MORPHOLOGICAL AND STOMATAL CHARACTERIZATION OF Heliconia chartacea VAR. SEXY PINK INDUCED POLYPLOIDY CARACTERIZAÇÃO MORFOLÓGICA E ESTOMÁTICA DE Heliconia chartacea VAR. SEXY PINK INDUZIDAS A POLIPLOIDIA

: The growth of the tropical flower market has demanded a consistent search for new varieties, primarily those endowed with an exotic profile, but that are also beautiful and durable. The genus Heliconia , naturally found in the Amazon region, is among the most prominent of tropical flowers. Looking to augment the genetic variability available in Heliconia chartacea var. Sexy Pink, biotechnological research was conducted with the application of colchicine to induce polyploidy in plants from this species. With that in mind, this study was undertaken to evaluate the establishment of plants in the field drawn from in vitro polyploidy induction assay and to determine the morphological and physiological characteristics of 38 H. chartacea var. Sexy Pink clones. The characterization analyzes were performed through 49 morphological descriptors and a stomatal density evaluation using microscopy. The genotype 35 exhibited the greatest morphological variations, with alterations in the position and coloring of the inflorescence, in addition to having the edges of the entire limbus. Genotype 18 featured the lowest amounts for plant height and inflorescence size, showing promise for research geared towards use in reduced environments. Some genotypes did not have any flowering and are recommended exclusively for landscape composition such as foliage, since their exotic characteristics allow for this. The genotypes that were evaluated displayed stomata with tetracytic morphology and guard cells that had no significant changes. However, genotypes with greater equatorial diameter and stomatal density were obtained in relation to the mother-plant. Overall, the induction of polyploidy allowed for clones to be obtained with a high variability for the characteristics of the leaf, pseudostem and inflorescence, with various attributes that confer a more efficient post-harvest management to some genotypes, in addition to favorable aspects for commercialized purposes as a cut flower.


INTRODUCTION
The floriculture industry has one of the highest growth rates in national and international markets (JUNQUEIRA; PEETZ, 2014). The world floriculture production is growing at a rate of 10 per cent per year (HARISHA, 2017). Over the last few years, tropical flowers have been receiving particular attention due to the features that favor their scale production, such as beauty, exoticism, colors, shapes, resistance to transport and postharvest durability (CARVALHO et al., 2012;PRIA, 2017). Yet, consumers seeking the latest innovations have forced plant producers and breeders to increasingly release new varieties to keep up with the demand in this market. In this regard, genetic breeding programs have been presented as a tool for diversifying and improving existing plant materials, particularly through using methods that increase the available genetic variability.
The artificial induction of polyploidy has been an important method for creating new varieties of plant species, and this area has been gaining ground in the floriculture industry with the increasingly efficient use of mutagenic agents. One of the most widely used chemical agents is colchicine. Many varieties boasting "sales-friendly" features have been developed through their use. For example, there are plants with characteristics that are resistant to environmental adversities, like in Ziziphus mauritiana, which is resistant to water stress (NTULI; ZOBOLO, 2008), as well as new flower phenotypes and ornamental plants, such as Oncidium flexuosum (VICHIATO et al., 2007), Dendrobium nobile (UNEMOTO et al., 2009) and Heliconia bihail (CAVALCANTE FILHO, 2011).
An increase in the chromosome number is usually illustrated through a higher cellular content, which can be reflected in an growth in the size of the plant organs and in a shift of its morphologies (WU et al., 2010). Thus, the association between the level of ploidy and morphological characteristics aids in identifying polyploid plants. The morphological characterization still allows the almost immediate adoption of plants with phenotypes that are pertinent to the producers and breeding programs, or even for differentiating and identifying potential genotypes that are to be designated as varieties. Because it is an easy and simple methodology, stomatal analysis has also been used for identifying supposed polyploids through the counting and comparative measurement of stomata (GŁOWACKA et al., 2010). Various polyploidy induction activities in vegetables used this characteristic to select polyploid candidates, like in Centella asiática (KAENSAKSIRI et al., 2011) and Lagerstroemia indica (WANG; LEI, 2012).
