Insecticidal activity of Piper aduncum oil: variation in dillapiole content and chemical and toxicological stability during storage

The insecticidal effect of the essential oil of Piper aduncum (EOPA), and of its constituent dillapiole [1-allyl-2,3-dimethoxy-4,5-(methylenedioxy) benzene] in particular, is well documented in the literature and can be associated with its interference with the enzymatic detoxification in arthropods. However, no data exist on the range of dillapiole content associated with insecticidal activity, which is necessary to establish reliable dose-activity parameters for a formulated product. The oil composition can also change during storage after distillation, mainly due to environmental factors such as light incidence, atmospheric oxygen and temperature, which can be deleterious to oil quality. In this study, EOPA subjected to different storage conditions over four years and its rectified fractions were submitted to bioassays to evaluate their insecticidal effect by topical contact and residual contact against Spodoptera frugiperda . Our objectives were to determine the relationship between dillapiole content and the insecticidal activity of EOPA, and to evaluate its chemical and toxicological properties over time under different conditions. Our results showed that EOPA was stable with respect to the dillapiole content and the toxicological effect against S. frugiperda under different storage conditions for four years. The overall chemical composition of the EOPA did not vary significantly among storage conditions. EOPA with dillapiole content ranging between 68% and 100% showed greater insecticidal toxicity by residual and topical contact against S. frugiperda larvae.


INTRODUCTION
The essential oil of Piper aduncum L. (EOPA) has an excellent yield (2.5 to 3.5%) in comparison with other essential oils, and is rich in dillapiole (31.5 to 91.1%), a allylphenol with a high oxygenation pattern (1-allyl-2,3dimethoxy-4,5-(methylenedioxy) benzene) (Maia et al. 1998).The insecticidal effect of EOPA, and of dillapiole in particular, is well documented in the literature and can be associated to its interference with arthropod enzymatic detoxification, as reviewed by Durofil et al. (2021), who reported 23 arthropods of agricultural and livestock importance controlled by P. aduncum compounds.
The composition and thus the biological activity of essential oils vary batch-to-batch (Morais 2009).Associative effects and different action forms are observed between major and minor component activities, as well as synergistic effects (Bakkali et al. 2008).These effects must be evaluated on a case-by-case basis for any particular oil composition.Arylpropanoids, like dillapiole, can act by inhibiting enzymatic pathways, while terpenic compounds can be neurotoxic or, due to their lipophilicity, facilitate the action of other oil constituents to cross insects cuticle (Afshar et al. 2017).
EOPA activity is associated with the presence of dillapiole (Estrela et al. 2006).However, there are no data on the relation between content and activity level of dillapiole.This knowledge is necessary to establish reliable dose-activity parameters for formulated products.The quantitative and qualitative composition of EOPA varies with genetic and geographic intraspecific variability, e.g., in the western Brazilian Amazon, EOPA can contain from 18 to 56 components (Andrade et al. 2009).
The essential oil composition can also change during storage, after distillation, mainly due to environmental factors such as light incidence, atmospheric oxygen and temperature (Turek and Stintzing 2012).The effect of these factors can be deleterious to oil quality, particularly due to the formation of hydroperoxides (Choe and Min 2006).Molecular rearrangement and thermal degradation were also observed in the absence of oxygen (Geier 2006).
Reliable and wide-scope data on essential oil storage are rare, and the accurate definition of shelf-life for most essential oils has not yet been established (Blitzke 2009;Turek and Stintzing 2012).When available, these data were obtained from oils rich in terpenic compounds, which are particularly volatile and reactive (Turek and Stintzing 2012).No data were found for arylpropanoid-rich oils, such as EOPA.Therefore, the objectives of this study were to establish the relation between the dillapiole content and the insecticidal activity of EOPA, and its chemical and toxicological stability under different storage conditions.

