Phenoplasticity of Essential Oils from Two Species of Piper (Piperaceae): Comparing Wild Specimens and Bi-Generational Monoclonal Cultivars

This study tested the hypothesis that “clonal chemical heritability is a crucial factor for the conservation of chemical uniformity of Piper essential oils in controlled monoclonal cultivation”. We asexually propagated first and second-generation clones of two medicinal and aromatic species, Piper gaudichaudianum Kunth and Piper mollicomum Kunth (Piperaceae), for use as experimental models since they show high chemical plasticity in the wild. Leaves from wild specimens of both species, and their respective cultivated specimens, were hydrodistilled in a Clevenger-type apparatus to produce essential oils (EOs). EOs were chemically characterised by GC-MS and GC-FID. The analysis identified 63 compounds in EO of P. mollicomum, which were predominantly monoterpenes, and 59 in EO of P. gaudichaudianum, which were predominantly sesquiterpenes. Evaluation of chemical diversity and oxi-reduction indices showed a loss of chemical homology across the intergenerational cline. Chemometric analysis indicated higher chemical plasticity between wild and intergenerational specimens of P. mollicomum, than for P. gaudichaudianum. EO compounds were significantly less oxidized throughout the generations in both species. Therefore, while clonal heritability is crucial to chemical homology, significant chemical plasticity is likely to occur when cultivated from wild specimens.


Introduction
In Brazil, species in the genus Piper are among the most versatile of the useful plants in the country [1]. For example, the two species Piper mollicomum Kunth and P. gaudichaudianum Kunth have extensive ritualistic and medicinal uses. They are known synonymously by the vernacular name "Jaborandi" by local herbalists [2][3][4]. It is not uncommon for the same name to be used for two or more taxa with similar appearances that are used interchangeably for the same end uses. This is corroborated by ethnobotanical studies of Jaborandi that highlighted overlap of use, in particular, in the initiation of people into religions of Brazilian African origin. Furthermore, both species are used as ingredients for the bath of "amaci", for aromatic drinks, or as an incense [5,6].
These two species of Piper have numerous therapeutic applications that are recognized in contemporary practice, as well as in history. Records made in 1888 describe the use of infructescences (the fruit clusters) from P. mollicomum for the treatment of diseases of the gastrointestinal tract, as well as for venereal diseases [7]. Modern records describe the use of its leaves in the treatment of liver diseases, for the relief of spine pain, in the reduction of intense menstrual flow [8][9][10][11][12], as well as antifungal [13], antibacterial [10,14], and antinociceptive [15] applications. Another organ that is used is the roots, which are extracted and applied as a local anesthetic for toothache in the form of an aqueous infusion [7,9,11,16]. The other species, P. gaudichaudianum, is similarly an anti-inflammatory agent, also used in the relief of toothache and in the treatment of liver diseases [17]. The pharmacological potential of the extract of this species is also corroborated by in vitro studies that demonstrate fungicidal, larvicidal, anti-inflammatory and analgesic activities [18][19][20].
Regarding essential oils (EOs), those obtained from the leaves of P. gaudichaudianum have shown in vitro larvicidal, insecticidal anti-inflammatory and cytotoxic activities [21]. There is also evidence of moderate antibacterial activity of EOs produced from the leaves of P. mollicomum [16]. Previous chemical studies show that the EOs from leaves of P. mollicomum and P. gaudichaudianum are predominantly terpenoid with traces of arylpropanoids, however their chemical constitutions vary according to their collection site, season, and time of day [15,[21][22][23][24][25].
Most of the medicinal plants in Brazil that are highly regarded have not yet been put into cultivation [26]. Consequently, the increasing use of medicinal plants that are wild harvested is putting the native populations under considerable pressure, and without intervention they may become threatened. This is already a significant problem globally, considering that 40% of the world's flora is at risk of extinction and/or genetic erosion, due to excessive biota harvesting [27]. In anticipation that such problems may affect the medicinal species P. mollicomum and P. gaudichaudianum, we initiated studies related to agronomic practices.
The cultivation of medicinal plants outside their niche habitat is unregulated in Brazil, but there are efforts to acquire scientific agronomic knowledge in small-scale cultivations that remain close to natural growing land areas. Indeed, there are a few cultivation approaches with medicinal species from the genus Piper [28][29][30]. However, the circulation of seedlings and seeds of medicinal plants around Brazil, either at the community level or regionally, antagonizes the understanding and standardization of chemical and biological phenotypes within and across species [31,32]. For example, it is well known that plants grown in non-natural environments can present qualitative and quantitative variations in the production of specialized metabolites, because the chemical phenotypes are not just interconnected with chemical heritability [33], since extrinsic factors are also significant.
In this regard, the concept of "chemical heritability" is used to define the ratio of chemical diversity over genetic diversity within a population [34,35]. Chemical heritability recognizes the influence of exogenous factors in bringing about chemical diversity within a population. Through different concepts and in practice, chemical heritability in plants can be further evaluated from sexual (seed) and asexual (clonal) propagation. In the first, genetic factors may still influence the chemical diversity that is measured from cultivated plants. However, in the second method of clonal propagation, the elimination of genetic diversity across replicates is the key to understanding the phenomena of chemical phenotypic plasticity, and obviously, the mechanisms of chemodiversity [36,37]. Thus, clonal propagation is used in the current study as the chosen method to answer the question: "Is clonal chemical heritability a crucial factor for the conservation of chemical and chemodiversity characteristics of Piper foliar essential oils in controlled cultivation?" To answer this question, this work aims to evaluate the clonal heritability of EO chemodiversity between wild and intergenerational specimens P. gaudichaudianum and P. mollicomum.

