Plant Soil (2010) 332:369–385 DOI 10.1007/s11104-010-0302-9 REGULAR ARTICLE Species adaptation in serpentine soils in Lesbos Island (Greece): metal hyperaccumulation and tolerance Elena Kazakou & George C. Adamidis & Alan J. M. Baker & Roger D. Reeves & Malinda Godino & Panayiotis G. Dimitrakopoulos Received: 4 November 2009 / Accepted: 22 January 2010 / Published online: 23 February 2010 # Springer Science+Business Media B.V. 2010 Abstract Serpentine (ultramafic) soils, containing relatively high nickel and other metal concentrations, present a stressful environment for plant growth but also a preferred substrate for some plants which accumulate nickel in their tissues. In the present study we focused on: (1) the relationships between serpentine soils of Lesbos Island (Greece) and serpentinophilic species in order to test their adaptation to the ‘serpentine syndrome’, and (2) the Ni-hyperaccumulation capacity of Alyssum lesbiacum, a serpentine endemic, Nihyperaccumulating species, recorded over all its distribution for the first time. We sampled soil and the most abundant plant species from the four serpentine localities of Lesbos Island. Soil and leaf elemental concentrations were measured across all the sites. Our results confirmed our hypothesis that serpentinophilic species are adapted to elevated heavy metal soil concentrations but restricting heavy metal concentration in their leaves. We demonstrated that different A. lesbiacum populations from Lesbos Island present differences in Ni hyperaccumulation according to soil Ni availability. Our results highlighted the understanding of serpentine ecosystems through an extensive field study in an unexplored area. Alyssum lesbiacum and Thlaspi ochroleucum emerge as two strong Ni hyperaccumulators with the former having a high potential for phytoextraction purposes. Keywords Alyssum lesbiacum . Heavy metals . Lesbos . Nickel hyperaccumulation . Thlaspi ochroleucum . Ultramafic soils Responsible Editor: Juan Barcelo. E. Kazakou (*) Montpellier SupAgro, UMR Centre d’Ecologie Fonctionnelle et Evolutive, CNRS UMR 5175, 1919 route de Mende, 34293 Montpellier, France e-mail: elena.kazakou@cefe.cnrs.fr G. C. Adamidis : P. G. Dimitrakopoulos Biodiversity Conservation Laboratory, Department of Environment, University of the Aegean, University Hill, 81100 Mytilene, Lesbos, Greece A. J. M. Baker : R. D. Reeves : M. Godino School of Botany, The University of Melbourne, Parkville, VIC 3010, Australia Introduction Species plant adaptation to different soil types has been recognized as a consequence of the strong natural selection imposed by ecological discontinuities (Wallace 1858). Among such examples of edaphic specialization, plant adaptation to serpentine (ultramafic) soils is an ideal system for studies in evolutionary ecology and satisfies the key requirements for addressing mechanistic questions of adaptive evolution in nature (Brady et al. 2005). Serpentine environments are typically inhospitable for many 370 plants because of three factors of their soil chemistry, collectively called the ‘serpentine syndrome’ (Kruckeberg 1984; reviewed by Kazakou et al. 2008): (a) low availability of calcium relative to magnesium; (b) deficiency of other essential macronutrients (P, N, K), and (c) high levels of potentially phytotoxic elements (Ni, Cr, Co, and sometimes Mn and/or Cu). Since the serpentine soil properties are disadvantageous for most plants, distinctive vegetation communities have evolved on serpentine soils (Brooks 1987). Species growing on these soils can be classified into two categories: (a) serpentine-tolerant plants, which withstand the serpentine conditions, but are also found more widely, and often show better growth elsewhere, and (b) serpentine-endemic plants, restricted exclusively to serpentine soil (Reeves et al. 1999). A very important facet of the serpentine problem is the understanding of the ecology of species tolerant to serpentine soils. As previously discussed, serpentinetolerant plants must endure a variety of adverse chemical conditions. The adaptive mechanism(s) that confer to plants tolerance to soils with limited quantities of Ca and high concentrations of Mg and heavy metals is still not well understood. Lee et al. (1977) compared the foliar concentrations of nine elements (N, P, K, Ca, Mg, Ni, Cu, Co, and Cr) in 12 plant species growing on and off serpentine soil in New Zealand. In each case, the foliar concentrations of Mg and Ni were significantly higher in plants growing on serpentine soil than in conspecifics growing on nonserpentine soil. Conversely, the concentration of Ca was significantly less for plants growing on serpentine soil. In several species, the concentration of Cu, Co, and/or Cr was also significantly higher in plants growing on serpentine soil. However, our current understanding of the serpentine-tolerant phenotype (‘syndrome’) is somewhat vague mostly because the majority of studies have compared ecotypes from serpentine and non-serpentine populations of one species (e.g. Freitas and Mooney 1996 for Bromus hordeaceus; Reeves and Baker 1984 for Thlaspi goesingense; Wright et al. 2006 for Collinsia sparsiflora) and there is a lack of field studies comparing ecotypes of many species grown on both serpentine and non-serpentine soils (but see Ater et al. 2000). One strategy of plant adaptation to serpentine soil is heavy metal and especially Ni hyperaccumulation. Hyperaccumulation is one of the most remarkable phenomena in nature and is known for Al, Cu, Co, Plant Soil (2010) 332:369–385 Mn, Ni and Zn. At least 400 Ni hyperaccumulators (defined as plants in which a nickel concentration of at least 1,000 mg kg−1 has been recorded in the dry matter of any above-ground tissue in at least one specimen growing in its natural habitat) are known to exist (Brooks 1998; Reeves and Baker 2000). A high proportion of Ni hyperaccumulators (85–90%) appear to be serpentine-endemic species (Reeves et al. 2007). It should be emphasized that Ni hyperaccumulation is quite rare, having so far been found in some 2% of species on serpentine worldwide but its biochemical and physiological interest and the potential applications of the hyperaccumulator species far outweigh this rarity (Kazakou et al. 2008). Part of the reason for the intense interest in Ni hyperaccumulators (which usually form only a small fraction of the serpentine flora) is their potential use in cleaning-up sites that are rich in nickel. A crucial question about Ni hyperaccumulation is whether the hyperaccumulator species responds to the heterogeneity of serpentine soils by showing intraspecific variation in foliar Ni levels. There has been no