3.1 Polyploidy and the genomic consequences in Agave
Many authors have described that the increase of genome size in polyploid species is accompanied by genetic changes, resulting in greater variability and genetic diversity in polyploid plants (Chen 2010; Wendel 2000; Moghe and Shiu 2014; Tamayo-Ordoñez et al. 2016a). The dendrogram obtained from the AFLP marker showed two clusters, cluster I having two subgroups denominated as A and B (Fig. 1). The subgroup A showed the clustering of hexaploid and diploid accessions of Agave angustifolia (AAM1-RO, AAM2-RO, AAM3-RO, AAC1-RO, AAC2-RO, AAC3-RO, CHA1-PCTY, CHA5-PCTY and CHA6-PCTY), and the subgroup B included the diploid accessions of A. tequilana (AT1-RO, AT2-RO, AT3-RO, 1444-2-PCTY, 1444-3-PCTY, and 0345b-PCTY; Fig. 1). Cluster II grouped the triploid and pentaploid accessions of A. fourcroydes. According to the global variability, the AFLP marker indicated a closer genetic proximity between A. angustifolia and A. tequilana.
The analysis of similarity indexes and percentages of polymorphism in accessions belonging to the same species (considering ploidy level as an important factor) indicated that Agave accessions containing polyploid varieties had low similarity indexes (0.76) compared to the results from the analysis including only diploid accessions like A. tequilana (2n = 2x = 60) and A. angustifolia (2n = 2x = 60) (0.83). The analysis of A. angustifolia (2n = 2x = 60) and A. angustifolia (2n = 6x = 180) showed percentages of polymorphism of 75 and a similarity index of 0.76, while accessions of A. fourcroydes (2n = 3x = 90) and A. fourcroydes (2n = 5x = 150) showed percentages of polymorphism of 79 and a similarity index of 0.73 (Table 1). Apparently, polyploidy does have an effect on genome size and genetic variability in Agave.
The analysis of correlation of alleles indicated that, the accessions of A. tequilana cultivated in the RO-CICY share a greater proportion of alleles, similar in the same proportion as A. angustifolia cultivated in both microclimates (RO-CICY and GB-PCTY). Likewise, the varieties of A. fourcroydes apparently contain a narrow genetic germplasm, reflected in the conservation of genetic material (Fig. 2).
Table 1. Polymorphism and index of genetic variation of Agave L.
|
Especie
|
Total number of individuals
|
Accessions (number of individuals evaluated by each accession)*
|
Total numbers of loci
|
Total numbers of polymorphic loci
|
Total numbers of common loci
|
Polymorphic percentageƗ
|
Index of similarity
|
Agave spp.
|
33
|
ATA-RO (3), AAM-RO (3), AAC-RO(3), AFK-RO(3), AFS-RO (3), AFY-RO (3), 1444-PCTY (2), 345b-PCTY (1), CHA-PCTY(3), KK-PCTY (3), SK-PCTY(3) and YK-PCTY (3)
|
201
|
182
|
19
|
90.54
|
0.66
|
Agave tequilana Weber. (2n = 2x = 60) and Agave angustifolia Haw. (2n = 2x = 60)
|
9
|
ATA-RO (3), AAM-RO (3), 1444-PCTY (2), and 345b-PCTY (1)
|
73
|
33
|
40
|
45.20
|
0.83
|
Agave tequilana Weber. (2n = 2x= 60)
|
6
|
ATA-RO (3), 1444-PCTY (2) and 345b-PCTY (1)
|
44
|
20
|
28
|
45.45
|
0.85
|
Agave angustifolia Haw. (2n = 2x= 60) and Agave angustifolia Haw. (2n = 6x= 180)
|
9
|
AAM-RO (3), AAC-RO(3) and CHA-PCTY(3)
|
82
|