Results
The EO composition and yield obtained from the leaves of P. gaudichaudianum wild (PGW), first (PGF) and second (PGS) generations, as well as P. mollicomum wild (PMW), first (PMF) and second (PMS) generations are shown in the Table 1. The EOs were slightly yellow in colour. The lowest yield was registered for PMF (0.12%), and the highest for PMW (0.86%). The yields of P. gaudichaudianum ranged from 0.13-0.18%. Table 1. Results of the analysis of essential oils from Piper mollicomum (PM), referred to as wild, first-and second-generation cultivars (PMW, PMF and PMS, respectively), and Piper gaudichaudianum (PG) referred to as wild, first-and second-generation cultivars (PGW, PGF and PGS, respectively).
A qualitative analysis of the EOs of both species is presented in the form of a Venn diagram ( Figure 1). The amount of overlap between chemical profiles across the three generations of the species is captured by this analysis. The profile of EOs from leaves of P. gaudichaudianum ( Figure 1A) demonstrates that there are 18 compounds (32.10%) common to the three generations studied. However, at the level of individual specimens, these 18 compounds represent 67.36%, 67.86% and 81.33% of the total number of compounds in the profile of PGW, PGF and PGS, respectively. Six compounds for PGF were unique to this generation: α-pinene, αylangene, amorpha-4,7(11)-diene, β-calacorene, 5-epi-7-epi-α-eudesmol, and 7-epi-α-eudesmol. However, they are only minor compounds (<1%). For PGS, only one compound was exclusive to this sample, again in low relative percentage (E-β-farnesene, 1.73%). In contrast, the 30 major compounds that account for more than 80% of the total profile, are shared between PGW and PGF. This commonality is not seen in the final propagated clone (PGS), which only shares eight compounds with its progenitor (PGF).

Chemical Variation in P. mollicomum
A total of 63 compounds were successfully identified in the EOs of wild P. mollicomum (PMW), corresponding to 98.34% of the whole mass. EOs are dominated by monoterpenes (oxygenated and non-oxygenated; 79.40%), with a fraction of sesquiterpenes (oxygenated and non-oxygenated; 18.60%). Over the generations in cultivation, the class of identified compounds gradually became more sesquiterpenoid (first generation (PMF): 25.91%/second generation (PMS): 45.25%), with evident reciprocal declining of the monoterpene composition (PMF: 56.07/PMS: 37.33%). In addition, the diversity of the identified compounds in cultivated specimens was considerably reduced, particularly from PMF to PMS.

Chemical Variation in P. mollicomum
A total of 63 compounds were successfully identified in the EOs of wild P. mollicomum (PMW), corresponding to 98.34% of the whole mass. EOs are dominated by monoterpenes (oxygenated and non-oxygenated; 79.40%), with a fraction of sesquiterpenes (oxygenated and non-oxygenated; 18.60%). Over the generations in  Figure 2 shows the results of the hierarchical cluster analysis (A) and the main component analysis (B) of chemical profiles from both the species and their intergenerational specimens. Figure 2 demonstrates that intergenerational differences between the clones was less of a discriminating factor, compared to interspecies differences. This clarifies that clonal heritability remains as a robust factor in chemical expression patterns. The first  Table 2 demonstrates several inorganic characters of the native soils from which the wild specimens were collected. Compared to the propagation soils of the clones, the wild soils had a significantly lower pH, a lower salt content and a higher mineral content. The influence of these abiotic factors has not been examined in any further detail.   Table 2 demonstrates several inorganic characters of the native soils from which the wild specimens were collected. Compared to the propagation soils of the clones, the wild soils had a significantly lower pH, a lower salt content and a higher mineral content. The influence of these abiotic factors has not been examined in any further detail.

