Introduction

Family Chenopodiaceae contains the largest number of C4 species among dicots (cf. Sage 2004). C4 evolved in subfamilies Chenopodioideae (Atriplex), Suaedoideae (Bienertia and Suaeda), Salicornioideae (Tecticornia), and Salsoloideae (various genera), with at least 12 independent origins (Kadereit et al. 2003; Kapralov et al. 2006; Akhani et al. 2007; Voznesenskaya et al. 2008). The diversity currently found in the family includes a number of Kranz anatomical types (six major classifications, some of which have further minor variations in anatomy), two anatomical forms of single-cell C4 functioning species, as well as the occurrence of two biochemical subtype C4 species, NAD-malic enzyme (NAD-ME) and NADP-malic enzyme (NADP-ME), and C3–C4 intermediate species (Carolin et al. 1975; Akhani et al. 1997; Pyankov et al. 1992; Pyankov et al. 2001; Voznesenskaya et al. 2001a; Kadereit et al. 2003; Edwards et al. 2004; Akhani et al. 2005). Most of this diversification in C4 traits occurred in Central Asian and Middle Eastern deserts, in a wide range of habitats from hyper-saline to extremely arid conditions (Butnik et al. 1991, 2001; Sage 2004; Akhani 2006).

Demonstration of the occurrence of C4 photosynthesis in individual chlorenchyma cells in Suaeda aralocaspica (Bunge) Freitag & Schütze and B. cycloptera Bunge ex Boiss., B. sinuspersici Akhani broke the paradigm that Kranz anatomy is required for its function in terrestrial plants (Voznesenskaya et al., 2001b; Edwards et al., 2004; Akhani et al., 2005). Function of C4 photosynthesis in leaves and cotyledons of these species (see Edwards et al. 2004), and surprisingly, in flowers of B. cycloptera and S. aralocaspica (Boyd et al. 2007), has been shown using a variety of methods. Differentiation to form C4 type chlorenchyma cells occurs as young leaves mature (Voznesenskaya et al. 2003, 2005). The spatial separation to form two cytoplasmic compartments, including dimorphic chloroplasts and other organelles in single mature chlorenchyma cells, is supported by a well-developed cytoskeleton (Chuong et al. 2006). In Bienertia, the peripheral cytoplasm (having one type of chloroplast) and a central cytoplasmic compartment (a ball-like structure having another type of chloroplast) are proposed to function analogous to mesophyll and bundle sheath cells in Kranz type C4 species. In S. aralocaspica, spatial separation of functions occurs because of dimorphic chloroplasts being oriented toward the distal and proximal ends of elongated mesophyll cells.

In C4 plants, atmospheric CO2 is converted to bicarbonate (by carbonic anhydrase) which is utilized by phosphoenolpyruvate carboxylase (PEPC), leading to the synthesis of C4 acids. Effective donation of CO2 from C4 acids to Rubisco in bundle sheath cells restricts discrimination against 13C by Rubisco resulting in more positive δ13C values than C3 plants (Farquhar 1983; Peisker and Henderson 1992). The δ13C values for C4 plants are approximately between −10‰ and −15‰; whereas, values for C3 species are typically between −24 and −30‰ (Akhani et al. 1997; Cerling 1999). δ13C values for herbarium specimens of B. cycloptera, collected from various sites, have values indicative of C4 photosynthesis (Voznesenskaya et al. 2002; Akhani et al. 2005). However, Freitag and Stichler (2002) reported values of leaves of B. cycloptera grown in greenhouse conditions ranging from −15.5 to −21.1‰ (with the latest developing leaves being most negative), whereas, the single-cell C4 functioning species S. aralocaspica and related Kranz type species had C4 type values. Voznesenskaya et al. (2002) also reported lower values in young leaves than in mature leaves of B. cycloptera from plants grown in growth chamber conditions. This raised the question as to whether by nature this species might function as a facultative C4 plant, i.e., in a C3, intermediate or C4 mode depending on environment (e.g., light level) or stage of development.

To test this hypothesis, during a complete growing season, we followed morphological development, and sampled and analyzed δ13C values of photosynthetic organs of B. cycloptera, eight other C4 species, and one C3 species belonging to Chenopodiaceae which co-occur in a saline biotope in north central Iran. Seeds were collected from the site for a number of the C4 species, and plants grown in a greenhouse; subsequently, the anatomy of cotyledons versus leaves, and the biochemical subtype of photosynthetic CO2 fixation, were determined.