Taking into account that part of the technologies applied in Brazil's flower production comes from other countries, research for tropical ornamental plants is still scarce, particularly with native species like Amazonian heliconias. These plants have an enormous potential for growth in the floriculture market, but the characterization of exotic germplasms and information on crop and post-harvest management deserves greater attention, even for more valuable plants like Heliconia chartacea var. Sexy Pink. Due to the need for ongoing research activities applied to the tropical flower and ornamental plants industry, Embrapa Amazônia Occidental coordinated a research project in 2009 titled "The development of technologies for heliconia production: a new niche market for the Amazon". The purpose of this project was to define criteria for the production of Heliconia chartacea var. Sexy Pink in the environmental conditions surrounding the city of Manaus in the State of Amazonas that involved agronomic and biotechnological studies. One of the main products it sought to generate was establishing a population of polyploid plants in a field that demonstrated early interesting and innovative characteristics for this species. Consequently, this project sought to characterize Heliconia chartacea var. Sexy Pink plants that are obtained through induced polyploidization for the morphological and physiological characteristics that are of commercial interest.
Therefore, this study was undertaken to evaluate the establishment of plants in the field drawn from in vitro polyploidy induction assay and to determine the morphological and physiological characteristics of 38 H. chartacea var. Sexy Pink clones.

MATERIAL AND METHODS
The genotypes used in the this study are linked to the collection of Heliconia chartacea var. Sexy Pink established in Embrapa Amazônia Ocidental's Experimental Field Headquarters in Manaus, Amazonas. These plants are derived from an induction of an in vitro polyploidy assay which consisted of the immersion of stem apices from a single mother-plant at different concentrations of colchicine (ranging from 0.01 to 0.10%).
After the mutation induction, the explants were cultured and multiplied in MS medium (MURASHIG;SKOOG, 1962) supplemented with 0.15 mg of AIA (indole-3-acetic acid) and 3 mg of BAB (6-benzylaminopurine) (RAIZER et al., 2017), in vitro rooted (QUISEN et al., 2013), followed by acclimatization in a greenhouse and planted in a field in March 2011. In May 2014, the collection with 38 genotypes was transferred to a new area divided by clumps and established with a spacing of 3.0m between rows and 1.5m between plants, arranged in random blocks with four replications per genotype. Fertilization and cultivation cleaning followed Embrapa's pre-established schedule and was based on recommendations for the cultivation of other tropical flowers (LUZ et al., 2005;SOUSA et al., 2009).
Monthly evaluations were conducted of morphological descriptors in the genotypes established in the field for two consecutive years, and the criteria used was based on forty seven qualitative descriptors and two quantitative descriptors (Table 1) LOGES et al., 2007). In order to have a more accurate identification of hue-related descriptors, a Munsell color chart was used. Visual assessments were made of the pseudo stem, leaves and inflorescences and the measurements were done with the aid of a digital caliper and tape measure in four plants from each genotype.
