Abstract
Background and Aims Thiophores, which are typically desert gypsophytes, accumulate high (2–6 % S dry weight) sulphur concentrations and may possess unique tolerance to environmental stress factors, e.g. sulphate/metal toxicity, drought and salinity. Little is known of the prevalence of the behaviour or the associated physiological aspects. The aim of this study was to (a) determine the prevalence of thiophore behaviour in a group of Australian xerophytes; (b) identify elemental uptake/storage characteristics of these thiophores; and (c) determine whether the behaviour is constitutive or environmental.
Methods The elemental composition of soils and the foliage of 11 species (seven genera) at a site in the Tanami Desert (NT, Australia) was compared and 13 additional Acacia species from other locations were examined for elevated calcium and sulphur concentrations and calcium–sulphur mineralization, thought to be particular to thiophores.
Key Results Acacia bivenosa DC. and 11 closely related species were identified as thiophores that can accumulate high levels of sulphur (up to 3·2 %) and calcium (up to 6.8 %), but no thiophores were identified in other genera occupying the same habitat. This behaviour was observed in several populations from diverse habitats, from samples collected over three decades. It was also observed that these thiophores featured gypsum (CaSO4·2H2O) crystal druses that completely filled cells and vascular systems in their dried phyllode tissues.
Conclusions The thiophores studied exhibit a tight coupling between sulphur and calcium uptake and storage, and apparently store these elements as inorganic salts within the cells of their foliage. Thiophore behaviour is a constitutive trait shared by closely related Acacia but is not highly prevalent within, nor exclusive to, xerophytes. Several of the newly identified thiophores occupy coastal or riparian habitats, suggesting that the evolutionary and ecophysiological explanations for this trait do not lie solely in adaptation to arid conditions or gypsiferous soils.
Keywords: Biomineralization, hyperaccumulation, thiophore, xerophyte, sulphur regulation, calcium regulation, Acacia bivenosa, Acacia robeorum, sulphur accumulation, ionic storage, phytoaccumulation, geobotany.
INTRODUCTION
Sulphur is an essential nutrient involved in protein synthesis, enzymatic function and redox control (Abrol and Ahmad, 2003), and terrestrial plants generally accumulate foliar concentrations of this element in the range 0·3–0·5 % S dry weight (d. wt) to support healthy growth (Abrol and Ahmad, 2003; Hawkesford et al., 2012). However, certain ‘sulphur-loving’ plants (thiophores; Duvigneaud and De Smet, 1968) are capable of accumulating comparably excessive (over an order of magnitude higher) sulphur concentrations in their tissues, a behaviour with parallels non-essential element hyperaccumulation in metallophytes (Brooks, 1998).
The thiophores identified to date, which accumulated foliar concentrations frequently in the range of 2–6 % S d. wt, are predominantly obligate gypsophytes occupying semi-arid or desert habitats (Duvigneaud and De Smet, 1968; Al-Ani et al., 1971; Ruiz et al., 2003). In these gypsophytes, thiophore behaviour is considered an adaptation to phytotoxic soil sulphate concentrations, affording these species an ecological niche and the opportunity to avoid competition (Ruiz et al., 2003; Palacio et al., 2007). However, He et al. (2012a) recently identified the desert shrub, Acacia robeorum Maslin, as a thiophore. Acacia robeorum grows on skeletal sand or sandy loam over granite, laterite or quartz (Maslin and van Leeuwen, 2008), indicating that this sulphur-accumulative behaviour may not be unique to gypsophytes. He et al. (2012b) proposed the numerous calcium- and sulphur-rich biominerals (probably calcium sulphates) observed in this species as a storage-tolerance mechanism for excess calcium and sulphate. Other proposed tolerance mechanisms for toxic cytosolic sulphur concentrations include vacuolar sulphate compartmentalization (Hawkesford and De Kok, 2007), enhanced protein assimilation (Ruiz et al., 2003) and salt gland excretion (Storey and Thomson, 1994). Sulphur uptake and metabolism are vital for the control of reactive oxygen species (ROS), and have been associated with protection against metal phytotoxicity (Rausch and Wachter, 2005; Na and Salt, 2011), drought and high salinity (Khan et al., 2008). Consequently, thiophores may possess unique tolerance to a range of environmental stress factors, and studies investigating these species are required to gain new insights into the evolutionary, ecological and functional aspects of plant stress tolerance.
Our knowledge of the prevalence of thiophore behaviour between and within genera is limited, as is our understanding of the related physiological and biochemical aspects. For this study, we hypothesized that thiophore behaviour is a constitutive trait occurring in desert-dwelling (xerophyte) species of several genera, and that characteristic mineralization would feature in these species after drying. To test this hypothesis, we compared the elemental composition of soils and foliage of dominant tree and shrub species sharing those soils at a site in the Tanami Desert (Northern Territory, Australia) that was the focus of exploratory biogeochemical surveys (Reid et al., 2009; Reid and Hill, 2010), as well as the foliage of species from the Stuart Shelf (South Australia) and the Great Sandy Desert (Western Australia). In addition, we compared samples of 11 species that are closely related to Acacia bivenosa DC. (the bivenosa ‘group’ – see the Materials and Methods), one of the species sampled during the biogeochemical surveys.
MATERIALS AND METHODS
Comparing soils and foliage of genera sharing the same growth substrate and conditions: plant materials and soil sampling
Fourteen tree and shrub species (eight genera) were sampled during surveys at three sites situated across Australia (Fig. 1): the Tanami Desert (20°17′S, 129°59′E), the Stuart Shelf (30°26′S 136°52′E) and the western edge of the Great Sandy Desert (22°20′S 122° 4′E). In the Tanami Desert (site 1), samples of nine species were collected on a 2 × 3 km grid with sampling locations spaced approx. 500 × 250 m apart. Regional sampling was conducted on the Stuart Shelf (site 2), where three species were sampled along four transects (trending NSEW) from sites (approx. 500 m2 quadrants) at 1, 3, 5, 10, 25, 50 and 100 km from the transect epicentre. In the Great Sandy Desert (site 3), samples of the two dominant tree/shrub species were collected along two bisecting 3 km transects with collections every 150 m. Table 1 provides details of the species and number of samples collected from each site (total of 512 samples).
