Reproductive biology of Kalanchoe laetivirens (Crassulaceae) in the edaphoclimatic conditions of Santa Catarina, Brazil

Crassulaceae of the Kalanchoe laetivirens species have ornamental potential, but are poorly studied, especially their reproductive biology and trophic resources made available to visitors. The aim of this study, therefore, was to characterize the reproductive system of K. laetivirens through analyses of floral morphology, trophic resources and floral visitors. Floral structures, pollen / egg ratio, in vitro pollen germination, pollen grain structure in scanning microscopy and the availability of trophic resources by the evaluation of potential and instant nectar were described. The survey of floral visitors was carried out, and the pollen and nectar collection frequencies were recorded. The results show that the species has tubular flowers, complete with tetramer calyx and corolla, eight epipetalous stamens and four pistils with pluriovulated ovaries. Based on pollen / ovule ratio, this species is optionally autogamous. Morphologically, the pollen grain is small, tricolporate and has radial symmetrywith in vitro germination facilitated in the presence of sucrose (20%). Visitors mostly collected pollen with the highest frequency of visits at 9:30 am. The production of instant nectar had little variation between the evaluated times. The results are relevant to the understanding of the reproductive biology of Kalanchoe laetivirens and its relationship with floral visitors.


Introduction
The Crassulaceae family is represented by vegetables, such as Kalanchoe laetivirens Descoings (1997), succulent plants originating in tropical areas, mainly in Africa. Some species of this genus are widely distributed in horticulture, with K.
Kalanchoe phytochemicals are widely described, but the most important highlight is in the area of plant physiology owingto the acidic metabolism of crassulaceae (Garcia, 1998;Milad, El-Ahmady & Singab 2014). However, research on the reproductive system of these plants is scarce, in particular their floral biology, production of trophic resources and interactions with possible floral visitors.
Plants have floral organs with morphological, phenological and physiological peculiarities that influence the reproductive process, with pollination being dependent on animals as the most efficient method of distribution (Rech, Agostini, Oliveira & Machado, 2014). The convergent adaptations between plants and animals, specifically floral visitors, or between plants and abiotic factors (water, wind) culminated in the emergence of many pollination syndromes (Van der Pijl, 1961).
To close the gap in the understudied reproductive system of K. laetivirens, the present study focused on its reproductive processes in the edaphoclimatic conditions of Santa Catarina, Brazil, through the prism of its morphology, resources and floral visitors.

Study sites and plant material
Field evaluations were carried out on days with stable weather (average temperature of 19.9 ° C, relative humidity of 77.16% and average wind speed of 43 km / h), in the city of Paulo Lopes (-27.966491 S, -48.660037 Research, Society andDevelopment, v. 10, n. 1, e27010111567, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i1.11567 3 The study was carried out with clones of a matrix plant of K. laetivirens identified by the botanical key of Shaw (2008) based on morphological characteristics and in studies of Smith (2020). Another reference in the identification of the species was a registered exsiccate (n ° 532) deposited on July 3, 1979 at the Herbarium Jardim Botânico of Rio de Janeiro.

Floral morphology of stamens and pistils
The morphological characterization of the flowers and the inflorescence of K. laetivirens were supported by the descriptions proposed by Gonçalves (2011) and Souza, Flores and Lorenzi (2013). Pollen grains were evaluated by scanning electron microscopy and described following the terminology standardized by Punt et al. (2007). Finally, measurements of stamens and pistils were performed with the aid of a caliper, using five flowers in balloon stage (adapted from Cortinóz et al., 2008).

Flowering scale
The classification of the flower's development phases was adapted to the initial proposal of criteria by Ducroquet and Hickel (1991), using the denominations shown in Table 1.  Ducroquet and Hickel (1991).