62
|
20
|
75.60
|
0.76
|
Agave fourcroydes Lem. (2n = 3x= 90) and Agave fourcroydes Lem. (2n = 5x= 150)
|
12
|
AFK-RO(3), AFS-RO (3), AFY-RO (3), KK-PCTY (3), SK-PCTY(3) and YK-PCTY (3)
|
142
|
113
|
29
|
79.57
|
0.73
|
Agave tequilana Weber. ‘Azul’ (2n = 2x = 60)
|
6
|
ATA-RO (3), 1444-PCTY (2) and 345b-PCTY (1)
|
48
|
20
|
28
|
41.66
|
0.85
|
Agave angustifolia Haw ‘Marginata’ (2n = 2x = 60)
|
3
|
AAM-RO (3)
|
68
|
33
|
35
|
48.52
|
0.81
|
Agave angustifolia Haw. ‘Chelem ki’ (2n = 6x = 180)
|
6
|
AAC-RO(3) and CHA-PCTY(3)
|
63
|
38
|
25
|
60.31
|
0.77
|
Agave fourcroydes Lem. ‘Kitam ki’ (2n = 3x= 90)
|
6
|
AFK-RO(3) and KK-PCTY (3)
|
103
|
51
|
52
|
49.51
|
0.81
|
Agave fourcroydes Lem. ‘Sacki ki’ (2n = 5x= 150)
|
6
|
AFS-RO (3) and SK-PCTY(3)
|
139
|
82
|
49
|
58.99
|
0.74
|
Agave fourcroydes Lem. ‘Yaax ki’ (2n = 5x= 150)
|
6
|
AFY-RO (3) and YK-PCTY (3)
|
101
|
57
|
44
|
56.43
|
0.77
|
*The specifications of the collections are described in the table S1. ƗPolymorphic percentage= (Total numbers of polymorphic loci/ Total numbers of loci)*100
Polyploidy is a very common phenomenon in species of angiosperms and vascular plants (Moghe and Shiu 2014; Wendel 2000). Nowadays, the ancestors of many polyploid plant species, like those in Agave, remain to be unknown. The genus Agave includes triploid (e.g., A. fourcroydes ‘Kitam ki’ 2n=3x=90) to octoploid species (e.g., A. datylio 2n=8x=240; Castorena-Sánchez, 1990), however, despite having species with a wide range of ploidy level, most studies –conducted in few species of the genus– have focused on cytogenetic characterization. In addition, it is suggested that due to the presence of bimodal karyotype with a basic chromosome number of n=30 (5 long acrocentric chromosomes and 25 small metacentric or submetacentric chromosomes; Moreno-Salazar et al. 2007; McKain et al. 2012; Palomino et al. 2015) the genus is possibly of allopolyploid origin (McKain et al. 2012).
Allopolyploidy can be associated with an increased number of gene copies and, therefore, it implies the generation of redundant genes. These genes, mainly generated by duplication, are involved in epigenetic regulation, become specialized to perform a complementary function or may eventually be lost. The functional partitioning and differences in expression patterns of these redundant genes are associated to changes (physiological and phenotypic) (Tamayo-Ordoñez et al. 2016) that confer advantages in allopolyploid species.
Otherwise, the genus Agave is of recent origin (6-8 Mya and 1.5-3 Mya), has a high index of species diversity (0.32 to 0.56 species per million years) relative to angiosperms in general (0.089-0.07 species per million years), and according to the model for evolution of rDNA regions proposed by Tamayo-Ordoñez et al. (2018), it is possible that certain genomic regions are undergoing an evolutionary stage of fixation of certain genes copies. This evolutionary process suggests that genes are subjected to genetic events including recombination, amplification, duplication, transposition, and gene loss that could have resulted in the high variability and diversity observed in the Agave genome.
