Introduction
Ilex paraguariensis (Aquifoliaceae), popularly known as yerba mate, is a forest tree species, native to South America, namely to Brazil, Argentina and Paraguay (Carvalho, Reference Carvalho2003). These three countries are solely responsible for virtually the entirety of the yerba mate produced in the world, producing 1,016,085 tons of the product in 2017 (FAO, 2020). Brazil alone produces more than 60% of the total yerba mate in the world and 80% of this amount is used for local consumption (Reichert et al., Reference Reichert, Friedrich, Cassol, Pensin, Mitsui, Donaduzzi and Cardozo Junior2013). This species possesses considerable economic importance, with broad market potential, ranging from the pharmaceutical and cosmetic industries to tea manufacturing (Dartora et al., Reference Dartora, Souza, Paiva, Scoparo, Iacomini, Gorin, Rattmann and Sassaki2013). Yerba mate products are becoming increasingly popular in other countries. Ilex paraguariensis teas and energy drinks are consumed in the United States, Germany and Syria and, in the last decade, the use of the plant has expanded to other countries such as Spain, Italy, Australia, France, Japan, South Korea and Russia, mainly due to its flavour and stimulating properties (Cardozo Junior and Morand, Reference Cardozo Junior and Morand2016).
Yerba mate shoots are rich sources of minerals, vitamins, phenolic compounds, alkaloids (caffeine, theobromine) and terpenes (carotenoids, saponins) (Heck and Mejia, Reference Heck and Mejia2007). Some of these compounds are associated with the antioxidant, anti-inflammatory, anti-tumour and anti-obesity activities of different extracts from I. paraguariensis (Bracesco et al., Reference Bracesco, Sanchez, Contreras, Menini and Gugliucci2011). Additional studies have evidenced that yerba mate products exert important antimicrobial (Martin et al., Reference Martin, Porto, Alencar, Glória, Corrêa and Cabral2013), neuroprotective (Branco et al., Reference Branco, Scola, Rodrigues, Cesio, Laprovitera and Heinzen2013), hypocholesterolemic (Bravo et al., Reference Bravo, Mateos, Sarriá, Baeza, Lecumberri, Ramos and Goya2014), hepatoprotective (Tamura et al., Reference Tamura, Sasaki, Yamashita, Matsui-Yuasa, Saku, Hikima, Tabuchi, Munakata and Kojima-Yuasa2013), diuretic and anti-rheumatic activities (Isolabella et al., Reference Isolabella, Cogoi, López, Anesini, Ferraro and Filip2010).
The generation of new information regarding pharmacological activities and the development of new market opportunities for yerba mate products evidence the need for the selection of more productive genotypes regarding biomass, secondary metabolite production or overall silvicultural performance to supply the growing demand for high-quality raw material. According to Santin et al. (Reference Santin, Benedetti, Barros, Fontes, Almeida, Neves and Wendling2017), the selection of parental plants from different sources allows for the identification and improvement of specific traits of I. paraguariensis, which can result in higher productivity and profitability. In this context, the phytochemical characterization of different genotypes is a fundamental step for breeding programmes and for the establishment of profitable yerba mate forests.
As important as the selection of productive genotypes is the ability to efficiently propagate such plants. Currently, plants used for the establishment of commercial yerba mate forests are mostly produced by seeds and originated from unspecialized nurseries, which ultimately results in low productivity (Santin et al., Reference Santin, Benedetti, Brondani, Reissmann, Orrutéa and Roveda2008). One viable alternative to overcome this issue is the development of protocols for large-scale cloning of highly productive genotypes (Wendling and Santin, Reference Wendling and Santin2015). Although there are studies on commercial propagation of I. paraguariensis by mini-cuttings (Wendling et al., Reference Wendling, Dutra and Grossi2007; Sá et al., Reference Sá, Portes, Wendling and Zuffellato-Ribas2018; Pimentel et al., Reference Pimentel, Lencina, Kielse, Rodrigues, Somalia and Bisognin2019, Reference Pimentel, Pedroso, Lencina, Oliveira and Bisognin2020), these are still scarce, especially regarding rejuvenation techniques to efficiently promote rhizogenesis on tissues from mature trees (Wendling and Santin, Reference Wendling and Santin2015).
The present study aimed to assess the chemical composition, antioxidant activity and rooting performance of yerba mate mini-cuttings from 15 different genotypes as well as to evaluate the correlations between the phytochemical components and the rhizogenic potential of the vegetative propagules.