Fig. 1.
Spatial distributions of the most dominant A. bivenosa group members, adapted from Moore (2005). Vegetation survey locations are indicated as site 1 (Tanami Desert), site 2 (Stuart Shelf) and site 3 (Great Sandy Desert).
Table 1.
Species name and authority, collection site location and number of individuals samples collected for elemental analysis (total of 512 samples), growth form and leaf form (Moore 2005)
| Species | Collection site | n | Growth form | Leaf form |
|---|---|---|---|---|
| Acacia ancistrocarpa Maiden & Blakely | 3 | 24 | Shrub (3 m) | Linear |
| A. aneura F. Muell. | 2 | 53 | Variable shrub-tree (1–14 m) | Terete to linear |
| A. bivenosa DC. | 1 | 56 | Shrub (2 m) | Elliptic |
| A. bivenosa DC. × sclerosperma F.Muell. subsp. sclerosperma (Maslin et al., 2010). | 3 | 10 | Shrub (4 m) | Linear to terete |
| A. coriacea subsp. sericophylla F.Muell. | 1 | 32 | Tree (3–7 m) | Linear |
| A. ligulata A.Cunn. ex Benth. | 2 | 48 | Shrub (2 m) | Linear to oblong |
| Atriplex vesicaria Heward ex Benth. | 2 | 43 | Shrub (<1 m) | Elliptic |
| Corymbia opaca (D.J.Carr & S.G.M.Carr) K.D.Hill & L.A.S.Johnson | 1 | 8 | Tree (15 m) | Lanceolate |
| Eucalyptus pachyphylla F.Muell. | 1 | 11 | Mallee (4 m) | Lanceolate |
| Grevillea striata R.Br. | 1 | 12 | Tree (3–15 m) | Linear |
| Hakea macrocarpa R.Br. | 1 | 40 | Tree (5 m) | Terete |
| Melaleuca lasiandra F.Muell. | 1 | 79 | Shrub (2–4 m) | Elliptic |
| M. glomerata F.Muell. | 1 | 17 | Shrub (1–2 m) | Linear |
| Triodia pungens R.Br. | 1 | 79 | Hummock (1 m) | Terete |
Collection sites were (1) Tanami Desert, (2) Stuart Shelf and (3) Great Sandy Desert.
Leaves and phyllodes (foliage of Acacia subgenus, Phyllodineae) were collected by hand while wearing powder-free latex gloves, replaced between samples to minimize cross-contamination, and stored in paper bags to minimize sample degradation. Care was taken to collect mature foliage with minimal signs of disease and surface dust contamination. Additional specimens of the A. bivenosa group and other related Acacia were obtained from herbaria and various field locations within Australia (Table 2).
Table 2.
Species name, sampling location and source of additional Acacia specimens collected for SEM analyses
| Species | Sample location | Source |
|---|---|---|
| A. ampliceps Maslin | 22°4'32''S, 119°40'29''E | Western Australian Herbarium |
| A. bivenosa DC. | 23°21'5''S, 119°43'48''E | Western Australian Herbarium |
| A. blakelyi Maiden* | 28°10'58''S, 114°49'22''E | Western Australian Herbarium |
| A. cupularis Domin | 35°36'S, 113°6'E | Cape Jervis, South Australia |
| A. didyma A.R.Chapman & Maslin | 26°38'18''S, 113°37'55''E | Western Australian Herbarium |
| A. ligulata A. Cunn. ex Benth. | 34°58'3''S, 138°37'47''E | Waite Arboretum, South Australia |
| 26°34'S, 113°36'E | Useless Loop, Western Australia | |
| A. myrtifolia Willd* | 34°22'19''S, 117°41'53''E | Western Australian Herbarium |
| A. robeorum Maslin | 20°51'51''S, 120°42'8''E | Western Australian Herbarium |
| A. rostellifera Benth. | 26°34'S, 113°36'E | Useless Loop, Western Australia |
| A. salicina Lindl. | 34°48'S, 138°39'E | Para Hills, South Australia |
| A. saligna H.L. Wendl.* | 29°42'30''S, 115°12'15''E | Western Australian Herbarium |
| A. sclerosperma F.Muell. | 20°43'23''S, 116°53'33''E | Western Australian Herbarium |
| A. startii A.R.Chapman & Maslin | 23°17'35''S, 113°54'43''E | Western Australian Herbarium |
| A. telmica A.R. Chapman & Maslin | 29°16'28''S, 115°13'43''E | Western Australian Herbarium |
| A. tysonii Luehm. | 29°15'40''S, 116°1'13''E | Western Australian Herbarium |
| A. xanthina Benth. | 31°56'S, 115°46'E | Bold Park, Western Australia |
Related Acacia species classified outside the A. bivenosa group are marked with an asterisk.
At site 1 (Tanami Desert), soils were sampled in tandem with the vegetation survey, both using the same sampling grid. The soils were taken from 2–10 cm below the surface using a plastic shovel and stored in resealable plastic bags for transport to the laboratory. At one location from this site, a soil profile was collected (sampled every 0·1 m from 0·0 to 1·1 m below the surface). The <75 μm soil fraction was prepared for elemental analysis by sieving oven-dried (48 h at 45 °C) soils using plastic sieves with nylon mesh, followed by homogenization and splitting (5 g sub-sample). The foliage samples were oven-dried and then finely ground using a rotating blade and stainless steel mill, and prepared as described by Hulme and Hill (2003).