The use in Agave of ISSR (Vargas-Ponce et al. 2009; Aguirre-Dugua and Eguiarte 2013), AFLP (Sánchez-Teyer et al. 2009) and SSAP (Bousios et al. 2007) molecular markers has indicated variability and genetic diversity between wild and cultivated populations of A. angustifolia, A. tequilana, and A. fourcroydes, among other species. The Agave accessions we analyzed showed high genetic variability. The accessions with the highest level of ploidy –like A. angustifolia ‘Chelem ki’ (2n = 6x = 180), A. fourcroydes ‘Sack ki’ (2n = 5x= 150), and A. fourcroydes ‘Yaax ki’ (2n = 5x= 150)– showed the lowest similarity indexes (<78) in comparison to the other analyzed accessions. Apparently, polyploidy has an effect on genome size and genetic variability in Agave, and the presence of these polyploid species within the genus could be a factor that contributes to the high species diversity index (0.32 to 0.56 species per million years) of the genus. Also, according to Good-Avila et al. (2006), the process of speciation in Agave has been coincident with increasing aridity in central Mexico, which suggests that high retrotransposition activity in response to water deficit may have had an important role in speciation. Tamayo-Ordoñez et al. (2016b) found that in some species of Agave there is differential regulation of genes associated with biotic and abiotic stress factors depending on their habitat and proposed that stressing environmental factors could have contributed to gene diversity that expresses as speciation and species’ adaptation.
Cultivation of some species of Agave since pre-Columbian time (Casas et al. 2016) and the involved pressure from anthropic domestication processes, artificial selection, and intensive cultivation has resulted in its species having different degrees of domestication (Colunga-GarcíaMarín et al. 2004). But, unfortunately, until now only a handful of species have been studied with the goal of obtaining evidence of their domestication and management processes. Studies made of the rDNA, NBS-LRR, and LEA genes in cultivated polyploid species of Agave have revealed that A. tequilana ‘Azul’ (2n = 2x = 60) –used for production of tequila– and A. fourcroydes ‘Kitam ki’ (2n = 3x = 90) –used for fiber production– have variants of these genes that could be involved in gene evolution, rRNA functionality, and activity of proteins participating in defense responses. This strongly suggests that anthropic pressure might have had negative effects on their variability, genetic diversity, and response to abiotic and biotic stress.
Consequently, it must be emphasized that morphological, molecular, and physiological studies be made before attempting to use polyploid Agave plants in order to achieve their sustainable use and to avoid negative consequences on the variability and genetic diversity of the genus. Also, based on the possible allopolyploid origin of Agave, its genome may contain valuable information that could be analyzed and used for genetic improvement of economical important crops exposed to biotic and abiotic stresses induced by global climate change (Tamayo-Ordoñez et al., 2018).
3.2 Implications of the Conservation of NADH, RbcL, PEPC and PEPCK enzymes in Agave
In CAM and C4 plants, the activity of key enzymes such as PEPC, PPDK, NAD (P) –ME, and PEPCK is much higher than in C3 plants. Even so, the cellular compartmentalization of said enzymes differs between CAM and C4 plants. CAM plants have evolved specific diurnal and nocturnal patterns of expression and regulation to accommodate the flux of carbon through gluconeogenesis and glycolysis necessary to meet the nocturnal demand for PEP and diurnal decarboxylation of four-carbon organic acids. The characterization of certain specific CAM genes (PEPC and PEPCK) that are related to the cyclic electron transport and chlororespiration (NADH), and to enzymes related to carbon fixation (RbcL) and their expression patterns in CAM plants, could clarify the molecular mechanisms subjected to evolution and expression of said enzymes.
In Agave, the basic enzymes and metabolites necessary for the optional functioning of CAM are not yet known, so in this work we isolated partial regions of the enzymes and studied their phylogenetic relationships with members of the classes Liliopsida and Magnoliopsida. Analysis of aa substitutions found in the sequence of the enzymes NADH, rbcL, PEPC, and PEPCK and its impact in the conformation of the 3D structures were included.
Phylogenetic analyses indicated that the NADH enzyme sequences from Agave spp. were more closely related with the families Asparagaceae and Agavoidaceae. The genus Allium, representative of the Amaryllidaceae family, formed a separate group (Fig. 3A). In the phylogenetic analysis of the RbcL, PEPC, and PEPCK enzymes we included accessions belonging to the class Liliopsida and Magnoliopsida. Sequences close to the family Agavaceae were not included due to the scarce information available. RbcL of Agave grouped with species in genera belonging to the class Liliopsida. Agave accessions were more closely related with accessions from the genus Oryza (Fig. 3B). The genera belonging to the class Magnoliopsida, formed a separate group.