Materials and methods

Field studies and plant material

Field studies were conducted in salt flats near Mardabad, in saline soils near the Rude-Shur River, which is located 40 km west of Tehran. The studies lasted from early spring (beginning of March) to the middle of autumn (end of October), 2003. The cotyledons, young leaves on seedlings, and leaves and assimilating organs later in the life cycle (floral leaves, bracts and bracteoles, and photosynthetic shoots) were sampled. The species included B. cycloptera Bunge ex Boiss., and eight Kranz type C4 species, Atriplex leucoclada Boiss., Suaeda arcuata Bunge, S. cochlearifolia Woloszczak ex Stapf, Climacoptera turcomanica (Litw.) Botsch., Halimocnemis rarifolia (C. Koch) Akhani, Halimocnemis mollissima Bunge, Halimocnemis pilifera Moq., and Petrosimonia glauca (Pall.) Bunge, and a C3 species, Atriplex verrucifera M. Bieb. (as a control), all belonging to Chenopodiaceae. They were growing either in association with B. cycloptera or in close proximity on salt flats and adjacent wastelands. Collections were made at regular intervals (~biweekly). At the time of collection, the developmental stage was noted and documented by herbarium vouchers and digital photography. All vouchers are present in the Botanical Biodiversity Research Laboratory of the School of Biology, University of Tehran.

Carbon isotope composition

For each species, on a given date samples were collected as follows with replicates dependent on availability: for cotyledons (n = 2–5); for other photosynthetic organs (leaves, floral leaves etc., n = 1–5). For a given date, replicates for each species were combined to form a single preparation for isotope analysis. When leaves were sampled during the growing season, generally the most newly developed and exposed leaves or bracts were taken. For determination of δ13C values, cotyledons, leaves, or photosynthesizing shoots (depending on their age and availability) were analyzed according to the standard procedure relative to PDB (Pee Dee Belemnite) limestone as the carbon isotope standard as described by Bender et al. (1973) and Osmond et al. (1975). One to two measurements were made of the isotope value. The samples were ground to a fine powder, and then 1–2 mg placed in a tin capsule and combusted in a Heraeus CHN Eurovector elemental analyzer (GSF National Research Center for Environment and Health, Munich, Germany). The resulting N2 and CO2 were separated by gas chromatography and admitted into the inlet of a Finnigan MAT Delta S (GSF) or Micromass Isoprime isotope ratio mass spectrometer (IRMS) (Washington State University) for determination of 13C/12C ratios (R). Each sample was analyzed once or twice. δ13C values were calculated where δ = 1000 × (Rsample/Rstandard − 1).

Material for anatomical studies

Seeds of Bienertia cycloptera, Suaeda arcuata, S. cochlearifolia, Halimocnemis rarifolia, H. pilifera and Petrosimonia glauca were collected from the study area and were germinated in Petri dishes at room temperature. Depending on the species, after 2–10 days, the young seedlings were transplanted into 10-cm diameter pots containing 10 parts commercial potting soil, one part clay, one part sand and 100 g gypsum, and grown for 3 weeks at room temperature under ~50 μmol m−2 s−1 photosynthetic photon flux density (PPFD). After 3 weeks, the plants were transferred to a greenhouse and grown under natural sunlight plus supplemental light (sodium vapor lamps providing a PPFD of 400 μmol m−2 s−1) with a maximum mid-day PPFD on clear days of 1750 μmol m−2 s−1, and 25°C day and 18°C night temperatures. Plants were watered once weekly with 20:20:20 Peters Professional fertilizer (1 g l−1) plus 150 mM NaCl per liter of water, or otherwise with water as needed.

Light microscopy

The cotyledons, and young and mature leaves, were fixed at 4°C in 2% (v/v) paraformaldehyde and 1.25% (v/v) glutaraldehyde in 0.05 M PIPES buffer, pH 7.2. The samples were dehydrated with a graded ethanol series and embedded in London Resin White (LR White, Electron Microscopy Sciences, Fort Washington, PA, USA) acrylic resin. Cross sections were made on a Reichert Ultracut R ultramicrotome (Reichert-Jung GmbH, Heidelberg, Germany). Semi-thin sections were stained with 1% (w/v) Toluidine blue O in 1% (w/v) Na2B4O7. The sections were photographed with a Nikon Optiphot-2 microscope coupled with a digital camera Muticam 2300 (School of Biology, University of Tehran). The anatomical features of Halimocnemis mollissima, Climacoptera turcomanica, and the two Atriplex species were studied using hand sections (not shown) to determine the type of anatomy.

Climate

The synoptical climatic data for 19 years (1985–2003) at station Karaj (35°55′N, 50°54′E, 1312.5 m), located at 25 km NW of the area, were obtained from the Iranian Meteorological Organization. Averages for the most important climatic data were obtained according to Walter’s method (Walter and Lieth 1967). More detailed data for 2003, the year of the study, are presented for comparison with the long-term, 19-year average.