Material and methods
Plant material and growth conditions
The genotypes evaluated in the present study, with the exception of BRS BLD Aupaba, are originated from a selective breeding programme (parent plants established from a provenance/progeny test of half-siblings) carried out in the city of Ivaí, Paraná state, Brazil. The genotypes from the breeding programme are named EC20, EC21, EC24, EC26, EC27, EC28, EC31, EC38, EC44, EC45, EC48, EC50, EC53 and EC54. Genotype BRS BLD Aupaba (from now on referred to as BRS) is a commercial yerba-mate cultivar used for reference (Wendling et al., Reference Wendling, Santin, Nagaoka and Sturion2017). Parental plants (10-year-old trees) were propagated by conventional stem cutting as described by Bitencourt et al. (Reference Bitencourt, Zuffellato-Ribas, Wendling and Koeler2009). Rooted stem cuttings from the selected genotypes were transplanted in June 2017 and kept as mini-stumps by periodical pruning in order to produce juvenile shoots. The mini-stumps were grown under greenhouse conditions (minimum and maximum temperatures of 13 and 37°C, respectively, with an average of 25.3°C and ≥85% air relative humidity) in the city of Colombo, Paraná State, Brazil (25°20′S and 49°14′W, 950 m asl). According to the Köppen classification, the climate of the region is temperate, Cfb type. The plants were grown at a density of 44 mini-stumps per square meter (15 cm × 15 cm spacing) in a semi-hydroponic channel system with sand beds (Wendling et al., Reference Wendling, Dutra and Grossi2007). Fertilization was performed by nutrient solution: NO3− (156.0 mg/l), NO4+ (50.0 mg/l), P (25.0 mg/l), K+ (200.0 mg/l), Ca2+ (200.0 mg/l), Mg2+ (45.0 mg/l), S (76.9 mg/l), B (1.5 mg/l), Cu2+ (0.1 mg/l), Fe2+ (5.0 mg/l), Mn2+ (1.0 mg/l), Zn2+ (0.7 mg/l), Mo2− (0.07 mg/l), delivered via drip fertigation system, three times a day at a total flow of 5 l/m2, according to the recommendations of Wendling and Santin (Reference Wendling and Santin2015).
Assessment of mini-cuttings rooting and sprouting
Mini-cuttings were taken from mini-stumps in April (autumn) 2019, from the middle region of the plant (basal and apical mini-cuttings discarded). The propagules were made 5 cm long, with a bevelled cut at their base and straight cut at the apex and contained a pair of whole leaves at their apical portion. The mini-cuttings were kept in a styrofoam box with water to avoid dehydration. Immediately before planting, the mini-cuttings were removed from the styrofoam box and had their bases (about a third of the stem length) immersed in a hydroethanolic solution (50% v/v) of indolebutyric acid (IBA) at 3000 mg/l (Wendling and Santin, Reference Wendling and Santin2015). After IBA treatment, the mini-cuttings were planted in 53 cm3 polypropylene conic-shaped containers, filled with pre-moistened commercial substrate Agrinobre® TN Mix 0,6 composed of sphagnum peat, expanded vermiculite, carbonized rice husk, dolomitic limestone, agricultural gypsum, NPK fertilizer and micronutrients. The chemical and physical characteristics informed by the manufacturer are the following: pH 5.0; electrical conductivity 0.6 (mS/cm); maximum humidity 55% (w/w); density of 150 kg/m3 and water retention capacity −140% (w/w).
After planting, the mini-cuttings were kept in an artificially climatized greenhouse (85% air relative humidity and 20–30°C temperature) with intermittent misting.
After 120 days from planting (Wendling and Santin, Reference Wendling and Santin2015), the following variables were assessed: rooting percentage (percentage of mini-cuttings with roots of at least 1 mm in length), number of roots per mini-cutting; average length of the three longest roots per mini-cutting (cm); percentage of callogenesis (non-rooted mini-cuttings with the presence of an undifferentiated mass of cells at their base); percentage of mini-cuttings showing both calluses and roots; percentage of alive mini-cuttings (mini-cuttings with neither roots nor calluses but with no necrotic tissue); percentage of leaf retention (mini-cuttings that kept the original leaves in the 120-day period), sprouting percentage; and percentage of mortality (necrotic mini-cuttings). The experiment was organized in a completely randomized design, with 15 treatments (genotypes), four replications and 10 mini-cuttings per plot.
