Advertisement

Plant and Soil

, Volume 418, Issue 1–2, pp 45–60 | Cite as

Natural variation reveals contrasting abilities to cope with alkaline and saline soil among different Medicago truncatula genotypes

  • Heithem Ben Abdallah
  • Hans-Jörg Mai
  • Ana Álvarez-Fernández
  • Javier Abadía
  • Petra Bauer
Regular Article

Abstract

Background and Aims

Abiotic stress conditions cause extensive losses to agricultural production worldwide. Salinity and alkalinity affect plant growth, photosynthesis and availability of nutrients including Fe. Many studies have described the response mechanisms of plants to single abiotic stress conditions. However, in the field, crops and other plants are routinely subjected to a combination of different abiotic stresses. Salinity and alkalinity are wide-spread in Tunisia, where Medicago truncatula occurs as a native species.

Methods

We established a growth system to study the combined effects of salinity and alkalinity conditions in laboratory conditions. We screened 11 Tunisian M. truncatula lines from the SARDI collection based on their phenotypic responses to the double stress.

Results

Salinity and alkalinity affected germination rates, shoot and root dry weights, pigment contents and root morphology parameters. We were able to select among the 11 investigated lines four sensitive and tolerant lines with different abilities to respond to the double stress. Tolerant and sensitive genotypes (two lines each) differed in root flavin contents, root flavin staining patterns and concentrations of root flavins in the nutrient solution.

Conclusions

Root architecture, flavin root localization in epidermal cells and flavin secretion are relevant tolerance mechanisms for salt and alkaline stress in M. truncatula. Pairs of contrasting lines from close origins were identified that will be useful tools to identify genes for the tolerance mechanisms.

Keywords

Medicago truncatula Root Iron deficiency Alkalinity Salinity Natural variation Flavin 

Abbreviations

Chl

chlorophyll

Rbfl

riboflavin

Introduction

Soil salinity is a global environmental issue affecting an area of approximately 830 million ha worldwide (Rengasamy 2016). Plant growth in saline soils is usually hampered due to osmotic and oxidative stress and also to toxic effects resulting from high ion concentrations (Rengasamy 2016). In sodic soils, clay dispersion and soil structural degradation result from high Na+ cation amounts and affect negatively plant growth (Rengasamy 2016). Moreover, sodic soils with a dominating presence of Na carbonate and calcareous soils are very alkaline (Rengasamy 2016). Alkalinity causes a low solubility of several other ions, including Fe. Many plant species display Fe deficiency symptoms, including leaf chlorosis and reduced biomass, when grown in a medium at neutral or alkaline pH, since Fe is essential for many enzymes involved in electron transfer processes in respiration, photosynthesis and N fixation (Briat et al. 2015). Iron deficiency also represents a serious problem in human nutrition worldwide (Naranjo-Arcos and Bauer 2016). Salinity and alkalinity are widespread in Tunisian soils, on one hand due to an occurrence of salt-affected soils in depressions and in the main “sebkhas” and “chotts” (dry alkali flats, usually near the sea, and interior salt lake areas, respectively), and on the other hand due to secondary salinity derived from low-quality irrigation water that is rich in dissolved salts.

Alfalfa (Medicago sativa) is a perennial forage with high yield and good nutritional quality, interesting for cropping practices in many agricultural areas, including some in North Africa (Basigalup et al. 2014). Its diploid annual relative Medicago truncatula is a major model legume and is used for molecular studies of physiological responses to stresses such as salinity and Fe deficiency (Rodríguez-Celma et al. 2013; Zahaf et al. 2012). This species presents a large diversity and geographic distribution in the Mediterranean basin and in particular in diverse Tunisian soils (Ellwood et al. 2006; Lazrek et al. 2009). M. truncatula shows a broad adaptation including tolerance to severe growth conditions, with some ecotypes characterized as displaying a high tolerance to salinity and calcareous soil conditions, and hence M. truncatula germplasm can be screened for molecular-ecological adaptation (Friesen et al. 2010; Friesen et al. 2014). M. truncatula natural variation lines and genome sequences have been successfully used for association genetics to pinpoint novel genetic loci and genes for variation of responses (http://www.noble.org/medicago/ecotypes.html). However, a successful approach for exploiting natural variation relies on the successful identification of lines with interesting properties, followed by combined comparative genomics and functional genomics analysis of the lines that have contrasting abilities, e.g., to show tolerance or susceptibility towards salt or Cd stress (Friesen et al. 2010; Rahoui et al. 2016; Zahaf et al. 2012). Such approaches can ultimately help to elucidate candidate genes and functional pathways that have been selected during ecological adaptation.

Salt tolerance relies to a large extent on the sequestration and secretion of cellular ions and the counteraction of osmotic effects and reduced transpiration (Hanin et al. 2016). Calcareous soils affect Fe uptake in dicotyledonous species, which mobilize Fe in the soil via proton extrusion and reduction of Fe(III) to Fe(II) (Abadia et al. 2011). The differences in the signaling responses and tolerance mechanisms between Fe deficiency and calcareous soil conditions are still not fully understood (Hsieh and Waters 2016). In Poaceae the up-regulation of Fe acquisition correlated with tolerance to saline-alkaline stress (Li et al. 2016), suggesting that Fe deficiency is a major problem on such soils. Recently, it has become evident that secondary compounds secreted by dicotyledonous plant roots upon Fe deficiency could play important roles in the mobilization of Fe upon alkaline and calcareous growth conditions. For example, Arabidopsis mutants can develop leaf chlorosis if the production and secretion of phenolic coumarin compounds in roots is perturbed (Fourcroy et al. 2014; Schmid et al. 2014). Medicago truncatula and sugar beet roots produce and secrete flavins, instead of phenolics, in response to Fe deficiency (Andaluz et al. 2009; Rodriguez-Celma et al. 2013; Rodríguez-Celma et al. 2011a; Rodríguez-Celma et al. 2011b; Susin et al. 1993), and it has been demonstrated that extracellular flavins allow Fe-deficient sugar beet roots to dissolve soil Fe(III)-oxides at high pH via reductive mechanisms (Sisó-Terraza et al. 2016b).