Elemental analyses of soils and foliage
Elemental analyses of the ground foliage and sieved soil were performed using inductively coupled plasma mass spectrometry (Elan 6000, Perkin Elmer) and atomic emission spectrometry (CIROS Radial, Spectro Analytical Instruments) after either sequential digestion in heated, concentrated nitric acid, followed by a concentrated hydrochloric–nitric acid mixture (1:1:1 HNO3:HCl:H2O v/v/v), or using a concentrated nitric acid–hydrogen peroxide mixture (4:1 HNO3:H2O2 v/v/v). Analytical accuracy and reproducibility were evaluated using quality control standards, analysed with each analytical batch, and duplicate splits taken from every 30th sample. The concentrations determined for quality control standards fell within the uncertainty of the expected values, and intrasample uncertainty was generally within the uncertainty of the analytical techniques (≤10 %).
Data processing
Statistical tests and data processing were performed with Sigmaplot 11 (Systat Software), IoGAS 5 (Reflex Instruments Asia Pacific) and JMP 12 (SAS Institute) software packages. Elemental data (foliar concentrations and ratios) for each species surveyed were compared for significant differences using the non-parametric Kruskal–Wallis analysis of variance (ANOVA) on ranks (Kruskal and Wallis, 1952) and pairwise Dwass–Steel–Chritchlow–Fligner tests, which are appropriate for data sets with variable population size, variance and frequency distribution (Hollander and Wolf, 1999). Phylogenetic mapping was performed using FigTree v1.4.2 (http://tree.bio.ed.ac.uk/software/figtree).
Comparing species in the Acacia bivenosa group
The phyllodes from species of the A. bivenosa group and several other Acacia (Table 2) were examined for a rare calcium–sulphur mineralization and elevated calcium and sulphur content, which was observed in the thiophore, A. robeorum (He et al., 2012b). At this stage, this was only performed on single samples of each species from herbarium specimens to determine if similar concentrations and mineralization were observed.
Species description
The A. bivenosa ‘group’ (Chapman and Maslin, 1992), which consists of 12 phylogenetically related Acacia (Leguminosae: Mimosoideae) species (Table 2; Fig. 8), grow across Australia in many contrasting terrains and habitats (e.g. desert sand dunes and plains, rocky slopes, riparian areas and coastal dunes). The most widely distributed are A. bivenosa, A. ligulata and A. salicina, which cover large regions of Australia (Fig. 1), and are briefly described here.
Fig. 8.
Phylogenetic relationship between the Acacia species included in this study. Species identified as thiophores are highlighted, and those groups/clades not including species of interest are shown collapsed (without taxa). Modified after Bui et al. (2014). Tree data are available from Treebase: http://purl.org/phylo/treebase/phylows/study/TB2:S13659.
Acacia bivenosa (two-nerved wattle) is a dense, glabrous shrub, 1–2 m in height/diameter that grows on a variety of soils but mostly sandy soils and in areas of seasonal surface water accumulation (Moore, 2005). This species has a series of shallow lateral roots and a single, thicker sinker root to depth (<2 m at site 1). Acacia bivenosa is common across much of northern Australia (Fig. 1). Acacia ligulata (umbrella bush) is common across much of Australia, and merges with A. bivenosa to the north and A. cupularis along the southern coast (Fig. 1), where it typically grows on the lower slopes of red sand dunes. Acacia ligulata is a variably bushy shrub to small tree, and is similar in form to A. bivenosa, and these species are able to hybridize. Acacia salicina (cooba) is common across eastern Australia; in particular the Murray–Darling basin where it grows along water courses (creek and river banks, besides dams and ephemeral wetland areas). Acacia salicina grows as a shrub to large tree, 3–10 m high, with long weeping branches, and commonly forms small colonies by root suckering.
Scanning electron microscopy, energy dispersive X-ray spectroscopy and X-ray diffraction analyses of phyllodes
Phyllodes of selected species (Table 2) were resin embedded (Epotek 301-2FL, Epoxy Technologies, MA, USA) and polished (diamond paste, 1 μm nominal particle size) to produce lateral sections, which were examined with a Philips XL-40 scanning electron microscope (SEM) and a Zeiss Ultra Plus field emission gun (FEG) SEM. Both are located at the Australian Resources Research Centre, Perth, Australia, and were fitted with Bruker XFlash 6 energy-dispersive X-ray spectrometers (EDS). Standard analytical conditions for the SEM were an accelerating voltage of 30 kV, a beam current between 1 and 3 nA, a pressure of 0·5 mBar, a 10 mm working distance and 30 s live count time. The FEGSEM has an accelerating voltage of 10 kV, a beam current between 0·3 and 2 nA and a working distance of 7 mm. Standardless quantification was performed with ZAF matrix correction and processed using the Bruker Esprit 1·9 EDS software package. For X-ray diffraction (XRD) analysis, phyllodes were mounted onto glass plates using double-sided adhesive tape and analysed directly with a Bruker D4 Endeavor, fitted with a Co tube, Fe filter and a Lynxeye position-sensitive detector. The measured 2θ range was 5–90 °, with a step size of 0·02 ° and a divergence slit of 1 °. Diffraction spectra were processed and compared with mineralogical databases using the Bruker DIFFRAC.SUITE EVA software package.
RESULTS
Elemental anomalism in the foliage of Acacia bivenosa, A. bivenosa × sclerosperma subsp. sclerosperma and A. ligulata
A comparison of the foliar elemental composition of the 14 species collected for this study (Table 3) revealed that the closely related species, A. bivenosa, A. ligulata and, to a lesser extent, A. bivenosa × sclerosperma subsp. sclerosperma, were elevated (much greater than the median) with respect to calcium and sulphur (Fig. 2A, B). Strontium concentrations, which were obtained for samples collected from site 1, were also high in A. bivenosa (Fig. 2C).
Table 3.