definitive study to demonstrate that serpentine soils can also have a local heterogeneity and, in the affirmative case, if this heterogeneity can affect the mineral element composition of the populations of a hyperaccumulator species. A field study at the population level can link theoretical research on the estimation of differences between levels of metal concentration in plants to the practical advantage of being a possible first step towards the selection of populations with the highest phytoextraction ability (Pollard et al. 2002). The Ni hyperaccumulator species include about 50 taxa out of 172 species in the genus Alyssum (Brassicaceae) (Brooks et al. 1979; Baker and Brooks 1989; Reeves and Baker 2000; Reeves 2006; Ghaderian et al. 2007a, b; Reeves and Adigüzel 2008). Alyssum lesbiacum (Candargy) Rech. f. is endemic to Lesbos Island, Greece (Strid and Tan 2002) and is well established as a Ni hyperaccumulator (Brooks et al. 1979; Reeves et al. 1997). Recent studies have identified the mechanism for its Ni uptake and hyperaccumulation (Krämer et al. 1996; Ingle et al. 2005) and the cellular compartmentation of Ni in its tissues (e.g. Krämer et al. 1997; Küpper et al. 2001; Smart et al. 2007) but its intraspecific variability along a heavy metal gradient has not been studied. In this work we report data obtained from a survey of soil and plant samples collected from serpentine Plant Soil (2010) 332:369–385 and non-serpentine habitats on Lesbos island (Greece) to evaluate: (1) levels of element concentrations on serpentine and non-serpentine soils and the response of species to elevated heavy metal concentrations. We tested the hypothesis that species tolerant to serpentine soils can be considered as ‘excluders’, i.e. plants that restrict transport of metals to the shoot, and maintain relatively low metal concentrations in the shoot over a wide range of soil metal concentrations (sensu Baker 1981); and (2) the response of populations of A. lesbiacum (sampled across its whole geographical distribution) to the variable serpentine soils of Lesbos Island. We hypothesized that A. lesbiacum would adapt its Ni hyperaccumulation ability according to Ni soil availability. Material and methods Study sites The distribution of serpentine substrata in Lesbos is depicted in Fig. 1. Four serpentine sites were selected in the following localities: Loutra, Ampeliko, Olympos and Vatera (Table 1; Fig. 1). The selection was based on the presence at each site of large populations of A. lesbiacum, and designed to give coverage of the altitudinal and geographic ranges shown by this species. These sites are open grassland, olive groves or understorey of sparse pine forest (Pinus brutia Ten.). In addition, four non-serpentine sites (where A. lesbiacum does not occur) were selected to serve as a control for comparison with the serpentine sites. All non-serpentine sites were located on the same geological substratum, i.e. alluvial plains. Soil sampling and analysis During May–June 2007, five soil samples were collected from 1 to 10 cm depth at each site. This depth corresponds to the major rooting zone of the herbs and small shrubs (Dimitrakopoulos and Schmid 2004; Reeves et al. 2007). Soil samples were initially air-dried, sieved to <2 mm, and stored at 4°C until analysis. Sub-samples of 4–5 g were ground to pass a 371 70-mesh sieve (<215 μm) and dried at 70°C. A further subsample of 0.30–0.35 g was weighed to ±0.0001 g and transferred to a polypropylene beaker on a water bath at 100°C for digestion with 20 ml of a 1:1 HF/HNO3 mixture. After the solution had been taken to dryness the residue was dissolved in 20 ml of conc. HCl, taken to dryness again, and the residue finally dissolved in 20.0 ml of warm 2 M HCl. A further dilution by a factor of 10–20 was generally necessary to give solutions with suitable concentrations of Fe (<1,000 mg/l) and other analyte elements for multielement analysis by inductively coupled plasma (ICP) emission spectroscopy. Plant species selection and analysis A total of 21 plant species (20 herbaceous species and the serpentine endemic small perennial shrub A. lesbiacum) from 9 families were selected (Appendix I). These species were the most abundant and made up more than 80% of the aboveground biomass of each community (Garnier et al. 2004). Leaf specimens from 10 individuals of each species were collected in May 2007. Samples of each species were collected from at least one serpentine site, depending on its presence and relative abundance at each site. No attempt was made to collect leaf material of each species from all of the non-serpentine sites. Leaf tissue was washed with double-distilled water and dried at 60°C. About 1 g of dried leaf tissue was set aside in paper bags for analysis. ICP was used for the determination of heavy metal concentrations (see Reeves et al. 2007 for a description of the analysis). Data analysis For all the variables measured, the distribution of values was tested for normality (Shapiro-Wilks test, α=0.05). The soil and plant heavy metal data were subjected to analysis of variance. For the soil metal data we performed one-way ANOVA with soil type (serpentine, non-serpentine) as the main factor. A principal components analysis (PCA) was performed to investigate the relationships between the serpentine sites in terms of soil element concentrations and to reveal any gradient across them. Differences in soil element concentrations among serpentine sites were also examined using one-way ANOVA. 372 Plant Soil (2010) 332:369–385 Fig. 1 Map of Lesbos Island showing serpentine (●) and non-serpentine (▲) localities of the present study The effects of species, soil type and their interaction on the plant metal concentrations were tested with a two-way ANOVA. Bivariate correlations between plant metal concentrations were evaluated using Pearson’s rank coefficient. Plant elemental concentrations of species restricted to serpentine habitats (i.e. A. lesbiacum and Thlaspi ochroleucum Boiss. and Heldr.) were analysed using one-way Table 1 Study sites on serpentine and nonserpentine substrata from Lesbos Island Locality ANOVA. The latter species is known on Lesbos only from the serpentine on Olympos (Strid and Tan 2002; Bazos and Yannitsaros 2004), although its locations elsewhere in Europe (Akeroyd 1993) and Turkey (Hedge 1965) cover a wider range of geology. If the effect of locality was significant, post hoc tests (Student-Newman-Keuls comparisons) were carried out to identify variations among the four localities. Coordinates Latitude (Ν) Longitude (Ε) 39° 05′ 46.4″ 026° 19′ 59.9″ Altitude (m) Orientation Slope (%) 361 SW 20 Serpentine sites Ampeliko