In this present work we investigated the interactive effects of salinity and calcareous/alkaline soil conditions on germination, plant growth, leaf pigment content, root morphology parameters and root flavin content and secretion in selected lines of M. truncatula, to identify and characterize lines sensitive and tolerant to these stresses. We made use of the natural diversity from Tunisia by studying the responses of 11 accessions derived from the South Australian Research and Development Institute (SARDI) collection to identify among them tolerant and susceptible lines. We suspect that these lines will be related to each other due to their close origin of collection in Tunisia. Our aim was to identify lines with contrasting stress response behavior to use them as tool for future molecular dissection.

Materials and Methods

Plant material

We used the wild type reference M. truncatula Jemalong A17 line, kindly sent by Jean-Marie Prosperi, INRA Montpellier. Eleven Tunisian M. truncatula lines were selected from the SARDI collection, kindly provided by Steve Hughes, South Australian Research and Development Institute (SARDI). This selection was based on the origin of the collected lines, using criteria such as soil texture and pH. The locations for collection of accessions covered a large area from the north to the south of Tunisia (Supplementary Fig. 1).

Plant growth

M. truncatula plants were grown under controlled conditions in a scientific incubator (Percival Scientific AR-36 L2, Perry, IA, U.S.A.), using day/night cycles of 16/8 h, 24 °C/18 °C, and a relative humidity of 70%. Seeds were mechanically scarified by placing them on a fine grit sand paper sheet and rubbing gently with another piece of sand paper, until visible signs of abrasion appeared. Seeds were sterilized in 10% Na hypochlorite for 8 min and then abundantly rinsed with distilled water.

For soil experiments, seeds were imbibed in sterile water for 2 days at 4 °C and then planted in soil. For salinity treatments, different concentrations of NaCl ranging from 75 to 500 mM were used for soil watering, as indicated in the text. For calcareous and alkaline conditions (shortly termed as alkalinity in the text) the soil was supplemented with different mixtures of CaCO3 and NaHCO3. These mixtures are abbreviated throughout the text as “n/m BIC”, with n and m being the amounts of CaCO3 and NaHCO3, respectively (in g kg−1 soil). BIC treatments ranged from 5/3 to 50/15 BIC, as indicated in the text. Combined soil salinity and alkalinity experiments consisted of combinations of salt and BIC treatments, as indicated in the text.

For the hydroponic system, sterilized seeds were germinated for 4 days at 20 °C in Petri dishes on constantly moistened filter paper. Four day-old seedlings were transferred to a half strength aerated nutrient solution for 4 days. Then, seedlings of similar size were selected and cultured in groups of six plants in one-L pots filled with full strength aerated nutrient solution. The composition of the full strength nutrient solution was: 1.5 mM Ca(NO3)2, 1.25 mM KNO3, 0.75 mM MgSO4, 0.5 mM KH2PO4 and 10 μM H3BO3, 50 μM Fe(III)Na-EDTA, 1 μM MnSO4, 0.5 μM ZnSO4, 0.05 μM (NH4)6Mo7O24 and 0.4 μM CuSO4. Five different treatments were applied during 8 days using nutrient solutions modified as follows: control (50 μΜ Fe(III)Na-EDTA; C), carbonate-induced Fe deficiency (50 μM Fe(III)Na-EDTA, 1 g L−1 CaCO3 and 10 mM NaHCO3; BIC), Fe deficiency (0 μM Fe(III)Na-EDTA; −Fe), salinity (50 μM Fe(III)Na-EDTA and 100 mM NaCl; NaCl) and combined stress (50 μM FeNa-EDTA, 1 g L−1 CaCO3, 10 mM NaHCO3 and 100 mM NaCl; BIC/NaCl). The pH was adjusted to 6.0 for the treatments C, −Fe and NaCl and to 7.0 for treatments BIC and BIC/NaCl. The nutrient solution was renewed every 5 days. Plants were analyzed according to Supplementary Fig. 1 and as indicated in the text.

Determination of the germination rate and dry weight

To assess the germination rate, we counted the number of plants with emerged cotyledons (Supplementary Fig. 1). To obtain the dry weight (DW), plant shoot and root material was dried in an oven at 70 °C for 48 h and weighed.

Photosynthetic leaf pigment measurements

Leaf pigments were extracted from fresh leaves (leaf number 2, Supplementary Fig. 1) in 80% acetone and assayed spectrophotometrically according to (Arnon 1949) using the TECAN infinite 200 PRO multimode reader at 645 nm, 663 nm and 470 nm wavelengths. Readings were then used to calculate total Chl and carotenoid contents in mg g−1 leaf fresh weight.

Root morphology

Root systems were excised and thoroughly cleaned from soil particles if plants were grown in soil. Then, roots were laid out on a scanner, and root architectural parameters were recorded using the scanned images with the help of the WinRHIZO software (Regent Instruments Inc., Québec, Canada).

Extraction, analysis and localization of flavins in roots

Flavins were extracted from 1 to 1.5 cm root tips dissected from the main root as described in (Rodríguez-Celma et al. 2011b). Root extracts were dried, re-suspended in mobile phase (85% methanol and 0.1% formic acid) and analyzed for flavins using high-performance liquid chromatography with photodiode array detection coupled in-line to time-of-flight mass spectrometry [HPLC-UV/VIS-MS(TOF)], as described (Sisó-Terraza et al. 2016a). The HPLC system (Alliance 2795, Waters, Mildford, MA, U.S.A) included an analytical HPLC column (Symmetry R C18, 15 cm × 2.1 mm i.d., 5 μm spherical particle size, Waters) protected by a guard column (Symmetry R C18, 10 mm × 2.1 mm i.d., 3.5 μm spherical particle size, Waters). A gradient mobile phase built with 0.1% (v/v) formic acid in water and 0.1% (v/v) formic acid in methanol was used. A PDA 2996 (Hsieh and Waters) and an MS(TOF) equipped with a electrospray ionization (ESI) source (micrOTOF, Bruker Daltonics Bremen, Germany) were used for flavin detection. Flavin quantification was carried out by external calibration using the peak areas at m/z of the corresponding [M + H]+ ions and a Rbfl calibration curve.

Flavin localization in roots was studied using fluorescence microscopy of ca. 100 μm root cross sections taken 1–1.5 cm above the tip of the main root, with excitation and emission wavelengths of 422 and 528 nm, respectively, on an Axio Imager 2 microscope (Zeiss, Jena, Germany).