Summary of foliar nutrient concentrations for each species collected during this study
| Species | Ca (%) | K (%) | Mg (%) | P (mg kg–1) | S (%) |
|---|---|---|---|---|---|
| Acacia ampliceps* | 4·47 | 0·744 | 0·741 | 760 | 2·15 |
| Acacia ancistrocarpa | 0·79 6 ± 0·15 | 0·747 ± 0·13 | 0·158 ± 0·02 | 603 ± 197 | 0·134 ± 0·03 |
| A. bivenosa × sclerosperma subsp· sclerosperma | 3·87 ± 0·67 | 0·968 ± 0·27 | 0·343 ± 0·05 | 772 ± 134 | 1·17 ± 0·34 |
| A. blakelyi* | 1·30 | 0·94 | 0·260 | 190 | 0·322 |
| A. aneura | 1·66 ± 0·48 | 0·926 ± 0·23 | 0·161 ± 0·03 | 977 ± 267 | 0·143 ± 0·01 |
| A. bivenosa | 5·14 ± 0·83 | 1·49 ± 0·39 | 0·507 ± 0·15 | 591 ± 168 | 2·57 ± 0·40 |
| A. coriacea | 2·57 ± 0·85 | 1·21 ± 0·38 | 0·387 ± 0·10 | 608 ± 197 | 0·159 ± 0·09 |
| A. didyma* | 5·49 | 2·11 | 0·533 | 417 | 4·04 |
| A. cupularis* | 3·43 | 0·485 | 0·200 | 550 | 2·21 |
| A. ligulata | 5·28 ± 0·63 | 1·11 ± 0·27 | 0·460 ± 0·08 | 621 ± 114 | 2·69 ± 0·38 |
| A. myrtifolia* | 1·02 | 0·442 | 0·067 | 121 | 0·150 |
| A. robeorum* | 8·91 | 0·297 | 0·709 | 625 | 5·69 |
| A. rostellifera* | 4·05 | 0·745 | 0·414 | 988 | 0·287 |
| A. salicina* | 4·12 | 1·20 | 0·262 | 1090 | 1·84 |
| A. saligna* | 1·70 | 0·600 | 0·624 | 523 | 0·484 |
| A. slcerosperma* | 4·68 | 0·636 | 0·402 | 589 | 1·75 |
| A. startii* | 5·33 | 0·936 | 0·444 | 546 | 3·60 |
| A. telmica* | 3·27 | 0·160 | 0·244 | 417 | 0·780 |
| A. tysonii* | 2·05 | 2·14 | 0·613 | 672 | 1·95 |
| A. xanthina* | 4·51 | 0·515 | 0·312 | 1160 | 0·940 |
| Atriplex vesicaria | 1·07 ± 0·25 | 1·99 ± 0·39 | 0·451 ± 0·08 | 702 ± 187 | 0·334 ± 0·04 |
| Corymbia opaca | 1·21 ± 0·46 | 1·01 ± 0·32 | 0·380 ± 0·12 | 531 ± 51 | 0·118 ± 0·02 |
| Eucalyptus pachyphylla | 1·10 ± 0·15 | 0·845 ± 0·15 | 0·194 ± 0·06 | 435 ± 63 | 0·114 ± 0·02 |
| Grevillea striata | 0·332 ± 0·12 | 1·03 ± 0·20 | 0·127 ± 0·03 | 516 ± 141 | 0·313 ± 0·09 |
| Hakea macrocarpa | 0·952 ± 0·34 | 1·25 ± 0·22 | 0·196 ± 0·04 | 455 ± 91 | 0·311 ± 0·12 |
| Melaleuca glomerata | 0·822 ± 0·18 | 0·454 ± 0·09 | 0·422 ± 0·09 | 505 ± 92 | 0·507 ± 0·10 |
| M· lasiandra | 1·05 ± 0·23 | 0·799 ± 0·17 | 0·371 ± 0·08 | 469 ± 82 | 0·282 ± 0·07 |
| Triodia pungens | 0·305 ± 0·07 | 0·871 ± 0·30 | 0·120 ± 0·03 | 555 ± 217 | 0·122 ± 0·02 |
Uncertainties are given as standard deviations.
Species samples collected from herbaria for SEM analyses are marked with an asterisk.
Fig. 2.
(A–C) Foliar sulphur, calcium and strontium concentrations of 14 species collected from three sites across Australia. (D and E) Scatter plots illustrating correlation between foliar calcium, sulphur and strontium. (D) Data for two species groups: group 1 consists of 11 species expressing positive linear correlation (Ca vs. S) and group 2 (A. coriacea, A. aneura and Corymbia opaca) showed independence between these elements.
The calcium and sulphur concentration in phyllodes (Table 3) of A. bivenosa (5·1 ± 0·8 % d. wt Ca, 2·6 ± 0·4 % d. wt S), A. ligulata (5·3 ± 0·6 % d. wt Ca, 2·7 ± 0·4 % d. wt S) and a third hybrid of two species grouped with A. bivenosa, A. bivenosa × sclerosperma subsp. sclerosperma (3·9 ± 0·7 % d. wt Ca, 1·2 ± 0·34 % d. wt S) were significantly (Table 3) elevated compared with all other species collected for this study (1·0 ± 0·7 % d. wt Ca, 0·22 ± 0·12 % d. wt S). Together, A. bivenosa and A. ligulata comprised the 90th percentile for calcium and sulphur across the whole data set. Their mean foliar concentrations were five times (2–17) that of the other species for calcium, and ten times (5–24) higher for sulphur (excluding A. bivenosa × sclerosperma subsp. sclerosperma).
An examination of data for all species revealed correlation between foliar calcium, sulphur and strontium (Fig. 2D, E). Two groups of species were identified, with respect to correlation between sulphur and calcium. The majority (group 1) presented positive linear correlation, whereas A. coriacea, A. aneura and Corymbia opaca (group 2) showed independence between these elements (Fig. 2D). The tissue concentrations of strontium, a biogeochemical analogue of calcium, found in A. bivenosa (587 ± 290 mg kg–1 d. wt) were also notably elevated compared with the other eight species collected at site 1 (45 ± 51 mg kg–1 Sr), although the differences were only statistically significant (Table 4) when compared with five of the other species, owing to the high variance in those data (Fig. 2C). An exponential-type relationship was observed between calcium and strontium values across all species (Fig. 2E).
Table 4.