The receptivity of stigma and the viability of pollen grains
The receptivity of the stigma was evaluated based on Dafni (1992), through the release of O2 (bubbling) in the reaction of hydrogen peroxide (H2O2, 10 volumes), with the papillae of the stigma of the pistils of three flowers in stages E, F1, F2 and F3 collected at the same time. In the event of bubbling, observed under stereomicroscopy (Feldmann Wild Lentz, Research, Society andDevelopment, v. 10, n. 1, e27010111567, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i1.11567 FWL -150, 100x), 25% receptivity was attributed to each of the four stigmas of each flower, generating a receptivity of 0 -100% per flower.
To assess pollen viability, the colorimetric method was used according to the methodology proposed by Dafni (1992).
The anthers of three flowers in stages E, F1, F2 and F3 were rubbed lightly over slides; afterwards, 2 μL of acetic carmine were added. Then, the material was covered with a cover slip, and after five minutes, the percentage of pollen grains stained was evaluated with the aid of an optical microscope (Studarlab 400x). In the case of closed anthers, it was necessary to leave them in Petri dishes for 24 hours at room temperature for dehiscence to occur.

Estimation of the number of pollen grains and egg count
Four anthers of five flowers in balloon (E) stage were randomly collected, and diluted in 0.5 mL of 85% lactic acid in 1.5 mL microtubes. Two 1.5 μL aliquots of each microtube were pipetted in a Neubauer Chamber and observed under an optical microscope (Studarlab400x) to count pollen grains. An average of pollen grains from the aliquots (X) was obtained, and the total number of pollen grains for the four anthers was estimated as N = X * 500 / 1.5 * 1/4 (N = number of grains of pollen) of the same flower. (Albuquerque Junior, Denardi, Dantas & Nodari, 2010;Kearns & Inouye, 1993).
For the egg count, the pistils of five flowers in balloon stage (E) were placed on the sticky face of a transparent adhesive tape fixed on a glass slide. The ovaries were opened with the aid of histological needles, and the eggs were gently extended on the tape, facilitating their counting with a stereomicroscope (Feldmann Wild Lentz, FWL -150, 100 x).
The pollen / egg (P / O) ratio was obtained according to Cruden (1977). The fraction between the average number of pollen grains and the number of eggs per flower made it possible to infer the type of reproductive system.

In vitro germination of pollen grains
Pollen grains of 12 F2 flowers (three flowers per treatment) were distributed directly from the anthers on the growth medium (1 g agar / 100 mL of distilled water; 40 mg.L -1 of boric acid) deposited on Petri dishes (10 mL). Four different sucrose concentrations of 0, 10, 20 and 40% were established, totaling four treatments, with five replicates each (adapted from Nunes et al., 2001). Subsequently, the Petri dishes were kept in an incubator with biochemical oxygen demand (BOD Electrolab) at 25 ° C. Germination was evaluated under an optical microscope (Studarlab 100 x) at 1, 4 and 8 hours after distribution of pollen grains in the growth medium.

Scanning electron microscopy (SEM)
Samples of androecium and gynoecium of flowers in F2 were collected and immediately transferred to polypropylene tubes (1.5 mL) containing Karnovsky's Fixative (2.5% glutaraldehyde, 2% formaldehyde in 0.1M sodium cacodylate buffer, pH 7.2). The samples remained submerged in the fixative and were refrigerated (4 ° C) for 48 h, followed by washing in 0.1M cacodylate buffer. Preparation of the samples for observation in SEM followed the standard dehydration protocol in increasing series of ethyl alcohol and distilled water (70 to 100%) and drying with HMDS (Hexamethyldisilazane). Subsequently, the samples were assembled in stubs with double-sided carbon tape and subjected to metallization in a gold sputtering machine (modelEM SCD 500 / LEICA). The analyses were performed in a scanning electron microscope (JEOL JSM-6390LV), with an electron acceleration voltage of 10 kV, and the images were captured digitally by the JEOL / EO1.0 program, using equipment provided by the Central Laboratory of Electron Microscopy (LCME -UFSC).

Floral visitors
The survey of floral visitors was carried out using the methods of Benevides, Gaglianone and Hoffmann (2009) and Polatto and Alves Junior (2008) with modifications. The evaluations were carried out from 6:30 am to 6:30 pm, considering a radius of 2 m between the plants. Floral visitors were determined through naturalistic observations lasting 10 min each hour, totaling four hours of observations on September 15 and 22, 2019. Visits were characterized as the visitor's permanence in the inflorescence from arrival to departure, regardless of the number of flowers visited. The floral visitor's behavior regarding the collection of nectar or pollen was also evaluated, considering only the first co lection per visit, given the difficulty of accounting for multiple collections from different visitors. To avoid interference with the number of visitors, specimens were not collected for identification.