Results

Establishment of growth conditions using the M. truncatula reference line Jemalong A17

First, we established a soil growth system for M. truncatula plants with saline and alkaline growth conditions, in order to evaluate plant morphological and physiological traits. Initially, we tested saline and alkaline conditions separately. For the saline condition, we grew seedlings of the reference line Jemalong A17 and irrigated the soil with different saline solutions, ranging from 75 to 500 mM NaCl, during a period of 15 days (Fig. 1). Treatment with 100 to 200 mM NaCl decreased the germination rates significantly by 10–30%, compared with the control and the 75 mM NaCl treatment (Fig. 1a). At 300 mM NaCl, a further significant drop (60%) in the germination rate was observed (Fig. 1a). At higher salinity rates, 400 and 500 mM NaCl, seed germination was completely inhibited during the tested time (Fig. 1a). Up to 150 mM NaCl, the root and shoot DWs as well as the lengths of the main root were similar in salt-treated and control plants (Fig. 1b, c). At 200 mM NaCl, a significant drop in biomass and main root length was observed, and effects were intensified significantly at 300 mM NaCl.
Fig. 1

Effect of salt stress in M. truncatula Jemalong A17. (A) Germination rate; (B) biomass production; (C) main root length in M. truncatula wild type (Jemalong A17) grown during 15 days in soil with different salt concentrations. Data are means ± SD; n = 4; means with the same letter are not significantly different at P ≤ 0.05 according to Analysis of variance (= ANOVA) and Fisher’s Least Significant Difference (= LSD) test

Next, we tested alkaline-induced changes of plant growth by mixing the soil with different BIC systems prepared with different amounts of CaCO3 and NaHCO3 (labeled as BIC n/m, with n and m being the amounts of CaCO3 and NaHCO3, respectively, in g kg−1 soil). Compared to the untreated control, the germination rate was reduced by 10, 30 and 80% upon supply of 10/6, 20/9 and 30/12 BIC, respectively, whereas at 50/15 BIC no germination was observed (Fig. 2a). A marked root DW reduction (30–50%) took place at 20/9 and 30/12 BIC, whereas the main root was significantly reduced in size (ca. 40%) at 30/12 BIC (Fig. 2c). The shoot DW was significantly reduced by 20–50% at 10/6, 20/9 and 30/12 BIC (Fig. 2b). These preliminary results were used to narrow down appropriate combinations of saline and alkaline environmental conditions in further experiments.
Fig. 2

Effect of bicarbonate (BIC) stress in M. truncatula Jemalong A17. (A) Germination rate; (B) biomass production; (C) main root length in M. truncatula wild type (Jemalong A17) grown during 15 days in soil with different BIC concentrations. Data are means ± SD; n = 4; means with the same letter are not significantly different at P ≤ 0.05 according to ANOVA and Fisher’s LSD test

We then tested ten combinations of NaCl and BIC (alkalinity) treatments in the soil growth system, choosing combinations of those single stress treatments that did not affect or affected only moderately plant growth (see Fig. 3). Germination rates were significantly decreased at 75 mM NaCl when supplemented with 10/6 and 20/9 BIC (by 10 and 20%), and at 100 and 150 mM when supplemented with 5/3 (by 10% each), 10/6 (by 10% each) and 20/9 BIC (by 20 and 40%), with 20/9 BIC causing the highest drop (by 40%) in germination at 150 mM NaCl (Fig. 3a). Similarly, shoot DWs dropped clearly at 75–150 mM NaCl when supplemented with 20/9 BIC (by 25, 70 and 70%) (Fig. 3b). Root DWs were also most affected under these conditions (Fig. 3b). Main root lengths were shorter at 100 and 150 mM NaCl in the presence of 20/9 BIC (by 35 and 60%) (Fig. 3c). Thus, we observed in several combined stress conditions a decline in the growth compared to single stresses. For screening experiments under combined stress, we decided to apply 10/6 BIC in combination with 150 mM NaCl, which was the highest BIC treatment tolerated by Jemalong A17.
Fig. 3

Effect of combined salt and bicarbonate (BIC) stress in M. truncatula Jemalong A17. (A) Germination rate; (B) biomass production; (C) main root length of M. truncatula wild type (Jemalong A17) grown during 15 days in soil with different salt and BIC concentrations. Data are means ± SD; n = 4; means with the same letter are not significantly different at P ≤ 0.05 according to ANOVA and Fisher’s LSD test

Effects of individual and combined stress in hydroponic system on wild-type Jemalong A17

A combined stress condition was also applied to Jemalong A17 using a hydroponic system. We observed that the effect of combined stress (1 g L−1 CaCO3, 10 mM NaHCO3, 100 mM NaCl) was as harmful as BIC alone on shoot DW, main root length and carotenoid content, but more harmful than effects of single stresses on root DW, number of secondary roots and Chl content (Fig. 4). An effect of the single NaCl treatment was only noted for the number of secondary roots (Fig. 4d). These data confirmed that the combination of salt and alkalinity stress had stronger and more pronounced effects on M. truncatula plants than the individual stresses.
Fig. 4

Effects of individual and combined stress in M. truncatula Jemalong A17 upon hydroponic growth. (A) Shoot DW; (B) root DW; (C) main root length; (D) number of secondary roots; (E) leaf pigment content; (F) plant images of M. truncatula WT (Jemalong A17) grown during 15 days in a hydroponic system under control (regular) conditions, salt (+ 100 mM NaCl), BIC (+ 1 g/L CaCO3 + 10 mM NaHCO3), and BIC/NaCl combined stress (+ 1 g/L CaCO3 + 10 mM NaHCO3 + 100 mM NaCl). Data are means ± SD; n = 3; means with the same letter are not significantly different with P ≤ 0.05 according to ANOVA and Fisher’s LSD test

Screening of 11 Tunisian M. truncatula lines of the SARDI collection under combined stress