Statistical significance (P-values) of means comparisons (Dwass–Steel–Chritchlow–Fligner method) between the foliar Ca, S and Sr concentrations of A. bivenosa, A. ligulata and A. bivenosa × sclerosperma subsp. sclerosperma, and each of the other species collected for this study
| Species | Site | n |
A. bivenosa |
A. ligulata |
A. bivenosa × slcerosperma |
||||
|---|---|---|---|---|---|---|---|---|---|
| Ca | S | Sr | Ca | S | Ca | S | |||
| A. bivenosa | 1 | 56 | 1·00 | 1·00 | <0·01 | <0·01 | |||
| A. coriacea subsp. sericophylla | 1 | 32 | <0·05 | <0·01 | 0·416 | <0·01 | <0·01 | <0·05 | <0·01 |
| Corymbia opaca | 1 | 8 | <0·01 | <0·01 | 0·766 | <0·01 | <0·01 | <0·05 | <0·05 |
| Eucalyptus pachyphylla | 1 | 11 | <0·01 | <0·01 | <0·05 | <0·01 | <0·01 | <0·05 | <0·01 |
| Grevillea striata | 1 | 12 | <0·01 | <0·01 | 1·00 | <0·01 | <0·01 | <0·05 | <0·01 |
| Hakea macrocarpa | 1 | 40 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 |
| Melaleuca lasiandra | 1 | 79 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 |
| M. glomerata | 1 | 17 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 |
| Triodia pungens | 1 | 79 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 |
| A. aneura | 2 | 53 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | |
| A. ligulata | 2 | 48 | |||||||
| Atriplex vesicaria | 2 | 43 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | |
| Acacia ancistrocarpa | 3 | 24 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·05 | |
| A. bivenosa × sclerosperma | 3 | 10 | <0·01 | <0·01 | |||||
Ratios for foliar concentrations of essential macronutrients (calcium, magnesium, potassium and phosphorus) were calculated, relative to sulphur, for each species in the survey. The ratios demonstrate that in almost every case, those species observed with elevated foliar sulphur and calcium (A. bivenosa, A. ligulata and A. bivenosa × sclerosperma subsp. sclerosperma) also showed a significant (Table 5) bias for sulphur uptake relative to other macronutrients (Fig. 3B–D). In contrast, the S:Ca ratios for these three species (0·304–0·511) fell within the range of all those collected (0·085–1·00: Fig. 3A).
Table 5.
Statistical significance (P-values) of means comparisons (Dwass–Steel–Chritchlow–Fligner method) between the foliar concentration ratios (S:K, S:Mg and S:P) of A. bivenosa, A. ligulata and A. bivenosa × sclerosperma subsp. sclerosperma, and each of the other species collected for this study.
| Species | Site | n |
A. bivenosa |
A. ligulata |
A. bivenosa×sclerosperma |
||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| S:K | S:Mg | S:P | S:K | S:Mg | S:P | S:K | S:Mg | S:P | |||
| A. bivenosa | 1 | 56 | <0·01 | 0·127 | 1·00 | <0·05 | <0·05 | <0·01 | |||
| A. coriacea subsp· sericophylla | 1 | 32 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 |
| Corymbia opaca | 1 | 8 | <0·01 | <0·01 | <0·01 | <0·05 | <0·01 | <0·01 | <0·05 | <0·05 | <0·05 |
| Eucalyptus pachyphylla | 1 | 11 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 |
| Grevillea striata | 1 | 12 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | 0·524 | <0·01 |
| Hakea macrocarpa | 1 | 40 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 |
| Melaleuca lasiandra | 1 | 79 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 |
| M· glomerata | 1 | 17 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | 1·00 | <0·01 | <0·05 |
| Triodia pungens | 1 | 79 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 |
| A. aneura | 2 | 53 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 |
| A. ligulata | 2 | 48 | <0·01 | 0·127 | 1·00 | <0·01 | <0·01 | <0·01 | |||
| Atriplex vesicaria | 2 | 43 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 |
| Acacia ancistrocarpa | 3 | 24 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 | <0·01 |
| A. bivenosa × sclerosperma | 3 | 10 | <0·05 | <0·05 | <0·01 | <0·01 | <0·01 | <0·01 | |||
Fig. 3.
Elemental ratios of foliar calcium (A), potassium (B), magnesium (C) and phosphorus (D) relative to sulphur, respectively.
Geochemistry of underlying sub-strata
The soils at site 1 were investigated at depth to establish the geochemical characteristics of the underlying sub-strata and, consequently, the root environment of the species sampled there. The soil at site 1 was red sand with small amounts of clay in depressions. These depressed areas also showed some indicative gypsum efflorescence, although quantitative XRD analyses were not performed. The soil profile (Fig. 4) comprised a top layer (15 cm) of organic-rich sand, followed by nodular carbonate (15–50 cm) with increasingly fine texture with depth. The calcium concentration, which was low at the surface, was markedly elevated where the carbonate was present. Magnesium increased with depth, reflecting the increasingly dolomitic nature of the carbonate. At 50 cm there was an obvious gypsum layer, corresponding to elevated sulphur concentrations at this depth. Median values for major parameters associated with soils collected corresponding to all vegetation samples at site 1 are: pH 6·3, electrical conductivity 12 μS cm–1, Ca 0·055 %, K 0·53 %, Mg 0·07 %, P 0·01 %, Na 0·025 % and S 0·006 %.
Fig. 4.
Total concentrations of calcium (A), sulphur (B) and strontium (C) throughout a soil profile at site 1 (Tanami Desert). Samples were collected at 0·1 m intervals from 0·0 to 1·1 m below the surface.