Chemodiversity and Micromolecular Analysis
The chemical data in Table 1 is evaluated using two 'indices', which is the Shannon index, and the recently developed oxy-reduction index (Ramos and Moreira Index: GM RO ) [21]. These evaluations demonstrated a trend in the context of clonal chemical heritability.
The Shannon index changed across the generations, from wild to the second generation for both species. Specifically for P. mollicomum the difference is from 4.14 to 3.67 (r 2 = 0.944) and for P. gaudichaudianum it went from 3.81 to 3.30 (r 2 = 0.784). These results suggest a loss of chemodiversity from wild to cultivated (first and second generations). This loss of chemical diversity is more distinguished for P. mollicomum than P. gaudichaudianum.
The evaluation of the oxy-reduction characteristics of the two EO mixtures for P. mollicomum showed values from −3.73 to −2.55 (r 2 = 0.9435) and for P. gaudichaudianum from −6.94 to −3.39. These results mean that EO mixtures are less oxidized from wild to cultivated (first and second generations); in other words, EO mixtures reduced with acclimatation. It is not clear if this acclimation was to the abiotic factors measured and summarized in Table 2, or if other factors are significant, such as water regime, light exposure, or others.

Discussion
The high sesquiterpene content of P. gaudichaudianum in the current study is in accordance with literature data for other species in Piperaceae [17,38]. Chemical studies of the same species are also in agreement, i.e., Peres et al., [39] analysed the EO from leaves of P. gaudichaudianum that was collected in Rio Grande do Sul (Brazil). They described E-nerolidol (22.4%) as the major constituent, and in minor relative percentages, E-caryophyllene (8.9%) and bicyclogermacrene (7.4%). That study also showed cytotoxic, genotoxic, and mutagenic effects that were correlated to the presence of E-nerolidol, α-humulene and E-caryophyllene [39].
Another paper on the same species and the same Brazilian State, demonstrated a very different chemical character of volatiles from P. gaudichaudianum. The profile was dominated by the arylpropanoid dillapiole and the sesquiterpene α-humulene [40]. That study was carried out over a year of sampling, registered less relative percentage of E-caryophyllene, however, compounds bicyclogermacrene and Z-eudesma−6,11-diene were not detected.
A study managed by our group [21] showed nine chemotypes for P. gaudichaudianum (i.e., δ-cadinene, 1-epi-cubenol, longipinanol, viridiflorol, α-humulene, E-caryophyllene, germacrene B, dillapiole and bicyclogermacrene), confirming this pronounced phenotypic chemical plasticity. Furthermore, we showed that EOs from P. gaudichaudianum from Rio de Janeiro are normally the bicyclogermacrene chemotype. In the current work we have demonstrated that the bicyclogermacrene chemotype is conserved between first and second generations from a wild sample, confirming that the phenotypic expression for the group located in the city of Rio de Janeiro is likely a genotypic characteristic.
In relation to P. mollicomum, linalool is the major compound of the EO in this study, having been reported in all studied samples (7.26-36.99%). This monoterpene is commonly found in some species of the genus Piper, i.e., P. aduncum [47]; P. jacquemontianum, P. multiplinervium, and P. darienense [48]. Linalool is well known for its biological activities, as reported in the literature, such as antifungal activity against Candida albicans [49], antibacterial activity against Staphylococcus aureus and Escherichia coli [50], its action against periodontopathic and cariogenic bacteria [51], and its trypanocidal activity against trypomastigote forms of Trypanosoma cruzi [52]. Thus, it is suggested that the cultivation of P. mollicomum may provide significant amounts of a native lavender EO for market interest. The essential oil can be used as a key fragrance in the development of new aesthetic products.
Linalool also has an interesting ecological backstory. Literature reports showed that grape species exposed to high levels of UV light demonstrate an increase in the biosynthesis of this monoterpene. It is possible that UV light interacts with some enzymes, such as linalool synthase, which is involved in the bioconstruction process of cyclic and acyclic monoterpenes [53]. In contrast, other studies demonstrated that linalool is decreased when the plant is exposed to high UV radiation, while other cyclic monoterpenes such as camphor and 1,8-cineole are preserved [54]. UV radiation is not the only factor that is implicated in chemical expression. For example, Blande and Glinwood [55] identified an increase in linalool content in the vegetative parts of plants exposed to stress, due to abiotic factors.
Another compound of interest that is present in the chemical profile of EO from P. mollicomum is benzyl benzoate. This compound is rare in EO; however, it has already been described for species in the genus Piper, i.e., in EO produced from P. retrofractum (14.40%) and P. sarmentosum (49.50%) [56].
Other studies on the EO of P. mollicomum describe a volatile mixture that differs from our results. Santos et al., [38] published an EO composition (leaves collected in the city of Paraty, Rio de Janeiro State, Brazil) richer in oxygenated sesquiterpenes (39.04%) in higher content than non-oxygenated (21.29%). However, our group [24] described similar results on the composition of the P. mollicomum EO collected in Rio de Janeiro city. We found 1,8-cineole and linalool as the main compounds of the volatile mixture. Thus, it is suggested that there are multiple chemotypes for the species, even within the State of Rio de Janeiro.
Qualitative statistical analysis (Venn diagram) serves as a tool to ensure the chemical uniformity or inconsistency of EOs, as the influence of minor components can be more than is expected, such as by conferring synergistic effects [57][58][59]. Thus, chemotypes and phenotypic plasticity are important considerations in the context of commercialization of essential oils with associated health claims [60][61][62][63].
