comparison of Na and K.

Moreover, EDXS cartography was performed for all the detected peaks of the X-ray spectra on root and leaf samples of A. halimus from the three sites to precise element localization in plant tissues. However, MM were under the threshold of detection for EDXS. EDXS maps of root distribution for Si and Al are not shown because there are probably soil elements located on the outer layer of the root even after washing root samples, and no detection in the root tissues was obtained by this method. We hypothesized that there is no preferential accumulation of these elements in a specific tissue differently from our previous results for nickel in root phloem of Grevillea exul (Rabier et al. 2008). However, in root of A. halimus, Ca, Mg and K were detected mostly in the cork both as concretions and diffuse layer (Fig. 3). The level of detected elements was by increasing order Ca < Mg < K. K was also present in the primary and secondary xylem. Na and Cl were diffused in the cork and sparingly in the xylem. Moreover, Na was more concentrated near the inner boundary of the cork. Input from directly deposited marine aerosols is probably the cause for a supplement of Ca, K and Mg (Whipkey et al. 2000). Because the macronutrients Ca, Mg and K were all found as diffuse layers and concretions in the roots, two kinds of interpretation can be put forward. On the one hand, a difference of soluble macronutrients and the saline ion penetrations results from their competition at the level of membrane transporters (Hu and Schmidhalter 2005; Kronzucker and Britto 2010). This imbalance might also affect the absorption of MM. On the other hand, the presence of concretions of Ca, Mg and K might also incorporate traces of MM via solid solution in the crystal lattice but under the detection level for EDXS. Thus, the mineral may act as a storehouse for toxic ions (Skinner and Jahren 2003). The results shown were obtained with root parts from Calanque de Saména (Fig. 3), but similar concretions of Ca and Mg were observed for the two others sites.
Fig. 3

Microphotographs of cross section of A. halimus roots from Calanque de Saména. ae Ca, K, Mg, Cl and Na distribution maps with preferential locations in root tissues. f SEM micrograph of backscattered electron signal. Scale bar 200 μm. c.c. central cylindrer, cryst. calcium crystal, ph2 secondary phloem, xy1 primary xylem, xy2 secondary xylem

Scanning electronic microscopy coupled with trace and major elements analysis did not enable to localize specific heavy metal deposition in the leaf tissues (Fig. 4) even for Fe that was in not negligible amount in leaves. This confirms the hypothesis of an element dispersion in all plant tissues and not a tissue-dependent storage in a same organ. However, accumulation of Na, K and Cl was revealed by this method in leaf salt bladders relative to the salt accumulation ability of this species together with proportionally lower amounts of Mg and S (Figs. 4 and 5). As for Na and Cl, the presence of highly detectable level of Mg and S is related to their abundance in seawater (Cotruvo 2005). Calcium is present in the leaves as druses crystals mainly made of calcium oxalate according to the Ca, O and C peaks of the X-ray point analysis spectra.
Fig. 4

Microphotographs of cross section of A. halimus leaves from Calanque des Trous. ae Ca, K, Mg, Cl and Na distribution maps with preferential locations in leaf tissues, i.e. salt bladders and external layers of mesophyll. f SEM micrograph of backscattered electron signal. Scale bar 200 μm. col. collenchyma, cryst. calcium crystal, ep. epidermis, sb. salt bladder, kr. Kranz cell, ph1 primary phoem, ph2 secondary phloem, xy1 primary xylem, xy2 secondary xylem

Fig. 5

SEM micrographs and elemental analysis of leaf cross sections of A. halimus. a Cross section of a leaf. b and c Shape and elemental analysis of a calcium oxalate crystal. d and e Shape of salt bladders. f Elemental analysis of salt bladders

3.3 Occurrence of Root Symbiotic Association

The fungal structures due to AM colonization were detected only in some of the root cuttings observed, and no information on the potential role of this symbiosis in MM sequestration was obtained. In two of the three sites, representative mycelia and vesicles were observed in the roots of all individuals, and the colonization percentage was up to 45 % (arbuscular mycorrhizal colonization percentage was 41 %, 45 % and not detected, for Calanque de Saména, Calanque des Trous and Cap Croisette, respectively). However, low occurrence of arbuscules was observed and may indicate that nutrient exchange between the plant roots and mycorrhizal fungi was low. These results confirm the previous observation by He et al. (2002) on the occurrence of endomycorrhizae in this plant species and confirm our previous results on polluted soils (Laffont-Schwob et al. 2011b). Furthermore, AM symbiotic associations may be of interest in rhizoremediation process as reviewed by Kamaludeen and Ramasamy (2008) and may contribute to lower element transfer into the aerial parts of plants (Christie et al. 2004). Otherwise, no DSE was observed in the root fragments contrariwise to the observations of Barrow and Aaltonen (2001) in Atriplex canescens.