Presence of calcium sulphate crystals in dry phyllode tissues
Polished sections of phyllodes from A. bivenosa were examined under SEM using a backscattered electron detector (BSE), which revealed numerous accumulations of relatively heavy elements, compared with the carbon-dominated tissue matrix (Fig. 5). Elemental analyses using EDS determined that these regions consisted of approx. 1:1 CaO to SO2 in most spot samples (Fig. 6A), which is consistent with calcium sulphate. The lack of significant carbon concentrations indicates that the majority of these calcium-rich accumulations were not calcium oxalate [Ca(C2O4)·H2O], which is the typical calcium storage biomineral in vascular plants (Franceschi and Nakata, 2005). However, oxalates were observed separately from the sulphates throughout the tissues and were readily distinguished from the sulphates by their distinctive morphology (Fig. 5F) and elemental composition. X-ray diffraction analysis was performed on air-dried A. bivenosa phyllodes (Mount Whaleback, Western Australia), which were also sectioned and confirmed to contain the same calcium–sulphur accumulations observed in samples from site 1. The analysis yielded a fingerprint pattern consistent with gypsum (CaSO4·2H2O) (Fig. 6B), in addition to peaks for whewellite [calcium oxalate, Ca(C2O4)·H2O] in air-dried samples.
Fig. 5.
Electron micrographs taken from polished phyllode longitudinal sections from Acacia bivenosa (A, B, F), A. ligulata (D, E), A. rostellifera (C) and A. xanthina (G, H); (F) shows calcium oxalate (Ox) and gypsum (Gy) crystals, which appear to lyse cells (Gyl). (G) Cross-section of the junction between the branchlet and petiole of an A. xanthina phyllode, showing branchlet wood (Wb), petiole tissue (Pe) and epidermis (Se) – shown in greater detail in (H).
Fig. 6.
(A) Representative energy-dispersive X-ray spectra of Ca–S accumulations observed in phyllodes of the A. bivenosa group species; (B) X-ray diffraction spectrum of bulk A. bivenosa phyllodes sampled at Mount Whaleback (Pilbara region, Western Australia), showing most peaks related to gypsum (g). The remaining well-resolved peaks identified calcium oxalate (whewellite) and some silica.
Polished phyllode sections from each of the other species in the A. bivenosa group (Table 2) were also examined, and the presence of calcium sulphate accumulations with similar morphology and distribution was confirmed in each species (Figs 5 and 7). The calcium sulphate crystals were of a broadly spherical to spheroidal morphology when confined within individual cells. Their morphology and distribution are consistent with cell vacuoles (Fig. 5C, D) but, in many cases, appear to fill cells completely and even extend beyond cell walls (Figs 5C, F and 7E, G). Sulphate was also observed occupying intercellular spaces (Fig. 5E) and along vascular tissues (Figs 5B, C, G, H and 7H) but were absent from lignified or epidermal tissues (Fig. 5G).
Fig. 7.
Electron micrographs taken from polished phyllode logitudinal sections from (A) Acacia ampliceps, (B) A. cupularis, (C) A. didyma, (D) A. salicina, (E) A. bivenosa × A. sclerosperma subsp. sclerosperma, (F) A. startii, (G) A. telmica and (H) A. tysonii. In each image, the brighter material is all either calcium sulphate or the strongly crystalline calcium oxalate (Ox).
DISCUSSION
This study demonstrates a constitutive trait of excessive calcium and sulphur accumulation in the desert shrub, A. bivenosa, together with the 11 closely related species of the A. bivenosa ‘group’. In contrast to the previously established calcium oxalate storage mechanisms observed in many species, these species develop extremely high calcium and sulphur concentrations within the cells and vascular system of its foliage, sufficient for gypsum to precipitate. This unusual behaviour was not observed in any of the other desert species of eight genera included in this study. These findings add significantly to our knowledge and understanding of the prevalence, constitutive nature and physiological aspects of thiophores, which are of particular interest from the perspective of stress tolerance and desert ecophysiology. We discuss these findings further in the following sub-sections.
Acacia bivenosa accumulates sulphur, calcium and strontium
Acacia bivenosa was clearly elevated in foliar calcium, strontium and sulphur (Fig. 2A–C) compared with the other shrub and tree species growing in the same soils, under the same conditions. Sulphur was the most notably accumulated element in A. bivenosa. While the other species surveyed had foliar sulphur concentrations falling within or below the accepted range of 0·3–0·5 % d. wt (Abrol and Ahmad, 2003; Hawkesford et al., 2012), A. bivenosa exceeded the upper range by a factor of five and the lower range by a factor of > 20. High foliar concentrations of sulphur, calcium and strontium in A. bivenosa could reflect a source of gypsum in the soils at site 1. However, the gypsum enrichment found there was very shallow (≤50 cm) and so it is very unlikely that the other species at this site would not have been exposed to any related elevation in calcium and sulphur at their roots. Consequently, the elevated foliar sulphur and calcium concentrations found in A. bivenosa mark a clear difference in the uptake and storage of these elements in this species, compared with those sharing the habitat.
We interpret the relatively high strontium concentrations found in A. bivenosa as an effect of high calcium uptake. Strontium is recognized as a biogeochemical analogue for calcium (Ichikuni and Musha, 1978; Pors Nielsen, 2004), and investigators have long acknowledged a relationship between their uptake in vascular plants (Williams and David, 1963; Hutchin and Vaughan, 1967). This relationship is reflected by a correlation between calcium and strontium in every species in this study (Fig. 2E), although the slope of the correlation differs for A. bivenosa. Franceschi and Schueren (1986) reported competition between strontium and calcium for incorporation in calcium oxalate biominerals, and we speculate that strontium uptake is related to calcium regulation in A. bivenosa, but further investigation is required to elucidate any such relationship.
The overall macronutrient balance of each species sampled in this study also reflects this trait of excessive sulphur uptake. The ratios of sulphur to magnesium, potassium and phosphorus were notably higher in A. bivenosa, compared with the other species (Fig. 3B–D). In contrast to these macronutrients, the Ca:S ratios were not significantly different (Fig. 3A) because both were concurrently elevated and the concentrations of these elements were correlated in the phyllodes (Fig. 2D). These observations suggest that the uptake and regulation of calcium and sulphur is tightly coupled in this species, an assertion that is supported by the large number of calcium sulphate accumulations present throughout the tissues of the dry phyllodes (Fig. 5).