The high phenotypic plasticity of P. mollicomum has been studied in wild samples [24], but this is the first study of plants in a controlled environment. This finding contributes mainly to the medicinal and ritualistic use of P. mollicomum in Afro-Brazilian religious communities that cultivate samples around their sacred spaces. The matrices are, generally, obtained from both wild and propagated specimens, as we did in this study. Therefore, with consideration to loss or change in chemical composition, the expected therapeutic effects may alter [5,6].
Total compounds in P. mollicomum samples varied significantly, whereas in P. gaudichaudianum samples it is quite constant. Current evidence from our group shows that different Piper species present different chemical plastic responses at the same collection site [21,24,26]. We also have registered that P. mollicomum showed a greater acclimatization response to abiotic factors from day-to-day [26] and P. gaudichaudianum throughout the hours of the day [21].
In summary, it was possible to determine that the diversity of compounds decreased in the first and second generations, for both species. This fact may be related to the cultivation environment, since outside of its natural habitat these plants may be under stress promoted by predation, asides from the different soil composition and micro-climate. These results may suggest that ecological interactions in the natural habitat are important for maintaining chemical diversity [57,58]. We emphasize that there are several factors that modify the chemical composition of EOs, being abiotic and biotic factors [24,70]. The age difference between the cultivated and wild specimens of the two studied species is a limiting factor and should be better investigated on a larger time scale. However, the different plastic responses registered mainly for P. mollicomum reflect the complexity of the biosynthetic pathways vs. biotic and abiotic factors that are involved. Our findings are new for these two species of Piper but have been reported before for other species. For example, Satureja hortensis L., popularly known as garden savoury, showed a composition rich in the monoterpene carvacrol in cultivated specimens, but it was rich in thymol in the wild specimens [71], which are two compounds from the same biosynthesis pathway. Our results demonstrated a similar type of difference in the EO composition from leaves of P. mollicomum, but it is a dichotomy of 1,8-cineole/linalool. In addition, it is known that selection leads to micro-evolutionary changes in the composition and diversity of specialized metabolites, which may occur over only a few generations in time scales relevant to ecological interactions and forms of cultivation [72][73][74][75]. The passing of generations in commercial crops may be an important criterion for determining permanent phenotypic characteristics in species, especially in the process of determining and differentiating chemotypes and genotypes, respectively. Therefore, this report characterized these changes based on this phenomenon in the propagation of P. gaudichaudianum and P. mollicomum.
The main concerns of the current study are the selection of the explant/matrices, cultivation conditions, medium composition, culture age, genotype, temporal variations, number of subcultures, use of phytohormones, and regeneration methods. These are aspects significant for evaluating the stability and genetic and epigenetic variation of plants regenerated through propagation [76,77]. There are two ways in which epigenetic pro-cesses can contribute to microevolution in natural populations. First, if heritable epigenetic variation is translated into phenotypic variation and into adjustment differences between individuals, then epigenetic processes can provide a second system of heritable variation for natural selection to act upon, such as one based on genetic variation. On the other hand, epigenetic variation, in contrast with genetic variation, can be altered directly by abiotic factors and consequently provide an additional route for accelerating evolutionary change [73,74]. These variations were observed quantitatively with the loss of chemodiversity with the two studied species of Piper. However, in the language of molecular biology the term "epigenetic inheritance" is used for mitotic and meiotic inheritance of epigenetic modifications. This term is confused in classical genetics and evolutionary biology in which the term inheritance is usually restrictive to the description of transgenerational inheritance through meiosis [78,79].
The adequacy of an experimental design can be a significant tool for the selection of chemical characteristics for use in human health. These can be an alternative to study epigenetic evolution in action, in which it is guaranteed that the genotypes of the same plant are subjected to different environments, their descendants are raised in a common environment for several generations; after which phenotypic and epigenetic differences are quantified and statistically compared. If we find that the descendants of those lineages that have been submitted to different environments remain phenotypically different, and at the same time, they show a significant shift in patterns of DNA methylation, gene, or protein expression, despite being still identical at the DNA level, this is going to be the best evidence for rapid evolution based on epigenetics. This field brings interesting perspectives to studies on natural products [73,77,[80][81][82][83]. To exemplify, associating this notion with human food consumption, a study evaluated the process of domestication of wheat, from three subspecies of Triticum turgidum L. (wild emmer, emmer, and durum wheat). The authors qualitatively investigated a mixture of 51 central metabolites. A reduction in unsaturated fatty acids was observed, while there was a decrease in amino acids characterized in secondary domestication (that of durum wheat) [73]. Loss of these two chemical markers was essential for improvement to the sensory characteristics of this species and in industrial application. Correlating with our results, chemical characteristics of the EO had less oxidized metabolites over the generations and loss of chemical diversity that reveal more about the acclimatization process [73,74]. We believe that these insights are initial postulates to decipher chemodiversity using Piper as a model, as it is a species of rapid propagation.