3.4 Plant Stress Biomarkers

3.4.1 Setting-Up of Reference Values

Twenty percent of variation of chlorophyll indices was observed between abaxial and adaxial healthy leaves under controlled conditions (Table 5). Moreover, reference levels of chlorophylls under controlled conditions were maximal in the adaxial side of healthy leaves and decreased of 80 and 20 % in the yellowed and the senescent leaves, respectively (Table 5). A. halimus leaves showed slight and slow senescence process when incubating detached leaves in the dark, i.e. no visual difference in chlorophyll loss was detected between non-senescent and senescent leaves after 15 days. Anthocyans are detected only on young buds or on yellowed leaves but are under the detection level for mature plants (Table 5). Reference levels of flavonols were maximal on the adaxial side of healthy and yellowed leaves, and lower in the abaxial side of healthy and yellowed leaves (Table 5). Leaf epidermal phenol indexes were maximal for dead and yellowed leaves. Nitrogen balance indices were significantly lower for yellowed leaves towards all the other conditions.
Table 5

Biomarkers, i.e. chlorophyll, flavonol, anthocyan and leaf epidermal phenol indices in A. halimus leaves from the three sites and reference values of multiplex indices for healthy and senescing leaves and dead shoots of A. halimus

Plant material

Chlorophyll index

Anthocyan index

Flavonol index

Leaf epidermal phenol index

Nitrogen balance index

Field measurements

 Shoots from Calanque de Saména

2.20 ± 0.13 c


0.69 ± 0.03 a

96.6 ± 121.9 a

6.17 ± 12.8 a

 Shoots from Calanque des Trous

2.73 ± 0.12 a


0.90 ± 0.03 a

134 ± 147 a

0.42 ± 0.06 a

 Shoots from Cap Croisette

2.28 ± 0.12 c


0.64 ± 0.03 a

65.1 ± 41.5a

0.60 ± 0.12a

Reference levels of phenological stages and sides of leaves

 ap. buds

1.8 ± 0.1 e

0.050 ± 0.003 c

0.44 ± 0.02 a

24.4 ± 0.8 b

0.65 ± 0.05a

 ab. h. leaves

1.8 ± 0.1 e


0.29 ± 0.02 b

12.1 ± 0.7 c

0.90 ± 0.12a

 ad. h. leaves

2.3 ± 0.1 c


0.74 ± 0.03 a

15.4 ± 0.9 c

0.42 ± 0.04a

 ab. y. leaves

0.37 ± 0.02 f

0.19 ± 0.02 b

0.50 ± 0.04 a

184 ± 10 a

0.12 ± 0.005b

 ad. y. leaves

0.43 ± 0.01 f

0.25 ± 0.01 a

0.7 ± 0.1 a

147 ± 12 a

0.08 ± 0.02b

 d. shoots

0.61 ± 0.02 f



290 ± 12 a

1.14 ± 0.04a

Reference levels of leaf senescence process

 t 0 ad. h. leaves

2.5 ± 0.2 a


0.52 ± 0.07 a

37 ± 9 a

0.75 ± 0.12a

 t 15 ad. d. s. leaves

2.01 ± 0.15 a


0.4 ± 0.2 b

24 ± 9 b

0.91 ± 0.34a

Plant health biomarker values are expressed as an average value for multiplex analyses (25 flashes per plant) and as the mean ± SE (N = 5) for each population. Means followed by different letter in a same column are significantly different at p ≤ 0.05