We examined 11 different M. truncatula genotypes (SARDI collection, lines 1–11) along with the reference line JA17 using the above selected combination of salt and alkalinity stresses in soil conditions (10/6 BIC in combination with 150 mM NaCl). We found that the germination rates were decreased by 10% or more under combined stress compared to the untreated controls up to day 10 in four out of the 11 lines, namely lines 1, 8, 9 and 10, and JA17 (Fig. 5a, hatched bars). The ratios of germination upon combined stress versus control differed slightly between the lines (Fig. 5a). Combined stress resulted in a significant decrease of root DWs in lines 2, 3, 4, 5, 7, 8, 9 and JA17 (Fig. 5b, hatched bars). The ratios of combined stress versus control were lowest for lines 2 and 4, indicating more than 80% reduction, and lines 7, 8, 9, indicating 30–50% reduction of root dry weights. Root DWs of lines 1, 6, 10 and 11 appeared fairly unaffected by the stress treatment (Fig. 5b, filled bars). Combined stress caused a significant drop of shoot DW in lines 1, 7, 9, 10 and 11 (Fig. 5c, hatched bars), but not for the other tested lines (Fig. 5c, filled bars). The ratio of combined stress versus control regarding shoot DW was lowest for line 9, indicating a 60% reduction (Fig. 5c). The total Chl content was significantly diminished upon the combined stress versus the control in line 4 (by more than 20%), but however increased in the lines 8, 10 and 11 (by 40%) (Fig. 5d, hatched bars). The carotenoid content was lowered upon the combined stress in lines 4 and 6 (by more than 60%) and again enhanced in lines 10 and 11 (by 40%) (Fig. 5d, hatched bars).
Fig. 5

Natural variation of Tunisian M. truncatula lines in response to combined alkaline and saline stress, effect on plant growth. Ratios of combined stress versus control of (A) germination rate; (B) root DW; (C) shoot DW; (D) pigment contents of 11 Tunisian M. truncatula lines (SARDI collection, lines 1–11) and Jemalong A17 (J17) grown during 15 days in soil with 10 g/kg CaCO3, 6 g/kg NaHCO3 and 150 mM NaCl. Values are the means of the ratios of individual stress values versus mean control values ± SD, that were calculated using the Taylor expansion method (Livak 1997); n = 4; means with the same letter are not significantly different at P ≤ 0.05 according to ANOVA and Fisher’s LSD test. Hatched bars indicate that the difference between treatment and respective control condition was significant, while in the case of filled bars it was not significant, as tested in pairwise comparisons by t-test with n = 4, P ≤ 0.05

Total root length and root surface were significantly increased in line 4 (hatched bars), and the root surface decreased in line 3 (hatched bar), while total root length and root surface remained unchanged in all other lines (filled bars) upon combined stress versus the respective control (Fig. 6a, b). The root diameter was significantly changed only in line 4 (decreased by 30%) in the stress versus the control condition (Fig. 6c, hatched bar). While the ratios of combined stress versus control did not vary significantly between the lines for the parameters total root length and root surface (Fig. 6a, b), some differences were noted for root diameter measurements for some lines. For example, lines 10 and 11 had a higher ratio compared to lines 4 and 9, indicating that despite of the non-significant changes of root parameter data within lines 10, 11 and 9, differences were better apparent when comparing between the lines.
Fig. 6

Natural variation of Tunisian M. truncatula lines in response to combined alkaline and saline stress, effect on root growth. Ratios of combined stress versus control of (A) total root length; (B) total root surface; (C) total root diameter of 11 Tunisian M. truncatula lines (SARDI collection, lines 1–11) grown during 15 days in soil with 10 g/kg CaCO3, 6 g/kg NaHCO3 and 150 mM NaCl. Values are the means of the ratios of individual stress values versus mean control values ± SD, that were calculated using the Taylor expansion method (Livak 1997); n = 4; means with the same letter are not significantly different at P ≤ 0.05 according to ANOVA and Fisher’s LSD test. Hatched bars indicate that the difference between treatment and respective control condition was significant, while in the case of filled bars it was not significant, as tested in pairwise comparisons by t-test with n = 4, P ≤ 0.05

Taken together, our results show that the 11 investigated M. truncatula lines from the SARDI collection show variations in the responses to combined salinity and alkalinity stress versus the control, that are apparent at the level of germination, biomass production, photosynthetic pigment contents and root morphology. For better comparison of the lines, we plotted the combined stress versus control ratios in a single table and calculated a mean response ratio from the four parameters, namely germination, biomass (average of root and shoot DWs), pigment content (average of total Chl and carotenoid contents), root morphology (average of total root length, diameter and surface) (Fig. 7). Three lines showed a high mean ratio level of more than 95% and were considered tolerant to the combined stress, namely lines 8, 10 and 11. Six lines gave a ratio below 86% and were considered sensitive to the combined stress, namely lines 1, 2, 4, 6, 7 and 9. Lines 3 and 5 with values inbetween were designated neutral.
Fig. 7

Summary assessment of Tunisian M. truncatula lines for responses to combined stress. The ratios for the different physiological and growth parameters of combined stress versus control (see Figs. 5, 6) were represented as percent values in a table. The mean percent ratio was calculated from four parameters, namely germination, biomass (average of root and shoot DWs), pigment content (average of total chl and carotenoid contents), root morphology (average of total root length, diameter and surface). The colour code indicates whether the lines behaved as tolerant, sensitive or neutral

Analysis of selected tolerant and sensitive M. truncatula lines to the combined stress in a hydroponic system

We selected four lines (7, 9, 10 and 11) to validate and re-examine the responses to the combined stress conditions, this time using hydroponic growth. The aim was to conduct additional studies to deduce potential mechanisms of tolerance.

In hydroponic growth, the total Chl contents of the sensitive lines 7 and 9 were reduced by 40–55% upon combined stress versus control, while that of the tolerant line 10 was decreased by only 30% (Fig. 8a, hatched bars). For tolerant line 11, no significant drop in Chl content was noted (Fig. 8a, filled bar). The sensitive lines showed a significant root DW reduction of ca. 65% compared to 30–35% reduction for the tolerant ones (Fig. 8b, hatched bars). Shoot DW, on the other hand, was significantly decreased in all lines by 20–30% upon the combined stress versus the control (represented by hatched bars), but no differences were apparent between the lines (Fig. 8b). The main root lengths were also found decreased upon the stress in all lines by about 40–45% (hatched bars) with no significant differences between the lines (Fig. 8c). However, with respect to the number of secondary roots a stronger decrease of 60 and 70% was noted for sensitive lines 7 and 9 compared to the tolerant lines 10 and 11 with a reduction of 25 and 30% (Fig. 8c).
Fig. 8