Preparation of yerba mate aqueous extract
From the mini-cuttings taken for the rooting experiment, subsamples of each plot (20–30 mini-cuttings including the leaves) were taken to prepare the aqueous extract used for the phytochemical characterization and antioxidant assessment. The full amount of plant material (20–30 mini-cuttings) was microwave dried oven for 5 min (1000 W, 2450 MHz) and subsequently ground in a cutting mill (classified as fine grinding – sieve 0.25 mm), and stored in the freezer (−20°C) until extraction.
To prepare the aqueous extract, 0.1 g of the plant material was placed in 2.5 ml microtubes with 2 ml of ultrapure water. The solution was vortexed for 30 s and incubated for 1 h, at 60°C and 450 rpm using a Thermomixer® equipment. The aqueous extracts were then filtered through a 0.22 μm filter and used to determine the antioxidant capacity by ABTS and DPPH methods, total phenolic contents and chromatographic analyses (caffeine and theobromine). The extraction of each plot was done in triplicate and the average value was used for statistical analyses.
Antioxidant capacity, total phenolics and protein content assessment
The antioxidant capacity of yerba-mate aqueous extracts was assessed on 2,2-diphenyl-1-picrylhydrazyl (DPPH) radicals, using the method described by Brand-Williams et al. (Reference Brand-Williams, Cuvelier and Berset1995) with modifications. An amount of 0.1 ml of yerba-mate aqueous extract was added to 3.9 ml of DPPH diluted in methanol (0.06 mmol/l). The solution was shaken and incubated (sheltered from light) for 30 min at room temperature (23 ± 2°C). After this period, absorbance was determined at 515 nm.
The antioxidant capacity of extracts by free radicals ABTS (2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) was determined by reacting 10 ml of ABTS (7 mmol/l) with 176 μl of potassium persulfate (140 mmol/l) in the dark at room temperature for 16 h. An aliquot of 1 ml of this solution was added in 100 ml of sodium acetate buffer (20 mmol/l) pH 4.5. The absorbance was adjusted to 0.7 ± 0.05 and 3 ml was added to 30 μl of the extract. After 2 h, the absorbance was then determined (734 nm) (RE et al., Reference Re, Pellegrini, Proteggente, Pannala, Yang and Rice-Evans1999; YIM et al., Reference Yim, Chye, Rao, Low, Matanjun, How and Ho2013) with minor modifications.
The antioxidant activity for both methods was estimated using a Trolox analytical curve (range of 0–1000 and 0–2500 mg/l, for DPPH and ABTS, respectively). The results were expressed as μmoles of Trolox equivalent per gram of dry weight (μmol/g).
Total phenolic compounds content was determined by the Folin-Ciocalteu spectrophotometric method as described by Singleton and Rossi (Reference Singleton and Rossi1965), with slight modifications. An aliquot of 0.1 ml of yerba-mate aqueous extract was added to a test tube followed by the addition of 0.5 ml of Folin-Ciocalteu reagent and 5 ml of distilled water. The solution was vortexed for 1 min and 2 ml of 15% sodium carbonate was added to the tube. The solution was vortexed again for 30 s and the volume was adjusted to 10 ml with distilled water. After 2 h of incubation (sheltered from light), the absorbance was measured at 760 nm. Gallic acid standard curve (0.25–10 mg/l) was used for quantification and the results were expressed as milligrams of gallic acid equivalents per gram of dried weight (mgGAE/g).
The total protein content was determined by the Kjeldahl conventional method (digestion, distillation and titration) (AOAC, 1984) being the protein content determined by multiplying the content of total nitrogen by factor 6.25 (universal factor). The results were expressed as mg of protein per gram of dry plant tissue (mg/g).
Assessment of caffeine and theobromine contents by liquid chromatography
Chromatographic analyses were performed in a liquid chromatograph (Shimadzu® – model), controlled by the LC Solution software and equipped with an automatic injector and UV detector (SPD-20A). A Shim-Pack CLC-ODS (M)® C18 column (250 × 4.6 mm, Ø 5 μm) was used, with a Shim-Pack CLC G-ODS® pre-column (10 × 4, 0 mm, Ø 5 μm), both from Shimadzu®. The volume of injection was 20 μl of the aqueous extract (Tomasi, Reference Tomasi2020).