Loutra 39° 02′ 36.8″ 026° 32′ 55.0″ 94 NW 13 Olympos 39° 04′ 33.3″ 026° 20′ 16.3″ 759 NW 15 Vatera 39° 01′ 57.1″ 026° 15′ 52.5″ 37 NE 6 026° 16′ 31.9″ 25 NE 24 Non-serpentine sites Ampeliko 39° 08′ 42.6″ Loutra 39° 01′ 19.8″ 026° 32′ 43.8″ 12 Olympos 39° 04′ 17.8″ 026° 20′ 34.5″ 728 Vatera 39° 01′ 38.6″ 026° 12′ 19.9″ 40 NE 2 S 4 NW 5 Plant Soil (2010) 332:369–385 The capacity of these species for accumulation of a specific metal is assessed using its accumulation factor (Baker et al. 1994) calculated as (mean leaf concentration) / (mean soil concentration). Statistical analyses were conducted with SAS (version 8; SAS Institute, Cary, NC, USA) and the R environment (R Development Core Team 2008). Results Soil elemental concentrations The ranges for all the elemental concentrations are presented in Table 2. There were systematic differences in soil composition between the two soil types (Table 2; Fig. 2). As expected, soil from serpentine sites had significantly higher Mg (Fig. 2b), Ni (Fig. 2f) and Mg/Ca quotient (Fig 2c) than those from non-serpentine sites. They were also richer in Fe (Fig. 2d), Co (Fig. 2h), Cr (Fig. 2i) and Mn (Fig. 2g) and marginally higher in Ca (P=0.088; Fig. 2a) and Zn (F=4.04, P=0.05), relative to the non-serpentine soils. Thus the serpentine soils of Lesbos can be regarded as Mg-rich rather than as Cadeficient. Non-serpentine soils had higher values of K (Fig. 2e), whereas there was a broad overlap in the variation of Na, P, Cu and Pb (P>0.05) concentrations of serpentine and non-serpentine sites. Our data showed that there is a gradient of increasing heavy metal concentration from the Vatera to Ampeliko localities. Ampeliko has the highest heavy metal concentrations (Ni, Co, Cr and Zn; Table 3), whereas Vatera has the lowest Ni concentration (Table 3; Fig. 5a for Ni) and the lowest Mg/Ca quotient (Tables 2, 3). Olympos has the highest Mg values (Tables 2, 3). The principal components analysis of the soil elemental concentrations suggests that 87.2% of the variance is accounted for by three principal component axes (Fig. 3). Axis 1 explains 58.6% of the variance; it describes sites characterized by high concentrations of heavy metals such as Ni, Mn, Zn, Co, Fe and Cr and sites with high concentrations of K and Na. Axis 2 explains 17% of the variance and is determined by high concentrations of Mg, while Axis 3 explains 11.6% and appears to be determined mainly by Ca status. 373 Differences in plant elemental concentrations between serpentine and non-serpentine soils For all measured plant elemental concentrations, the species effect was significant (Fig. 4; Appendix II). The ANOVA showed that the soil effect was also significant for all the measured leaf elemental concentrations except for K, Na and P (0.02 < F< 2.29, Appendix II). Overall, leaves of species growing on serpentine soils were significantly richer in Mg (Fig. 4b), Cr, and Ni (Fig. 4e) and had significantly higher Mg/Ca quotients (Fig. 4c). Only three species (Dactylis glomerata, Hordeum bulbosum and A. lesbiacum) were sampled at all four serpentine sites (the results for A. lesbiacum are shown in the following section). Plants from nonserpentine sites had higher leaf mean Ca concentrations (Fig. 4a), Zn (Fig. 4f), Fe (Fig. 4c) and Cu (Fig. 4d). Correlations between elements in plants from different soil types Species ranking for all the variables except for Cr, Fe, Ni and Zn was similar for serpentine and nonserpentine soils (Table 4). For the serpentine soils, positive correlations were found between leaf Ca and Mg, Ni and Ca, Ni and Cr, and Zn and Cr. Leaf K was positively correlated with Na in both soil types and with P only in the serpentine soils (Table 4). Nickel hyperaccumulators from serpentine soils The results of the analyses of variance for Ca, Mg, Fe, K, Na, Cr, Mn, Ni, P and Zn from A. lesbiacum specimens collected from the four serpentine localities are shown in Table 3. Leaf Ni concentrations are very high at all sites, ranging from 1,818 mg kg−1 at Loutra to 23,650 mg kg−1 at Ampeliko. The Mg/Ca quotient is remarkably low, ranging from 0.04 at Olympos to 0.28 at Ampeliko; no significant differences were detected between localities. The Ni accumulation factor was highest at Vatera (Fig. 5c), indicating that the lower soil Ni concentration at this site has not led to a corresponding reduction in the plant’s availability to accumulate Ni. The factor was similar at the other three sites. 374 Table 2 The minimum, maximum and mean (S.E. in parentheses) of soil elemental concentrations at serpentine and non-serpentine soils on Lesbos Metals Serpentine soils Non-Serpentine soils Vatera Min / Max Loutra Mean (SE) Ampeliko Min / Max Mean (SE) Min / Max Olympos Mean (SE) Ca (%) 0.28–2.07 0.85 (0.34) 1.09–1.3 1.202 (0.04) 0.55–0.63 0.60 (0.01) Fe (%) 4.35–5.71 4.75 (0.26) 7.65–8.48 8.12 (0.18) 9.42–10.7 10.0 (0.26) Min / Max 0.75–1.02 8.9–10.01 All sites Mean (SE) Min / Max Mean (SE) 0.9 (0.05) 0.15–1.16 0.62 (0.1) 9.58 (0.21) 1.93–7.56 4.62 (0.86) K (%) 0.35–0.39 0.37 (0.01) 0.27–0.38 0.33 (0.02) 0.24–0.35 0.29 (0.02) 0.24–0.29 0.27 (0.01) 0.22–0.63 0.40 (0.05) Mg (%) 1.06–5.96 2.67 (0.98) 3.7–4.6 4.14 (0.15) 2.65–5.32 3.95 (0.52) 7.18–9.21 8.41 (0.36) 0.6–1.93 1.13 (0.16) 8.82–24.6 14.5 (3.2) 19.75–31 23.6 (2.06) 68–136 99.2 (13.34) 58.64–102 76.2 (7.45) 1.88–10 2,040 (54.8) 30.8–172 81.9 (18.24) 23.1 (1.04) 7.5–51.3 24.7 (5.71) 1,419 (98.9) Co (mg kg−1) Cr (mg kg ) 143–501 263 (69.8) 1,034–1,185 1,118 (30.2) 2,456–7,065 Cu (mg kg−1) 19.2–25.1 21.4 (1.3) 36.5–41.3 38.13 (0.84) 33.6–41.8 Mn (mg kg−1) 737–876 805 (25.4) 872–973 925 (19.7) 1,501–2,304 1,913 (134) 1,228–1,780 Na (mg kg−1) 359–556 508 (37.6) 411–624 511 (42.8) 289–366 326 (13) 301–394 Ni (mg kg−1) 210–577 337 (72.6) 1,134–1,234 1,197 (16.8) 2,694–3,740 3,326 (202) 1,748–2,392 642–1,191 −1 P (mg kg ) 326–583 449 (49.4) 483–695 582 (33.6) Pb (mg kg−1) 22.9–27.6 25.3 (0.97) 21.2–37 30.9 (3.4) −1 4,863 (859) 36.9 (1.42) 918 (97) 25.1 1,894–2,234 21.4–27 4.31 (1.45) 200–1,355 745 (160) 368 (16.9) 408–614 471 (18.1) 1,948 (113) 37.9–165 84.4 (15.48) 479–571 531 (15.8) 20.74–34.6 22.64 (1.91) 215–1,141 640 (132.25) 22.5–28.3 26.4 (1.01)) Zn (mg kg−1) 57.4–70.6 61.4 (2.45) 60.8–70.8 66.5 (1.8) 110–148 132 (7.3) 71.3–88 78.5 (3.06) 31.7–95.3 61.4 (9.16) Mg/Ca 2.87–4.03 3.34 (0.19) 2.97–3.73 3.45 (0.14) 4.8–8.87 6.5 (0.8) 7.01–12.27 9.57 (0.85) 1.18–6.11 2.20 (0.46) Plant Soil (2010) 332:369–385 Plant Soil (2010) 332:369–385 1,0 (a) 375 Fsoil= 3.11ns 6 (b) Fsoil= 20.2*** 7 Fsoil= 12.9** 6 5 0,8 (c) 5 0,4 soil Mg/Ca soil Mg (%) soil Ca (%) 4 0,6 3 4 3 2 2 0,2 1 0 0,0 (d) Fsoil= 14.6** 0,5 8 0,4 6 0,3 4 0 Serpentine soil K (%) soil