Distillation and rectification of EOPA
Piper aduncum plants were harvested in April 2016 at a 0.5-ha production area at Embrapa Acre (10º15'57"S, 67º42'17"W), state of Acre, northern Brazil.Plants were cut 0.4 m above the ground, the leaves were removed and dried in an oven at 20°C to 35°C until reaching 30% of moisture.In accordance with Brazilian legislation, access to native germplasm was authorized by the Ministry of the Environment (SisGen licenses: processes # 02001.006140/2011-78; 02000.000644/2013-56; 02000.000460/2013-96; 02000.002056/2014-38).
The essential oil was obtained by steam distillation of the leaves for approximately four hours in a 200-L still (ERCITEC, Bauru -SP, Brazil) in batches of 150 kg.The oil obtained was dried over anhydrous magnesium sulphate and had a dillapiole content of 78%.The distilled oil was rectified by fractional distillation in a 30-L still (ERCITEC, Bauru-SP, Brazil).Thirteen fractions, distilling from 270 °C to 350 ºC under atmospheric pressure (760 mmHg), were collected at 1-hour intervals, with increasing concentrations of dillapiole of 12.3, 22.1, 32.2, 42.1, 52.3, 62.4, 67.3, 72.1, 77.3, 82.2, 87.2, 94.1 and 99.8%.The fractions were stored in a domestic refrigerator at 6 °C until the beginning of the experiment in May 2016.

Analysis of EOPA and its fractions
The EOPA and its rectified fractions were analyzed by gas chromatography in an Agilent 7890A gas chromatograph (GC) fitted with a 7693B automatic sampler and a flame ionization detector (FID).Samples were diluted to 1% with dichloromethane and 1 µL was injected at 250 ºC in split mode (1:50).The FID was operated at 280 ºC.Separation of the components was obtained with a HP-5MS fused silica capillary column (5%-phenyl-95%-methyl-silicone, 30 m x 0.25 mm x 0.25 µm).Hydrogen was used as carrier gas (1.5 ml min -1 ).Oven temperature was programed from 60 to 240 ºC at 3 ºC min -1 .Quantification was based on the area (area %) from the signal of the FID normalized with an internal standard.All analyses were made in triplicate.All compounds > 0.1 area % were quantified in the oil composition, while compounds < 0.1 area % were considered as trace elements.
Analyses by mass spectrometry (GC-MS) were performed on an Agilent 5975C mass selective detector coupled to an Agilent 7890A gas chromatograph with the same column, temperatures and injection conditions as above.Helium was used as carrier gas at 1 mL min -1 .The mass detector was operated in electron ionization mode (70 eV), at 3.15 scans sec -1 , with mass range from 40 to 450 u.The transfer line was kept at 240 ºC, the ion source at 230 ºC and the analyzer at 150 ºC.

AMAZONICA
The identification of the oil components was performed by comparison of their mass spectra with those from the Wiley Registry of Mass Spectral Data (McLafferty 1994) or NIST databases (NIST 2011), as well as their linear retention indices (LRI), calculated according to Van Den Dool and Kratz (1963), after the injection of a homologous series of hydrocarbons (n-C 7 -C 26 ) in the same conditions as above, and compared to literature data (Joulain and König 1998;Adams 2007).

Degradation of EOPA
The experimental methodology to evaluate the degradation of EOAP was adapted from Turek and Stintzing (2012).
We tested four storage conditions: (a) exposure to ultraviolet radiation in a UV chamber; (b) exposure to direct sunlight; (c) indoors at room temperature at Embrapa Acre; and (d) uncovered in a domestic refrigerator.Eight vials containing 5 ml of EOPA with 78% dillapiole were used for each treatment (four colorless vials and four amber glass vials).A control consisted of four amber glass vials covered with aluminum foil stored in a refrigerator, to eliminate any interference of the refrigerator light when opening the refrigerator door.In total, 36 vials were used.The samples were coded according to Table 1.
The exposure period for all storage conditions was four years, from April 2016 to April 2020.Each year (at days 0, 360, 720, 1080 and 1440) one vial of each color was removed for chemical characterization and toxicological assays, using a sub-sample of 2 ml from each vial for each analysis.
A UV chamber was adapted from a laminar flux chamber fitted with an Actinica Philips model TLD15W/03 ultraviolet lamp (450 mm, 15 W, range: 380-480 nm), for simulation of UV-A and UV-B radiation.The lamp was kept lit permanently.An air conditioning system kept the temperature at an average 25.9 ºC (minimum 22.8 ºC, maximum 29 ºC).The indoors samples were kept in a 20-m 2 room with no temperature control.Average temperature in the room was 26.7 ºC (minimum 19.7 ºC, maximum 30.3 ºC).Direct exposure to sunlight was approximately 12 h day -1 (6:00 am to 6:00 pm), with average temperature of 29.8 ºC (minimum 22.6 ºC, maximum 50.7 ºC).Temperature data were obtained using a data logger (Escort RH iLog, range -40 to 70º C) for each experimental condition.