The separation of compounds in the aqueous extract was performed at 30°C with a flow rate of 0.5 ml/min. The mobile phases consisted of a gradient elution of water and acetic acid (99.9: 0.1 v/v, solvent A) and acetonitrile (100%, solvent B). The gradient elution programme was: 0–15 min (3% B), 15–20 min (3–20% B), 20–40 min (20% B), 40–45 min (20–30% B), 45–55 min (30−100% B), 55–75 (100% B), 75–80 (100–3% B) and 80–95 (3% B). The wavelength for compounds detection was set at 280 nm.
The identification and quantification of 1,3,7-trimethylxanthine (caffeine) and 3,7-dimethylxanthine (theobromine) was performed by means of an analytical curve of their standards at a range of 0–1.0 mg/ml and 0–0.5 mg/ml, respectively. The results were expressed in mg of compound per gram of dry sample (mg/g).
Statistical analyses
The variances were tested for homogeneity by the Bartlett test (P < 0.05). The variables whose variances were shown to be homogeneous were subjected to analysis of variance (P < 0.01 and P < 0.05) and those showing significant differences by the F-test had their means compared by the Scott-Knott test. Percentage of living mini-cuttings and percentage of mortality did not meet the assumptions of normality and homogeneity of variances and were, therefore, transformed by the equation Log10 (x + 10) to proceed with analysis of variance. Pearson correlations were performed using the same software.
Results
Mini-cuttings performance
Significant differences were observed among the genotypes for all the variables related to mini-cuttings rooting, survival and sprouting. Mini-cuttings from genotype EC24 had the highest average rooting percentage (72.5%), although not statistically different from BRS, EC21, EC54, EC20, EC27 and EC45 (Table 1).
Table 1. Percentage of mini-cuttings with roots (R), number of roots per cutting (NR), average length of roots per mini-cutting (RL), percentage of mini-cuttings with calluses and no roots (C), percentage of mini-cuttings with both calluses and roots (CR), percentage of alive mini-cuttings (A), percentage of dead mini-cuttings (D) and percentage of leaf retention (LR) of 15 genotypes (Gen.) of Ilex paraguariensis after 120 days in greenhouse
Means followed by the same letter in the columns do not differ statistically by the Scott-Knott test at 5% probability.
As for the average number of roots per mini-cutting, up to 10-fold differences were observed among the genotypes. EC24, BRS, EC21, EC54, EC20, EC45, EC31 and EC48 had statistically more roots per mini-cutting (means ranging from 5.0 to 7.26) than the other genotypes. A similar trend was observed for average roots length, with the exception of EC48, which presented relatively short roots (average 0.96 cm) (Table 1).
Genotype EC21 presented no mini-cuttings with calluses only, but had 30% of the propagules showing both calluses and roots, not differing statistically from EC24, BRS, EC21, EC54, EC27 and EC45, and all being superior to the other genotypes on this last variable. EC50 presented no mini-cuttings with simultaneous formation of roots and callus. For mini-cuttings with only calluses, EC24, BRS, EC20, EC31 and EC44 had the lowest values, not differing statistically among themselves. Regarding sprouting percentages, all genotypes reached no more than 10% (data not shown), with no clear differences among the different genetic materials.
Most of the genotypes with the highest rooting percentages showed lower percentages of alive mini-cuttings (mini-cuttings with neither roots nor calluses but with no necrotic tissue), except for EC27 and EC20, which presented relatively high percentages of both rooted and alive mini-cuttings. As for mortality rates, EC24 presented no dead mini-cuttings. EC54, EC20, EC27, EC45 and EC28 had lower mortality rates than the other genotypes, ranging from 10 to 15% and not differing statistically among themselves.
Genotypes presenting the lowest rooting percentages (EC53, EC28, EC38, EC31, EC26, EC48, EC50, EC44) also showed the lowest leaf retention percentages (mini-cuttings that kept the original leaves in the 120-day period), except for EC53 and EC28, which had relatively low rooting percentages (30 and 20%, respectively) and high leaf retention rates (both at 70%). EC31 and EC26 had the lowest leaf retention percentages among all the assessed genotypes (12.5 and 20%, respectively) (Table 1).
Chemical composition
The genotypes EC 20, EC50, EC45, EC44 and EC31 had the highest total phenolic contents, not differing statistically among themselves. The lowest total phenolic contents were observed on mini-cuttings from genotypes BRS, EC24 and EC27 (Table 2).