Fe (%) 10 Non serpentine 2 (e) Non serpentine Serpentine (f) Fsoil= 4.80* 2500 0,2 0,1 0 (g) Non serpentine 70 1400 1500 1000 0 Serpentine Fsoil= 7.48* Fsoil= 19.3*** 500 0,0 Serpentine Non serpentine 2000 soil Ni (mg.kg-1) Serpentine 1600 1 (h) Non serpentine Serpentine Fsoil= 7.37* 3000 (i) Non serpentine Fsoil= 9.87** 60 2500 1000 800 600 400 50 soil Cr (mg.kg -1) soil Co (mg.kg-1) soil Mn (mg.kg -1) 1200 40 30 20 0 0 Serpentine Non serpentine 1500 1000 500 10 200 2000 0 Serpentine Non serpentine Serpentine Non serpentine Fig. 2 Elemental concentrations in the soil at serpentine (black bars) and non-serpentine sites (white bars) in Lesbos. F-values for ANOVA testing the effect of soil are indicated. ***P<0.001; **P<0.01; *P<0.05. All results are means of original data ± SE Thlaspi ochroleucum was found on Lesbos only at the Olympos serpentine site, with Ni concentrations of 499–3,331 mg kg−1 (mean 1,277 mg kg−1, n=10), in line with its previously reported status as a Ni hyperaccumulator. However, this was significantly lower than the Ni levels of 5,860–21,360 mg kg−1 (mean 11,020 mg kg−1, n=20) in A. lesbiacum from the same area (F=86.19, P<0.001; Fig. 5b). 376 Table 3 Results of oneway ANOVA tests (F-values and probabilities) for the effects of locality on soil elemental concentrations and for Alyssum lesbiacum elemental concentrations for the four sites on Lesbos Island. Results of the post hoc tests for the differences between the four sites are presented Plant Soil (2010) 332:369–385 Soil concentrations Alyssum lesbiacum leaf concentrations Metals Locality post hoc Ca 1.97ns Mg 18.2*** O>L,A,V Mg/Ca 24.5*** O>A>L,V Fe 109*** A,O>L>V K 9.4** V,L>A,O Cu 57.3*** L,A>O,V Mn 35.7*** A>O>L,V Co 26.9*** A,O>L,V Na 9.9** L,V>O,A Ni 109*** A>O>L>V Cr 21.4*** A>O,L>V P 12.9*** A>L,O,V Zn 58.2*** A>O,L>V Ca 6.96** V,L,A>O Mg 11.7*** V,A,L>O Fe 3.55* V,L>A,O K 0.87ns Cu 0.21ns Mn 12.0*** ns, not significant. A = Ampeliko; L = Loutra, O = Olympos, V = Vatera Na 0.89ns Ni 12.8*** P 1.19ns ***P<0.001; **P<0.01; *P<0.05 Zn 7.44*** L>V>A,O A>O,V>L A,O>L,V Discussion Plants adaptation on serpentine soils of Lesbos: did excluders exist? The first hypothesis tested in the present study was that species tolerant to serpentine soils can be considered as ‘excluders’ maintaining relatively low metal concentrations in their leaves even when soil metal concentrations are very high. In order to test this hypothesis, we determined metal concentrations of all the serpentine soils of Lesbos and leaf metal concentrations of the most abundant species. The Fig. 3 Principal components analysis combining data on soil metal concentrations from four serpentine sites on Lesbos. The first two axes that account for 75.6% of the total variance are depicted Fig. 4 Mean values and SE of elemental concentrations in the plants; a Ca, b Mg, c Mg/Ca, d Cu, e Ni and f Zn. Black bars: serpentine soils, white bars: non-serpentine soils. F-values for ANOVA testing the effect of species, soil and their interaction are indicated. Species are ranked according to the highest elemental concentration in serpentine soils. ***P<0.001; **P< 0.01; *P<0.05, ns, not significant. (ANOVA for Ni was realised without metal values for both A. lesbiacum and T. οchroleucum) b (c) (a) TORINODO HIRSINCA PLANLAGO TRIFANGU LAGOCUMI CREPCOMM ANAGARVE TRIFCAMP TRIFARVE SANGMINO FILAERIO PHLEPRAT LOLIRIGI CYNOECHI TRACDIST HORDBULB AEGIBIUN DACTGLOM BROMCOMM ALYSLESB THLAOCHR Plant Soil (2010) 332:369–385 80x103 60x103 20x103 0 2,0 (e) Fspecies = 88.6*** Fsoil = 74.9*** Finteraction = 7.4*** F species = 15.5*** F soil = 156*** F interaction = 11.4*** Fspecies = 9.98*** Fsoil = 95*** Finteraction = 9.7*** BROMCOMM DACTGLOM CYNOECHI LOLIRIGI THLAOCHR HORDBULB TRACDIST AEGIBIUN CREPCOMM HIRSINCA SANGMINO PHLEPRAT TRIFANGU TRIFARVE ANAGARVE FILAERIO PLANLAGO TORINODO LAGOCUMI TRIFCAMP ALYSLESB 1,8 1,6 1,4 1,2 1,0 0,8 0,6 0,4 0,2 0,0 120 100 80 60 40 20 0 TORINODO LAGOCUMI FILAERIO TRIFARVE LOLIRIGI TRIFCAMP CYNOECHI CREPCOMM PHLEPRAT SANGMINO ANAGARVE TRIFANGU HIRSINCA PLANLAGO DACTGLOM BROMCOMM AEGIBIUN TRACDIST HORDBULB Ca (mg.kg-1) Mg/Ca 40x103 Ni (mg kg-1) -1 Mg (mg kg ) 18000 16000 14000 12000 10000 8000 (b) (d) Fspecies = 25.6*** Fsoil = 114*** Finteraction = 6.4*** Fspecies = 10.7*** Fsoil = 61.2*** Finteraction = 3.7*** HIRSINCA TORINODO CREPCOMM TRIFANGU PLANLAGO SANGMINO LAGOCUMI ANAGARVE TRIFARVE LOLIRIGI TRIFCAMP CYNOECHI FILAERIO PHLEPRAT TRACDIST HORDBULB DACTGLOM BROMCOMM AEGIBIUN ALYSLESB THLAOCHR 6000 0 16 14 12 10 8 6 4 (f) Fspecies = 17.8*** Fsoil = 47.4*** Finteraction = 14.7*** BROMCOMM ANAGARVE AEGIBIUN FILAERIO TRIFARVE HIRSINCA TRIFANGU CREPCOMM TRIFCAMP PLANLAGO CYNOECHI SANGMINO LAGOCUMI LOLIRIGI TORINODO DACTGLOM PHLEPRAT HORDBULB TRACDIST ALYSLESB THLAOCHR 2 0 600 500 400 300 200 100 0 CREPCOMM FILAERIO ANAGARVE LAGOCUMI PLANLAGO CYNOECHI TRIFCAMP TRIFARVE TORINODO HIRSINCA TRACDIST LOLIRIGI SANGMINO AEGIBIUN PHLEPRAT BROMCOMM TRIFANGU HORDBULB DACTGLOM ALYSLESB THLAOCHR 4000 Cu (mg.kg-1) 2000 Zn (mg.kg-1) 377 378 Plant Soil (2010) 332:369–385 Table 44 Pearson Pearson correlation correlationcoefficients coefficients between between measured measured plant elemental concentrations within and eachnon-serpentine soil type (serp: serpentine; correlations between serpentine soils for each non-serp: plant elemental non-serpentine) concentrations +(n=21within for serpentine each soilandtype n=19 (serp: for non-serpentine). In the first are column Pearson(A.correlations elemental concentration presented lesbiacum between and T. serpentine; and non-serp: non-serpentine) for serpentine serpentine non-serpentine soils for +(n=21 each elemental concentration are presented (A. lesbiacum and T. were ochroleucum elemental ochroleucum elemental concentrations not used in this and n=19 for were non-serpentine). In analysis) the first column Pearson concentrations not used in this analysis) Plant concentrations Soil type Ca serp Mg 0.93*** non-serp Mg serp 0.75*** non-serp K serp K serp Cr serp serp serp Ni serp P serp Zn serp Ni P Zn ns ns ns 0.53* ns 0.47* ns ns ns ns ns ns ns ns 0.66* ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns 0.55** ns ns ns ns ns 0.45* ns 0.81*** ns ns ns ns ns ns ns ns ns 0.453* 0.51* ns ns ns ns ns ns ns ns ns ns ns ns ns 0.67** ns 0.47* 0.55* ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns 0.62** ns ns ns ns ns ns ns ns 0.57* ns ns 0.77*** 0.66** ns 0.52* ns non-serp Mn Mn ns non-serp Fe Fe ns non-serp serp Cu 0.73*** non-serp Cu Cr ns non-serp Na Na 0.74*** non-serp ns non-serp ns 0.52* non-serp ns ns ns ns non-serp ns, not significant ***P<0.001; **P<0.01; *P<0.05 serpentine localities of