In vitro toxicological effect of EOPA
After fractional distilation, the 13 EOAP fractions with different dillapiole contents were submitted to bioassays to evaluate the insecticidal effect by topical contact and by residual contact in May 2016.The same assays were also performed each year with the storage-conditions samples.Larvae of Spodoptera frugiperda (JE Smith, 1797), which is a common agricultural pest in tropical and subtropical regions, were used as a target insect.
The toxicological evaluations were carried out at the Entomology Laboratory of Embrapa Acre and followed the methodology of Estrela et al. (2006).Third instar larvae were used in all bioassays (breeding authorization by SISBIO license # 13464-2).
The experimental parameters for the bioassays were determined in preliminary tests following the methodology of Robertson et al. (2016) using a completely randomized design with four repetitions per treatment.Each replicate consisted of 10 insect larvae in a Petri dish.For all 13 EOPA fractions and the yearly samples of EOAP from the storage-condition treatments, we used doses of 0.2 mL EOPA per larva for residual contact and 1 µL for topical contact, followed by 24 h without feeding.After determining the overall response range from concentrations that caused nearly zero to nearly 100% mortality of larvae, narrower response ranges were determined, following the methodology described by Finney (1971).Seven concentrations were selected through this methodology for the final residual and topical contact bioassays.
In the final residual and topical contact bioassays, the selected concentrations and doses were used for all 13 fractions and annual sub-samples of storage-condition treatmentswith four repetitions per treatment.Each replicate consisted of 10 insect larvae in a Petri dish.For the assay on topical contact, 1 µL of the test sample was applied on the dorsal side of the larva´s pronotum with the aid of a graduated micro syringe (Al-Sarar et al. 2006).For the assay on residual contact, a filter paper of 5 cm diameter impregnated with 0.2 mL of the test sample was dried in a fume hood for 5 min until the solvent had completely evaporated (Estrela et al. 2006) and then placed in a Petri dish that received a larva.In both assays, acetone solvent was used as a negative control, and the treated larvae were individualized and left without food in Petri dishes (5.0 cm × 1.5 cm) and placed and maintained in a thermoelectric refrigerated incubator at 25° C ± 2 °C, 70 ± 5% relative humidity, and 12 h photophase.After 24 hours larval mortality was assessed.

Statistical analysis
The observed mortality in the toxicological assays was corrected for natural mortality using Abbott´s correction (Abbott 1925).
For the determination of LC 50 and LD 50 , respectively the concentration (residual contact) and dose (topical contact) more likely to result in a 50% mortality of the larvae, concentration-mortality curves and confidence intervals (95% CI) were determined by Probit analysis using the SAS program (SAS Institute 2001).To test the goodness-of-fit, Pearson's chi-square test (χ 2 ) was used with a significance level of 5%.
The overlap among the confidence intervals (95% CI) for LC 50 and LD 50 was used to define the significance of the differences among the treatments according to Van Frankenhuyzen ( 2019).The overlap of the confidence intervals was determined visually and by comparing the maximum and minimum values of each IC within each experiment (13 fractions of EOPA with different dillapiole concentrations and the annual subsamples of storage-condition treatments).
The mortality-response data of the S. frugiperda larvae as a function of dillapiole content of the 13 EOPA fractions were fitted to a quadratic function and the range of dillapiole contents that did not differ significantly in larvae mortality was defined from the inflection points of the curve.The analysis was done with SYMBOLAB (2020) (EqsQuest Ltd).