Protein extraction

Total soluble protein from different samples was extracted from greenhouse-grown species using a buffer containing 100 mM Tris–HCl, pH 7.3; 1 mM EDTA; 10 mM MgCl2; 15 mM β-mercaptoethanol; 20% (v/v) glycerol; 1 mM NaF; 50 mM KH2PO4; 1 mM phenylmethylsulfonylfluoride; 1 mM dipyridyldisulfide; 10 μg ml−1 leupeptin and 10 μg ml−1 chymostatin. The samples were ground completely in a cold mortar and centrifuged at 10,000g for 10 min at 4°C. Samples were diluted in 0.25 M Tris–HCl, pH 7.5; 2% (w/v) SDS; 0.5% (v/v) β-mercaptoethanol and 0.1% (v/v) bromophenol blue and boiled for 2 min for SDS-PAGE. Protein concentration was determined using the Bradford reagent (Biorad) using bovine serum albumin as standard.

Gel electrophoresis and Western blots

SDS-PAGE was performed in 8% (w/v) polyacrylamide gels according to Laemmli (1970). Proteins were visualized with Coomassie blue or electroblotted onto a nitrocellulose membrane for immunoblotting according to Burnette (1981). Anti-Amaranthus viridis PEPC (1:200 dilution, Colombo et al. 1998); anti-maize 62 kDa NADP-ME (1:200, Saigo et al. 2004); anti-cucumber phosphoenolpyruvate carboxykinase (PEP-CK, 1:200, Walker et al. 1995); anti-Amaranthus hypochondriacus αNAD-ME (1:1,000, Long et al. 1994); and anti-spinach Rubisco large subunit (1:10,000, courtesy B. McFadden), were used. Bound antibodies were located by linking to AP-conjugated goat anti-rabbit IgG according to the manufacturer’s instructions (Sigma).

Results

Study area, vegetation, and climate

The area is a saline flat biotope near the Rude-Shur (salty river), located ca. 40 km West of Tehran, 25 km SE of Karaj (35°43′26″N, 50°44′27″E, 1166 m) (Fig. 1). The saline flats around the Rude-Shur area are partly wasteland from former cultivated ground, which is subjected to temporary inundation and runoff water during the rainy season. The soil is alkaline and composed of clay with salinity levels ranging from 4.9 to 78 ds m−1. The vegetation is dominated by annual C4 halophytic communities with scattered Tamarix shrubs, indicative of a high water table in the area, and patches of perennial hemi-cryptophytic communities dominated by Atriplex verrucifera (see Akhani et al. 2003, and Fig. S1 for details on the vegetation and soil characterization). The long roots of most of the annual species (personal observations, H. Akhani) may reach the wetter parts of the soil horizon during the dry season. The vegetation period begins toward the end of March and ends in November for most halophytes growing in the area. However, this area is also rich in pseudohalophytes and ephemerals which benefit from lower soil salinity caused by leaching due to winter and early spring rainfalls. Most of the pseudohalophytes complete their life cycles, or die, due to increasing temperatures from May onward.

Fig. 1
figure 1

Relief map of the study area showing its geographical location in North-Central Iran (lower inset left). Right side inset shows the climatic diagram of Karaj station based on an average over 19 years. ■, average mean daily temperature; ▲, average monthly precipitation. The data above the diagram shows elevation of the area from sea level, mean annual temperature (°C), average annual precipitation (mm), and geographical coordinates. Dotted area in the diagram represents the dry period and vertically hatched areas indicate the wet period. The black area on the x-axis represents months with mean daily minimum temperature below 0°C and shaded areas show months with absolute minimum temperature below 0°C, i.e., with early or late frosts. S: location of sampling; C: location of climatic station

Full size image

The diagram in Fig. 1 shows the averages of important climatic factors over 19 years. The area has received an average of 239.5 mm annual precipitation with an annual mean daily temperature of 14.9°C. Much of the precipitation occurs during late autumn, winter, and early spring, with the two maximums during December and March. From April to November, the area is arid with almost no rainfall from June to September. These conditions, combined with low humidity, ranging on average from 32 to 67%, indicate an extreme continental climate. Based on observations during this study, the area was found to be very windy due to interaction between cool mountain and warm desert conditions. Table 1 shows data for nine major factors for 2003. The April precipitation of 80 mm during 2003 was exceptionally high, compared to the 19-year average, dropping quickly to 18.3 mm in May. In 2003 the seasonal change in mean daily temperatures, and the essential lack of precipitation in June through September, was similar to conditions averaged over 19 years. The large difference (29.2°C) between the lowest and the highest temperatures recorded during this period is indicative of the great variation between night and day temperatures, and also the rapid increase of temperature in the summer.