Effect of combined bicarbonate and salt stress (BIC/NaCl) in selected sensitive and tolerant M. truncatula lines. Relative reduction under combined stress versus control of (A) total chlorophyll contents; (B) root and shoot DWs; (C) main root length (MRL) and number of secondary roots (NSR). Lines 7 and 9 were selected as sensitive, whereas lines 10 and 11 as tolerant to the stress. M. truncatula plants were grown during 15 d in a hydroponic system under control and BIC/NaCl combined stress (+ 1 g/L CaCO3 + 10 mM NaHCO3 + 100 mM NaCl) conditions. Values are the means of the ratios of individual stress values versus mean control values ± SD, that were calculated using the Taylor expansion method (Livak 1997); n = 3; means with the same letter are not significantly different at P ≤ 0.05 according to ANOVA and Fisher’s LSD test. Hatched bars indicate that the difference between treatment and respective control condition was significant, while in the case of filled bars it was not significant, as tested in pairwise comparisons by t-test with n = 3, P ≤ 0.05

Finally, we investigated whether Rbfl accumulation in the root tips could be a relevant trait for tolerance to the stress treatments as described previously for Fe deficiency (Rodriguez-Celma et al. 2011b). Two major flavins were found in the root extracts, Rbfl and 7-hydroxy-Rbfl, confirming previous results (Rodríguez-Celma et al. 2011b). Lines 7 and 11 had low Rbfl contents in root tips, whereas they were comparatively higher in lines 9 and 10 (Fig. 9a). However, a difference between sensitive and tolerant lines was found for 7-hydroxy-Rbfl, which was detected (in amounts much lower than those of Rbfl) only upon the double stress treatment and only in the sensitive lines 7 and 9 (Fig. 9b). Another finding was that the distribution of flavins in root sections close to the main root tips visualized using fluorescence microscopy differed between sensitive and tolerant lines. While the sensitive lines 7 and 9 accumulated fluorescent compounds in the stele and adjacent inner root regions upon double stress, the tolerant lines 10 and 11 showed intense staining in the epidermis and to a weaker extent in the stele (Fig. 9c). A relatively homogeneous distribution of fluorescence was present upon the control treatments, with the staining being stronger in line 10 (Fig. 9c), thus matching the quantitative measurements shown above (Fig. 9a). Furthermore, the 7-hydroxy-Rbfl concentrations in the nutrient solution were larger in the tolerant lines 10 and 11 than in the sensitive ones 7 and 9, whereas the amounts of Rbfl present in three of the lines were small (7, 10 and 11; Fig. 10). The difference between tolerant and sensitive lines was best described by the percentage of 7-hydroxy-Rbfl of the total flavins in nutrient solution, which was 91–100% in the tolerant lines and 0–30% in the sensitive ones.
Fig. 9

Effect of combined bicarbonate and salt stress (BIC/NaCl) on flavin concentration and cellular distribution in root tips of sensitive and tolerant M. truncatula lines. (A) Rbfl and (B) 7-Hydroxy-Rbfl concentrations in root tips; (C) flavin fluorescence distribution in root cross sections, from left to right line 7, 9, 10 and 11. Plants were grown as described in Fig. 8 and the root tips of main roots were used for analysis. Data are means ± SD; n = 5; means with the same letter are not significantly different with P ≤ 0.05 according to ANOVA and Fisher’s LSD

Fig. 10

HPLC-MS (TOF) chromatograms of flavins isolated from nutrient solutions after growth of sensitive and tolerant M. truncatula lines. Rbfl and 7-Hydroxy-Rbfl chromatograms were extracted at m/z 377.15 and 379.12, respectively, with a precision of 0.05 m/z units, showing peaks corresponding to these flavins. Plants were grown as described in Fig. 8

In summary, the physiological and morphological analysis of hydroponically grown plants confirmed that lines 7 and 9 reacted in a more sensitive way to the combined BIC/NaCl treatments than the tolerant lines 10 and 11. Root morphology as well as flavin composition and the cellular distribution and secretion of flavins represent potential mechanisms that could contribute to the adaptation to the combined stress, whereas Chl contents and plant biomass are likely direct consequences of the genotype tolerance and sensitivity.

Discussion

In this study, we show that salinity and alkalinity affected different growth parameters of M. truncatula plants, and that both factors combined led in an additive manner to a failure of germination and in a synergistic manner to an increase of stress symptoms in terms of decreased biomass production. Using Tunisian genotypes of the SARDI M. truncatula collection, we found natural variation in tolerance to the combined saline and alkaline stress condition. This natural diversity allowed us to identify tolerant and susceptible lines. Differences in flavin distribution patterns, flavin concentrations in the growth medium and root morphology support that these traits could be key mechanisms contributing to stress tolerance.

Combined salinity and alkalinity conditions lead to additive and synergistic stress effects

Stress symptoms were assessed using different growth parameters, ranging from germination to biomass production, photosynthetic pigment content and root architecture. Germination is a very critical period in the plant’s life cycle and seeds are under the control of intrinsic (hormonal) and extrinsic (temperature, light and water) factors (Fenner and Thompson 2005). Germination is affected by many biotic and abiotic factors, and both salinity and alkalinity are abiotic factors that restrict germination of alfalfa seeds (Gao et al. 2011; Guan et al. 2009). Under the effect of salt stress of 100 mM or more NaCl, the low water potential of the soil likely prevented proper imbibition and seed germination. Salt toxicity may in addition have been detrimental to cells. Likewise, we found a decrease of the germination rate with increasing BIC concentrations, starting at 10/9 BIC and being most severe starting at 50/18 BIC. Certainly, the detrimental effects on germination were not due to the actual concentrations of Na ions but rather caused by the elevated soil pH. High pH leads to reduced solubility and absorption of micronutrients, and this effect is also relevant in the apoplast of cells. Thus, the perturbation of remobilization and circulation of nutrients is the likely cause for the drastic reduction of germination upon high pH. Biomass production depends on photosynthesis and a proper utilization of macro- and micronutrients. Salt negatively affects ion usage, e.g., K and Ca (Cramer et al. 1991), whereas alkaline conditions cause a reduction in Fe and P utilization (Lucena 2000). When investigating the combined stress, we found that an effect of 5/3, 10/6 or 20/9 BIC on germination was similar at 75 and 100 mM NaCl (Fig. 3a) and in the absence of salt (Fig. 2a). Hence, a clear additive or synergistic effect of the double stress was not present. However, at the higher salt concentration of 150 mM NaCl, an additional 20/9 BIC treatment resulted in an additive stress reaction (40% reduction in germination with respect to the 20 and 30% reduction with single stresses (compare Fig. 1a, 2a, 3a). With 10/6 BIC, 150 mM NaCl just resulted in a 10% germination decrease, which was not more than in the single salt condition (20%) and similar as in the single BIC condition (10%) (compare Fig. 1a, 2a, 3a). With regard to biomass production and main root length, 75 mM NaCl did not intensify the BIC-induced decrease in DW. Also, 5/3 and 10/6 BIC together with 100 or 150 mM NaCl were not sufficient to decrease further the biomass. Only in the presence of 20/9 BIC, 100 and 150 mM NaCl resulted in a more severe biomass and root length decrease than in the single stresses. For example, 150 mM NaCl did not affect significantly biomass and root length, while 20/9 BIC caused a 20% reduction in shoot DW and a 30% reduction in root DW. However, the combined 150 mM NaCl and 20/9 BIC treatment resulted in a > 50% decrease in shoot DW and a > 75% decrease in root DW, indicating a synergistic effect of the double stresses on biomass production (compare Fig. 1c, 2c, 3c). The results also showed that M. truncatula Jemalong A17 tolerated the simultaneous presence of 150 mM NaCl and 10/6 BIC. Conditions like that could resemble saline soil areas in Tunisia. These soils include salts (NaCl and Na2SO4) and alkali components (NaHCO3 and Na2CO3), and may have salt soil concentrations of 0.7% or higher, and a pH in the range of 8.5 to 10.0 (Skujinš 1984). Higher combined salt and alkaline growth conditions have been observed in other systems (Guan et al. 2009; Rabhi et al. 2007; Yousfi et al. 2007).