Ampeliko, Loutra and Olympos have total soil concentrations of Ni, Cr, Co and Mn that fall into the expected range of typical ultramafic soils (see: e.g. Reeves et al. 1997, 1999; Freitas et al. 2004; Ghaderian et al. 2007a, b; Oze et al. 2008). For Vatera, the serpentine soil is lower in concentrations of elements such as Cr and Ni and may indicate only a partial ultramafic origin. However, this site fulfils the selection criteria used in this study, as it supports large populations of the serpentine-endemic species A. lesbiacum (Bazos and Yannitsaros 2004). The concentrations of K, P and Ca are also low in these soils (Reeves et al., 2007) but not significantly different from non-serpentine soils in our study (except for K). The Mg/Ca quotient is higher in serpentine relative to non-serpentine soils but much lower than the ultramafic soils of Iran (e.g. Ghaderian et al. 2007a, b), California (Oze et al. 2008) or Cuba (Reeves et al. 1999). However, comparable Mg/Ca quotients have been measured in ultramafic soils of Goiás in Brazil (Reeves et al. 2007). Differences in heavy metal concentrations (Table 4) and Mg/Ca quotients are notable between the four serpentine sites here (as in the study of Arianoutsou et al. (1993) in California), probably attributable to weathering and leaching processes combined with biological activity at each location (Alexander et al. 2007). The differences in plant elemental concentrations qualitatively mirror the differences in the soil, especially for Ni and Cr. In our study, leaf Ni concentrations were within the range of values typical for serpentine floras as reported by Brooks (1987) and Plant Soil (2010) 332:369–385 379 4000 16000 (a) Flocality = 109*** a 14000 b 2000 c 1000 a Alyssum lesbiacum Thlaspi ochroleucum b 12000 Leaf Ni (mg kg-1) Soil Ni (mg.kg-1) 3000 (b) Flocality = 12.8*** b 10000 8000 c 6000 4000 d 2000 0 0 Loutra Vatera Olympos Ampeliko 35 (c) Vatera Olymbos Ambeliko Alyssum lesbiacum Thlaspi ochroleucum 30 Ni Accumulator factor Loutra 25 20 15 10 5 0 Loutra Vatera Olympos Ampeliko Fig. 5 a Average values and SE of soil Ni concentrations (mg kg−1) in the four serpentine localities. Different letters (a, b, c, d) signify significant differences (P < 0.05) among localities. b Mean leaf Ni concentrations (mg kg−1) in Alyssum lesbiacum and Thlaspi ochroleucum specimens collected from serpentine sites (T. οchroleucum was found only at the Olympos locality). Significant differences (P<0.05) among localities in leaf Ni concentrations of A. lesbiacum are indicated by different letters (a, b, c). c Ni accumulation factor for A. lesbiacum and T. ochroleucum in the localities they were found. F-values for ANOVA testing the effect of locality are indicated. ***P<0.001; **P<0.01; *P<0.05 Reeves (1992). A very small number of species studied (1 from 19, omitting A. lesbiacum and T. ochroleucum) had Ni concentrations in their leaves higher than 60 mg kg−1, considered to be the threshold of physiological evidence of toxicity in plants of serpentine habitats (see discussion in Kazakou et al. 2008), confirming that no nickelhyperaccumulating but tolerant species exist in serpentine soils of Lesbos island. So the hypothesis that serpentine tolerant species are adapted to high heavy metal soil concentrations but restrict heavy metal accumulation in their leaves was confirmed excepted for Torilis nodosa. For foliar macronutrient concentrations, species from non-serpentine soils had much lower Mg/Ca quotients than species from serpentine soils, which can be explained by the higher foliar Ca and lower Mg concentration in non-serpentine species. It is noteworthy that all the species studied growing in serpentine soils (except Bromus commutatus) had 380 foliar mean Mg/Ca quotients <1, suggesting that these species preferentially take up Ca. This pattern is quite unusual for plants grown in serpentine soils with high Mg/Ca quotients, considered to be a principal cause of serpentine infertility (Proctor and Woodell 1975; Brooks 1987). This pattern was also found by Lombini et al. (1998) in serpentine species from Italy and can be explained by the hypothesis of selectivity: species growing on serpentine soils possess a mechanism for limiting the uptake of Mg and have a greater absorption capacity for Ca, and thus the lower available Ca is better utilized (Walker 1954). The inter-element correlations across species growing on serpentine soils in our study showed a significant correlation between Ni and Ca but no correlation between Ni and Mg contrary to previous studies (Ater et al. 2000; Shewry and Peterson 1976). Moreover, foliar Ni was not correlated with Zn but we found a correlation between Ni and Cr. It has been suggested that heavy metals are antagonistic to the uptake of other elements (Brooks and Yang 1984). In our study we found that that even if soil Zn and Mn were higher in the serpentine soils, species from non-serpentine soils exhibited higher Zn and Mn foliar concentrations. We assumed that species growing in serpentine soils generally act as excluders of Zn and Mn sensu Baker (1981) in that they restrict transport of metals to the shoots and maintain relatively low metal concentrations in leaves even at high soil concentrations. A notable exception to this behaviour, however, is Thlaspi caerulescens, which can be found on serpentine soils with remarkably high Zn concentrations (>1,000 mg kg−1) in addition to its Ni hyperaccumulation, even though the soil Zn concentrations in serpentines are usually in the normal range (Reeves et al. 2001). Intraspecific variation of Ni hyperaccumulation according to a Ni soil gradient The second hypothesis tested was that A. lesbiacum would modify its ability to accumulate Ni according to soil Ni availability. Our data showed that differences in leaf Ni concentrations of A. lesbiacum were detected between the different study locations (Table 3; Fig. 5b). Nickel concentrations were always >1,000 mg kg−1 across all locations confirming A. Plant Soil (2010) 332:369–385 lesbiacum ability to hyperaccumulate the metal (Brooks et al. 1979). What emerges is that different A. lesbiacum populations presented important differences in Ni-hyperaccumulation. The maximal leaf Ni concentration measured in this study was 23,650 mg kg−1 (c. 2.4%), very close to the value given by Brooks et al. (1979) using herbarium materials. A. lesbiacum is, therefore, one of about 15 species belonging to the genus Alyssum that have shown some Ni values above c. 2% (Fig. 6; Brooks and Radford 1978; Brooks et al. 1979; Freitas et al. 2004; Ghaderian et al. 2007a, b; Reeves and Adigüzel 2008). Our hypothesis was confirmed for three of the four