Table 1 Climatic data from station Karaj during 2003
Full size table

Taxonomy, comparative morphology

A list of the 10 chenopod species studied, their photosynthetic anatomical types, and biochemical subtypes among the C4 species, their life forms, distributions, and previous reports on δ13C values are given in Table 2. Pictures and descriptions of live specimens of all species at different developmental stages are shown in Fig. S1 (Supplementary material available at Photosynthesis Res online). The two Atriplex species are subshrubs belonging to subfamily Chenopodioideae, tribe Atripliceae. The remaining species are annual plants with succulent leaves belonging to subfamily Suaedoideae, tribes Bienertieae (Bienertia) and Suaedeae (Suaeda), and subfamily Salsoloideae, tribe Caroxyloneae (Climacoptera, Halimocnemis s.l. and Petrosimonia). The C4 species A. leucoclada is widespread, occurring in warmer parts of SW Asia in many disturbed saline habitats. The C3 species A. verrucifera is widely distributed in temperate deserts and semi-arid areas of SW Asia and has a more northerly range. The range of other species is centered in the Iranian temperate deserts and salines, and radiates into Central Asia as the Irano-Turanian elements. The morphological development and phenology of annual species are more or less similar. All species germinate during March; during April, May, and June, they are densely leafy, and thereafter begin flowering. As flowers develop during late June and early July, the larger vegetative leaves begin to senesce, and are replaced by shorter floral leaves and bracts.

Table 2 List of species studied, their anatomical type, C4 subtype, life form, distribution, and previous records of δ13C values from herbarium field collected specimens. Abbreviations: AT: anatomical type
Full size table

Anatomy of leaves and cotyledons

The types of leaves of the C4 chenopod species studied fall into five groups, which are summarized in Table 2. (i) Bienertia cycloptera, with Bienertioid type anatomy, has semi-terete linear-cylindrical leaves with three layers of chlorenchyma tissue internal to the epidermal cells and 2–3 layers of water storage cells in the central part of the leaf (Fig. 2a). (ii) Suaeda arcuata has cylindrical leaves with Salsinoid type anatomy consisting of a palisade mesophyll layer adjacent to the epidermal tissue and a bundle sheath (Kranz) layer surrounding the aqueous tissue, with vascular bundles imbedded in the center (Fig. 2b). (iii) Suaeda cochlearifolia, with Schoberiod type anatomy, has flat, succulent, ovate leaves and a unique Kranz anatomy with centrifugal chloroplasts in the bundle sheath cells surrounding the vascular bundles (Fig. 2c). (iv) The remaining five C4 species in tribe Caroxyloneae have Salsoloid type anatomy with semi-terete, cylindrical to filiform leaves. The leaves are composed of an epidermal layer, followed by a single layer of palisade chlorenchyma cells and a layer of Kranz cells, illustrated with Halimocnemis pilifera, H. rarifolia, and Petrosimonia glauca in Fig. 2d–f. (v) Atriplex leucoclada has laminate leaves with Atriplicoid type anatomy in which an interrupted Kranz layer encircles the central and lateral vascular bundles. On both the adaxial and abaxial sides, there is one layer of palisade chlorenchyma with large cells, followed by a second layer of smaller palisade chlorenchyma, which encircles the Kranz layer. A. verrucifera is a C3 species having laminate leaves and non-Kranz anatomy (not shown).

Fig. 2
figure 2

Light microscopy of anatomical cross sections of leaves of Bienertia cycloptera and some other chenopods at the study site. (a) Bienertia cycloptera, (b) Suaeda arcuata, (c) Suaeda cochearifolia, (d) Halimocnemis pilifera, (e) Halimocnemis rarifolia, and (f) Petrosimonia glauca. Abbreviations: C: chlorenchyma cells of single-cell C4 functioning species; AC: aqueous cells; VB: vascular bundle, KC: Kranz cells, HY: hypodermis; PC: palisade chlorenchyma. Scale bar for a is 200 μm; all other panels have the same scale bar as c which is 100 μm

Full size image

The C4 species in the study can be classified into two main groups with respect to type of anatomy of cotyledons. In one group cotyledons have Kranz type anatomy, with each species having the same form of C4 anatomy in cotyledons as that appears in leaves, which includes B. cycloptera, A. leucoclada, and the two Suaeda species. This is illustrated with S. cochlearifolia with cotyledons having Schoberoid type anatomy (Fig. 3a), similar to that of leaves (Fig. 2c). In the other group, which consists of species in tribe Caroxyloneae, the cotyledons have C3 type anatomy (illustrated for H. pilifera, H. rarifolia, and P. glauca, Fig. 3b, c), while leaves of these species have Salsoloid type Kranz anatomy (Fig. 2d–f).