Root morphology and flavin cellular distribution are potential mechanisms for tolerance to combined stress in the natural diversity of M. truncatula

Our work revealed two potential mechanisms that may contribute to the tolerance.

The first mechanism is related to root architecture. Single and double stress treatments always had a negative effect on biomass and root dry weight. The even smaller reduction of the root DW in the tolerant vs. sensitive lines upon the combined stress was likely caused by the smaller decrease in the number of lateral roots. Despite of the stress, tolerant lines thus seemed better suited to sustain root growth. Therefore, the previously reported arrest of root growth to alleviate stress periods (Bailey-Serres and Voesenek 2010) does not seem to be the relevant adaptive mechanism in our studied case. Future studies can be conducted to measure the impact of root architectural changes on stress tolerance. For example, it would be interesting to find out whether the roots of tolerant plants express those genes at higher level that code for proteins, enzymes and transporters involved in the vacuolar storage and exclusion of salt. Alternatively, the root of tolerant plants might show signs for induced nutrient mobilization to deal with potential deficiencies upon the calcareous growth condition. Interestingly, we observed upon establishment of the growth conditions that the 5/3 BIC treatment was a positive stress and stimulated slightly but significantly root dry weight, while higher BIC treatments were a negative stress and decreased root biomass. Previously, it was described that the plant hormone abscisic acid (ABA) resulted in a similar root growth dose response curve in M. truncatula. Low ABA concentrations promoted and high ABA concentrations inhibited root growth (Ariel et al. 2010). ABA is involved in salt stress responses and perhaps it mediates also calcareous stress responses.

A second key mechanism relies on the allocation of flavins in root epidermal cells and flavin secretion to the growth medium, and specifically the secretion of the Rbfl derivative 7-hydroxy-Rbfl. The total concentrations of flavins in the root extracts did not allow differentiating between tolerant and sensitive lines. However, in the two tolerant lines examined, flavin fluorescence was preferentially located in the epidermal cells, the root contents of 7-hydroxy-Rbfl were undetectable, and 7-hydroxy-Rbfl was the major flavin compound in the nutrient solution. In contrast, in the two sensitive lines examined flavin fluorescence was preferentially allocated to inner parts of the root, 7-hydroxy-Rbfl was detected in the root, and 7-hydroxy-Rbfl was either undetected or a minor flavin component in the nutrient solution. These results are very interesting and match the previous observations that 7-hydroxy-Rbfl can be a major flavin in the extracts especially under low Fe conditions (Rodriguez-Celma et al. 2011b). Our findings strongly support that the allocation of Rbfl in epidermal cells and the secretion (extracellular allocation) of oxygenated Rbfl derivatives are crucial for tolerance to alkalinity stress in M. truncatula and that there is genetic diversity in this trait.

The relevance of the accumulation of flavins in the epidermal cells under Fe deficiency conditions is in line with the hypothesis that flavins in root cells constitute a potent Fe reduction system in combination with the plasma membrane ferric chelate reductase (López-Millán et al. 2000). In root tips of Fe-deficient sugar beet, accumulated (oxidized) flavins can act as an electron shuttle between reduced pyridine nucleotides in the cell (produced by the high mitochondrial activity fueled by HCO3 fixation via phosphoenolpyruvate carboxylase) and the plasma membrane ferric chelate reductase (López-Millán et al. 2000). Indeed, Fe deficiency leads to increases in phosphoenolpyruvate carboxylase, root flavins and ferric chelate reductase activity in M. truncatula (Andaluz et al. 2009). Flavin production and secretion at high pH is ecologically meaningful if plants are grown under alkaline and calcareous soil where Fe deficiency occurs. Besides the two effects on flavins and root architecture it is well conceivable that additional mechanisms of tolerance to the double stress come into play, like e.g. soil acidification as a means to mobilize nutrients as well as salt tolerance reactions based on exclusion and detoxification of salts.

Extracellular flavins enable the establishment of a long-distance electron transfer chain between the reducing power in the root and soil Fe(III)-oxides (even at high pH), resulting in the formation of soluble Fe species for the plasma membrane Fe uptake system (Sisó-Terraza et al. 2016a). The Fe(III)-oxide reduction rates were similar for Rbfl and Rbfl-sulfates but were strongly dependent on the flavin concentration (Sisó-Terraza et al. 2016b). Therefore, the secretion of flavin species with a high solubility (e.g., hydroxylated ones) can constitute an ecological advantage to mine Fe from calcareous soil. For instance, the solubility of Rbfl-sulfates (the major flavin component secreted by sugar beet roots) is one to two orders of magnitude higher than that of Rbfl (Susin et al. 1993). In this context, coumarins secreted by A. thaliana roots, effective in dissolving soil Fe oxides at high pH values, also include a number of hydroxylated species that become predominant at high pH (Siso-Terraza et al. 2016a).