Species Propagation Protocol
Wild cuttings were collected from orthotropic branches of standard species, in three (n = 3) nodes, which were grown in a greenhouse at the Center for Social and Environmental Responsibility of the Botanical Garden of Rio de Janeiro, in the city of Rio de Janeiro, State of Rio de Janeiro (Brazil) in February 2017. Cuttings were disinfected with sodium hypochlorite at a concentration of 0.5% for 5 min, then washed with water, and finally dried with paper to remove excess moisture. The proximal ends of the cuttings were inserted 5 cm deep in 1 L plastic tubes, with the commercial substrate Tropstrato HT HORTALIÇAS ® for a period of 60 days.
The material was maintained with automatic irrigation and shading of 70%, until the formation of the root system, enabling transplantation. After rooting, the cuttings were transplanted into 10 L pots. When reaching 0.5 m, new cuttings were removed for planting the second generation, repeating the same procedure. The design was randomized, with twelve replications per species, and in triplicate. In January 2018, the two species and their surviving generations in cultivation reached length of approximately 90 cm. At this time, leaves were harvested to produce EOs.
Soil samples were obtained from Tijuca National Park where the two species were collected, considering five points near the base of each specimen, as described by Arruda et al., [84]. Chemical analysis of the soil, as well as of the commercial substrate Tropstrato HT HORTALIÇAS ® used in cultivation, was carried out by the Brazilian Agricultural Research Corporation-Soil Division (EMBRAPA-SOILS) ( Table 1) according to Teixeira et al., [85].

Essential Oil Production
Leaves (100 g) were obtained from wild and cultivated species (1st and 2nd generation) of P. mollicomum and P. gaudichaudianum. Hydrodistillation was achieved in a Clevengertype apparatus [86]. Fresh leaves were comminuted manually, with the aid of scissors, and placed in a 2L glass round bottom flask containing 700 mL distilled water. The flask was subjected to heating until boiling, and the procedure was continuous for 2 h [24,26]. After completion of the process, the pure EOs were separated from the aqueous phase, dried with anhydrous sodium sulfate, and stored in closed dark amber bottles in a freezer at −20 ºC until the time of analysis. Yields were calculated by the ratio of the volume in mL of oil and the weight in g of the fresh plant material used in the extraction, multiplied by 100, to express in percentage content [24,26]. The experiment was carried out in triplicate.