Distribution, morphology and proposed formation of calcium sulphate accumulations in A. bivenosa group phyllodes
Many of the sulphate accumulations observed in the A. bivenosa group phyllodes had a distribution consistent with the loci of cellular vacuoles. Although vacuolar compartmentalization is a known sulphate regulation mechanism (Hawkesford and De Kok, 2007), many cells appeared to be completely filled with calcium sulphate (Figs 5 and 7). Our broad understanding of sulphur uptake, transport and storage indicates that this element is primarily taken up by the roots as dissolved sulphate, then stored in the vacuoles of root and xylem parenchymal cells, and transported via the xylem to leaves, where the sulphate (S6+) is again stored in the vacuoles or reduced to sulphide (S2–) and incorporated into cysteine (i.e. protein synthesis) (Takahashi, 2010). In general, the majority of plant tissue sulphur (80–90 %) can be attributed to protein as cysteine, methionine and reduced glutathione (Khan et al., 2008). However, the crystalline calcium sulphate observed in phyllodes during the present study can only have been precipitated from relatively large quantities of inorganic sulphur in the cells. This suggests that protein synthesis is not the primary driver of sulphur uptake nor the primary sulphur storage mechanism for the thiophores described in this study.
Evidence for thiophore behaviour in other Acacia species
Samples of each of the other species from the A. bivenosa group exhibited high foliar calcium and sulphur concentrations (Table 3), as well as similar crystalline calcium sulphate accumulation in terms of distribution, abundance and morphology (Figs 5 and 7). In particular, a relatively large (n = 48) sample of A. ligulata provided elemental data very similar to those presented for A. bivenosa (Table 3; Fig. 2). This strongly suggests that these phylogenetically closely related species (Fig. 8) are also thiophores and that this trait is constitutive and inherited. Examination of phyllodes from several Acacia species outside the group (marked with an asterisk in Table 2), with the exception of A. robeorum, did not reveal the same calcium sulphate presence or such elevated calcium and sulphur levels (Table 3), indicating that sulphur and calcium accumulation is not a consistent behaviour throughout the genus, nor is it exclusive to the A. bivenosa group. The current evidence also indicates that sulphur accumulation appears to be independent of substrate and is not confined to individual Acacia populations, as similar observations were made from samples obtained from locations across much of the continent (Table 2), and were collected during different seasons, over three decades (1980–2013). The broad range of landscapes that this group of species are able to colonize are unlikely to all contain significant amounts of gypsum. For instance, A. ligulata was predominantly sampled from red sand dunes (non-gypsiferous) for this study.
He et al. (2012b) reported elevated sulphur accumulation in A. robeorum, which grows under similar conditions to A. bivenosa, and related sulphur accumulation to biogenic calcium sulphate precipitation (i.e. biomineralization). Our findings match those results in spite of a different sample preparation technique (critical point drying and tissue fracture). We support this assertion with the observation of the same sulphate mineralization characteristics, as discussed for A. bivenosa, in samples of A. robeorum examined during this study. As such, the mineralization found in these two species appears to relate to the same trait. Although A. robeorum was omitted from the most recent and exhaustive molecular phylogeny data set available for the genus (Bui et al., 2014), those species possessing the greatest morphological and distributional affinity with A. robeorum are A. synchronica and A. victoriae (Maslin, 1998; Maslin and van Leeuwen, 2008). The likely phylogenetic relationship between the newly identified A. bivenosa group thiophores and A. robeorum is illustrated by the locations of A. victoriae and A. synchronica in Fig. 8. Interestingly, this trait was not observed in two species (A. blakelyi and A. myrtifolia) that share a greater degree of common ancestry with the A. bivenosa group, compared with A. robeorum, suggesting that the evolutionary aspects of thiophore behaviour in the Acacia genus are complex and merit further investigation.
For this study, we set out to test the hypothesis: thiophore behaviour is a constitutive genetic trait occurring in desert-dwelling species of several genera, and that a characteristic mineralization would feature in these species. Our findings confirm that thiophores do possess a constitutive trait, which is inherited by several Acacia species, but is not common to many important xerophytes found in Australian desert habitats, nor exclusive to xerophytes. Several of the 12 species we have identified as thiophores occur in coastal (A. xanthina, A. rostellifera and A. cupularis) or riparian (A. salicina) habitats, which suggests that the evolutionary and ecophysiological explanations for this trait do not lie solely in adaptation to arid conditions. These newly identified thiophores are also recognized by characteristic foliar calcium–sulphur mineralization (identified as gypsum). These observations add to our understanding of thiophore nutrient uptake and storage, although the associated mechanisms (i.e. the coupling between calcium and sulphur) require further investigation. Future work on these plants may enhance our understanding of plant adaption to stressful environments, and potentially lead to applications for revegetating degraded habitats and contaminated soils.
ACKNOWLEDGEMENTS
We wish thank, in particular, Kingsley Dixon, Jason Stevens, Todd Erickson, Bruce Maslin, Steve Hill, Anna Petts, Lisa Worrall, Dirk Kirste, Brad Pillans and John Joseph. We also thank Tenten Pinchand, Rodney King, Keryn Welk and David Kaczan for sample preparation, Joe Miller and Elizabeth Bui for assistance with phylogenetic mapping, and Aaron Stewart, Alistair White and two anonymous reviewers for their assistance in improving the manuscript. We are grateful to Acme Laboratories, Vancouver for the analyses performed. This work was supported by the Cooperative Research Centre for Landscape Evolution and Mineral Exploration, AMIRA International, Newmont Asia Pacific, Tanami Gold NL and a CSIRO OCE Postdoctoral Fellowship [WA13/01407].