RESULTS
There was little variation in the oil composition of the control sample, even after four years of storage (Table 2).The main variation was observed for monoterpenes (α-pinene, β-pinene, α-phellandrene and β-ocimene), which had lower concentrations after 1440 days than at the beginning of the storage period.This can be explained by the slow evaporation of these very volatile components.
During storage, as the lighter compounds evaporated, the relative proportional areas changed, and the concentrations of several sesquiterpenes, oxygenated sesquiterpenes and even arylpropanoids, originally present in the oil as trace compounds (< 0.1%), rose above the 0.1% threshold and were included in the EOPA composition.The initial sesquiterpene quantification (peaks 5-9) did not change expressively during storage, but there was a small increase in the percentage of oxygenated sesquiterpenoids, such as caryophyllene oxide (1.0 to 1.4%) and humulene epoxide II (trace to 0.1%), which is compatible with the aging of the oil.The small reduction in the relative amount of dillapiole was likely due to the increase in quantifiable compounds over time.Therefore, when kept under refrigeration (thus in the dark), in a hermetically closed bottle, regardless of the color of the bottle, the EOPA was chemically stable for at least four years, similarly to the control.
Monoterpenes were lost under the more adverse storage conditions, and after four years they were found only in the samples kept under refrigeration, including the control (Table 3).The sesquiterpene content decreased under nonrefrigerated storage, mainly in the colorless flasks exposed to UV and sunlight, particularly for (E)-caryophyllene, α-humulene, bicyclogermacrene and germacrene D. Among the oxygenated sesquiterpenoids, the content of alcohols (nerolidol, spathulenol, viridiflorol) varied little among treatments (Table 3).
The epoxides such as caryophyllene oxide and humulene epoxide generally increased throughout time, with a higher increase in the samples exposed to highest radiation (UVCC and SUNC).A small decrease observed in pentadecane content can be associated to evaporation in the samples kept outside refrigeration (UVCA, UVCC, WARA, WARC, SUNA and SUNC).No relevant variation was observed in arylpropanoid content (Table 3, Supplementary Material, Tables S1,S2, S3).
The confidence intervals of LD 50 and LC 50 values largely overlapped among treatments and over time (Figures 1 and  2), indicating that the toxicity of 4-year stored EOPA against S. frugiperda larvae did not differ significantly from the fresh EOPA, which reflects the little variation observed in the EOPA composition over the storage time.
The fitted curve of LC 50 values for dillapiole fractions had a high coefficient of determination (R² = 0.9414) (Figure 3).The inflection point corresponded to a dillapiole content of 77.8% (LC 50 = 0.0055 µL cm -2 ).There was no significant difference in larval mortality in the range of 67 to 82% dillapiole, corresponding to an overlap of the confidence intervals of LC 50 of 0.5586 -0.0071 µL cm -2 and 0.0065 -0.0074 µL cm -2 , with p > 0.05 for Pearson's chi-square test (determined by Probit analysis).This means that the concentration of dilapiol in the range of 67 to 82% has a linear toxicological response to residual contact against S. frugiperda larvae.The model indicated that concentrations of dillapiole between 68% and 88% promote the highest residual contact toxicity to this insect.
The fitted curve of LD 50 values for dillapiole fractions also had a high coefficient of determination (R² = 0.9609) (Figure 4).The inflection point corresponded to a dillapiole content of 100.0%(LD 50 = 0.0024 µL mg -1 of insect weight).There was no significant difference in larval mortality in the range of 67 to 87% dillapiole, corresponding to an overlap of the confidence intervals of LD 50 of 0.0032 -0.0037µL mg -1 of insect weight and 0.0029 -0.0035 µL mg -1 of insect weight, with p > 0.05 for Pearson's chi-square test (determined by Probit analysis).The concentration of dilapiol in the range of 67 to 87% also had a linear toxicological response to topical contact against S. frugiperda larvae.The model indicated that ACTA AMAZONICA concentrations of dillapiole between 82% and 100% promote the highest topical contact toxicity to this insect.

DISCUSSION
Previous data suggest that dillapiole is sensitive to degradation when stored (Tisserand and Young 2013), however, our results showed that, even after four years under the most adverse conditions, such as exposure to UV radiation, sunlight and high temperatures (above 40°C), the dillapiole content of EOPA was preserved.Therefore, storage periods of up to four years can be considered safe for the dillapiole content of EOPA shelf-life.In addition, caryophyllene oxide and humulene epoxide content increased with time, especially in the samples exposed to radiation in colorless glass, in agreement with the role of epoxides as typical markers for old and oxidized oils (Turek and     ACTA AMAZONICA Some studies have described loss of insecticidal efficacy of essential oils after storage.For example, essential oil of camphor, Cinnamomum camphora (L.)J. Presl and thyme, Thymus serpyllum L. stored in the presence of light and oxygen showed a toxic effect against larvae of Aedes aegypti L. for only two weeks (Amer and Mehlhorn 2006), indicating rapid degradation of the oxygenated monoterpenes when exposed to light (Misharina et al. 2003).
The minor changes observed in oil composition throughout time did not impact negatively the toxicity of EOPA.The stability of the arylpropanoids was enough to maintain the toxicity, as the observed losses of monoterpenes during the 4-year storage did not contribute to the occurrence