Fig. 3
figure 3

Light microscopy of cross sections of cotyledon leaves of four selected species illustrating the occurrence of Kranz or C3 type anatomy. (a) Suaeda cochlearifolia (C4), (b) Halimocnemis pilifera (C3), (c) Halimocnemis rarifolia (C3), and (d) Petrosimonia glauca (C3). KC: Kranz cells. Scale bars for all panels 100 μm

Full size image

Western blots of some key enzymes involved in photosynthesis

Western blots (Fig. 4) clearly show that the five representative species/genera analyzed in this study have high levels of PEPC along with Rubisco, characteristic of C4 species. All species show immunolabeling for the C4 acid decarboxylase NAD-ME. There was no immunolabeling with antibody for NADP-ME (lanes 2–6), while lane 1 shows immunolabeling with the control Zea mays, a known NADP-ME type C4 species. There were no immunoreactive bands with antibody for PEP-CK (lanes 2–6), while lane 1 shows reactivity with the control Spartina alternifolia, a known PEP-CK type C4 species (for other controls, see Voznesenskaya et al. 2006). All C4 species in the study have NAD-ME subtype biochemistry.

Fig. 4
figure 4

Western blot analysis of total protein extracted from leaves revealed with antibodies raised against Amaranthus hypochondriacus αNAD-ME, recombinant 62 kDa NADP-ME from Zea mays leaves, cucumber PEP-CK, spinach Rubisco large subunit, and A. viridis PEPC. For Bienertia cycloptera (2), Suaeda cochlearifolia (3), Halimocnemis rarifolia (4), Halimocnemis pilifera (5), and Petrosimonia glauca (6) 5 μg of protein was loaded for Western blots for Rubisco and 20 μg for Western blots for other peptides. For the Western blots for NAD-ME, Rubisco and PEPC, selected bands from use of molecular weight markers are shown in lane (1). For the Western blots with antibodies against NADP-ME, there was no reactivity (lanes 2–6); lane (1) shows reactivity with antibody to NADP-ME using extracts from leaves of Zea mays, a control NADP-ME type C4 species. Western blots with antibodies against PEP-CK showed no reactivity (lanes 2–6); lane (1) shows reactivity with antibody to PEP-CK using extracts from leaves of Spartina alterniflora, a control PEP-CK type C4 species. Molecular masses of the immunoreactive bands are shown on the right

Full size image

Carbon isotope variation

Cotyledons

δ13C values for cotyledons (Table 3) can be grouped into three categories: (i) A. verrucifera, with a value of –26.7‰, clearly lies in the C3 range. (ii) The three species of Caroxyloneae, with values between −20.4‰ in H. mollissima and −22.7‰ in P. glauca, have C3-like values. (iii) Species belonging to Suaeda and Bienertia, have C4 type values which range from −11.8‰ in S. arcuata to −15.0‰ in B. cycloptera.

Table 3 Carbon isotope composition of cotyledons of Bienertia cycloptera and associated C4 and C3 species
Full size table

Leaves

The seasonal pattern of change in carbon isotope composition, and the seasonal average δ13C value for each species, based on samples (leaves, and later in the life cycle floral leaves and bracts) taken between April and October are shown in Fig. 5. Leaves of A. verrucifera have a C3 type isotope value (seasonal average −27.0‰), similar to that of its cotyledons. Leaves of all other species have C4 type isotope values, with seasonal averages ranging from δ13C of −11.3‰ in H. mollissima to −15.9‰ in A. leucoclada. Among the C4 species examined, the lowest values were observed in genus Halimocnemis (with an average δ13C for three species of −12.6‰). The highest seasonal variation based on the standard deviation (Fig. 5) was in Suaeda arcuata (±1.61) and the lowest in H. mollissima (±0.43).

