Adaptative Strategies in Gymnocalycium Species (Cactaceae) and the Presence of Ectomycorrhizae Associated with Survival in Arid Environments

The Cactaceae family makes use of different strategies, both physiological and biochemical, for anatomical adjustments that allow them to grow and reproduce in arid environments. Morphological studies of Gymnocalycium have been scarce, and the anatomy and phytochemistry are still largely unknown. The aim of the present work was to analyze the structural, physiological, and biochemical features of Gymnocalycium marianae and G. oenanthemum, two endemic species of arid regions in Argentina. The anatomic structure, biomass, and photosynthetic pigments, as well as phenolic compound contents, were analyzed in the stem, spine, and root of both species. G. marianae showed stems with deeper substomatal chambers and a more developed photosynthetic tissue than G. oenanthemum. The spines of G. oenanthemum showed higher biomass, thicker epidermal and subepidermal cell walls, and a higher content of phenolic compounds than those of G. marianae. Ectomycorrhizae were observed for the first time in roots in both species. Roots of G. marianae showed high colonization, biomass, and content of phenolic compounds. Both species showed abundant mucilaginous fibers in the stem and root. Finally, these results show the strategies associated with the survival in xeric environments of two cacti species at risk of extinction. They could be useful for the development of ex situ conservation programs.


Introduction
The Cactaceae family is native to the American continent and comprises approximately 200 genera and 2500 species [1]. The genus Gymnocalycium has around 60 species and 7 subgenus; it is distributed in Paraguay, Brazil, Bolivia, Uruguay, and Argentina [2,3]. The province of Catamarca in Argentina has around 14 species of this genus; four of them are endemic, i.e., Gymnocalycium baldianum (Speg.) Speg. and G. marianae Perea; O. Ferrari, Las Peñas & R. Kiesling (subgenus Gymnocalycium); G. stellatum Speg. occultum Ssp. Fric ex H. Till & W. Till (subgenus Trichomosemineum); and G. oenanthemum Backeb. (subgenus Microsemineum). Each of these species has a limited distribution and inhabits specific climatic conditions, i.e., mean annual temperatures and precipitation that range between 9.5 and 19.7 • C and between 380 and 670 mm, respectively [4]. The species used in the present work, G. oenanthemum and G. marianae, grow in arid phytogeographic

Results
Due to the scarce information on morphoanatomy and growth of Gymnocalycium, our studies aimed at making a histological description of the stem, spine, and root of both plant species in relation to biomass distribution, photosynthetic pigments, and phenolic compound content.

Plant Morphology
The fruits and 3.5-year-old plants of G. marianae and G. oenanthemum are shown in Figure 1. Fruits of both taxa were fleshy with rounded scales with a whitish-green edge on the surface, globose and dark green in G. marianae and oval and light green in G. oenanthemum (Figure 1a,b). G. marianae showed grayish-green stems with abundant fine, rigid, straight, dark spines at the base ( Figure 1c). Instead, G. oenanthemum showed dark green stems with thick spines bent back towards the plant body ( Figure 1d).
Stem length was 20% larger in G. marianae than in G. oenanthemum, whereas the stem diameter in the latter was 11% larger (Table 1). Both species showed an axonomorphic main root with lateral and adventitious roots, close to the neck, although thicker in G. marianae (Figure 1c,d). The root length did not present significant differences between both species (Table 1). Stem length was 20% larger in G. marianae than in G. oenanthemum, whereas the stem diameter in the latter was 11% larger (Table 1). Both species showed an axonomorphic main root with lateral and adventitious roots, close to the neck, although thicker in G. marianae (Figure 1c,d). The root length did not present significant differences between both species (Table 1). Values are mean ± SD; different letters in the same arrow indicate significant differences by Student's t test, p < 0.05.

Photosynthetic Pigments
The photosynthetic pigment content of stems in both species is shown in Table 1. Chl a and Chl b content in G. marianae was higher than in G. oenanthemum by 47% and 31%, respectively. These different percentages result in differences in the Chl a/Chl b ratio  Values are mean ± SD; different letters in the same arrow indicate significant differences by Student's t test, p < 0.05.