Table 2. Total phenolic compounds (TP), antioxidant capacity (ABTS and DPPH methods), and protein, caffeine and theobromine contents of mini-cuttings from different genotypes of Ilex paraguariensis
Means followed by the same letter in the columns do not differ statistically by the Scott-Knott test at 5% probability.
The antioxidant activity measured by the ABTS method presented values ranging from 1550.29 to 2790.41 μmol/g, although no statistically significant differences were observed among the genotypes. On the other hand, the antioxidant activity assessed via the DPPH method was significantly affected by the genotypes. Genotypes BRS, EC24, EC54, EC27, EC53, EC28 and EC38 had the lowest DPPH scavenging capacities, not differing among themselves (Table 2).
Genotype EC31 had the single highest contents of both protein and caffeine (194.01 and 28.26 mg/g, respectively), followed by EC27 (156.84 and 21.29 mg/g of protein and caffeine, respectively), which did not differ statistically from EC50 for proteins. Genotype BRS, on the other hand, had the single lowest protein content (125.02 mg/g) and also the lowest caffeine content, although not statistically different from EC28 on this variable. The lowest value for theobromine was observed in genotype EC26 (1.43 mg/g) and the highest on EC44 (5.89 mg/g).
Correlation analysis
Statistically significant positive correlation (0.93, P < 0.01) was observed among rooting percentage and percentages of mini-cuttings with both calluses and roots. Rooting percentage was also positively correlated with number of roots, roots length and leaf retention percentage. Rooting was negatively correlated with mortality percentage, total phenolic compounds, antioxidant activity by the ABTS method and theobromine content (Table 3).
Table 3. Person's correlation coefficients between rooting-related variables and chemical composition of mini-cuttings from 15 genotypes of Ilex paraguariensis
R, rooting (%); NR, number of roots per mini-cuttings; RL, average roots length per mini-cuttings (cm); D, dead mini-cuttings (%); CR, mini-cuttings with both calluses and roots (%); LR, leaf retention (%); TP, total phenolic compounds (g/g); AB, ABTS antioxidant activity (g/g); DP, DPPH antioxidant activity (g/g); P, protein content (mg/g); Th, theobromine (g/g); Cf, caffeine (g/g); ns, non-significant.
**Significant at 1% probability level. *Significant at 5% probability level.
Number of roots per mini-cutting was positively correlated with roots length and percentage of mini-cuttings with both calluses and roots and negatively correlated with theobromine contents. Theobromine contents were negatively correlated with a number of rooting-related variables and positively correlated with mini-cuttings mortality. Mortality percentage was also positively correlated with total phenolics and antioxidant activity by the ABTS method.
Leaf retention was positively correlated with mini-cuttings with both roots and calluses as well as with rooting percentage. Negative correlation was observed for mini-cuttings leaf retention and mortality and protein contents. Protein contents presented a positive correlation with caffeine levels of the yerba mate mini-cuttings.
Total phenolics besides the abovementioned negative correlation with mini-cuttings rooting presented a positive correlation with antioxidant activity assessed by both ABTS and DPPH methods.
Discussion
Significant variability was observed regarding the rhizogenesis potential of mini-cuttings from different genotypes of yerba mate. Such information is relevant for the vegetative propagation of the species, since each genotype may demand specific methodologies for the productions of high-quality transplants. Genotype effects on rooting of I. paraguariensis mini-cuttings were observed in previous studies and factors such as seasonality (better rooting in summer, fall or winter depending on the genotype), types of mini-cuttings and use of plant growth regulators (up to 3000 mg/l IBA) were pointed out as potential strategies for those presenting lower rhizogenic potential, since genotypes respond differently to environmental stimuli (Wendling and Santin, Reference Wendling and Santin2015; Pimentel et al., Reference Pimentel, Lencina, Kielse, Rodrigues, Somalia and Bisognin2019; Mayer et al., Reference Mayer, Nienow and Tres2020).