PEPC of Agave grouped with members of the order Poales (Oryza, Aegilops, Tillandsia, and Ananas). Some plant accessions belonging to genera in the class Liliopsida (Asparagus, Oryza, Setaria, Aegilops, Hordeum, and Triticum) grouped with members of genera in the class Magnoliopsida (Herrania, Theobroma, Gossypium, Jatropha, Manihot, Ricinus, Populus, Curcubita, and Arabidopsis; Fig. 4A). Presence of isoforms of the PEPC enzyme in C3 and C4 plants has been widely described in plants (Paulus et al. 2013), and it is possible that the formation of two different groups including aa sequences of PEPC from Oryza and Aegilops is a consequence of the presence of PEPC isoforms in these genus. For this part, PEPCK sequences of Agave that grouped with members of genera from the class Liliopsida (Elaeis, Phoenix, Musa, Panicum, Sorghum and Asparagus), were more closely related with the genus Asparagus. Members of genera in the class Magnoliopsida formed a separate group (Fig. 4B).
The chloroplast NADH dehydrogenase-like complex (NDH) is comprised of many subunits. The plastid genomes of flowering plants also have 11 genes (ndhA–ndhK; Ifuku et al., 2011); a subunit specific to photosynthetic NDH is NdhL (dehydrogenase subunit L). In this study we identified a partial region of dehydrogenase subunit L in Agave that showed high conservation when compared with other members of the Asparagaceae family. Analysis of aa substitutions found in NADH enzyme from the Agavaceae and Amaryllidaceae families, indicated that Agavaceae presents 2% of substitutions from V (117) to L, 12% from V (298) to A, and 12% from A (301) to V (Table S4). In the family Amaryllidaceae we found 33% of substitution from V (117) to I, 33% from V (298) to I, and 33% from A (301) to E (Table S4). Analysis of the effect of aa substitutions on 3D structures indicated that these changes do not affect the tertiary structure of the NADH enzyme. Our three-dimensional (3D) models of NADH dehydrogenase of the families Asparagaceae and Amaryllidaceae had a sequence identity of >40% with NADH- quinone-oxidoreductase subunit L (Fig. 5 and Table S5). 3D models of NdhL (320aa) from the families Asparagaceae (Agave) and Amaryllidaceae (Allium) showed structures similar with the NADH-quinone-oxidoreductase subunit L described in E. coli (Efremov and Sazanov 2011) (Fig. 5). In both families, it was possible to identify the formation of nine alpha helices (Fig. 5B-D and Table S5).
This results, suggesting that in the analyzed plant accessions of Agave with eight million years of evolution (Tamayo-Ordoñez et al. 2018) the dehydrogenase subunit L enzyme presents structural conservation. The importance of conservation in this enzyme, lies in that the chloroplast NADH dehydrogenase-like (NDH) complex mediates cyclic electron transport and chlororespiration, which in angiosperms further associates with photosystem I (PSI) to form a super-complex (Yamori et al. 2015).
For this part, in our analysis of point mutations of PEPC enzyme, total aa substitutions (100% change) were detected in 22 aa residues from Liliopsida and in 28 aa residues from Magnoliopsida. This result indicated that 18% of the PEPC enzyme sequence analyzed differed between classes (Table S7). These differences found in the aa substitutions affect the grouping support between the aa sequences of the PEPC enzyme in the classes Liliopsida and Magnoliopsida, which was previously reflected in the phylogenetic tree of this enzyme (Fig. 4A). In Agave, The conservation of A and R in the substrate binding ((PWIF(A/S)WTQR) and inhibitory site ((DLLEGDPYLKQ(R/G)IRLRDSYIT)) of PEPC enzyme, was 100%, which suggests according to Paulus et al. (2013), possibly, it is of the C3-type in Agave and other members of to the classes Magnoliopsida and Liliopsida. Also, 3D models of PEPC from Liliopsida (genera Agave and Ananas) and Magnoliopsida (genera Arabidopsis and Gossypium) showed similar structures (identity >85% with the phosphoenolpyruvate carboxylase (PEPC) of A. thaliana model) (Fig. 6 and Table S5); in both classes it was possible to identify the formation of 11 alpha helices (Fig. 6B-D and Table S5).