Photosynthetic Pigments
The photosynthetic pigment content of stems in both species is shown in Table 1. Chl a and Chl b content in G. marianae was higher than in G. oenanthemum by 47% and 31%, respectively. These different percentages result in differences in the Chl a/Chl b ratio between both plant species, where G. marianae showed a higher value, namely, close to 3, due to its higher Chl a content. In G. marianae, the carotenoid content was 1.6-fold higher than in G. oenanthemum (Table 1).

Biomass Distribution
The FW and DW in both species were higher in stems than in roots and spines (Table 2). Total DW in G. marianae was 20% higher than in G. oenanthemum. This latter species showed a high percentage of water in all its organs (Table 2). Regarding the dry weight distribution (DWD), although the stem showed a similar percentage, i.e., around 76%, in both species, it was substantially different in roots and spines (Table 2); G. marianae and G. oenanthemum showed 13.0% and 9.5% DWD in roots and 10.8% and 15.5% DWD in spines, respectively. The DW/FW ratio was higher in all organs of G. marianae compared to those of G. oenanthemum.

Ectomycorrhizae
The lateral roots of 3.5-year-old plants obtained in a greenhouse exhibited noticeably short and swollen branches, corresponding to simple-type structures of ectomycorrhizae ( Figure 2a). The number of structures observed was 19.28 ± 3.75 and 6.44 ± 2.23 in G. marianae and in G. oenanthemum, respectively (p < 0.0001). Both a thin mantle and a Hartig net corresponding to the intercellular ectomycorrhizae mycelium were detected in lateral root cross-sections in both species (Figure 2c-e). The important finding of ectomycorrhizae in plants grown in greenhouses led to the analysis of roots in both plant species in their habitats; the same symbiotic structures were found ( Figure 2b).
The mantle was a 5-40 µm thick and irregular prosenchymatous structure. Mostly, the Hartig net penetrates into the first layer of the cortical parenchyma (Figure 2c-h, indicated with arrows).

Root Anatomy
The primary root structure showed an epidermis made up of one layer of cells, a parenchymal cortex made up of two layers of cells, and the endodermis. Prismatic calcium oxalate crystals were visualized in the cortex of G. oenanthemum only (Figure 2f, inset). The vascular cylinder showed a diarch stele and a multilayered pericycle (Figure 2f). The adventitious roots were structurally identical to the lateral roots, presenting triarch, tetrarch, and pentarch steles.
The secondary structure of roots showed different degrees of development. Early stages with primary tissues, i.e., epidermis, cortex with Hartig net, and endodermis, still attached to the developing periderm were more noticeable (Figure 2g,h). In the secondary structure, periderm, i.e., suber, phellogen, and phellodermis, cortical parenchyma with 6-9 cellular layers, phloem, and xylem, continuous due to cambium activity, and parenchymatical medulla were observed (Figure 2i). Vascular bundles in the phloem ( Figure 2i) and mucilaginous fibers in xylem tissue (Figure 2i, inset) were noticed.

Root Anatomy
The primary root structure showed an epidermis made up of one layer of cells, a parenchymal cortex made up of two layers of cells, and the endodermis. Prismatic calcium oxalate crystals were visualized in the cortex of G. oenanthemum only (Figure 2f, inset). The vascular cylinder showed a diarch stele and a multilayered pericycle (Figure 2f). The adventitious roots were structurally identical to the lateral roots, presenting triarch, tetrarch, and pentarch steles. Ectomycorrhizal intercellular hyphae (Harting net) in cortical parenchyma (red color) (c,f-h). Thin and ultrathin cross-sections denoting the fungus mantle (d,e). Early secondary root structures in different degrees of development (g,h). Secondary root structure with vascular bundles in the phloem and detail of mucilaginous fibers in the xylem (inset) (i). G. marianae (a-e,g-i); G. oenanthemum (f). c, cambium; cp, cortical parenchyma; ep, epidermis; hn, Hartig net; m, mantle; mf, mucilaginous fibers; p, phloem; pa, parenchyma; pe, periderm; vb, vascular bundles; x, xylem. The white and black bars represent the scale.