serpentine sites: the Ampeliko site with the highest soil and leaf Ni concentration, the Olymbos site with intermediate soil and leaf Ni concentration and Loutra with the lowest. Only at the Vatera site did A. lesbiacum present a relatively high Ni concentration even with a very low soil Ni concentration, so resulting in the highest Ni accumulation factor. The mechanisms responsible for Ni uptake and accumulation by A. lesbiacum are related to free histidine production (Krämer et al. 1996; Ingle et al. 2005). Krämer et al. (1996) found that Ni concentration was linearly related to free histidine occurring in the xylem of A. lesbiacum when it was exposed to a gradient of Ni concentrations. Ingle et al. (2005) revealed that constitutively enhanced histidine biosynthesis in A. lesbiacum is the main cause of the production and maintenance of high levels of free histidine in its roots despite the increased shoot concentration of histidine when it is exposed to nickel. Considering the concentrations of Cr, Fe and Mn in the leaves of A. lesbiacum, values for plants growing on serpentine substrata on Lesbos are low. These results are in accordance with those from studies conducted on serpentine habitats worldwide (e.g. Lombini et al. 1998; Proctor et al. 2004; Ghaderian et al. 2007a; but see Freitas et al. 2004 for A. serpyllifolium). Alyssum lesbiacum has a low Mg/Ca quotient but it falls into the range of values for other Alyssum species (e.g. Lombini et al. 1998; Shallari et al. 1998; Ghaderian et al. 2007a) or serpentine species (e.g. Reeves et al. 1999). This is because the concentration of Ca in Alyssum is remarkably high (as in the other plant of the Brassicaceae in our study, Hirschfeldia incana), especially considering the low soil Ca. This is a consequence of the unusual ability Plant Soil (2010) 332:369–385 381 Alyssum cassium Alyssum eriophyllum Alyssum serpyllifolium lusitanicum Alyssum peltarioides subsp. virgatiforme Alyssum masmenaeum Alyssum samariferum Alyssum lesbiacum Alyssum cypricum Alyssum floribundum Alyssum pterocarpum Alyssum dudleyi Alyssum pinifolium Alyssum corsicum Alyssum murale murale var. murale Alyssum crenulatum Alyssum davisianum Alyssum peltarioides peltarioides Alyssum dubertretii Alyssum callichroum Alyssum constellatum Alyssum pateri Alyssum caricum Alyssum chondrogynum Alyssum oxycarpum Alyssum murale murale var. haradjianii Alyssum markgraphii Alyssum cilicicum Alyssum huber-morathii Alyssum bertolonii Alyssum robestrianum Alyssum heldreichii Alyssum trapeziforme Alyssum discolor Alyssum argenteum Alyssum syriacum Alyssum janchenii Alyssum troodii Alyssum akamasicum Alyssum sibiricum Alyssum anatolicum Alyssum longistylum Alyssum penjwinensis Alyssum giosnanum Alyssum murale Alyssum smolicanum Alyssum virgatum Alyssum obovatum Alyssum euboeum Alyssum fallacinum Alyssum inflatum Alyssum alpestre Alyssum tenium Alyssum condensatum Alyssun bracteum Alyssum singarense Present study 0 10000 20000 30000 40000 50000 60000 70000 -1 Ni in plant dry matter (mg.kg ) Fig. 6 Maximum Ni concentration (mg kg−1) measured in specimens of Alyssum species that meet the definition of Ni hyperaccumulators (data from: Brooks and Radford 1978; Brooks et al. 1979; Freitas et al. 2004; Ghaderian et al. 2007a, b; Reeves and Adigüzel 2008). The A. lesbiacum value is from the present study 382 of Alyssum species to accumulate high Ca concentrations, even from soils with the low Ca/Mg quotients that are characteristic of serpentines, as previously noted in several studies (e.g. in Reeves et al. 1997; Ghaderian et al. 2007b).The relationship between Ni and Ca uptake in Alyssum species occurring on serpentines has been extensively investigated (e.g. Li et al. 2003). Gabbrielli and Pandolfini (1984) suggested that for the Italian endemic A. bertolonii, internal Ca and Mg concentrations possibly counteract Ni toxicity or enhance Ni tolerance, although the physiological mechanisms involved are still unknown. Measurements of above-ground biomass at the time of peak standing crop, of the time necessary to achieve this biomass (species life cycle), as well as of plant density are required to evaluate the potential of A. lesbiacum for phytoextraction purposes (Raskin and Ensley 2000; Prasad and Freitas 2003). However, optimal agronomic management practices in both the soil and crop system are required to achieve effective phytoextraction by each hyperaccumulating species (Li et al. 2003). For example, the efficiency of Ni phytoextraction may be affected by changes in soil pH or parameters of soil fertility (e.g. Robinson et al. 1997; Li et al. 2003). Agronomic practices have not been applied to the majority of hyperaccumulator species and therefore the maximal amount of plant biomass that can be achieved under favourable conditions (climate, nutrient levels) remain unknown (Reeves 2006). Thlaspi ochroleucum is a rare plant of the Aegean region: the study site near the Mt Olympos summit is the only reported location on Lesbos (Strid and Tan 2002; Bazos and Yannitsaros 2004), while populations have also been mapped on the islands of Thasos and Evvia (Strid and Tan 2002). Although Mt Olympos is included in the list of Natura 2000 sites of Lesbos (GR4110005), no major conservation initiatives have been carried out for the protection of its species and habitats. Thlaspi ochroleucum was recorded as a Ni hyperaccumulator from serpentine sites (Reeves and Brooks 1983). With regard to Zn accumulation, T. ochroleucum was reported with 6,310 mg kg−1 Zn from Uludağ (Bursa province, Turkey) (Reeves and Brooks 1983). High Zn levels (1,740–4,130 mg kg−1) in this species also occur from Zn-rich soils surround- Plant Soil (2010) 332:369–385 ing old mine sites on Thasos (Kelepertsis and Bibou 1991). There has been some ambiguity about the way in which the Zn-accumulating ability of this species should be described (Kelepertsis and Bibou 1991; Shen et al. 1997), as it does not reach the 10,000 mg kg−1 threshold suggested by Baker and Brooks (1989) as defining Zn hyperaccumulation. On present evidence it should be regarded as a hyperaccumulator of Ni and a strong accumulator of Zn, as it certainly shows remarkable accumulation of Zn from Zn-rich soils. The serpentine soils of Olympos on Lesbos do not have high Zn concentrations. However, the Zn concentrations of 223–728 mg kg−1 (mean 434 mg kg−1, n=10) in T. ochroleucum found here are higher than those in any other species in this study. On the serpentine soils of Olympos, hyperaccumulation of Ni by T. ochroleucum is observed, although the levels are not always greater than the 1,000 mg kg−1 threshold, and the mean is only slightly above this