Essential Oil Analysis
Chemical characterization and quantification of the EOs of P. mollicomum and P. gaudichaudianum were carried out by Gas Chromatography (GC) coupled to Mass Spectrometry (MS) and GC coupled to the Flame Ionization Detector (GC-FID), respectively [24,26]. The EOs were diluted in dichloromethane (HPLC grade, TEDIA, Rio de Janeiro, Brazil) before analysis (1 mg/mL-1000 ppm). A sample of 1 µL of this solution was injected direct (splitless) into an HP AGILENT GC 6890 coupled to selective mass detector from the AGILENT MS 5973-N series, in which the injector temperature was set at 270 • C. Mass spectral ionization energy was set at 70 eV. An HP-5MS capillary column [AGILENT J & W; GC columns (Santa Clara, CA, USA)] with 30 m × 0.25 mm i.d. × 0.25 µm particle size was used. The chromatography conditions were 60 to 240 • C at 3 • C/min, totalling 60 min. Helium (~99.99%) was used as carrier gas at 1.0 mL/min in constant flow rate, and mass spectrometer operated at mass range from m/z 40 to 600 atomic mass units (u) [24,26].
GC-FID was obtained using a solution of 1 µL EO in dichloromethane (HPLC grade, TEDIA, Brazil, 1 mg/mL-1000 ppm), which was injected under the same analytical conditions described for GC-MS, except for the carrier gas used (hydrogen) with a constant flow rate of 1.0 mL/min. The retention times (Rt) of the compounds were measured in min, without correction, and used to calculate the linear retention indices (RI), that were obtained from the injection of a homologous series of n-alkanes (C 8 -C 25 ; Sigma-Aldrich, Brazil), under the same analytical condition of the sample. Relative percentage areas were obtained from the GC-FID analysis [24,26,87]. Quantification was accomplished using sensitivity values determined from calibration curves that were developed with the use of external standards [24,26].
For compound identification, the mass spectral fragmentation pattern of the constituents was compared to the commercial GC-MS library NIST 98 and WILEY 7n [24,26] and the RI values were compared with those published in the literature [88]. In addition, co-injection with authentic standard was carried out wherever possible as described previously [21].

Evaluation of Chemodiversity and Micromolecular Parameters
To evaluate the chemical diversity of the EOs, to facilitate comparison between wild and intergenerational cultivated species, the Shannon index was used [89], as follows: Although Shannon's index is normally used to determine species diversity in an ecosystem, by treating each compound identity as a species and its relative abundance as the total number of species, chemical diversity could be calculated. In these equations, P i is equivalent to the proportional abundance of the respective compound, which is obtained by dividing the quantity as determined by GC-FID by the total number of compounds identified in the sample, of which i is that number.
To characterize the micromolecular oxidation-reduction of the EO mixture, Ramos and Moreira's index for mixtures was applied [21]. The equation is given below: In these equations, the N OR is the weighted oxidation state of the substance of interest and it is obtained by multiplying by the quantitative value of the substance found in the sample and divided by the number of carbon atoms in the molecular skeleton (n). The GM OR is then obtained by the sum of N OR of all substances in the mixture, divided by the number of identified substances (N IA ) in the sample. A lower GM OR indicates that the mixture has a lower average oxidation state by comparison with a sample that has a higher GM OR [21].

Statistical Analysis
All data on the percentage of compounds in the EO were reported as mean ± standard deviation for three independent experiments (extraction). For the qualitative statistical analysis, the data referring to the constituents of the EO were submitted to the Venn diagram to verify the degree of similarity between the samples. Statistical significance was assessed using the Tukey test (ANOVA by Tukey HSD post hoc test). The chemometric analysis, principal component analysis (PCA) and hierarchical analysis (HCA) were used to assess the variance between the EO of wild and intergenerational cultivated species. The results were processed using STATISTICA software version 10 (StartSoft Inc., Tulsa, OK, USA).

Conclusions
The cultivation of P. mollicomum drastically changed its volatile composition. Therefore, before cultivating P. mollicomum it is necessary to identify the chemotype of the specimen and establish a study to determine the growing conditions and the factors that influence the essential oil production. P. gaudichaudianum showed almost the same volatile compounds in wild and cultivated specimens, which suggests that this plant tends to be less externally influenced. Our findings shed some light on issues that matter in the management, use and conservation of these two species of Piper, widely used as medicinal applications, in addition to contributing to the ecological and chemophenetic knowledge of these Piperaceae.