Stem Anatomy
Stem cross-sections of both species showed round shapes; obtuse lobes with elongated centers and sharp vertices in G. marianae ( Figure 3a) and rounded in G. oenanthemum (Figure 3c) could be seen. Vascular bundles, some scattered and others distributed as a ring, were found in the cortex (arrows in Figure 3a,c).    Square, as well as oblong, cells with sinuous walls were found in the epidermis, in paradermal view. The cuticle was smooth in G. marianae (Figure 3b) and strongly striated with papillae in G. oenanthemum (Figure 3d). Stomatal structure (paracytic-type) was similar in both species (Figure 3d, inset); however, size and density were higher in G. marianae than in G. oenanthemum (Table 1). Glandular trichomes with a multicellular uniseriate foot and unicellular head were observed in both taxa (Figure 3e). Figure 3f,g shows stem cross-sections. Both species showed an epidermis consisting of one cell layer, with a thick and smooth cuticle in G. marianae, but a thin and striated one in G. oenanthemum. G. marianae showed greater thickness in the cell walls of the subepidermal collenchyma. Parenchyma with 4-5 and 5-7 layers of square or oblong cells with 20-23 and 4-5 elongated cell layers and 20-25 and 17-19 circular cell layers was found in G. marianae and G. oenanthemum, respectively. The deep substomatal chambers were observed, extending throughout 2-3 subepidermal lamellar collenchyma layers in both species, whereas parenchyma was observed in up to 5-6 layers in G. marianae and in 1-2 layers in G. oenanthemum.
Both elongated and round idioblasts containing mucilage in the outer area of the cortex (Figure 3h), alongside collateral vascular bundles distributed in a ring with cambium differentiation in some cases (Figure 3i), were found in both species. Also, dispersed bundles in the medulla (not shown) and in the cortex (Figure 3a,c), medullar parenchyma (Figure 3i), as well as xylem with numerous mucilaginous fibers, were found ( Figure 3j).

Spine Anatomy
Glandular trichomes with pluricellular uniseriate feet and unicellular heads could be seen at the spine bases of both species. Trichomes in G. marianae were of two types: the thin, elongated, smooth-cuticled ones (Figure 4a), and the thick and short ones with a striated cuticle (Figure 4b). G. oenanthemum showed thin and elongated trichome clusters with a smooth cuticle (Figure 4c). Square, as well as oblong, cells with sinuous walls were found in the epidermis, in paradermal view. The cuticle was smooth in G. marianae (Figure 3b) and strongly striated with papillae in G. oenanthemum (Figure 3d). Stomatal structure (paracytic-type) was similar in both species (Figure 3d, inset); however, size and density were higher in G. marianae than in G. oenanthemum (Table 1). Glandular trichomes with a multicellular uniseriate foot and unicellular head were observed in both taxa (Figure 3e). Figure 3f,g shows stem cross-sections. Both species showed an epidermis consisting of one cell layer, with a thick and smooth cuticle in G. marianae, but a thin and striated one in G. oenanthemum. G. marianae showed greater thickness in the cell walls of the subepidermal collenchyma. Parenchyma with 4-5 and 5-7 layers of square or oblong cells with 20-23 and 4-5 elongated cell layers and 20-25 and 17-19 circular cell layers was found in G. marianae and G. oenanthemum, respectively. The deep substomatal chambers were observed, extending throughout 2-3 subepidermal lamellar collenchyma layers in both species, whereas parenchyma was observed in up to 5-6 layers in G. marianae and in 1-2 layers in G. oenanthemum.
Both elongated and round idioblasts containing mucilage in the outer area of the cortex (Figure 3h), alongside collateral vascular bundles distributed in a ring with cambium differentiation in some cases (Figure 3i), were found in both species. Also, dispersed bundles in the medulla (not shown) and in the cortex (Figure 3a,c), medullar parenchyma (Figure 3i), as well as xylem with numerous mucilaginous fibers, were found ( Figure 3j).