In summary, the combined stress affected most of the tested M. truncatula lines. Tolerant and sensitive lines were collected from various parts of Tunisia, and their geographic distribution might be correlated with their degree of evolutionary relatedness to some extent. Further studies at the genomics level may provide answers to this. The selected double stress-tolerant and sensitive lines can be used as contrasting pairs for comparative RNA-Seq experiments to search for adaptive changes in gene regulation and gene sequences. Also, the relevance for tolerance to stress of the flavin allocation in the epidermal root cells and the secretion of 7-hydroxy-Rbfl by roots should be validated in further studies with M. truncatula and the molecular basis for this mechanism explored in detail. These tolerant genotypes may be cultivated on salt-affected soils prone to alkalinity-induced Fe deficiency. In future experiments the interplay of Na and Ca and resulting effects on soil structure could be further explored using natural soils.

Notes

Acknowledgements

This work was supported partly by a DAAD Stibed stipend to H. B. A. provided through the Heinrich Heine University, and Spanish MINECO projects AGL2013-42175-R and AGL2016-75226-R (co-financed with FEDER) to J. A. and A. A.-F.

Supplementary material

11104_2017_3379_Fig11_ESM.gif (100 kb)
Supplementary Fig. 1

(GIF 99 kb)

11104_2017_3379_MOESM1_ESM.tif (1.6 mb)
High resolution image (TIFF 1597 kb)