Fig. 5
figure 5

Histogram showing seasonal variation of δ13C values in Bienertia cycloptera and nine other Chenopodiaceae species studied. Samples were taken from leaves, floral leaves, bracts, and shoots. The seasonal average δ13C values are given at the upper left side of each diagram. Values not indicated above each histogram represent normal leaves; other cases include: A—Floral leaves or bracts and bracteoles; B––upper leaves; C—Flowering or fruiting shoots; and D—Uppermost inflorescence branch (green flowers, very small floral leaves, and stem)

Full size image

Discussion

The genus Bienertia is composed of two species which perform C4 photosynthesis without Kranz anatomy. B. cycloptera is distributed in temperate salty deserts in Iran and the surrounding areas, whereas its vicariant, B. sinuspersici, inhabits warm saline and coastal areas around the Persian Gulf and the Gulf of Oman (Akhani et al. 2005). The ecology of B. cycloptera shows that it associates with C4-dominated annual chenopod communities (Akhani et al. 2003). In the study area, it is common and associated with several other C4 species, Climacoptera turcomanica, Halimocnemis rarifolia, Petrosimonia glauca, and Suaeda cochlearifolia, which are growing on clay soils. In close proximity there are several other Chenopodiaceae species, including the C4 species Halimocnemis mollissima and H. pilifera, which occur in drier and less saline soils, the C4 species Atriplex leucoclada, growing in disturbed areas, and the C3 species Atriplex verrucifera, growing on highly saline and heavy clay and wet soils. The domination by NAD-ME C4 subtypes in this community (Table 2) might be related to availability of nitrogen from edaphic conditions with nitrified soils due to previous agricultural activities. Studies on C4 grasses indicate NAD-ME type species have a lower nitrogen use efficiency (photosynthesis and biomass produced per unit leaf N) than NADP-ME type species suggesting the latter might be more competitive on N-poor soils (Ghannoum et al. 2005; Akhani and Ghasemkhani 2007).

Carbon isotope values of B. cycloptera relative to other species in this study

The discovery of single-cell type C4 photosynthesis in family Chenopodiaceae raised the question of whether these plants are facultative (C3 or C4 depending on environmental conditions) or obligate C4 plants (See Introduction). From studies of plants grown in controlled environments, during leaf development the chlorenchyma cells have a developmental transition from a “C3 default” mode with one type of chloroplast, to differentiation to form two types of chloroplasts and two cytoplasmic compartments for C4 photosynthesis (Voznesenskaya et al. 2003, 2005). Also, the question has been raised as to whether the single-cell C4 system in Bienertia may have greater leakage of CO2 from site of donation of CO2 from C4 acids to Rubisco than Kranz type plants. In either case, development of C3 type chlorenchyma, or less efficiency in function of C4, should lead to more negative δ13C values than typical for C4 plants. With B. cycloptera cultivated in growth chambers, younger leaves had more negative δ13C values (by 2–4‰) than mature leaves, indicating there is a stage of development where there is incomplete function of C4 photosynthesis (Voznesenskaya et al. 2002, 2005). Freitag and Stichler (2002) found with B. cycloptera cultivated under greenhouse conditions that the δ13C values ranged from approximately −15.5‰ to values as low as −21‰ in the uppermost leaves in the fall, which they suggested may be due development under low light conditions. Studies have shown that discrimination against fixing 13CO2 does occur in C4 plants under lower light, and the point at which this occurs may be species-dependent (Henderson et al. 1992, 1998; Kubasek et al. 2007). These results raised the question of whether by nature, the leaves may function in a C3 or intermediate mode under some conditions, or whether they are programmed to develop C4 chlorenchyma independent of environmental conditions.

Among the nine C4 species in this study, the seasonal average δ13C values ranged from −11.3 to −15.9‰ (Fig. 5). The results show that B. cycloptera is performing C4 photosynthesis during the life cycle in this natural habitat, having an annual seasonal average δ13C of −15.2‰. This is consistent with results from herbarium specimens of Bienertia collected from various habitats and several countries, which indicate function of C4 photosynthesis, where δ13C values of samples (leaves, stems, flowers, roots) for B. cycloptera were on average −13.9‰ (n = 16), and −13.8‰ (n = 7) for B. sinsupersici (Akhani et al. 2005).