Spine Anatomy
Glandular trichomes with pluricellular uniseriate feet and unicellular heads could be seen at the spine bases of both species. Trichomes in G. marianae were of two types: the thin, elongated, smooth-cuticled ones (Figure 4a), and the thick and short ones with a striated cuticle (Figure 4b). G. oenanthemum showed thin and elongated trichome clusters with a smooth cuticle (Figure 4c).   In spine cross-sections, both species showed a circular shape with an epidermis made up of one cellular layer and a smooth cuticle. Papillose epidermises without and with a thickened outer periclinal wall were observed in G. marianae and G. oenanthemum, respectively (Figure 4d,e). The cortex was formed by numerous layers of sclerenchyma with or without lignification in the outermost layer (Figure 4e). The dissociation of the spine showed that the sclerenchymal tissue is formed by fibers (Figure 4f), macrosclerids, and segments of short vessels with areolate pits in G. marianae (Figure 4g) and helical thickening in G. oenanthemum (Figure 4h).

Phenolic Compounds
Total phenol compound content was higher in stems than in roots and spines in both plant species; the content of these compounds was higher in ethanolic extracts than in aqueous extracts ( Figure 5). In spine cross-sections, both species showed a circular shape with an epidermis made up of one cellular layer and a smooth cuticle. Papillose epidermises without and with a thickened outer periclinal wall were observed in G. marianae and G. oenanthemum, respectively (Figure 4d,e). The cortex was formed by numerous layers of sclerenchyma with or without lignification in the outermost layer (Figure 4e). The dissociation of the spine showed that the sclerenchymal tissue is formed by fibers (Figure 4f), macrosclerids, and segments of short vessels with areolate pits in G. marianae (Figure 4g) and helical thickening in G. oenanthemum (Figure 4h).

Phenolic Compounds
Total phenol compound content was higher in stems than in roots and spines in both plant species; the content of these compounds was higher in ethanolic extracts than in aqueous extracts ( Figure 5). Stem phenolic compound contents in G. oenanthemum were 3.07 mg GAE/g DW (ethanolic extract) and 3.55 mg GAE/g DW (aqueous extract); in G. marianae, however, the phenolic content was lower than in G. oenanthemum (32 and 23%, respectively) ( Figure 5a). The total phenolic compound content of ethanolic extracts from spines was 37% higher in G. oenanthemum compared to G. marianae (Figure 5c). The G. marianae root ethanolic extract showed a higher content of total phenolic compounds than the G. oenanthemum root extract by 42% (Figure 5e). Stem phenolic compound contents in G. oenanthemum were 3.07 mg GAE/g DW (ethanolic extract) and 3.55 mg GAE/g DW (aqueous extract); in G. marianae, however, the phenolic content was lower than in G. oenanthemum (32 and 23%, respectively) ( Figure 5a). The total phenolic compound content of ethanolic extracts from spines was 37% higher in G. oenanthemum compared to G. marianae (Figure 5c). The G. marianae root ethanolic extract showed a higher content of total phenolic compounds than the G. oenanthemum root extract by 42% (Figure 5e).
The flavonoid content was lower in the spines and roots than in the stem. However, the flavonoid content was higher in G. marianae stem than in G. oenanthemum stem (Figure 5b).