One aspect that appears to be related to high rooting percentages is the ability of yerba mate mini-cuttings to keep their leaves. In the present study, the genotypes presenting the highest rooting percentage generally showed the highest percentages of leaf retention, a relationship that was statistically confirmed by the positive correlation observed between these two variables. Leaves are frequently associated with the ability of vegetative cuttings to accumulate carbohydrates during the rooting period as well as the production of endogenous auxins and rooting co-factors (Hartmann et al., Reference Hartmann, Kester, Davies Junior and Geneve2011; Belniaki et al., Reference Belniaki, Rabel, Gomes and Zuffellato-Ribas2018; Gomes et al., Reference Gomes, Machado, Miola and Deschamps2018). In a qualitative study, Pimentel et al. (Reference Pimentel, Pedroso, Lencina, Oliveira and Bisognin2020) reported that mini-cuttings of yerba mate genotypes that present small quantities of carbohydrate (starch) tend to present poor rooting performance. In a previous study with conventional stem cuttings of I. paraguariensis, Tarragó et al. (Reference Tarragó, Sansberro, Filip, López, González, Luna and Mroginski2005) had also observed a high positive correlation between leaf retention and rooting capacity. The present study reinforces this concept for cloning of the species as we show leaf retention is also important for more rejuvenated propagules (mini-cuttings) and not only conventional stem cuttings.
Although there is a difference between roots initiation and root growth, in the present study, genotypes presenting the highest rooting percentages also showed the highest number and length of roots, as evidenced by the positive correlations among these variables (Table 3). In general, the greater the number of roots, the greater is the absorption of nutrients and water by plants, resulting in better growth rates (Navroski et al., Reference Navroski, Nicoletti, Lovatel, Pereira, Tonett, Mazzo, Meneguzzi and Felippe2016) and better survival rates after transplanting (Tsakaldimi et al., Reference Tsakaldimi, Ganatsas and Jacobs2013). Roots length varied from 0.32 to 4.0 cm in the present study, with values similar to those reported by Sá et al. (Reference Sá, Portes, Wendling and Zuffellato-Ribas2018) in a study with mini-cuttings of I. paraguariensis.
The percentage of mini-cuttings with calluses ranged from 0 to 62.5% and did not present a clear relationship with mini-cuttings rooting or mortality. On the other hand, the percentage of mini-cuttings with both calluses and roots presented a clear positive correlation with rooting percentage, number and length of roots. According to Fachinello et al. (Reference Fachinello, Hoffmann and Nachtigal2005), the processes of callogenesis and adventitious rhizogenesis roots are generally influenced by the same factors, and can occur simultaneously, without, however, having a direct relationship.
Mini-cuttings from yerba mate were previously reported to present direct adventitious rhizogenesis (Stuepp et al., Reference Stuepp, Bitencourt, Wendling, Koehler and Zuffellato-Ribas2017), meaning they will develop roots regardless of previous callus formation. In addition, it is also reported that callus formation can be detrimental to rooting in the species (Stuepp et al., Reference Stuepp, Bitencourt, Wendling, Koehler and Zuffellato-Ribas2017). However, in the present study, the formation of calluses did not appear to be detrimental for rooting and, for genotypes EC28 and EC44, all rooted mini-cuttings presented callus formation before the emission of adventitious roots, evidencing indirect rhizogenesis. On the other hand, for genotypes EC20 and EC50, respectively 64 and 100% of rooted mini-cuttings did not have rhizogenesis preceded by callus formation. Therefore, it can be inferred that the direct or indirect rhizogenesis is genotype-depended in I. paraguariensis.
Considering the mini-cuttings long time of permanence in the greenhouse, a relatively high percentage of alive mini-cuttings (non-rooted and without calluses) was observed in some genotypes (up to 32.5%). Wendling and Santin (Reference Wendling and Santin2015) stated that 120 days under greenhouse conditions is the maximum period recommended for mini-cuttings of I. paraguariensis due to economic aspects, among other reasons. Alive mini-cuttings, along with the relatively high percentage of non-rooted mini-cuttings with calluses (up to 62.5%), evidences that mortality (usually related to dehydration of herbaceous mini-cuttings) was not the main reason for the lack of rooting in some genotypes of I. paraguariensis. Therefore, environmental stimuli such as plant growth regulators in higher concentrations than the one used in the present study (3000 mg/l indolebutyric acid) may be a strategy to increase rooting percentages on genotypes with both lower rhizogenic potential and mortality. Sá et al. (Reference Sá, Portes, Wendling and Zuffellato-Ribas2018) reported increases in rooting percentages of mini-cuttings of yerba mate when using indolebutyric acid up to the concentration of 8000 mg/l.