To visualize the differences among H. chartacea genotypes based on the qualitative and quantitative descriptors, a dendrogram was designed through an analysis of clusters of the UPGMA type (Unweighted Pair Group Method using Arithmetical Averages) with the help of the statistical program R (R DEVELOPMENT CORE TEAM, 2017). The dissimilarity matrix was obtained using the Gower algorithm (1971) and the suitability of the cluster analysis to the original data was evaluated by the coefficient of co-optic correlation.  Obtained a height of 30 cm from the soil Color of pseudostem ("CPC") Greenish-yellow (2.5 GY 8/12); Light green (2.5 GY (8/8); Medium green (5GY 6/6); Dark green ( Internal color of the sheaths ("CIB") Purple (5R 3/10); Reddish-Pink (2.5 R 7/2); Green (5GY 6/6); Brown (5YR 3/4); Transparent

Inflorescence
Positioning of the inflorescence in relation to the stem ("PIH") Upright; Drooping Length of inflorescence ("CI") Short (10 cm ˂); Average (between 11 to 40 cm); Large (˃ 41 cm) Length of the 2 nd bract ("CB") Small (10 cm ˂); Average (between 11 to 40 cm); Large (˃ 41 cm) Width of the second bract ("LB") Narrow (up to 10 cm); Median (of 20 to 30 cm); Large (more than 31 cm) Angle of the 1 st bract in relation to the peduncle ("ABP") ˃ 90°; 75°-90°; 45°-60°; 15°-30°; 0° Twisting the spine ("TR") Present; Absent Visualization of the spine ("VR") Visible; Covered by bracts Ovary: Color of the ripe fruit ("COM") White; Cream; Purplish blue Flowering period ("PF") Short (˂ 5 months); Medium (between 5 to 8 months); Long (˃ 8 months) To verify the existence of differences in the structures of a leaf blade, a clump of genotype with a good phytosanitary quality was selected, and three fully expanded leaves (third leaf) were collected. From the mid-part of each leaf, using a circular hole-puncher, a sample was taken of foliar tissue with 1 cm diameter. It was later placed in a petri dish and immersed in commercial hypochlorite solution until full depigmentation (24-48 hours). Next, using a soft-bristle brush, the abaxial and adaxial portions of all samples were separated, they were also stained with 1% toluidine blue and the blades were assembled with the addition of one drop of Canada balsam (Canada turpentine).
The size of the stomas and stomatal density (DE -number of stomatas per unit area) of both parts of the leaf, abaxial and adaxial, were analyzed under an optical microscope and measured under a 40X lens, observing fifteen fields of view for each sample. Based on a micrometer scale for the eyepiece used, the stomatal density was determined through the ratio between the number of stomata and the area of the real field visible from the 40X lens (0.14 mm 2 ).
Slides were prepared with toluidine blue in order to evaluate the polar diameter (PD) and equatorial diameter (QD) of the stomata. They were observed in a clear chamber, in accordance with the Labouriau technique (CASTRO et al., 2009). The stomatal functionality ("FUN") was calculated from this data, given by the ratio between the polar diameter and equatorial diameter (CASTRO et al., 2009). The averages obtained for all variables were compared through the Scott Knott test at 5% probability.

RESULTS AND DISCUSSION
The clustering analysis based on morphological descriptors and Gower algorithm allowed for three distinct clusters to form (Table 1). Group 2 gathered most of the genotypes, while genotype 35 was the only one remaining alone in group 1, thus demonstrating that it had morphological characters distinct from the others.   There was no presence of hairiness in the pseudo stem in any of the genotypes. The presence of hairs is an important characteristic in the production of heliconia, especially an excess, because it has a negative effect on post-harvest handling and transport that requires additional cleaning, because the presence of hairs leads to the retention of debris .
The plant's height was over 1.5m in all genotypes (Table 2), which could hamper the handling and preparation of flower stems during post-harvest, because large stems are more prone to toppling or being ruptured. However, this is a characteristic that has already been described for this variety and the values obtained for the length of the floral stems are within what the market requires (ABBOTT et al., 2010). Clone 18 displayed the lowest plant height along with reduced inflorescences, which can be better used in landscaping projects that have a floristic composition in reduced environments.
For the diameter of the pseudo stem (DPS), there was a significant variation among the genotypes, ranging from 23.32 (genotype 5) to 58.22mm (genotype 26). According to Farias et al. (2013), the DPS warrants attention due to its influence on flower resistance, handling, selection, packaging and post-harvest durability. A study conducted by Castro et al. (2007) with 30 heliconia genotypes, classified H. chartacea flower stems as large, indicating a high degree of difficulty in handling, and requiring special care for packaging and transportation.