Discussion
The present work showed adaptive and acclimation strategies in G. marianae and G. oenanthemum in arid regions. The presence of ectomycorrhizae was demonstrated for both Gymnocalycium species, both in those grown in greenhouses and in those collected from their natural habitats. Ectomycorrhizae were reported in some members of the families Pinaceae, Fagaceae, Salicaceae, Betulaceae, Myrtaceae, and Fabaceae, but this is the first report in Cactaceae [7,16]. Up to date, only arbuscular mycorrhizal fungi associations have been found in species of Cactaceae of the genera Opuntia, Trichocereus, Gymnocalycium, and Acanthocalycium [9], and Pachycereus, Machaerocereus, and Lemaireocereus [8]. Ectomycorrhizae (basidiomycete and ascomycete) have evolved independently from saprotrophic ancestors and retained some of their functional genes, thus leading to differences in the mode and capacity of organic matter decay [17].
The association root-mycorrhizae is central for plant nutrition and water capture, as well as environmental stress responses in some plant species [18][19][20]. Thus, the ectomycorrhizal symbiosis detected in the species of Gymnocalycium could result in an adaptation mechanism for tolerance to xerophytic environments.
The fungal colonization was higher in G. marianae than in G. oenanthemum, a fact that can be linked to the higher root biomass and the higher phenolic compound content. The increase of these compounds was also observed in the interaction between silver birch and ectomycorrhizae [21]. The higher root biomass of G. marianae reveals a greater sink force of photosynthates, which could be a consequence of the symbiotic association. This, in turn, can be linked with the higher development of photosynthetic parenchyma in this plant, i.e., number of cell layers, higher density of stomata, and photosynthetic pigment content, in comparison to G. oenanthemum. High stomatal density in G. marianae was also reported by Perotti et al. [4]. Moreover, other differential traits of G. marianae have been detected in the stem: cuticle thickness, subepidermal collenchyma with thick cell walls, and major carotenoids and flavonoid content could protect the stem against winds, desiccation, and high solar radiation.
A very interesting finding in G. marianae was the presence of substomatal chambers of great depth, which extend to the fifth to sixth layer of parenchyma. This fact would correspond to a particular adaptive strategy for water conservation in xeric environments. It has been reported that substomatal chambers cross only up to the "hypodermis" in species of Cipocereus (Cactaceae) [12] and extend into the first layer of the cortical parenchyma in Opuntia species [22]. The control of guard cell opening and closing, as well as the depth of the substomatal chambers, is of great importance to the efficiency of water use. Guard cell openings in CAM plants appear to be unresponsive to blue light; instead, they present a strong sensitivity to the vapor pressure deficit (VPD) in the plant-atmosphere system [23]. In this sense, the participation of the very deep substomatal chambers of G. marianae could play a fundamental role in the regulation of VPD.
Water conservation is also associated with the presence of mucilage-containing idioblasts and the numerous mucilaginous fibers detected in both species. The function of mucilaginous fibers in the xylem tissue of stems and roots with secondary growth has not been thoroughly studied to date. Thus, they would participate in regulating the flow of water through the plant by changes in mucilage hydric potential and could absorb and retain water when the stomata are closed and release the water towards the stomata when they are open, i.e., during the night or by VPD regulation. Thus, one of the proposals of the present work is to consider mucilaginous fibers as adaptive structures required for the flow of water through xylem in CAM plants.
The G. oenanthemum spines showed major biomass, thicker cell walls, i.e., epidermis and subepidermis, and higher phenolic compound content than those of G. marianae, a fact that could be associated with defense against herbivores. Vessels with helical thickening were present in the sclerenchymatic tissue of G. oenanthemum spines. This could be related to the flexibility required during the formation of the curved spines. G. marianae showed vessels with areolate pits and straight spines. Helical thickenings present in tracheids are known to provide more organ flexibility compared to tracheids with areolate pits [24]. The helical thickening or areolate pits found in the vessels, characters associated with the morphology of the spines, could indicate that both species have different adaptations for defense against herbivores or other environmental factors.
The papillae on the G. oenanthemum stem surface, as well as the trichomes found between ribs and at the base of the spines in both species, could help retain and take atmospheric water inside the plant. Papillae are rare in cacti, except in species of Pterocactus, and they are considered an adaptation to arid environments [25,26].

Habitat Traits of Species under Study
G. marianae grows in Andalgalá, Catamarca, Argentina (27 • 40 S, 66 • 01 W), between 1600 and 1800 m.a.s.l. The plant population is located on the northern slope of Sierra de Ambato, on undulating hills exposed to both water and wind erosion. The soil is composed of coarse sediment and rocky outcrops of metamorphic origin. The mean annual precipitation is around 380 mm, with a marked dry season from May to September, with less than 10 mm per month. The rains are concentrated in the warm season, i.e., January to March. The mean annual temperature is around 9.5 • C. Frosts occur from April to October-November [27]. G. marianae grows in an ecotone of mount and prepuna with scarce vegetation. Other plants found in the region are Prosopis nigra (algarrobo negro) and, less frequently, Condalia microphila (piquillín pispa) and Larrea cuneifolia (jarilla macho) (Figure 6a,b). The G. marianae habitat region was inhabited by the Incas until the mid-16th century, as evidenced by archaeological pieces [28] found in the area.
tations for defense against herbivores or other environmental factors.
The papillae on the G. oenanthemum stem surface, as well as the trichomes found between ribs and at the base of the spines in both species, could help retain and take atmospheric water inside the plant. Papillae are rare in cacti, except in species of Pterocactus, and they are considered an adaptation to arid environments [25,26].