value. Since T. ochroleucum in all its occurrences has both lower biomass and lower levels of Ni accumulation than those of A. lesbiacum, its potential for phytoextraction and phytomining is clearly inferior to that of the latter species. Conclusions Our results demonstrate that species tolerant to serpentine soils of Lesbos can be considered as ‘excluders’ maintaining relatively low metal concentrations in their leaves even when soil total metal concentrations are very high. Alyssum lesbiacum and Thlaspi ochroleucum emerge as two strong Ni hyperaccumulators with the former having a high potential for phytoextraction purposes. Acknowledgements We would like to thank S. Koukoulas for the creation of the GIS-based map depicted in Fig. 1, N. Fyllas for the PCA analysis in the R environment and his comments on the manuscript, and A.Y. Troumbis because he gave us the opportunity to collaborate. This paper forms part of the Innovative Actions 2000–2006—North Aegean 2nd Project (BIOBUS: Biodiversity Resources for Innovative Business Development), EU-DG Regional Policy, decision CCI 2005 GR 16 0 PP 005 [E(2005)5523–13/12/2005]. Plant Soil (2010) 332:369–385 383 Appendix I Table 5 Locations for plant leaf material sampling in Lesbos Serpentine sites Vatera Non-serpentine sites Loutra Ampeliko Olympos ✓ ✓ ✓ Family Genus Species Poaceae Aegilops biuncialis Primulaceae Anagallis arvensis Poaceae Bromus commutatus ✓ Asteraceae Crepis commutata ✓ Poaceae Cynosurus echinatus Poaceae Dactylis glomerata ✓ Apiaceae Filago eriocephala ✓ Brassicaceae Hirschfeldia incana Poaceae Hordeum bulbosum Apiaceae Lagoecia cuminoides Poaceae Lolium rigidum Poaceae Phleum pratense Plantaginaceae Plantago lagopus Rosaceae Sanguisorba minor Apiaceae Torilis nodosa Poaceae Trachynia distachya ✓ Fabaceae Trifolium angustifolium ✓ Fabaceae Trifolium campestre Fabaceae Trifolium arvensis Brassicaceae Alyssum lesbiacum Brassicaceae Thlaspi ochroleucum ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ Appendix II Table 66 Results Resultsof of the the two-way two-way ANOVA ANOVA (F-values (F-values and probabilities) and for the on effects of species, soil (serpentine and non-serpentine) substrata Lesbos. The ANOVA was performed without metal probabilities) and their interaction for theoneffects elemental of species, concentrations soil (serpentine of plant collected and from thefor serpentine non-serpentine substrata on Lesbos. The values both A. and lesbiacum and T. οchroleucum. Differences non-serpentine) ANOVA was performed and theirwithout interaction metal onvalues elemental for both concentrations A. lesbiacum and T. οchroleucum. Differences the serpentine and nonbetween the serpentine and between non-serpentine soils are also of plant collected frompresented the serpentine and non-serpentine serpentine soils are also presented Metals Species Soil Species × Soil Differences Ca 88.6*** 74.9*** 7.40*** Non-serp>Serp Non-serp>Serp Fe 12.9*** 9.3** 7.60*** K 18.8*** 0.02ns 2.99*** Mg 25.6*** 114*** 6.40*** Serp>Non-serp Cr 0.6ns 3.98* 0.52ns Serp>Non-serp Cu 10.7*** 61.2*** 3.70*** Non-serp>Serp Mn 27.1*** 26.4*** 4.40*** Non-serp>Serp 384 Plant Soil (2010) 332:369–385 Table 6 (continued) Metals Species Soil Species × Soil Differences Na 22.4*** 2.29ns 4.10*** Ni 9.98*** 95*** 9.70*** P 17.2*** 0.1ns 8.10*** Zn 17.8*** 47.4*** 14.7*** Non-serp>Serp Mg/Ca 15.5*** 156*** 11.4*** Serp>Non-serp Serp>Non-serp ns, not significant; Non-serp: non-serpentine soils; serp: serpentine soils ***P<0.001; **P<0.01; *P<0.05 References Akeroyd JR (1993) Thlaspi. In: Flora Europaea, vol 1, 2nd edn. Cambridge University Press, Cambridge, pp 384–388 Alexander EB, Coleman RG, Keeler-Wolf T, Harrison SP (2007) Serpentine geoecology of Western North America: geology, soils and vegetation. Oxford University Press, Oxford Arianoutsou M, Rundel PW, Berry WL (1993) Serpentine endemics as biological indicators of soil elemental concentration. In: Markert B (ed) Plants as biomonitors. VCH, Weinheim, pp 179–189 Ater M, Lefèbvre C, Gruber W, Meerts P (2000) A phytogeochemical survey of the flora of ultramafic and adjacent normal soils in North Morocco. Plant Soil 218:127–135 Baker AJM (1981) Accumulators and excluders—strategies in the response of plants to heavy metals. J Plant Nutr 3:643– 654 Baker AJM, Brooks RR (1989) Terrestrial higher plants which hyperaccumulate metallic elements—a review of their distribution, ecology and phytochemistry. Biorecovery 1:81–126 Baker AJM, Reeves RD, Hajar ASM (1994) Heavy metal accumulation and tolerance in British populations of the metallophyte Thlaspi caerulescens J. and C. Presl (Brassicaceae). New Phytologist 127:61-68 Bazos I, Yannitsaros A (2004) Floristic reports from the island of Lesvos (Greece) I. Dicotyledones: Aceraceae to Guttiferae. Edinb J Bot 61:49–86 Brady KU, Kruckeberg AR, Bradshaw HD (2005) Evolutionary ecology of plant adaptation to serpentine soils. Annu Rev Ecol Evol Systemat 36:243–266 Brooks RR (1987) Serpentine and its vegetation: a multidisciplinary approach. Dioscorides, Oregon Brooks RR (1998) Biogeochemistry and hyperaccumulators. In: Brooks RR (ed) Plants that hyperaccumulate heavy metals. CAB International, Wallingford, pp 95–118 Brooks RR, Radford CC (1978) Nickel accumulation by European species of the genus Alyssum. Proc Royal Soc Lond B 200:217–224 Brooks RR, Yang XH (1984) Elemental levels and relationships in the endemic serpentine flora of the central Dyke, Zimbabwe and their significance as controlling factors for the flora. Taxon 33:392–399 Brooks RR, Morrison RS, Reeves RD, Dudley TR, Akman Y (1979) Hyperaccumulation of nickel by Alyssum Linnaeus (Cruciferae). Proc Royal Soc Lond B 203:387–403 Dimitrakopoulos PG, Schmid B (2004) Biodiversity effects increase linearly with biotope space. Ecol Lett 7:574– 583 Freitas H, Mooney H (1996) Effects of water stress and soil texture on the performance of two Bromus hordeaceus ecotypes from sandstone and serpentine soils. Acta Oecol 17:307–317 Freitas H, Prasad MNV, Pratas J (2004) Analysis of serpentinophytes from north-east of Portugal for trace metal accumulation—relevance to the management of mine environment. Chemosphere 54:1625–1642 Gabbrielli R, Pandolfini T (1984) Effect of Mg+2 and Ca+2 on the response to nickel toxicity in a serpentine and nickel accumulating species. Physiol Plant 62:540–544 Garnier E, Cortez J, Billès G, Navas M-L, Roumet C, Debussche M, Laurent G, Blanchard A, Aubry D, Bellmann A, Neill C, Toussaint JP (2004) Plant functional markers capture ecosystem properties during secondary succession. Ecology 85:2630–2637 Ghaderian SM, Mohtadi A, Rahiminejad MR, Baker AJM (2007a) Nickel and other metal