In considering species-to-species variation in carbon isotope composition of biomass of C4 plants, there are several factors which can cause differences in δ13C values. C4 plants have more positive δ13C values than C3 plants because the C4 cycle concentrates CO2 around Rubisco in a semi-closed compartment. Not all the CO2 delivered is fixed by Rubisco; leakage of a percentage of the CO2 generated from decarboxylation of C4 acids (ϕ) results in more negative δ13C values (with a given enzymatic capacity of the C4 and C3 cycles), as predicted according to theory by Farquhar (1983) and as shown experimentally by Henderson et al. (1992). Variations between species may be due to intrinsic differences (biochemistry and diffusive resistance to CO2 leakage and environmental factors). From estimates of leakage from online measurements of discrimination with a number of C4 species, an average of ~20% for ϕ values was obtained by Henderson et al. (1992). A study by Kubasek et al. (2007), using both online measurements and dry matter analyses of δ13C values to estimate leakage in a number of C4 species, showed a range in ϕ approximately from 20 to 40% (Kubasek et al. 2007). In addition to the effect of CO2 leakage on δ13C, the relationship between the CO2 concentration in the intercellular air space of the leaf and that in the atmosphere (Ci/Ca ratio), which is influenced by stomatal conductance, can also affect isotope discrimination in C4 plants. With a ϕ value of 20%, decreasing the Ci/Ca ratio is predicted to lead to more negative δ13C values (maximum of ~4‰ by decreasing Ci/Ca from 1 to 0). With leakiness values of ~40%, a change in Ci/Ca has no effect on discrimination. With higher ϕ values, where the C4 system is very ineffective in concentrating CO2 around Rubisco, decreasing Ci/Ca values (e.g., due to increased diffusive resistance to CO2 by drought) would have a reverse effect with less discrimination against 13CO2 (more positive δ13C values), as observed in C3 plants (Farquhar 1983; Henderson et al. 1992). A study on Sorghum showed discrimination varies between different lines because of differences in the Ci/Ca ratio; lines having lower Ci/Ca ratios had increased discrimination (Henderson et al. 1998). In this case, differences in discrimination are not because of differences in the efficiency of C4 photosynthesis (e.g., due to variation in ϕ).

δ13C values in dry matter tend to be more negative (e.g., δ13C values 1–2‰) than values obtained from online measurements of discrimination during photosynthesis. This occurs for non-photosynthetic reasons, which are not well defined, but they include fractionation occurring during respiration, and during biosynthesis (e.g., synthesis of lignin and lipid) (Henderson et al. 1992; Kubasek et al. 2007). Thus, differences in the discrimination following fixation of CO2 could contribute to variation in the δ13C values between species, and to seasonal changes in δ13C within a species.

δ13C values relative to features of C4 species in the study

Cotyledons

In family Chenopodiaceae, there are two groups of C4 plants, some having C4 anatomy/photosynthesis in both leaves and cotyledons, others having C4 photosynthesis/anatomy in leaves and C3 photosynthesis/anatomy in cotyledons (Butnik 1974; Voznesenskaya et al. 1999; Pyankov et al. 2000b; Pyankov et al. 2001; Akhani and Ghasemkhani 2007). In this study, there is a relatively high frequency of species in tribe Caroxyloneae at the study site, in which the cotyledons have C3 type and the leaves C4 type anatomy (e.g., Climacoptera turcomanica, Halimocnemis spp. and Petrosimonia glauca). The temperate conditions which occur early in the life cycle in this habitat, would minimize photorespiratory losses of CO2 in cotyledons performing C3 photosynthesis. The first measurements of δ13C on the various species were on cotyledons in the spring (see Table 3). In the three species analyzed in tribe Caroxyloneae, the δ13C values of the cotyledons were more positive than typical values for C3 plants. The values range from −20.4‰ in H. mollissima, to −22.7‰ in P. glauca. In plants which have C4 photosynthesis in leaves, the carbon exported to sink tissue, including developing seeds, will have C4 type isotope composition. Following germination, as seed reserves (carrying the C4 type isotope signature) are utilized (normally photosynthate is transferred from cotyledons to leaves, not vice versa) and as C3 type cotyledons of the above species are performing C3 photosynthesis, the isotope values should shift toward C3 type values. This could explain the “intermediate” isotope values of cotyledons in species in tribe Caroxyloneae. Also, in a study of several C4 species in a saline biotope in Golestan National Park, NE Iran, C3 type cotyledons had C3-like carbon isotope values: Haloxylon ammodendron (−22.2‰), Noaea mucronata (−23.8‰), Suaeda microphylla (−23.7‰), and Pyankovia brachiata (−23.3‰) (Akhani and Ghasemkhani 2007).

Leaves, bracts, shoots

As indicated, the nine C4 species in the study have average seasonal differences in δ13C values ranging from −11.3 to −15.9‰ (a range of ~4.5‰, Fig. 5). All nine species are classified as NAD-ME type C4 plants (Table 2, Fig. 4); so, variation in δ13C is not considered to be because of differences in the type of C4 biochemistry. All of the species grow in open flat habitats, and except for possible differences in soil salinity and water availability in the microhabitats, there are no other apparent abiotic differences which may affect δ13C values.