ND not detected, udl under detection level, ap. buds apical buds, ab. h. leaves abaxial side of healthy leaves, ad. h. leaves adaxial side of healthy leaves, ab. y. leaves abaxial side of yellowed leaves, ad. y. leaves adaxial side of yellowed leaves, d. shoots dead shoots, t 0 ad. h. leaves time zero adaxial side of healthy leaves, t 15 ad. d. s. leaves adaxial side of excised healthy leaves after 15 days of dark-induced senescence

3.4.2 Field Measurements

A plant stress status of the individuals from the three populations was given by non-destructive measurements of chlorophyll and flavonoid indices on the field. Twenty percent of variation of chlorophyll indices in leaves was observed between the collection sites (Table 5) but were in the same order of variation as between chlorophyll indices of abaxial and adaxial healthy leaves under controlled conditions (Table 5). Anthocyan indices of plants from the three sites were under the detection level. Neither the flavonol indices nor the leaf epidermal phenol indices or the nitrogen balance indices were significantly different between the sites, although the mean values were the lowest at Cap Croisette.

Thus, the main conclusion that can be drawn is that the mean values of in situ chlorophyll indices for the three sites correspond to the reference indices for healthy leaves and let suppose a MM tolerance of this species. This is in accordance with the visual observations that no significant percentage of yellowed or necrotized leaves of A. halimus was observed on the three sites. Contrariwise to most plant species (Becker and Apel 1993; Thimann 1980), A. halimus leaves showed slight and slow senescence process when incubating detached leaves in the dark, i.e. no visual difference in chlorophyll loss was detected between non-senescent and senescent leaves after 15 days. The anthocyan index is not a good marker for A. halimus stress status because it was detected only on young buds, fruits or on yellowed leaves but was under the detection level for mature plants. Globally, significant differences for all plant biomarkers were observed for the different physiological stages and the leaf side. The difference of UV absorbing pigments between the leaf adaxial and the abaxial sides is in accordance with the results of several authors on different plant species (Cerovic et al. 2008; Bidel et al. 2007). Flavonols were not detected into the dead shoots. A linear regression expression was calculated to describe the change of flavonol indices of individual plants as a function of Pb concentration of their mycorhizospheric soil: flavonol index = 0.65378 + 0.0002479 × Pb mg kg−1 soil. The probability was p = 0.163, and a weak correlation of R2 = 0.37 was obtained between flavonol index and Pb soil concentrations. Therefore, the flavonol index is weakly linked to the soil pollution levels. Previous results on the protective mechanism of the antioxidant activities on barley (Hordeum vulgare) leaf senescence test (Arnao and Hernández-Ruiz 2009) and on Quercus ilex resprouts after fire (El Omari et al. 2003) suggest that the presence of antioxidant activities in Mediterranean plant species protects them against a variety of reactive oxygen species (ROS)-producing stresses. Globally, these non-destructive values for the three sites were close to those of the adaxial healthy leaves. This lead to conclude to a good health status of the three A. halimus populations. However, since the different phenological stages of the references are simultaneously present on a same individual and because the plants from each site were submitted to an heterogeneous heavy metal and As distribution, this contributes to the distribution of the values and the lack of clear significant statistical difference between the sites. Some explanation may also be obtained from the literature, i.e. the scavenging of reactive oxygen under stressed environments is one of the suggested roles of flavonoids in plants (Türkan and Demiral 2009). Adaptation of plant cells to high NaCl salinity involves osmotic adjustment and the compartmentalization of toxic ions, whereas an increasing body of evidence suggests that high salinity also includes oxidative stress (Khavari-Nejad et al. 2006; Türkan and Demiral 2009). The three A. halimus populations are submitted to sea sprays; however, the coupled effect of salt and trace metals and metalloids may differ in Cap Croisette from both other sites. A next step would be to characterize the genetical variability of these three populations with the aim to distinguish in- and extrinsic factors affecting A. halimus resistance to MM and particularly to As.

These results confirmed that some flavonoids are implicated in stress tolerance of A. halimus such as salt and drought and may be extended to other stresses such as metal contamination on the studied sites.

Most of our results are congruent with a lower effect on A. halimus ecophysiological status in the less polluted site, i.e. Cap Croisette. However, no drastic effects of MM pollution on spontaneous populations of this species were noticed on the field.