Habitat Traits of Species under Study
G. marianae grows in Andalgalá, Catamarca, Argentina (27°40′ S, 66°01′ W), between 1600 and 1800 m.a.s.l. The plant population is located on the northern slope of Sierra de Ambato, on undulating hills exposed to both water and wind erosion. The soil is composed of coarse sediment and rocky outcrops of metamorphic origin. The mean annual precipitation is around 380 mm, with a marked dry season from May to September, with less than 10 mm per month. The rains are concentrated in the warm season, i.e., January to March. The mean annual temperature is around 9.5 °C. Frosts occur from April to October-November [27]. G. marianae grows in an ecotone of mount and prepuna with scarce vegetation. Other plants found in the region are Prosopis nigra (algarrobo negro) and, less frequently, Condalia microphila (piquillín pispa) and Larrea cuneifolia (jarilla macho) ( Figure  6a,b). The G. marianae habitat region was inhabited by the Incas until the mid-16th century, as evidenced by archaeological pieces [28] found in the area.
G. oenanthemum grows in Ambato, San Fernando, and Capayán, three districts in Catamarca between 500 and 1000 m.a.s.l. The plant material was collected from the east slope of Sierra de Ambato, Ambato (28°16′ S, 65°52′ W). The soil is rocky and of metamorphic origin, with abundant superficial organic matter. The mean annual precipitation is about 670 mm, with rainy and dry periods from October to March and from April to September, respectively. The mean annual temperature is 15 °C, with frosts occurring from April to October [29]. The ecoregion corresponds to Chacoan grassland, with abundant herbaceous vegetation and stray cattle. The plants are either exposed to full sunlight or protected by shrubs, coarse grasses (Stipa sps.) and trees such as Vachellia caven (churqui) (Figure 6c,d). G. oenanthemum grows in Ambato, San Fernando, and Capayán, three districts in Catamarca between 500 and 1000 m.a.s.l. The plant material was collected from the east slope of Sierra de Ambato, Ambato (28 • 16 S, 65 • 52 W). The soil is rocky and of metamorphic origin, with abundant superficial organic matter. The mean annual precipitation is about 670 mm, with rainy and dry periods from October to March and from April to September, respectively. The mean annual temperature is 15 • C, with frosts occurring from April to October [29]. The ecoregion corresponds to Chacoan grassland, with abundant herbaceous vegetation and stray cattle. The plants are either exposed to full sunlight or protected by shrubs, coarse grasses (Stipa sps.) and trees such as Vachellia caven (churqui) (Figure 6c,d).

Plant Material
Ripe fruits of 12 plants of both G. marianae and G. oenanthemum were collected from their natural habitats in Andalgalá and Ambato, respectively, during January and February 2017 (Figure 1a,b). The seeds were extracted from the fruits, washed, dried, and stored at room temperature. The seeds were germinated and grown for one year in plastic trays with commercial organic substrate "Bertinat" and kept under greenhouse conditions; temperatures in the greenhouse ranged between 10 and 35 • C and relative humidity between 40 and 60% in winter and summer. Then, the plants were transferred to a speeding tray containing peat for them to continue growing for up to 3.5 years. Then, 25 plants of each species were randomly collected, and the roots were carefully washed. These plants, which were used for all the assays, were freeze-dried and powdered for phytochemical analysis. The fresh plants were used for morphological characterization.
In order to examine whether the plants grown in the habitat present mycorrhizal symbioses, an independent assay was carried out. For this purpose, roots were collected from 5 individuals of both G. marianae and G. oenanthemum coming from Andalgalá and Ambato, respectively. FAA (formaldehyde: acetic acid: 50% ethanol, 5:5:90, v:v:v) was used to fix the roots until the assays were carried out.

