P. rivinoides has the liturgical name "Èwé Boyí Funfun" according to ethnobotanical surveys in Brazil, and its use in cleansing baths in religious rituals originated in Africa. P. rivinoides has been used medicinally to treat wounds, ulcers, vaginal discharge, bleeding, and even oral problems. Furthermore, this plant is sold in folk markets and collected from wild specimens (Barros, 2010; Mandarino and Gomberg, 2009; Albuquerque et al., 2007).

Previous research with the extracts of this species revealed that the hexane fraction contains a high concentration of bioactive benzofuran neolignans (eupomatenoids and conocarpan) and pharmacological potential against Candida albicans and Leishmania amazonensis (Moreira et al., 2016; Marques et al., 2014).

Literature data on EO, obtained by hydrodistillation from fresh leaves of P. rivinoides, have shown a great chemical diversity rich in monoterpenes, sesquiterpenes, and arylpropanoids, which may vary depending on the collection site (Bernuci et al., 2016; Perigo et al., 2016; Leal et al., 2019). Aspects of the pharmacological potential of these EO indicate their antinociceptive (Souza et al., 2014), antifungal (Bernuci et al., 2016; Leal et al., 2019), antimicrobial (Leal et al., 2019), and cytotoxic (Fonseca et al., 2020; Machado et al., 2021) activity.

Recently, the apoptotic effect of β-pinene (main component) and EO of P. rivinoides in oral squamous cell carcinoma (Machado et al., 2021) was described. This study confirmed the plant’s long history of traditional medicinal use in the treatment of oral problems (Barros, 2010). Essential oil research is of interest in the production of herbal raw materials due to their functions as medicines, cosmetics, and other products. Because of the low cost of extraction, processing is advantageous in low-income areas and allows for large-scale production with low operating costs, which would become profitable through incentive programs and public project funding (Araújo et al., 2021).

Despite the fact that P. rivinoides is widely used in traditional medicine, research on the seasonal variation profile of EO from this plant is limited. It is well known that biotic and abiotic factors influence the large chemical phenotypic plasticity of Piperaceae species, and that these variations in the composition of plant chemistry can lead to changes in their respective biological potential (Monzote et al., 2010; Salehi et al., 2019).

Based on the foregoing, this study aims to evaluate the effect of seasonality on the chemical profile of EO of wild specimens of the medicinal plant P. rivinoides collected from two different sites in fragments of the Atlantic Forest with varying altitudes and climatic patterns. Determining the chemical nature of EO of P. rivinoides in different months over a year and at two different collection sites can help us understand the chemotaxonomics and chemophenetics of the plant and the genus Piper. Furthermore, this study provides tools for determining the best time to collect plant material that can provide an EO with high-quality secondary metabolites and, consequently, biological activities.

Plant

Samples of fresh leaves of wild P. rivinoides were collected from two specimens in different regions of the Atlantic Forest in Rio de Janeiro State: (a) Tijuca National Park, Rio de Janeiro/RJ (PRR) (S 22° 57' 09. 94"; E 43° 01 '17. 19", elevation 205 m (permit SISBIO n. 57296-1)  and (b) Serra da Tiririca State Park, Niterói/RJ (PRN) (S 22° 97' 18.24"; E 43° 24' 23.78", elevation 28 m (INEA permit n° 07/ 0002.4729/ 2019). Dr. Elsie Franklin Guimares performed the botanical identification of the plant material at the Research Institute of the Botanical Garden of Rio de Janeiro, and the herborized specimens were deposited in the Herbarium of the College of the State of Rio de Janeiro (HRJ) under voucher numbers 13404 for PRR and 13403 for PRN. In each season of 2019, one specimen was collected twice: summer (January and February), autumn (April and May), winter (July and August), and spring (October and November), always in the morning between 9:00 and 10:00 am. Data on abiotic factors, including average temperature (°C), humidity (%), and precipitation (mm), were obtained from the Brazilian Institute of Metrology and Research (INMET) and are presented in Figure 1.

Compounds identification and statistical analysis

Retention indices (RI) were calculated from the retention time of a homologous series of saturated aliphatic hydrocarbons (C₈-C₂₈, Sigma-Aldrich, Brazil) obtained with CG-FID under the same analytical conditions as EO (Dool and Kratz, 1963). Compounds in the EO were identified by comparing their mass spectra with database records (WILEY 7n, NIST) and by comparing the calculated RIs with those from the literature (Adams, 2017). Furthermore, co-injection with authentic standards was performed wherever possible, as described previously. All data on the percentage of compounds in the EO were reported as mean ± standard deviation for three independent experiments (extraction).