Mortality, however, was an issue for some genotypes such as EC31, which presented up to 65% of mini-cuttings with necrotic tissues after 120 days in the greenhouse. Given the interesting results on chemical attributes (mainly high contents of caffeine, proteins and total phenolics) and antioxidant capacity (DPPH), further strategies to improve the survival of mini-cuttings from this genotype are required. Environmental aspects such as relative humidity, substrate chemical and physical characteristics as well as seasonality effects are some of the factors that may be related to the survival and adventitious rooting (Hartmann et al., Reference Hartmann, Kester, Davies Junior and Geneve2011).
The present study is the first report on the phytochemical composition of 15 different yerba-mate genotypes using mini-cuttings, including BRS, which has been previously studied only regarding elemental composition and agronomic performance (Wendling et al., Reference Wendling, Santin, Nagaoka and Sturion2017; Barbosa et al., Reference Barbosa, Motta, Consalter, Poggere, Santin and Wendling2018). Great variation was observed among the genotypes; therefore, different industrial applications may be suited for each genotype based on their distinct characteristics and their field performance, which is still to be studied and may differ from those observed under greenhouse conditions.
For total phenolic compounds, contents varied from 7.4% (74 mg/g) to 11.2%. These contents are similar to those reported by Cardozo Junior et al. (Reference Cardozo Junior, Ferrarese-Filho, Filho, Ferrarese, Donaduzzi and Sturion2007), when evaluating 16 yerba mate progenies from different locations in the south of Brazil (values ranging from 7.9 to 9.6%). Interestingly, total phenolic contents of the genotypes assessed in the present study were superior to those reported by Souza et al. (Reference Souza, Oldoni, Cabral and Alencar2011), who evaluated the samples of commercial yerba mate teas and reported total phenolic contents up to 6.9%. It is noteworthy that in addition to the genetic factors, several environmental factors influence the content of the compounds (seasonality, temperature, water and nutrient availability, radiation) (Verma and Shukla, Reference Verma and Shukla2015; Croge et al., Reference Croge, Cuquel and Pintro2020) as well as the leaf development stage (Blum-Silva et al., Reference Blum-Silva, Chaves, Schenkel, Coelho and Reginatto2015). However, as the present experiment was conducted under the same conditions, the variation in the contents of secondary metabolites and rooting performance found among the genotypes is probably attributed to a higher extent to their genetic components.
In regards to antioxidant capacity, some differences were observed between the two methods. The genotypes did not vary statistically according to the ABTS assay, but two groups were distinct when using DPPH. Differences between these methods have been discussed in the scientific literature and are mainly attributed to lower reactivity of DPPH radical and different reactions with specific compounds in plant extracts (Zhao et al., Reference Zhao, Fan, Dong, Lu, Chen, Shan and Kong2008; Mareček et al., Reference Mareček, Mikyška, Hampel, Čejka, Neuwirthová, Malachová and Cerkal2017). Similar to the present report, a study evaluating antioxidant assays on extracts from 12 species with different contents of phenolic compounds concluded that ABTS was, on average, the least sensitive method (Chaves et al., Reference Chaves, Santiago and Alías2020). It is convenient to emphasize, however, that both ABTS and DPPH methods presented a positive correlation with total phenolic compounds, reinforcing the fact that phenolic compounds act protectively against oxidative stress both through their ability to donate hydrogen or electrons and because they are stable radical intermediates (Chaves et al., Reference Chaves, Santiago and Alías2020). Positive correlation between this class of compounds and the antioxidant capacity was also observed by Duarte et al. (Reference Duarte, Tomasi, Helm, Amano, Lazzarotto, Godoy and Wendling2020) for yerba mate genotypes.
Protein levels in the present study ranged from 12.5 to 19%, higher than those reported by Braghini et al. (Reference Braghini, Carli, Bonsaglia, Junior, Oliveira, Tramujas and Tonial2014), who obtained maximum contents of 11.83% when assessing five brands of yerba mate tea. Chiesa et al. (Reference Chiesa, Schlabitz and Souza2012) evaluated the effect of adding yerba mate leaves to cereal bars and concluded that the species has the potential to meet market requirements, providing products with higher protein contents.