Instant and potential nectar
Based on the methodology proposed by Dafni (1992)

10 Statistical analysis
The data collected in relation to pistil and stigma morphometry and pollen grain germination were subjected to parametric analysis of variance and Tukey test (p = 0.05) to compare the averages in R software.

Results
The flowers of Kalanchoe laetivirens have a green to rosacea color ( Figure 1A), pedicellated, without hypanthium, complete, actinomorphic and have a cyclic arrangement of the floral pieces. In addition, they are tubular, tetrameres, dichlamydeous, heteroclamids, gamosepalous and gamopetalouswith valve prefloration ( Figure 1B). The androecium was classified as diplostemonous, isodynamic, epipetalous, with stamens included up to balloon stage, basifixed anthers that have longitudinal dehiscence and powdery dispersion of pollen grains ( Figure 1C). The gynoecium isapocarpous, tetracarpellar and unilocular ( Figure 1D), parietal placentation, semi-infertile, numerous ovules ( Figure 1E) with terminal insertion of the stylet and undivided stigma. The nectary is located between the stamens and carpels. The flowers are collected in cymose-type inflorescences called pleiochasium and, under the environmental conditions of this study, did not form seeds.   Ducroquet and Hickel (1991) and described in Table 1.

Source: Authors (2019).
Analysis of the scanning photomicrographs indicated that the pollen grains of K. laetivirens have a reduced size (10 -25 µm), are isopolar, tricolporate, prolatespheroidal and radially symmetrical. It was possible to distinguish the nectaries at the base of the carpels ( Figure 3A) and to confirm the longitudinal dehiscence of the anthers ( Figure 3B), as well as highlight the agglomeration of the pollen grains enabling characterization of their dispersion form ( Figure 3C) asstriated-corrugated tectum ( Figure 3D) and ornamentation classified as "Sedum hispanicum-type" according to Qaiser, Perveen and Sarwar (2015).  Pollen tube emission occurred within 8 hours, regardless of the treatment used. Pollen germination in vitro was affected by the concentration of sugar in the growth medium during the evaluated periods, except for pollen submitted to a concentration of 20% sucrose, the germination of which was above 90% in the first hour of evaluation and remained constant for all periods evaluated ( Table 2). Irrespective of the presence of sugar, it should be noted that more than half of the pollen grains started to germinate in the first hour of evaluation.  (Fig. 5). The potential nectar volume was 17.33 μL (± 7.33 μL), and the average sucrose concentration was 9.24 ° Brix Research, Society and Development, v. 10, n. 1, e27010111567, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i1.11567 10 (± 2.25 ° Brix), whereas for instant nectar, it ranged between 9.14 and7.1 μL and 13.36 and12.98 ° Brix ( Figure 5) over the evaluation period.