3D models indicated that the PEPCK from Agave had a sequence identity >45% with the phosphoenolpyruvate carboxykinase (PEPCK) described for the E. coli model (Fig. 7A and Table S5). 3D models of PEPCK from the classes Liliopsida (genera Agave and Asparagus) and Magnoliopsida (genera Arabidopsis, Brassica, and Arachis) showed similar structures between them, and it was possible to identify the conservation of aa residues that allow the union of Mg+, Ca2+, and pyruvic acid –important for the function of the enzyme (Sudom et al. 2013; Fig. 7B-F and Table S5). The ATP binding site could not be identified in the plant accessions that we analyzed.
C4 photosynthesis has evolved independently more than 62 times, including 7500 species in 19 families, or 3% of the flowering plant species (Deng et al. 2016). Phosphoenolpyruvate carboxylase (PEPC; EC 4.1.1.31) has shown plays a key role in the carbon metabolism of C4 and CAM plants, and markedly improves photosynthetic efficiency and water use efficiency (Driever and Kromdijk 2013). Deng et al. (2016) analyzed 60 available published plant genomes and delimited that the PEPC family consists of three distinct subfamilies (PPC-1, PPC-2, and PPC-3). The monocot CAM –or C4-related PEPC– originated from the PPC-1M1 clade and WGD may increase the number of copies of the PEPC gene, suggesting the formation of more isoforms of the PEPC in Plants CAM (Fan et al., 2013; O’Leary et al., 2011). It is possible that the broad amino acid changes (18%) found between the classes Magnoliopsida and Liliopsida, reported in this study, may be due to the presence of isoforms in the accessions representative of each class.
D-Ribulose 1,5-bisphosphate carboxylase/oxygenase (EC 4.1.1.39; Rubisco) catalyzes the initial steps of photosynthetic carbon reduction and photorespiratory carbon oxidation (Bracher et al. 2017). Land plants, have hexadecameric type I rubisco with an approximate molecular mass of 550 kDa composed of eight large (L;55 kDa) and eight small (S;15 kDa) subunits (L8S8). A. thaliana produces four distinct small-subunit isoforms (RbcS1A, RbcS1B, RbcS2B and RbcS3B) (Valegård et al. 2018). In the analysis of RbcL enzyme, it was possible to identify unique substitutions between members of the classes Liliopsida and Magnoliopsida. Total aa substitutions (100% change) were detected in aa residues such as such as S/Q (1), R/K(22), I/V (37), Q/E (68), and S/M (Table S6). Our 3D models indicated that the RbcL from Agave had a sequence identity of 99% with the ribose bisphosphate carboxylase small chain (Fig. 8A and Table S5) described in the O. sativa model (Matsumura et al. 2012). 3D models of RbcL from genera in the classes Liliopsida (Oryza, Agave, and Aegilops) and Magnoliopsida (Tragopogon and Arachis) showed similar structures. In both classes, we were able to identify the formation of four alpha helices and two beta folds (Fig. 8B-D and Table S5). In Aegilops and Tragopogon, we observed that the length of the first alpha helix was lower compared to the reference model (Matsumura et al. 2012). The differences found between these tertiary structures can be derived from the presence of different isoforms that code for the small subunit of RbcL, as it has been described A. thaliana (Valegård et al. 2018).
Because Rubisco catalyzes the rate-limiting step of C3-photosynthesis, many studies have been carried out for future improvement of Rubisco activity by genetic modification to increase the productivity of crop plants (Borland et al. 2011; 2014). The regulation of rubisco activity goes hand in hand with a negative coordination of PEPC regulation, suggesting a complex coregulation of both carboxylases compete for CO2 during the early morning hours. According to the results obtained, it seems that the enzymes rbcL, PEPCK, PEPC and NADH are conserved between the classes Liliopsid and Magnoliopsida, however the presence of amino acid changes (>18%) in the enzymes PEPC and rbcL, could indicate the presence of isoforms, and genes that could be functionally specialized to better perform their catalytic activity. The participation of these two enzymes is related to photosynthetic efficiency and water use efficiency, suggesting that they are in constant evolutionary change according to the conditions of water limitation, high temperatures, increase of CO2 as a result of global warming, which we are currently experiencing.