Morphological Characterization, Photosynthetic Pigments, and Biomass
The diameter and height of the stem, as well as the root length, were recorded using a digital caliper.
Chlorophyll and carotenoids were extracted from 30 mg of fresh stem with 2 mL of dimethyl sulfoxide for 12 h at 45 • C. The chlorophyll a (Chl a), chlorophyll b (Chl b), and carotenoid contents were calculated from absorbances at 665, 649, and 480 nm according to Wellburn [30] and expressed as µg g −1 FW.
The root, stem, and spine were freeze-dried, and both the fresh and dry weights (FW and DW) were determined. DW/FW ratio, water content, and DW distribution percentage (DWD) [(∑ DW root, stem, and spine/DW each organ) × 100] were calculated.

Ectomycorrhizae Morphology
The morphological types of ectomycorrhizal roots were determined according to Agerer [31] from habitat and greenhouse grown plants. The number of ectomycorrhizal branches in lateral roots was evaluated through observations using a stereomicroscope (Olympus Co, Olympus SZX7, Tokyo, Japan).

Anatomical Analysis
The anatomical analysis was performed on fresh plant material, namely roots, stems, and spines, fixed in FAA. Free-hand cross-sections were made from the equatorial zone of the stem and spine and serial sections from all the radical systems. To characterize the anatomical structure, a successive double astra blue-safranin stain was used and subsequently mounted in water-glycerine (1:1, v/v) [32,33]. Spine dissociation was performed using the Schultze technique [32].
Stoma types were determined by using the Dilcher classification [34]. The length, width, and density of stomata were quantified.
The observations were made by using a stereomicroscope (Olympus Co, Olympus SZX7, Tokyo, Japan) and a light microscope (Carl Zeiss, Axiostar Plus, Göttingen, Germany) coupled to a digital camera (Canon A620, Power Shot 7.1 MP).
The slides analyzed were deposited at the Institute of Plant Morphology of Miguel Lillo Foundation, Tucumán, Argentina.

Transmission (TEM) and Scanning (SEM) Electron Microscopy
The ultrathin sections of the root were mounted on copper grids and contrasted with uranyl acetate and lead citrate [37] for observation with a transmission electron microscope (Carl Zeiss, EM109, NTS GmbH, Oberkochen, Germany).
Stem surface micromorphology was analyzed. The fragments fixed in Karnovsky's solution were dehydrated with alcohol and acetone. Next, the preparations were critical-point dried in liquid CO 2 , sputter-coated with gold-palladium, and observed with SEM (Zeiss Supra 55VP, Carl Zeiss, Oberkochen, Germany). SEM micrographs were taken with a digital camera (Olympus SP350, 8.0 MP, Tokyo, Japan). SEM and TEM studies were performed at Centro Integral de Microscopía Electrónica (CIME-CONICET, Tucumán, Argentina).

Soluble Phenolic Compound Determination
The samples were freeze-dried and ground. The soluble phenolic compounds were extracted in ethanol 80 • and water for 1 h at room temperature by using different DW/solvent ratios, as follows: 1:4 for stem and 1:12 for spine and root in ethanolic extracts, and 1:18 for stem, 1:12 for spine, and 1:13 for root in aqueous extracts. Total phenol content was determined according to Singleton et al. [38]. Results were expressed as mg gallic acid equivalents/g DW (mg GAE/g DW). Total flavonoids were estimated by Woisky and Salatino [39]. Results were expressed as mg of quercetin equivalents/g DW (mg QE/g DW).

Statistical Analysis
Statistical differences in morphological features (n = 20), biomass (n = 20), pigment photosynthetic content (n = 5), and phenolic content (n = 5) were determined by analysis of variance (ANOVA) followed by Tukey's test (p < 0.01). The anatomical analyses were performed using five plants from each species. The stomata density was quantified for n = 10 fields at × 40 objective lenses, with three replications for each one. Both the length and width of the stomata were evaluated (n = 50). The density and size of stomata and the number of ectomycorrhizal structures were expressed as mean ± standard deviation and compared by using Student's t tests.

Conclusions
The present study shows a detailed anatomical description strongly correlated with physiological aspects of two Gymnocalycium species. Different adaptive strategies are found in Gymnocalycium species that allow them to survive in xeric environments. To our knowledge, this is the first report on the presence of ectomycorrhiza in Cactaceae. It is essential to further the study of adaptation mechanisms of plants resistant to extreme environments, taking into account the global changes of the planet and the risk of extinction of plant species due to human activities.