Both the diameter and stem length features are of highly important for the flower's resistance to wind, as stems that have larger diameters are also more rigid. According to Albuquerque et al. (2010), this is due to the carbon reserve contained in the stem that is used to prolong the potential longevity of the flowers. In this regard, Hermans et al. (2006) and Castro et al. (2007) claim that the longer the rod length and diameter, the greater the post-harvest durability. One explanation for the variation in the diameter of the stems is that it is due to the chromosome duplication in the evaluated genotypes.  The presence of large amount of waxiness was constant in most of the evaluated genotypes. One noted exception was genotype 35, which featured waxiness in the younger pseudo stems and, as they developed, lost this characteristic. Another attractive feature of genotype 35 was the change in the leaf limbus (Table 3).  It was the only one that exhibited the edges of the entire limbus and had more vivid coloring, characteristics that increase the possibilities of using this genotype in landscaping, as it becomes something that is more visually attractive ( Figure   3). The presence of wax on the upper and lower surfaces of the leaf blade was also observed exclusively in the individuals of this genotype, which gives them extra protection against water loss from transpiration. Among the descriptors corresponding to the leaf area (Table 3), there was no variation for the leaf position, presence of spots on the limbus of the sapling, shape of the petiole base, edge of the base and the shape of the petiole edge. In this case, all genotypes exhibited arched, non-spotted leaves with a narrow petiole with upright and closed edges and lacking scariosum base. Assessments of the length of the third leaf (CFO) ranged from 57.56 (genotype 40) to 94.50cm (genotype 34) and only genotype 6 displayed leaves with a width greater than 30cm. By and large, the leaf descriptors presented great variability, however, a series of characterizations are needed to standardize these for the producer and, as a result, the end consumer because parameters such as beauty, exoticism and ideal coloring are subjective.
Data shown in Table 4 again shows that the genotype 35 was detached from the others in relation to the position of the inflorescence. H. chartacea var. Sexy Pink normally exhibits drooping inflorescences and the genotype 35 produced erect and more compact inflorescences that are less than 40 cm in length which, according to Albuquerque et. al. (2010) would classify this genotype as "suitable" for commercialization as a cut flower. The other genotypes would require individual packaging according to their size, which would require careful handling to avoid the rupturing the axis where the bracts are inserted.
These genotypes also presented twisting of the spine, which makes the post-harvest preparation process excessively slow, including the use of protection for the bracts. All these resources include an additional component to the sale price.  NID  NID  NID  NID  NID  NID  NID  NID  NID  NID  NID  NID   37  NID  NID  NID  NID  NID  NID  NID  NID  NID  NID  NID  NID 40  NID  NID  NID  NID  NID  NID  NID  NID  NID  NID  NID  NID  For the inflorescences descriptors, there was no polymorphism for the length of the 2nd bract (average), width of the second bract (narrow), angle of the 1st bract relative to the peduncle (>90°), visualization of the soine (visible), apex shape (asymmetric), number of bracts in ripe inflorescence (average), perigone base color (cream), perigone lobes color (orange-yellow), presence of anthocyanin in perigonium (absent), presence of pollen (absent), ripe fruit color (purplish blue).
The length of the bract classified as medium (11 to 40 cm) is significant because it usually represents the most prominence in flower arrangements and, therefore, in customer appreciation. Because there were no changes in this characteristic, we can assert that there was no loss of one of the main commercial features in this species.
The absence of hairiness in the bracts is a positive characteristic because the presence of hair interferes greatly with the handling, preparation and transport. The damage to the handling is due to the fact that it occurs to the formation of a protective layer in the bracts, which can minimize the benefits of the conditioning treatment done for remove the heat from the field and restore the turgescence of harvested flowers. The hair layer can also host small insects, insecticide residues and dust particles, leading to intense handling and preparation and resulting in increasing production costs. Another negative aspect of the presence of hairiness with regards to transport is that the friction of the inflorescences causes hair to fall out in some bracts, damaging the appearance of the inflorescence and compromising its quality. Additionally, Broschat and Donselman (1983) warn of the necessity of packages that remain dry for suitable transport, and the hariness allows the bracts to be stored in excessive humidity, and can be transitioned to the packaging.