It is important to mention that the grouping of Agave with monocots in the order Poales, suggests that the genus possibly has genetic material that allows it to tolerate environments where abiotic stresses are extreme, as described for that order (Linder and Rudall 2005). Bouchenak-Khelladi et al. (2014) described that CO2-concentrating mechanisms counteract the effects of low atmospheric CO2 and reduce phototranspiration. It is believed that the parallel evolution of C4 and CAM photosynthesis in Poaceae, Cyperaceae, and Bromeliaceae is an adaptation to changes in atmospheric CO2 concentrations. Combinations of extrinsic and intrinsic factors might have played a role in shifts in diversification rates and may explain the variation in species richness in Poales. In the genus Agave the richness in diversity is not yet fully known, so the physiological exploration of a larger number of species could help us to know if the changes in atmospheric CO2 concentrations during the evolution of the genus allowed the diversification of the C3, C4, and CAM responses.
3.3 Differences in stomata and transcriptional regulation of CAM-related genes in Agave L.
With the aim of assessing the physiological changes that could be observed in stomata and that could impact the dynamics of CO2 exchange and transpiration, we focused on determining the opening and closing of stomata by scanning electron microscopy in the two Agave species with different levels of ploidy (A. angustifolia and A. fourcroydes) at 19:00 h, 23:00 h, 3:00 h and 7:00 h. Differences in the opening and closing of stomata on the abaxial and adaxial leaf surfaces were observed (Fig. 9 and Table 2).
Table 2. Features of stomata in Agave L.
|
Characteristics*
|
A. angustifolia Haw. ‘Marginata’ (2n=2x=60)Ɨ▲
|
A. angustifolia Haw. ‘Chelem ki’ (2n=6x=180)Ɨ▲
|
A. fourcroydes Lem. ‘Kitam ki’ (2n=3x=90)Ɨ▲
|
A. fourcroydes Lem. ‘Sac ki’ (2n=5x=150)Ɨ▲
|
Leaf section
|
Adaxial/Abaxial
|
Adaxial/Abaxial
|
Adaxial/Abaxial
|
Adaxial/Abaxial
|
Stomata area (µM2)
|
997±56a/1161±32a
|
2080±240b/2059±167b
|
1106±66 a/1149±63a
|
1621±73b/1619±54b
|
Stomata density (number mm-2)*
|
58±3a/47±2a
|
32±2b/34±2b
|
45±1a/44±1a
|
36±2b/38±1b
|
Guard cells area (µM2)
|
19:00 PM: 853±6a/850±4a 23:00 PM: 859±38a/889±40a 3:00 AM: 1,161±50b/792±62a 7:00 AM: 954 ±18a/1,000±52a
|
19:00 PM: 1001±3a/1135±61b 23:00 PM: 1475±9d/1470±23d 3:00 AM: 1310±61d/1380±83d 7:00 AM: 1448±35d/1516±60d
|
19:00 PM: 915±33a/919a±44c 23:00 PM:769±3c/756±13c 3:00 AM: 873±18a/889±33a 7:00 AM: 887±30a/1002±30a
|
19:00 PM: 1071±1b/1254±46b 23:00 PM: 1140±64b/1179±39b 3:00 AM: 1342±2d/1152±18b 7:00 AM: 1051±26a/998±47a
|
Suprastomatic cavity area (µM2)
|
19:00 PM: 77±4a/330±17c 23:00 PM: 50±0a/317±33c 3:00 AM: 231±28c/306±7c 7:00 AM: 159 ±9b/ 216±20c
|
19:00 PM: 383±59c/510±17c 23:00 PM: 865±37d/1683±19e 3:00 AM: 450±9c/740±1d 7:00 AM: 630 ±23d/955±2c
|
19:00 PM: 101±3a/225±29c 23:00 PM: 291±15c/294±36c 3:00 AM: 364±8c/326±12c 7:00 AM: 181±3b/398±2c
|
19:00 PM: 555±43d/319±17c 23:00 PM: 390±9c/377±8c 3:00 AM: 350±28c/576±67d 7:00 AM: 170±4b*/153±4b*
|
ƗLowercase letters (a, b, c, d or e) indicate significantly different values (Student’s t test; p<0.05).