Principal component analysis (PCA) and hierarchical analysis (HCA) graphs were constructed for chemometric analysis, in which the chemical composition of EO was treated as operational taxonomic units. For this purpose, the percentage values (% area) were extracted from the CG data, transformed into a sinusoidal arc of the root of p, and converted into a matrix, excluding unidentified compounds (Catinella et al., 2021). The Euclidean distance was combined with the UPGMA (Unweighted Pair Group Method with Arithmetic Mean) method to generate the dendrogram. Multiple comparisons between the main components of EO and abiotic factors (temperature, humidity, precipitation) were performed using Pearson correlation (p < 0.05). All results were processed using STATISTICA version 10 software (StartSoft Inc., Tulsa, USA).

Essential oil yield

A total of 16 colorless samples were obtained from EO. These results are presented in Table 1. The yields found in this study for wild P. rivinoides ranged from 0.16 to 1.36 (w/w) (Table 1), higher than others previously reported in the literature (0.05 (Souza et al., 2014) and 0.63 (Perigo et al., 2016)). Monthly yields varied similarly at both collection sites, with no site dominating (p > 0.05) (Figure 2).

There was a considerable increase  in  the EO  yield  in the spring, which coincides with the reproductive phase of the plant (flowering). A tendency to increase yield was also noticed as the temperature decreased. So, a lower quantity of EO was obtained in the hottest months of the year (summer). Similar results were found for other Piperaceae species such as Piper mosenii C. DC and Piper gaudichaudianum Kunth (Del Quiqui et al., 2019). Other reports in the literature also corroborate these findings (El Beyrouthy et al., 2015; Marques et al., 2014).

Pearson's analysis did not show correlations between EO yield and precipitation and temperature. However, there was a moderate correlation between EO yield and humidity for both locations (r = 0.5954 PRN and r = 0.6588 PRR), but significant only for the results of PRN (p < 0.05). According to the literature, in humid environments where water is not a limiting factor, as in rainforests, there is an increase in the rate of plant transpiration, which may explain this correlation. When the rate of metabolism increases the rate of volatilization decreases, then the EO yield is directly affected (Queiroz et al., 2017). AHC of bioclimatic data suggested that general volatile differentiation in P. rivinoides from Rio de Janeiro essentially does not represent an adaptive response to bioclimatic, or geological conditions. Therefore, there was no clear pattern of correlation between climate and EO yield for P. rivinoides, suggesting the biggest connection to the vegetative cycle than climatic variations, a fact previously described for other plants (El Beyrouthy et al., 2015; Shadi and Saharkhiz, 2016).

In the spring, which coincides with the reproductive phase of the plant (flowering), there was a significant increase in EO yield. A tendency to increase yield was also observed when temperatures decreased. Thus, a lower amount of EO was obtained during the hottest months of the year (summer). Similar results were found for  other  Piperaceae  species  such as P. mosenii C. DC and P. gaudichaudianum Kunth (Del Quiqui et al., 2019). Other reports in the literature also confirm these results (El Beyrouthy et al., 2015; Marques et al., 2014).

Pearson analysis showed no correlations between EO yield and rainfall and temperature. However, there was a moderate correlation between EO yield and moisture for both sites (r = 0.5954 PRN and r = 0.6588 PRR), but it was significant only for the results of PRN (p < 0.05). The literature states that in humid environments where water is not a limiting factor, such as rainforests, plant transpiration rate increases, which could explain this correlation. When the metabolic rate increases, the rate of volatilization decreases, affecting the EO yield directly (Queiroz et al., 2017). The AHC of bioclimatic data suggests that the overall differentiation in volatility in P. rivinoides from Rio de Janeiro is essentially not an adaptive response to bioclimatic or geological conditions. Therefore, there was no clear pattern of correlation between climate and EO yield in P. rivinoides, suggesting that the greatest link is to the vegetative and reproductive cycles than to climatic variation, a fact previously described for other plants (El Beyrouthy et al., 2015; Shadi and Saharkhiz, 2016).

Chemical profile of the essential oils

The relative percentage of compounds is shown in Table 1. Forty-one compounds were identified in summer, 37 in fall, 24 in winter, and 21 in spring, representing 98.74, 97.56, 97.58, and 98.56%, respectively, of the total compounds identified.