The period for outputting the first inflorescence varied greatly between the genotypes. Genotypes 12 and 41 exhibited inflorescence up to eight months after planting (seven and eight months, respectively), while the other genotypes showed a longer flowering period, flowering only after eight months of planting. Genotypes 13, 20, 36, 37 and 40 did not output inflorescences at the end of the two years of the evaluation period. Heliconias are cut tropical flowers and the floral stem is the endproduct of commercial significance that is used in arranging and preparing bouquets. Hence, the largest number of flower stem represents a lower cost of production for the producer, along with higher competitiveness and greater profitability. According to Broschat and Donselman (1983), the neotropical species of heliconia produce terminal inflorescences after outputting 4 to 5 leaves, which was not noted in these genotypes. One use for plants that did not output inflorescences would be in constructing landscapes as foliage, since its exoticity enables this purpose.
Stomata of the genotypes offered tetrapeptic morphology and guard cells without significant changes. However, several genotypes displayed greater equatorial diameter and stomatal density compared to the mother-plant (Figure 4 and Table  5). These differences in the size and number of stomata per area have been common characteristics featured on several plant materials submitted to polyploidization, and are used as basic characteristics in indentifying polyploid individuals in various studies (CAVALCANTE FILHO, 2011;GŁOWACKA et al., 2010;WANG;LEI, 2012). On average, the stomatal density ("SD") was 94.73% higher in the abaxial part compared to the adaxial part ( Table 5). The adaxial epidermis' SD may be less responsive to environmental variations due to radiation being directly incidental upon it (CASTRO et al., 2009) and transpiration occurs mainly through the epidermis of the abaxial face (STRECK, 2003). 1 Averages followed by the same letter in each column pertain to the same group according to the Scott and Knott grouping criterion at 5% probability; 2 Coefficient of experimental variation.
The highest SD were observed in genotypes 1 and 14 for the adaxial part of the leaves, whereas genotypes 10, 12 and 14 were the ones with the highest values for the abaxial part. The lowest values were found in genotypes 7 and 35 for the adaxial part and in genotypes 7 and 13 for the abaxial part of the leaves, differences caused by the induced mutation. Castro et al. (2009) relate that a higher number of stomata per unit of leaf area may confer a high adaptability to environments that are dry or have a lack of available water. Therefore, a higher SD can enable a stomatal opening in a shorter period of time, permitting an adequate capture of CO 2 and reducing the time in which these stomata become open. This reduces transpiration and allows for the plants to better adapt to conditions in which water is scarce (MIGLIORANZA; OLIVEIRA, 2013).
A higher stomatal functionality (FUN) may be associated with reduced transpiration because the stomata becomes more elliptical (CASTRO et al., 2009;BATISTA et al., 2010), the reduced transpiration may also be associated with a greater SD, which is frequently observed under conditions with a greater amount of radiation and less available water (SOUZA et al., 2007;CASTRO et al., 2009). Therefore, among the varieties with the highest FUN and SD values, genotypes 1, 12 and 14 stand out, and the best averages were obtained by genotype 14, both for the adaxial as well as the abaxial part, showing that this genotype presents the most functional stomata.

CONCLUSIONS
The induction of polyploidy in H. Chartacea var. Sexy Pink allowed for clones to be obtained with a high variability for leaf characteristics, pseudostem and inflorescence, with various attributes that confer a more efficient postharvest management to some genotypes, in addition to favorable aspects for commercialized purposes as a cut flower.
The presence of genotypes with changes in the position, coloring and size of the inflorescence, in addition to modifications in the leaf limb that left them with whole edges, showed promise for future projects geared towards the production of cut flowers and landscaping in H. chartacea var. Sexy Pink.
The evaluated genotypes present stomata with tetracytic morphology and guard cells that do not have significant changes. However, some genotypes show a larger equatorial diameter and stomatal density in relation to the mother-plant.
The next steps of this research are the analysis of the DNA content by flow cytometry for polyploidy identification of the genotypes.