In conclusion, considering the low trace element translocation in the aerial parts certainly involving exclusion mechanisms along with the plant health diagnosis and a deep root-system, A. halimus appears as a good candidate for phytostabilization of trace metals and metalloids and notably As in salt-affected contaminated soils of the Mediterranean littoral. Under controlled conditions, similar levels of Zn, Cu and Pb in shoot parts of various Algerian populations of A. halimus have been previously detected (Lotmani et al. 2011). These authors confirmed that these elements were preferentially stored in the root parts rather than transferred in the aerial parts of the plants. Moreover, further study on salt effect on heavy metal absorption by A. halimus, endomycorrhized or not, may be of interest. Previous studies of antagonism of metal versus salt on Atriplex genus concentrated on root absorption (Lefèvre et al. 2009) or adsorption on dead tissues (Sawalha et al. 2009). Our observations lead to suppose that leaf absorption of elements present in sea spray might also play a role in metal tolerance of A. halimus submitted to metal-polluted soils on the seashore. The apparent high MM resistance of A. halimus could be linked to tolerance mechanisms both at the level of the root by actions of salinity levels and symbiotic associations and at the level of leaf surface and bladders.

Observing the wide spontaneous occurrence of this plant species on polluted salt-affected soils, its use for phytostabilization maybe of interest. However, the biogeographical status of A. halimus in Southern France is still discussed. If some coastal and insular populations are probably native, as suggested by the genetic results of Ortíz-Dorda et al. (2005), this plant species is also frequently planted and naturalized along the French Mediterranean coast. Therefore, its invasive potential has to be determined before an in situ use even if this species is currently suggested as a substitute of the highly invasive Baccharis halimifolia L. for cultivation and ornamental purposes (Filippi and Aronson 2010; Heywood and Brunel 2008). Our field study was on a protected area included in the National Park of Calanques, South-East of France, and our results suggest to combine scientific approaches in ecology and plant physiology when dealing with pollution in protected areas.


The authors thank Thomas Devenoges, Soumia Djilalli and Cécile Evrard for their technical assistance of root preparation and endomycorrhizal colonization counting and Laurent Vassalo and Carine Demelas for their analytical assistance for trace and major element measurements. Many thanks to Alma Heckenroth, who conceived the map of the collection sites, and to Frédéric Médail for helpful discussion on the biogeographic distribution of A. halimus. The authors are grateful to Lidwine Le Mire Pécheux from the National Park of Calanques for helpful discussion on environment management. This study was funded by the French Research National Agency (ANR Marséco 2008 CESA 018) and supported by the National Innovative Cluster on Risks Management. This research has been made possible by participation in the EU COST Action FA 0901 favouring links between scientists of various countries on halophytes. Isabelle Laffont-Schwob is grateful to Tim Flowers and James Aronson that gave her the opportunity to participate to this COST Action.

Copyright information

© Springer International Publishing Switzerland 2014

Authors and Affiliations

  • Jacques Rabier
    • 1
  • Isabelle Laffont-Schwob
    • 1
  • Anca Pricop
    • 1
    • 2
  • Ahlem Ellili
    • 1
    • 3
  • Gabriel D’Enjoy-Weinkammerer
    • 1
  • Marie-Dominique Salducci
    • 1
  • Pascale Prudent
    • 2
  • Brahim Lotmani
    • 4
  • Alain Tonetto
    • 5
  • Véronique Masotti
    • 1
  1. 1.Aix Marseille Université, CNRS, IRD, Avignon UniversitéInstitut Méditerranéen de Biodiversité et d’Ecologie marine et continentale (IMBE)Marseille cedex 03France
  2. 2.Aix Marseille Université, CNRS, Laboratoire de Chimie de l’Environnement, FRE 3416Marseille cedex 3France
  3. 3.Unité de Physiologie et Biochimie de la Tolérance au Sel des Plantes, Faculté des Sciences, Département de BiologieUniversité de Tunis El ManarTunisTunisia
  4. 4.Laboratoire Protection des végétaux, Unité culture in vitroUniversité A. Ibn Badis de MostaganemMostaganemAlgeria
  5. 5.Aix Marseille Université, PRATIMMarseille cedex 3France

Personalised recommendations