Increase in the relative percent content of b-pinene in January (44.30%) and July (44.39%) and decrease in the percent content of limonene in the same months (7.08% and 3.07%, respectively) were highlighted in the Serra da Tiririca-Niterói/RJ collecting area (PRN). This variation was not observed in the specimens from the Tijuca/RJ collecting area (PRR), which even showed greater uniformity in the content of α-and β-pinene and greater variation in the content of limonene (2.32% in February to 9.09% in January).

Chemical analysis in the different seasons showed that the intraspecific variation in the EO of leaves of P. rivinoides collected in the two study areas was very low (Figure 3A-B). PCA analysis showed that it is not possible to distinguish the two areas or the four seasons with respect to the biosynthesis of compounds present in EO. This  is confirmed by the Euclidean distance formed in the dendrograms (Figure 4A  and B), where clusters can only separate samples with a higher content of α-pinene from samples with a higher content of β-pinene and/or limonene, without separating the two study areas. This implies that P. rivinoides exhibits some stability of chemical phenotype in the two different collection areas. Although many studies have reported variation in the chemical profile of EO from the same plant species in different collection areas (Marques et al., 2019; Queiroz et al., 2017; Morshedloo et al., 2018), it is also possible that the variation (chemotype) is small, implying that intrinsic biotic factors are more influential than abiotic factors (Elechosa et al., 2017; Souza et al., 2017).

The results of the present study differ slightly from other results available in the  literature, which describe a volatile fraction for P. rivinoides composed exclusively of monoterpenes, with a α-pinene (73.2%) and a lower β-pinene (5.2%) for samples collected in the state of São Paulo/Brazil (Perigo et al., 2016). In another study with wild-collected P. rivinoides in the state of Paraná/ Brazil, the sesquiterpenes bicyclogermacrene (11.8%) and α-humulene (10.0%) were registered as the maincomponents, with a lower relative proportion of α-pinene (4.4%) and and β-pinene (3.7%) (Bernuci et al., 2016). On the other hand, this study agrees with the studies of Souza et al. (2014), who describe as main constituents the monoterpenes α-pinene(32.9%) and b-pinene (20.9%), besides limonene (3.22%). It is worth noting that the collection sites of the present study are fragments of the Atlantic Forest, as well as the specimen cultivated by Souza et al. (2014), whose specimen originated from the Atlantic Forest, suggesting a possible ecotype or chemotype for this species.

It is also notable that metabolites from the acetate-mevalonate or methylerythritol phosphate pathway in the EO are exclusive to both plants studied in the present study. Secondary metabolites from, for example, shikimate was not detected, although they are common in Piperaceae and have been identified in other EO samples of P. rivinoides. The EO of this plant, collected from the Bom Jesus Biological Reserve in the Brazilian state of Paraná, had a chemical composition completely different from the others previously described. The arylpropanoid  E-isoelemicin  was  the   main   constituent (40.81%), followed by non-oxygenated monoterpenes (16.88%) and non-oxygenated sesquiterpenes (12.45%).  In this study, the compound α-pinene, β-pinene and limonene were not detected (Leal et al., 2019).

The presence of α-pinene and β-pinene in a high relative percentage in EO of P. rivinoides from Rio de Janeiro state is significant because studies conducted with these terpenes have shown various biological activities, including activity against infectious bronchitis virus (IBV) with IC50 of 0.98 ± 0.25 and 1.32 ± 0.11 mM for α-pinene and β-pinene, respectively (Yang et al., 2011).

In addition α-pinene showed a dose-dependent growth inhibitory  effect  on  hepatocellular  carcinoma  cell  lines (BEL -7402) with an inhibition rate greater than 79% for all concentrations tested (Chen et al., 2015). A large apoptotic potential or cytotoxic effect against tumor cell line has been reported in the literature for β-pinene (Fonseca et al., 2020; Machado et al., 2021). An assay evaluating nematocidal activity against Bursaphelenchus xylophilus described that among 97 compounds evaluated, the monoterpenes (+)-α-pinene and (-)-α- pinene showed the highest inhibition with 83% and 80%, respectively (IC₅₀ 0.64 and 18.03 mM, respectively). For (+)-β-pinene and (-)-β-pinene, the inhibitory effect was moderate at 39% and 35%, respectively (IC₅₀ was not described) (Kang et al., 2013).