There are five anatomical forms of C4 represented among the nine C4 species, which could influence diffusive resistance to CO2 and the fraction of leakage (ϕ). Factors which could influence leakage include thickness and composition of bundle sheath cell walls in Kranz type species, distances for diffusion of CO2 in the liquid phase for leakage relative to fixation by Rubisco, and, in NAD-ME type C4 species, the position of mitochondria (the site of C4 acid decarboxylation) relative to Rubisco-containing chloroplasts (von Caemmerer and Furbank 2003). However, the average seasonal range of δ13C values of leaves among the C4 species in the study does not correlate with differences in anatomical types. The leaves of B. cycloptera, with single-cell C4 photosynthesis, had a seasonal average of −15.2‰, which is more positive than A. leucoclada with Atriplicoid type Kranz anatomy (−15.9‰) and very close to P. glauca with Salsoloid type Kranz anatomy (−14.8‰). Climacoptera turcomanica (Salsoloid type) had a seasonal average value of −14.2‰. S. arcuata (Salsina type) and S. cochlearifolia (Schoberia type) had values of −14.1 and −14.3‰, respectively.

The most positive average seasonal δ13C values among the species investigated were observed in Salsoloid Kranz type species belonging to Halimocnemis s.l., with a seasonal average for three members of −12.6‰ (H. mollissima (−11.3‰), H. pilifera (−12.6‰), and H. rarifolia (−13.8‰)). This result is in accordance with previous studies of a larger number of species belonging to Halanthium, Physandera, and Halarchon showing an average δ13C value of −12.1‰ for 11 species sampled (Akhani et al, 1997). All these species, which belong to Halimocnemis s.l., grow in xerohalophytic communities on very dry and mostly gypsum soils. Thus, the most important factor influencing δ13C values in these species may be their degree of xerophytism compared with other species at the study site. Mechanistically, why these species have more positive δ13C values is not clear; it could be due to lower ϕ or differences in capacity of the C4 versus C3 cycle.

Seasonal variation in carbon isotope composition

Besides differences between the nine C4 species in seasonal average δ13C values, within each species there is temporal variation, which resulted in individual values ranging approximately from −11 to −17‰. In the single-cell C4 species B. cycloptera, there is relatively low seasonal variation in δ13C values, which indicates performance of C4 photosynthesis in leaves throughout the life cycle. The standard deviation over the season for B. cycloptera of ±0.99 is lower than for Kranz type S. arcuata and similar to those of S. cochlearifolia, H. pilifera, and C. turcomanica, (see Fig. 5). Our regular sampling from early spring to late autumn of Bienertia, eight other C4, and one C3 member of Chenopodiaceae, demonstrates that Bienertia shows temporal variation of δ13C values similar to a number of C4 species having Kranz anatomy (Table 3, Fig. 5).

The seasonal patterns showed that in most species δ13C values of leaves tended to be most positive in late May and/or early June and more negative in early spring and late summer. During early growth (April–May) there is rainfall and temperatures are cooler, late May and early June temperatures are warmer, while in July and thereafter conditions are hot and dry. In young leaves of B. cycloptera sampled in April the average δ13C was −16.4‰. These C4-like values may suggest a more leaky C4 system early in the growing season in young leaves (greater overcycling by the C4 pathway), partial direct fixation of atmospheric CO2 by Rubisco, and/or increased post-CO2 fixation discrimination through biosynthesis. The more negative values which occur in a number of C4 species late in the growing season (between August and October) (B. cycloptera, A. leucoclada, H. pilifera, H. rarifolia, P. glauca, and S. cochlearifolia), may be related to warm temperatures combined with low precipitation reported in 2003 in the area (Table 1). In September, the decrease in δ13C in B. cycloptera (−16.8‰) is similar to that in A. leucoclada and S. cochlearifolia (Fig. 5). These climate conditions could result in decreased stomatal conductance and lower Ci/Ca ratios which, as noted earlier, can increase carbon isotope discrimination against 13CO2. In S. arcuata, rather negative isotope values were obtained from bracts and fruiting branches in October (δ13C of −18.6‰ which is ~4‰ more negative than values in samples collected earlier in the month). Since leaves were showing signs of senescence at this time, either increased ϕ values (CO2 leakage) or increased respiratory losses could lead to more negative values.

In conclusion, B. cycloptera is a single-cell functioning C4 species, which shows C4 type carbon isotope values in nature with a seasonal pattern of values similar to those of other co-occurring C4 species having Kranz type of anatomy. This, and values for both species of Bienertia from herbarium specimens from different sites, indicates that they perform C4 photosynthesis in nature. There is no support for the hypothesis of Freitag and Stichler (2002) that this is a facultative C4/C3 species. During culture under certain conditions, the rate of development of mature C4 type chlorenchyma cells may be slower, resulting in more negative carbon isotope values, particularly under limiting light. The study by Chuong et al. (2006) shows that the spatial separation of organelles in mature C4 type chlorenchyma cells of Bienertia is supported by an extensive cytoskeleton network and vacuolar development, which may limit potential for reversions to C3 type chlorenchyma in response to climatic conditions.