References

  1. Abadia J, Vazquez S, Rellan-Alvarez R, El-Jendoubi H, Abadia A, Alvarez-Fernandez A, Lopez-Millan AF (2011) Towards a knowledge-based correction of iron chlorosis. Plant Physiol Biochem 49:471–482.  https://doi.org/10.1016/j.plaphy.2011.01.026 CrossRefPubMedGoogle Scholar
  2. Andaluz S, Rodríguez-Celma J, Abadía A, Abadía J, López-Millán A-F (2009) Time course induction of several key enzymes in Medicago truncatula roots in response to Fe deficiency. Plant Physiol Biochem 47:1082–1088CrossRefPubMedGoogle Scholar
  3. Ariel F, Diet A, Verdenaud M, Gruber V, Frugier F, Chan R, Crespi M (2010) Environmental regulation of lateral root emergence in Medicago truncatula requires the HD-Zip I transcription factor HB1. Plant Cell 22:2171–2183.  https://doi.org/10.1105/tpc.110.074823 CrossRefPubMedPubMedCentralGoogle Scholar
  4. Arnon DI (1949) Copper enzymes in isolated chloroplasts. Polyphenoloxidase in Beta vulgaris. Plant Physiol 24:1CrossRefPubMedPubMedCentralGoogle Scholar
  5. Bailey-Serres J, Voesenek LA (2010) Life in the balance: a signaling network controlling survival of flooding. Curr Opin Plant Biol 13:489–494.  https://doi.org/10.1016/j.pbi.2010.08.002 CrossRefPubMedGoogle Scholar
  6. Basigalup D, Irwin J, Mi F, Abdelguefri-Laouar M (2014) Perspectives of alfalfa in Australia, China, Africa and Latin America. The journal of the International Legume Society: Legume Perspectives 9–10Google Scholar
  7. Briat J-F, Dubos C, Gaymard F (2015) Iron nutrition, biomass production, and plant product quality. Trends Plant Sci 20:33–40CrossRefPubMedGoogle Scholar
  8. Cramer GR, Epstein E, Läuchli A (1991) Effects of sodium, potassium and calcium on salt-stressed barley. Physiol Plant 81:197–202CrossRefGoogle Scholar
  9. Ellwood S, D’souza N, Kamphuis L, Burgess T, Nair R, Oliver R (2006) SSR analysis of the Medicago truncatula SARDI core collection reveals substantial diversity and unusual genotype dispersal throughout the Mediterranean basin. Theor Appl Genet 112:977–983CrossRefPubMedGoogle Scholar
  10. Fenner M, Thompson K (2005) The ecology of seeds. Cambridge University PressGoogle Scholar
  11. Fourcroy P, Siso-Terraza P, Sudre D, Saviron M, Reyt G, Gaymard F, Abadia A, Abadia J, Alvarez-Fernandez A, Briat JF (2014) Involvement of the ABCG37 transporter in secretion of scopoletin and derivatives by Arabidopsis roots in response to iron deficiency. New Phytol 201:155–167.  https://doi.org/10.1111/nph.12471 CrossRefPubMedGoogle Scholar
  12. Friesen ML, Cordeiro MA, Penmetsa RV, Badri M, Huguet T, Aouani ME, Cook DR, Nuzhdin SV (2010) Population genomic analysis of Tunisian Medicago truncatula reveals candidates for local adaptation. Plant J 63:623–635CrossRefPubMedGoogle Scholar
  13. Friesen ML, von Wettberg EJ, Badri M, Moriuchi KS, Barhoumi F, Chang PL, Cuellar-Ortiz S, Cordeiro MA, Vu WT, Arraouadi S (2014) The ecological genomic basis of salinity adaptation in Tunisian Medicago truncatula. BMC Genomics 15:1CrossRefGoogle Scholar
  14. Gao ZW, Zhu H, Gao JC, Yang CW, Mu CS, Wang DL (2011) Germination responses of Alfalfa (Medicago sativa L.) seeds to various salt-alkaline mixed stress. Afr J Agric Res 6:3793–3803Google Scholar
  15. Guan B, Zhou D, Zhang H, Tian Y, Japhet W, Wang P (2009) Germination responses of Medicago ruthenica seeds to salinity, alkalinity, and temperature. J Arid Environ 73:135–138.  https://doi.org/10.1016/j.jaridenv.2008.08.009 CrossRefGoogle Scholar
  16. Hanin M, Ebel C, Ngom M, Laplaze L, Masmoudi K (2016) New Insights on Plant Salt Tolerance Mechanisms and Their Potential Use for Breeding. Front Plant Sci 7:1787.  https://doi.org/10.3389/fpls.2016.01787 CrossRefPubMedPubMedCentralGoogle Scholar
  17. Hsieh EJ, Waters BM (2016) Alkaline stress and iron deficiency regulate iron uptake and riboflavin synthesis gene expression differently in root and leaf tissue: implications for iron deficiency chlorosis. J Exp Bot 67:5671–5685.  https://doi.org/10.1093/jxb/erw328 CrossRefPubMedPubMedCentralGoogle Scholar
  18. Lazrek F, Roussel V, Ronfort J, Cardinet G, Chardon F, Aouani M, Huguet T (2009) The use of neutral and non-neutral SSRs to analyse the genetic structure of a Tunisian collection of Medicago truncatula lines and to reveal associations with eco-environmental variables. Genetica 135:391–402CrossRefPubMedGoogle Scholar
  19. Li Q, Yang A, Zhang WH (2016) Efficient acquisition of iron confers greater tolerance to saline-alkaline stress in rice (Oryza sativa L.) J Exp Bot 67:6431–6444.  https://doi.org/10.1093/jxb/erw407 CrossRefPubMedPubMedCentralGoogle Scholar
  20. Livak K (1997) ABI prism 7700 sequence detection system user bulletin# 2 relative quantification of gene expression. ABI company publicationGoogle Scholar
  21. López-Millán AF, Morales F, Andaluz S, Gogorcena Y, Abadía A, De Las RJ, Abadía J (2000) Responses of sugar beet roots to iron deficiency. Changes in carbon assimilation and oxygen use. Plant Physiol 124:885–898CrossRefPubMedPubMedCentralGoogle Scholar
  22. Lucena JJ (2000) Effects of bicarbonate, nitrate and other environmental factors on iron deficiency chlorosis. A review. J Plant Nutr 23:1591–1606CrossRefGoogle Scholar
  23. Naranjo-Arcos MA, Bauer P (2016) Nutritional Deficiency. Chapter 4: Iron Nutrition, Oxidative Stress, and Pathogen Defense. INTECH Open Access PublisherGoogle Scholar
  24. Rabhi M, Barhoumi Z, Ksouri R, Abdelly C, Gharsalli M (2007) Interactive effects of salinity and iron deficiency in Medicago ciliaris. C R biol 330:779–788CrossRefPubMedGoogle Scholar
  25. Rahoui S, Martinez Y, Sakouhi L, Ben C, Rickauer M, El Ferjani E, Gentzbittel L, Chaoui A (2016) Cadmium-induced changes in antioxidative systems and differentiation in roots of contrasted Medicago truncatula lines. Protoplasma: 1–17Google Scholar
  26. Rengasamy P (2016) Soil Salinization. OXFORD RESEARCH ENCYCLOPEDIA, ENVIRONMENTAL SCIENCE.  https://doi.org/10.1093/acrefore/9780199389414.013.65
  27. Rodriguez-Celma J, Lattanzio G, Grusak MA, Abadia A, Abadia J, Lopez-Millan AF (2011a) Root responses of Medicago truncatula plants grown in two different iron deficiency conditions: changes in root protein profile and riboflavin biosynthesis. J Proteome Res 10:2590–2601.  https://doi.org/10.1021/pr2000623 CrossRefPubMedGoogle Scholar
  28. Rodriguez-Celma J, Lin WD, Fu GM, Abadia J, Lopez-Millan AF, Schmidt W (2013) Mutually exclusive alterations in secondary metabolism are critical for the uptake of insoluble iron compounds by Arabidopsis and Medicago truncatula. Plant Physiol 162:1473–1485.  https://doi.org/10.1104/pp.113.220426 CrossRefPubMedPubMedCentralGoogle Scholar
  29. Rodriguez-Celma J, Vazquez-Reina S, Orduna J, Abadia A, Abadia J, Alvarez-Fernandez A, Lopez-Millan AF (2011b) Characterization of flavins in roots of Fe-deficient strategy I plants, with a focus on Medicago truncatula. Plant cell physiol 52:2173–2189.  https://doi.org/10.1093/pcp/pcr149 CrossRefPubMedGoogle Scholar
  30. Schmid NB, Giehl RF, Doll S, Mock HP, Strehmel N, Scheel D, Kong X, Hider RC, von Wiren N (2014) Feruloyl-CoA 6′-Hydroxylase1-dependent coumarins mediate iron acquisition from alkaline substrates in Arabidopsis. Plant Physiol 164:160–172.  https://doi.org/10.1104/pp.113.228544 CrossRefPubMedGoogle Scholar
  31. Sisó-Terraza P, Luis-Villarroya A, Fourcroy P, Briat JF, Abadía A, Gaymard F, Abadía J, Álvarez-Fernández A (2016a) Accumulation and Secretion of Coumarinolignans and other Coumarins in Arabidopsis thaliana Roots in Response to Iron Deficiency at High pH. Front Plant Sci 7:1711.  https://doi.org/10.3389/fpls.2016.01711 CrossRefPubMedPubMedCentralGoogle Scholar
  32. Sisó-Terraza P, Rios JJ, Abadía J, Abadía A, Álvarez-Fernández A (2016b) Flavins secreted by roots of iron-deficient Beta vulgaris enable mining of ferric oxide via reductive mechanisms. New Phytol 209:733–745.  https://doi.org/10.1111/nph.13633 CrossRefPubMedGoogle Scholar
  33. Skujinš J (1984) Microbial ecology of desert soils. Springer, Advances in microbial ecologyCrossRefGoogle Scholar
  34. Susín S, Abián J, Sánchez-Baeza F, Peleato ML, Abadía A, Gelpí E, Abadía J (1993) Riboflavin 3′-and 5′-sulfate, two novel flavins accumulating in the roots of iron-deficient sugar beet (Beta vulgaris). J Biol Chem 268:20958–20965PubMedGoogle Scholar
  35. Yousfi S, Mahmoudi H, Abdelly C, Gharsalli M (2007) Effect of salt on physiological responses of barley to iron deficiency. Plant Physiol Biochem 45:309–314CrossRefPubMedGoogle Scholar
  36. Zahaf O, Blanchet S, De Zélicourt A, Alunni B, Plet J, Laffont C, De Lorenzo L, Imbeaud S, Ichanté J-L, Diet A (2012) Comparative transcriptomic analysis of salt adaptation in roots of contrasting Medicago truncatula genotypes. Mol Plant 5:1068–1081CrossRefPubMedGoogle Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  • Heithem Ben Abdallah
    • 1
  • Hans-Jörg Mai
    • 1
  • Ana Álvarez-Fernández
    • 2
  • Javier Abadía
    • 2
  • Petra Bauer
    • 1
    • 3
  1. 1.Institute of BotanyUniversity of DüsseldorfDüsseldorfGermany
  2. 2.Department of Plant Nutrition, Aula Dei Experimental Station, Consejo Superior de Investigaciones Científicas (CSIC)ZaragozaSpain
  3. 3.Institute of BotanyHeinrich-Heine UniversityDüsseldorfGermany

Personalised recommendations