uptake and accumulation by species of Alyssum (Brassicaceae) from the ultramafics of Iran. Environ Pollut 145:293–298 Ghaderian SM, Mohtadi A, Rahiminejad R, Reeves RD, Baker AJM (2007b) Hyperaccumulation of nickel by two Alyssum species from the serpentine soils of Iran. Plant Soil 293:91–97 Hedge IC (1965) Thlaspi L. In: Davis PH (ed) Flora of Turkey and the East Aegean Islands, vol 1. Edinburgh University Press, Edinburgh, pp 333–341 Ingle RA, Mugford ST, Rees JD, Campbell MM, Smith JAC (2005) Constitutively high expression of the histidine biosynthetic pathway contributes to nickel tolerance in hyperaccumulator plants. Plant Cell 17:2089–2106 Kazakou E, Dimitrakopoulos PG, Baker AJM, Reeves RD, Troumbis AY (2008) Hypotheses, mechanisms and trade-offs of tolerance and adaptation to serpentine soils: from species to ecosystem level. Biol Rev 83:495–508 Kelepertsis AE, Bibou A (1991) Heavy metal contamination of soils at old mining sites on Thasos Island, Greece. Environ Geochem Health 13:23–28 Plant Soil (2010) 332:369–385 Krämer U, Cotter-Howells JD, Charnock JM, Baker AJM, Smith JAC (1996) Free histidine as a metal chelator in plants that accumulate nickel. Nature 379:635–638 Krämer U, Grime GW, Smith JAC, Hawes CR, Baker AJM (1997) Micro-PIXE as a technique for studying nickel localization in leaves of the hyperaccumulator plant Alyssum lesbiacum. Nucl Instrum Methods Phys Res 130:346–350 Kruckeberg AR (1984) California serpentines: flora, vegetation, geology, soils and management problems. University of California Press, Berkeley Küpper H, Lombi E, Zhao F-G, Wieshammer G, McGrath SP (2001) Cellular compartmentation of nickel in the hyperaccumulators Alyssum lesbiacum, Alyssum bertolonii and Thlaspi goesingense. J Exp Bot 52:2291–2300 Lee J, Brooks RR, Reeves RD, Boswell CR, Jaffré T (1977) Plant-soil relationships in a New Caledonian serpentine flora. Plant Soil 46:675–680 Li Y-M, Chaney RL, Brewer EP, Angle JS, Nelkin J (2003) Phytoextraction of nickel and cobalt by hyperaccumulator Alyssum species grown on nickel-contaminated soils. Environ Sci Technol 37:1463–1468 Lombini A, Dinelli E, Ferrari C, Simoni A (1998) Plant-soil relationships in the serpentinite screes of Mt Prinzera (Northern Apennines, Italy). J Geochem Explor 64:19–33 Oze C, Skinner C, Schroth AW, Coleman RG (2008) Growing up green on serpentine soils: biogeochemistry of serpentine vegetation in the Central Coast Range of California. Appl Geochem 23:3391–3403 Pollard AJ, Powell KD, Harper FA, Smith AC (2002) The genetic basis of metal hyperaccumulation in plants. Crit Rev Plant Sci 21:539–566 Prasad MNV, Freitas H (2003) Metal hyperaccumulation in plants—biodiversity prospecting for phytoremediation technology. Electronic Journal of Biotechnology 6 (Available online at: http://ejbiotechnology.ucv.cl/content/vol6/issue3/ full/6/index.html) Proctor J, Woodell SRJ (1975) The ecology of serpentine soils. Adv Ecol Res 9:255–365 Proctor J, Alexeeva-Popova NV, Kravkina IM, Yurtsev BA, Drozdova IV, Kataeva MN (2004) Arctic ultramafics: new investigations on Polar Ural vegetation. In: Boyd RS, Baker AJM, Proctor J (eds) Ultramafic rocks: their soils, vegetation and fauna. Science Reviews, Herts, pp 121–135 Raskin I, Ensley BD (eds) (2000) Phytoremediation of toxic metals: using plants to clean up the environment. Wiley, New York R Development Core Team (2008) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, AU. Available at: http:// www.R-project.org Reeves RD (1992) The hyperaccumulation of nickel by serpentine plants. In: Baker AJM, Proctor J, Reeves RD (eds) The vegetation of ultramafic (serpentine) soils. Intercept, Andover, pp 253–277 Reeves RD (2006) Hyperaccumulation of trace elements by plants. In: Morel JL, Echevarria G, Goncharova N (eds) Phytoremediation of metal-contaminated soils. NATO 385 Science Series: IV: earth and environmental sciences, vol 68. Springer, New York, pp 25–52 Reeves RD, Adigüzel N (2008) The nickel hyperaccumulating plants of the serpentines of Turkey and adjacent areas: a review with new data. Turk J Biol 32:143–153 Reeves RD, Baker AJM (1984) Studies on metal uptake by plants from serpentine and non-serpentine populations of Thlaspi goesingense Halácsy (Cruciferae). New Phytol 98:191–204 Reeves RD, Baker AJM (2000) Metal-accumulating plants. In: Raskin I, Ensley BD (eds) Phytoremediation of toxic metals. Wiley, New York, pp 193–229 Reeves RD, Brooks RR (1983) European species of Thlaspi L. (Cruciferae) as indicators of nickel and zinc. J Geochem Explor 18:275–283 Reeves RD, Baker AJM, Kelepertsis A (1997) The distribution and biogeochemistry of some serpentine plants of Greece. In: Jaffré T, Reeves RD, Becquer T (eds) Ecologie des milieux sur roches ultramafiques et sur sols metallifères. ORSTOM, Nouméa, pp 205–207 Reeves RD, Baker AJM, Borhidi A, Berazaín R (1999) Nickel hyperaccumulation in the serpentine flora of Cuba. Ann Bot 83:29–38 Reeves RD, Schwartz C, Morel JL, Edmondson J (2001) Distribution and metal accumulating behaviour of Thlaspi caerulescens and associated metallophytes in France. Int J Phytoremediat 3:145–172 Reeves RD, Baker AJM, Becquer T, Echevarria G, Miranda ZJG (2007) The flora and biogeochemistry of the ultramafic soils of Goiás state, Brazil. Plant Soil 293:107–119 Robinson BH, Brooks RR, Howes AW, Kirkman JH, Gregg PEH (1997) The nickel hyperaccumulator plant Alyssum bertolonii as a potential agent for phytoremediation and phytomining of nickel. J Geochem Explor 59:75–86 Shallari S, Schwartz C, Hasko A, Morel JL (1998) Heavy metals in soils and plants of serpentine and industrial sites of Albania. Sci Total Environ 209:133–142 Shen ZG, Zhao FJ, McGrath SP (1997) Uptake and transport of zinc in the hyperaccumulator Thlaspi caerulescens and the non-hyperaccumulator Thlaspi ochroleucum. Plant Cell Environ 20:898–906 Shewry PR, Peterson PJ (1976) Distribution of chromium and nickel in plants and soil from serpentine and other sites. J Ecol 64:195–212 Smart KE, Kilburn MR, Salter CJ, Smith JAC, Grovenor CRM (2007) NanoSIMS and EPMA analysis of nickel localisation in leaves of the hyperaccumulator plant Alyssum lesbiacum. Int J Mass Spectrom 260:107–114 Strid A, Tan K (2002) Flora Hellenica, vol 2. A.R.G. Gantner Verlag K.G, Ruggell Wallace AR (1858) On the tendency of varieties to depart indefinitely from the original type. Journal of the Proceedings of the Linnean Society (Zoology) 3:53–62 Walker RB (1954) The ecology of serpentine soils: II. Factors affecting plant growth on serpentine soils. Ecology 35:259–266 Wright JW, Stanton ML, Scherson R (2006) Local adaptation to serpentine and non-serpentine soils in Collinsia sparsiflora. Evol Ecol Res 8:1–21