▲Values are averages of triplicates obtained from analysis of 30 stomata in 3 leaves of three plants of each accession ± standard error
*Values of suprastomatic cavity area (µM2) at 7:00AM were obtained from stomata that remained open and represent only 25% of the total stomata observed
Comparing different ploidy level accessions from the same species of Agave, those with higher ploidy level (A. angustifolia ‘Chelem ki’ and A. fourcroydes ‘Sac ki’) have a lower stomatal density (32±2 and 52±3) and their stomata are 1.5 to 2 times larger in comparison to their lower ploidy level counterparts (A. angustifolia ‘Marginata’ and A. fourcroydes ‘Kitam ki’) (Fig. 10 and Table 2). Analyses of guard cell area (μM2) showed larger size in polyploid plants of the species A. angustifolia 'Chelem ki' and A. fourcroydes 'Sac ki,' coinciding with stomata size (Fig. 9 and Table 2). Stomatal indices according to ploidy numbers have also been related to adaptation to stress (Balao et al. 2011; Jordan et al. 2015; Males and Griffiths 2017).
Analysis of suprastomatic cavity area (µM2) showed important changes between the polyploid species. A. angustifolia ‘Chelem ki’ showed a greater suprastomatic cavity area (1683±19 µM2) compared to A. angustifolia ‘Marginata’ (317±33 µM2) at 23:00 h (Fig. 11 and Table 2). Lower values of suprastomatic cavity area in A. angustifolia ‘Marginata’ were observed at 19:00 h and 23:00 h. Suprastomatic cavity area in the abaxial section of the leaf did not show significant differences along the temporal course, however, for A. angustifolia ‘Marginata,’ the suprastomatic cavity area was higher on the abaxial section of the leaf in comparison to the stomata located on the adaxial part of the leaf (Fig. 11).
In the hexaploid species A. angustifolia 'Chelem ki', high values of suprastomatic cavity area were observed at 23:00 h in both sections of the leaves (adaxial: 865±37 µM2 and abaxial: 1683±19 µM2; Table 2). The lowest values of suprastomatic cavity on the adaxial and abaxial surfaces were observed at 19:00 h. Similar to what we found in A. angustifolia 'Marginata,' the suprastomatic opening on the abaxial surface was always greater compared to the same on the adaxial surface.
Behavior of stomata opening in A. fourcroydes showed patterns different to those observed in A. angustifolia. The pentaploid species A. fourcroydes 'Sack ki' showed the highest values of suprastomatic cavity area at 3:00 h (576±67 μM2) and 19:00 h (555±43 μM2), both on the abaxial and on the adaxial surfaces. Suprastomatic cavity area was not as large as that reported in the hexaploid species A. angustifolia 'Chelem ki' (abaxial: 1683±19 μM2). In addition, in the adaxial leaf section, 75% of the stomata analyzed proved to be closed and the aperture of the remaining 25% indicated low values of suprastomatic cavity area (adaxial: 170±4 μM2 and abaxial: 153±4 μM2) (Table 2). Triploid A. fourcroydes ‘Kitam ki’ showed few significant differences along time, indicating that in this species the opening of the stomata is maintained longer; which is different to what was observed in A. angustifolia Haw 'Chelem ki', in which immediately after the stomatal opening at 23:00 h., suprastomatic cavity area decreases drastically (Fig. 9 and Table 2)
Relative expression of the PEPC gene showed significant differences, raising its expression at 3h and 23h (Fig. 12). The highest values in PEPC expression were identified at 23 h in the accessions cultivated in the BGR-RO (Fig. 12A), coinciding with the highest values of suprastomatic cavity area. Similar trends were found in the expression of the PEPCK enzyme, which also indicated higher values at 3h and 23h, in plants belonging to both microclimates (Fig. 13). Highest values of RbcL expression were observed at 15h for both plantations (BGR-RO and GB-PCTY; Fig. 14). Regarding NADH, there were no significant differences during the evaluated period (data not shown).