Regarding caffeine, the contents reported in the present study are higher than those described by Braghini et al. (Reference Braghini, Carli, Bonsaglia, Junior, Oliveira, Tramujas and Tonial2014) (2.45%) for different brands of yerba mate tea and by Cardozo Junior et al. (Reference Cardozo Junior, Ferrarese-Filho, Filho, Ferrarese, Donaduzzi and Sturion2007), when evaluating leaves from 16 yerba mate progenies (up to 1.66%). It is important to highlight that in the above studies, the authors used only leaves for the analyses and in the present study mini-cuttings were used. Compared to the values reported in the literature, it appears that the genotypes assessed in the present study have high stimulating potential, an interesting property for the pharmacological industry. Among the assessed genotypes, EC31 was among the top values – if not the single highest – for a set of phytochemical variables (total phenolics, DPPH antioxidant activity, proteins and caffeine contents). Plants with high levels of phenolic compounds and caffeine are advantageous for the mate tea industry, since these compounds are responsible for the antioxidant and stimulating properties (Bastos et al., Reference Bastos, Fornari, Queiroz and Torres2006), consecutively providing greater added value to the final product.
As for the relationship between rooting-related variables and chemical characteristics, there were negative correlations between rooting percentage and total phenolic compounds (−0.59, P < 0.05) and ABTS antioxidant activity (−0.58, P < 0.05). These correlations are possibly justified by the predominance of monophenols in some yerba mate genotypes, since they act in stimulating the degradation of indole acetic acid (IAA), by increasing the activity of the enzyme IAA-oxidase (Musacchi, Reference Musacchi1994; Lima et al., Reference Lima, Biasi, Zanette, Zuffellato-Ribas, Bona and Mayer2011).
Although high levels of phenolic compounds in commercial yerba mate are desired for industrial applications, in the present study, the contents of such compounds showed a significant positive correlation with the mortality of mini-cuttings (0.55, P < 0.05). Casagrande Junior et al. (Reference Casagrande Junior, Bianchi, Strelow, Bacarin and Fachinello1999) hypothesize that the oxidation of phenolic compounds promotes higher mortality on stem cuttings. Accordingly, Tarragó et al. (Reference Tarragó, Filip, Mroginski and Sansberro2012) reported that high levels of soluble phenolics in parent plants of yerba mate negatively affect non-rejuvenated stem cuttings rooting due to the products of oxidation, causing the browning and death of the propagules. However, because the phenolics are such a diverse class of natural products, more targeted analyses (on specific types of phenolic compounds) will be needed to establish a precise explanation on the negative effects of rooting.
Given the diverse response of genotypes regarding rooting percentage and chemical composition, specific vegetative propagation protocols will need to be developed for each genotype, especially those presenting high levels of phytochemical compounds with potential for pharmaceutical and tea industries, considering the chemical profile will remain stable under field conditions, which need to be determined in future studies. We emphasize that the present study only assessed rooting and chemical profile in one season of the year (autumn), which was previously reported as the most favourable for rooting yerba mate cuttings (Stuepp et al., Reference Stuepp, Bitencourt, Wendling, Koehler and Zuffellato-Ribas2017). A different study by Sá et al. (Reference Sá, Portes, Wendling and Zuffellato-Ribas2018) pointed out that spring may be the most favourable season to collect vegetative propagules of yerba mate, while Pimentel et al. (Reference Pimentel, Lencina, Pedroso, Somavilla and Bisognin2017) did not report statistically significant differences between different seasons of the year for propagule survival. This variation may be explained by the genotype-dependent effect as pointed out by Mayer et al. (Reference Mayer, Nienow and Tres2020) and needs to be addressed in future studies for the genotypes described in the present study.
Conclusions
The adventitious rooting and the phytochemical profile of yerba mate mini-cuttings are genotype-dependent. Leaf retention is a relevant factor in the rooting of yerba mate mini-cuttings and the levels of total phenolic compounds, antioxidants and theobromine present in mini-cuttings are negative components to the adventitious rooting of I. paraguariensis. Genotype EC31 shows high contents of total phenolics, proteins, caffeine and high antioxidant activity, as well as the potential for propagation via mini-cuttings. Genotypes such as EC50 (high levels of total phenolics, proteins and antioxidant activity) and EC44 (high contents of theobromine) present interesting chemical profile, however have a poor rooting performance.
Acknowledgements
The authors acknowledge the funding received from the Brazilian Federal Agency for Support and Evaluation of Graduate Education (CAPES) – Finance Code 001 for the first author and DOC_PLENO/proc. n° 88881.129327/2016-01 for the fourth author. The authors also thank the Federal University of Paraná and EMBRAPA-Forests for the technical support and structure for the development of the study.