Studies with EO of Baccharis reticulata Ruiz & Pav. α-pinene, dillapiole and β-pinene, showed excellent antibacterial potential against multidrug-resistant strains of Staphylococcus aureus (MICs ≥256 μg/mL), Pseudomonas aeruginosa (MICs ≥1024 μg/mL), and Escherichia coli (MICs ≥ 1024 μg/mL). The authors concluded that the effect was related to the presence of α-pinene in high concentrations in the sample (Freitas et al., 2020). Other studies also highlight the antimicrobial potential of terpinene and β- pinene monoterpenes (Freitas et al., 2020; Nguyen et al., 2022; Huong et al., 2019; Leite et al., 2007).

There are numerous other studies in the literature describing the biological potential of α and β-pinene monoterpenes (Medeiros et al., 2017; Pereira et al., 2017; Da França et al., 2015; Rivera-Yañez et al., 2017).

All in all, it is evident that the EO medicinal plant P. rivinoides from fragments of the Atlantic Forest in Rio de Janeiro State has shown potential for biological activities, such as the previously published apoptotic effect of β- pinene on oral squamous cell carcinoma (Machado et al., 2021). However, the activity of an essential oil must be evaluated rather than the activity of some of its constituents to avoid antagonisms or synergies between them. For this reason, the overall relationship between the potential uses of this plant and its bioactivity is still very relative and further studies are needed to confirm its bioactivity.

Furthermore, it is also worth highlighting the significant increase in the levels of limonene in PRN (3.07 - 14.05%) and linalool in PRR (2.32 - 9.09%). This is very important because pure limonene and high levels EO of this compound are bradycardia and antiarrhythmic (Nascimento et al., 2019), antifungal (Viriato, 2014; Cai et al., 2019), molluscicidal (Barros et al., 2020), antibacterial and antibactericidal (Yatoo et al., 2021; Hazrati et al., 2020), antioxidant (Yatoo et al., 2021; Hazrati et al., 2020), antiparasitic activities, among others (Vieira et al., 2018). Antimicrobial (Herman et al., 2016), insecticidal (Estrada et al., 2019), anti-inflammatory (Wu et al., 2014), cytotoxic (Chen et al., 2019), antidepressant (Dos Santos et al., 2018), larvicidal (Fujiwara et al., 2017), anesthetic, and sedative (Pereira et al., 2018; Mirghaed et al., 2016) effects have been reported for linalool.

Pearson analyses were also performed to evaluate the correlation of the main compounds and chemical classes with temperature (°C), humidity (%), and precipitation (mm). The results are described in Table 2. 

The monoterpens α-pinene and β-pinene, and limonene do not appear to be affected by climatic factors at either collection  site.  The  biosynthesis  of  monoterpenes  and sesquiterpenes also does not seem to be affected by climatic abiotic factors evaluated for Serra da Tiririca State Park (PRN). Other results were found for the specimen from Tijuca National Park (PRR), where the percentage content of monoterpenes showed a negative correlation with temperature (r = -0.7813; p < 0.05), while the content of sesquiterpenes was positively correlated with temperature (r = 0.7236; p < 0.05) and precipitation (r = 0.6055; p < 0.05). For the other components of EO, there was a moderate positive correlation between bicyclogermacrene content and temperature for both sites. However, it is known that interactions resulting from biotic factors with this species at this site may be an important factor in this analysis (Salazar et al., 2016; Whitehead et al., 2021).

The results presented here highlight the biological potential of EO, which was extracted from P. rivinoides being suitable for medicinal use in the treatment of various problems due to its chemical composition, which is consistent with its traditional medicinal use. It is suggested that P. rivinoides EO has low chemical phenotypic variability in both areas, despite environmental and seasonal differences. This chemical stability of EO is little observed in Piperaceae species that show great phenotypical chemistry variability.  This is a good thing because it means that the EO's biological activities will continue to be active even in different months with different abiotic influences, making P. rivinoides a potential source of bioactive compounds. The seasonality of chemical variations suggests that the phenological cycle has a greater influence on biosynthesis than abiotic factors. Based on the results presented and the literature, we suggest possible chemotypes or ecotypes for P. rivinoides. However, a more in-depth study is needed to confirm them  by  examining  EO  samples  from  different regions of Brazil in the field and in cultivation.

The authors have not declared any conflict of interests.

The authors thank CAPES, FAPERJ, PROEP - CNPq (407845/2017-8) for their financial support.