Discussion
The morphological characteristics observed for the floral structures of K. laetivirens are those established by Milad, El-Ahmady and Singab (2014), and they are essential in the botanical identification of the species and in the construction of the flowering scale. Characterization of the flowers helps in the understanding of their morpho-functionality, consequently providing phylogenetic clues to the adaptations to which the plant organs were subjected throughout evolution (Oliveira, 1991). They can also help predict floral visitors and possible pollinators for the species, according to the interactions of the known pollination syndromes, which, in some cases, are extremely specialized, defined and well synchronized (Alves-dos-Santos, Silva, Pinheiro & Kleinert, 2016).
Flowers are one of the consumer products in the ornamental plant market, standing out as an economic activity that generates employment and income for micro and small producers across the country, with a significant participation of rural female labor (Junqueira & Peetz,., 2014). In addition, knowledge of floral biology, together with the reproductive mechanisms of plant species, is essential for the establishment of plant breeding programs, allowing selection of the most appropriate technique for use (Allard, 1971). The establishment of breeding techniques for ornamental species in Brazil is a priority strategy for the development of this important sector of the economy (Cardoso, 2013).
The pollen grains of K. laetivirens fit the general pollen morphology of crassulaceae, as described by Qaiser, Perveen and Sarwar (2015) whose compendium of characteristics facilitates the distinction of pollen types by genera, given the heterogeneity between them. P/O ratio and the same size of the stamens and pistils observed in the evaluated flowers indicate that the specie can be classified as facultative and homomorphic autogamous, respectively (Cruden, 1977;Karasawa, 2009).
The flowers of K. laetivirens have characteristics similar to those of angiosperms with the capability of selfpollination since they present homogamy with viable pollen release just when the stigma is receptive (Rashid & Singh, 2000).
Aligned with the lack of physiological and morphological characteristics, such as heterogamyand heteromorphism, the abundance of pollen in longitudinal dehiscence anthers favors self-pollination (Costa, Krupek & Krawczyk, 2015) The germination of pollen grains in vitro was greater than 50% in the first hour of evaluation, regardless of the concentration of sugars in the culture medium, ensuring their viability. The germinative behavior of pollen grains depends on the osmotic balance between them and the culture medium, which is related to the carbohydrate source for the growth of the pollen tube (Stanley & Linskens, 1974), in this case, facilitated by the concentration of 20% sugar. Additionally, the results obtained on the pollen aspects of K. laetivirens indicate that this species produces viable pollen and in large quantities with conditions conducive to the formation of seeds. It is noteworthy that pollen is not nutrition-intensive since they can be germinated in vitro in all concentrations of growth media evaluated.
Seed formation was not observed in the evaluated plants. It is possible that this occurred because of the selfincompatibility of the specimens used, considering that they were clones, which can be confirmed in future controlled pollination tests. Some species of Kalanchoe have the defective LEC1 gene, which prevents the formation of viable seeds; thus, they reproduce asexually by the formation of seedlings in the margins of the leaf blade, all coordinated by regulators of organogenesis and embryogenesis (Garcês et al., 2007). It is possible that K. laetivirens was a carrier of this gene, considering that the emission of seedlings was observed in the leaf blade, as previously described.
Visitors may prefer pollen over nectar by the ease of collection since it is a source exposed in these flowers,decreasing the energy cost of foraging (Rech, Agostini, Oliveira, & Machado, 2014). It is also possible that the composition of the nectar of this plant has selectivity over the collectors (Adler, 2000). In addition, the local flora influences the dynamics of visits. The average values for the volume of potential and instantaneous nectar of K. laetivirens are similar to those of Echeveria gibbiflora (Parra, Vargas & Eguiarte, 1993); however, the average sugar concentration (° Brix) found was three to five times lower than that in this species or in the nectar of K. daigremontiana (Herrera & Nassar 2009). The volume of nectar is related to pollination syndrome. The ideal volume is one which attracts visitors, but does not satisfy them in a single flower, thus expanding visitation and, hence, the probability of cross-pollination (Wolf, 2006).
The composition of the nectar and the available quantity of this resource affect the dynamics of floral visits, even if foraging occurs in less favorable conditions (Polatto, Chaud-Netto & Alves-Junior, 2014). Thus, environmental conditions are extremely important in the frequency of visits throughout the day, as they interfere with the ecophysiology of the plant, altering the production of resources such as nectar (Silva, Dutra, Nucci & Polatto, 2013). Environment is also an influence, for example, the thermoregulation of insects and, consequently, their flight activity, depending on body sizes (Teixeira & Campos, 2005). Therefore, both environmental conditions and resources favored visitation at 9:30 am within the diversity of visitors which frequented the flowers, similar to that reported by Pegoraro and Neto (2005).
From the results obtained, the sexual reproduction of K. laetivirens is a wasted energy investment for the survival of the species since seeds are not formed. Despite this, it is evident that its flowers are an important source of food (pollen and nectar) for visitors. This reproductive pathway proved to be only slightly viable for K. daigremontiana and K. pinnata, which, although producing seeds, have a low germination rate, opting instead for vegetative propagation (Herrera & Nassar, 2009;Leon, Herrera & Guevara, 2016). Such vegetative capacity is significant in the genus Kalanchoe, but both reproductive pathways, sexual and asexual, guaranteed the geographical expansion of these vegetables, even in places with inhospitable conditions (Allorge-Boiteaij, 1996).

Conclusion
The floral biology of Kalanchoe laetivirens is morphologically complex, but does not have functionality for reproductive purposes under the conditions of this study, and further evaluations are necessary to determine the causes deficit in the formation of seeds. However, the capacity of these plants to provide trophic resources to their floral visitors in the environments in which they inhabit stands out.