LITERATURE CITED
- Abrol YP, Ahmad A, eds. 2003. Sulphur in plants. Dordrecht, The Netherlands: Springer Science + Business Media. [Google Scholar]
- Al-Ani T, Habib I, Abdulaziz A, Ouda N. 1971. Plant indicators in Iraq, part II: mineral composition of native plants in relation to soils and selective absorption. Plant and Soil 35: 29–36. [Google Scholar]
- Brooks R. 1998. Plants that hyperaccumulate heavy metals. Egham, UK: CABI. [Google Scholar]
- Bui EN, Thornhill A, Miller JT. 2014. Salt- and alkaline-tolerance are linked in Acacia. Biology Letters 10: 20140278. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chapman A, Maslin B. 1992. Acacia miscellany: 5. A review of the A. bivenosa group (Leguminosae: Mimosoideae: section Phyllodineae). Nuytsia 8: 249–283. [Google Scholar]
- Duvigneaud P, De Smet SD. 1968. Essai de classification chimique (éléments minéraux) des plantes gypsicoles du bassin de l’Ebre. Bulletin de la Société Royale de Botanique de Belgique 101: 279–291. [Google Scholar]
- Franceschi VR, Nakata PA. 2005. Calcium oxalate in plants: formation and function. Annual Review of Plant Biology 56: 41–71. [DOI] [PubMed] [Google Scholar]
- Franceschi VR, Schueren AM. 1986. Incorporation of strontium into plant calcium oxalate crystals. Protoplasma 130: 199–205. [Google Scholar]
- Hawkesford MJ, Horst W, Kichey T, et al. 2012. Functions of macronutrients In: Marschner P, ed. Marschner’s mineral nutrition of higher plants. Amsterdam: Elsevier, 135–190. [Google Scholar]
- Hawkesford MJ, De Kok LJ, eds. 2007. Sulfur in plants: an ecological perspective. Dordrecht, The Netherlands: Springer. [Google Scholar]
- He H, Bleby TM, Veneklaas EJ, Lambers H, Kuo J. 2012a. Precipitation of calcium, magnesium, strontium and barium in tissues of four Acacia species (Leguminosae: Mimosoideae). PLoS one 7: e41563. doi:10.1371/journal.pone.0041563. [DOI] [PMC free article] [PubMed] [Google Scholar]
- He H, Bleby TM, Veneklaas EJ, Lambers H, Kuo J. 2012b. Morphologies and elemental compositions of calcium crystals in phyllodes and branchlets of Acacia robeorum (Leguminosae: Mimosoideae). Annals of Botany 109: 887–896. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hollander M, Wolf DA. 1999. Nonparametric statistical methods. Chichester: Wiley. [Google Scholar]
- Hulme K, Hill S. 2003. River Red Gums as a biogeochemical sampling medium in mineral exploration and environmental chemistry programs in the Curnamona Craton and adjacent regions of NSW and SA In: Roach IC, ed. Advances in regolith. CRC LEME, 205–210. [Google Scholar]
- Hutchin M, Vaughan B. 1967. Relation between calcium and strontium transport rates as determined simultaneously in the primary root of Zea mays. Plant Physiology 42: 644–650. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ichikuni M, Musha S. 1978. Partition of strontium between gypsum and solution. Chemical Geology 21: 359–363. [Google Scholar]
- Khan NA, Singh S, Umar S, eds. 2008. Sulfur assimilation and abiotic stress in plants. Berlin: Springer. [Google Scholar]
- Kruskal WH, Wallis A. 1952. Use of ranks in one-criterion variance analysis. Journal of the American Statistical Association 47: 583–621. [Google Scholar]
- Maslin BR. 1998. Acacia miscellany 6. Review of Acacia victoriae and related species (Leguminosae: Mimosoideae: Section Phyllodineae). Nuytsia 8: 285–309. [Google Scholar]
- Maslin B, van Leeuwen S. 2008. New taxa of Acacia (Leguminosae: Mimosoideae) and notes on other species from the Pilbara and adjacent desert regions of Western Australia. Nuytsia 18: 139–188. [Google Scholar]
- Maslin B, van Leeuwen S, Reid J. 2010. Wattles of the Pilbara. Perth: Department of Environment and Conservation. [Google Scholar]
- Moore P. 2005. A guide to plants of inland Australia. Sydney, Australia: Reed New Holland. [Google Scholar]
- Na G, Salt DE. 2011. The role of sulfur assimilation and sulfur-containing compounds in trace element homeostasis in plants. Environmental and Experimental Botany 72: 18–25. [Google Scholar]
- Palacio S, Escudero A, Montserrat-Martí G, Maestro M, Milla R, Albert MJ. 2007. Plants living on gypsum: beyond the specialist model. Annals of Botany 99: 333–343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pors Nielsen S. 2004. The biological role of strontium. Bone 35: 583–588. [DOI] [PubMed] [Google Scholar]
- Rausch T, Wachter A. 2005. Sulfur metabolism: a versatile platform for launching defence operations. Trends in Plant Science 10: 503–509. [DOI] [PubMed] [Google Scholar]
- Reid N, Hill SM. 2010. Biogeochemical sampling for mineral exploration in arid terrains: Tanami Gold Province, Australia. Journal of Geochemical Exploration 104: 105–117. [Google Scholar]
- Reid N, Hill SM, Lewis DM. 2009. Biogeochemical expression of buried gold mineralization in semi-arid northern Australia: penetration of transported cover at the Titania Gold Prospect, Tanami Desert, Australia. Geochemistry: Exploration, Environment, Analysis 9: 267–273. [Google Scholar]
- Ruiz JM, López-Cantarero I, Rivero RM, Romero L. 2003. Sulphur phytoaccumulation in plant species characteristic of Gypsiferous soils. International Journal of Phytoremediation 5: 203–210. [DOI] [PubMed] [Google Scholar]
- Storey R, Thomson W. 1994. An X-ray microanalysis study of the salt glands and intracellular calcium crystals of Tamarix. Annals of Botany 73: 307–313. [Google Scholar]
- Takahashi H. 2010. Regulation of sulfate transport and assimilation in plants. International Review of Cell and Molecular Biology 281: 129–159. [DOI] [PubMed] [Google Scholar]
- Williams CH, David DJ. 1963. The uptake of calcium and strontium by plants from some Australian soils. Soil Research 1: 185–202. [Google Scholar]