ACTA
AMAZONICA of an additive effect on toxicity.Similarly, oils from Piper with higher content of arylpropanoids were more active against larvae of A. aegypti (Morais et al. 2007) Terpenoids have an insecticidal effect by inhibiting acetylcholinesterase, which is associated with a neurotoxic effect in insects (da Silva et al. 2017).Likewise, the increase of sesquiterpenes such as germacrene D may have an insecticidal effect, as this compound is an inhibitor of esterases and glutathione S-transferases (Ribeiro 2012), which could increase lethality in synergy with other compounds present in the EOPA.
Dillapiole concentrations above 88% were less effective against S. frugiperda larvae then the medium-high concentrations, regardless of the route of contamination.This can be an advantage in the sense of simplifying the process to obtain a commercial product.This could be a result of a synergistic or additive interaction of minor compounds in the composition of EOPA with the dillapiole, contributing to the insecticidal effect, as has been observed in other studies on the evaluation of essential oils on insect control (Hummelbrunner and Isman 2001;Liu 2006;Gillij et al. 2008;Pavela 2008;Singh et al. 2009;Pavela 2014).
As dillapiole was the major compound of EOPA throughout the storage period, the retention of toxicity may be associated with a higher insecticidal activity of arylpropanoids, as pointed out by da Silva et al. (2017).Furthermore, the occurrence of other arylpropanoids in EOPA in addition to dillapiole, such as myristicin and elemicin, throughout the storage period, enables the inhibition of the three main families of detoxifying enzymes (P450 monooxygenases, esterases and glutathione S-transferases), allowing the increase in lethality of EOPA via synergistic interaction among its compounds (Bernard et al. 1993;Shankarganesh et al. 2009;Liu et al. 2014).Piper oil with dillapiole as major component (54.7%) had similar larvicidal activity against A. aegipty compared to enriched oil (98.9%) -36.0 ppm and 42.9 ppm, respectively (Navarro et al. 2013).This may suggest a more important role in the definition of additivity or synergy in interactions between the major arylpropanoids present in EOPA.
Regardless of the contamination pathway, the larvicidal activity of EOPA mediated by dillapiole against S. frugiperda was optimal at concentrations of 82%-88%, which is the range indicated for the standardization in the prospection of future commercial products.Batch standardization may be achieved by indirect determination of the dillapiole content of EOPA through its refractive index, as proposed by Pateira et al. (1999), to control the limits for maximum insecticidal activity.

CONCLUSIONS
We established the relation between the dillapiole content and the insecticidal activity of EOPA against S. frugiperda larvae, and its chemical and toxicological stability when submitted to different storage conditions.EOPA dillapiole content was stable and its chemical composition did not vary significantly among different storage conditions over four years.The toxicological effects of EOPA on S. frugiperda larvae by topical and residual contact were not altered during this storage period.Dillapiole contents in the EOPA between 68% -88% and 82% -100% promote, respectively, optimal and stable residual and topical contact toxicity against S. frugiperda larvae.Further experiments should determine the efficacy of EOPA against multiple insect species, in order to generalize the trends of the observed phenomenon.Our results suggest that

SUPPLEMENTARY MATERIAL (only available in the electronic version)
Fazolin et al.Insecticidal activity of Piper aduncum oil: variation in dillapiole content and chemical and toxicological stability during storage

Figure 2 .
Figure 2. Lethal dose of Piper aduncum essential oil subjected to different storage conditions by topical contact on Spodoptera frugiperda caterpillars.Symbols are the mean of four replicates and bars the 95% CI obtained by Probit analysis.CONT = sample stored in a domestic refrigerator, amber glass flask covered with aluminium foil; REFA = sample kept in a refrigerator at 6 ºC in an amber flask; REFC = sample kept in a refrigerator at 6 ºC in a colorless flask; SUNA = sample kept under sunlight exposition (approximately 12 hours per day) in an amber flask; SUNC = sample kept under sunlight exposition (approximately 12 hours per day) in a colorless flask; INAF = sample kept indoors at ambient temperature in an amber flask; INCF = sample kept indoors at ambient temperature in a colorless flask; UVCA = sample under UV light in an amber flask; UVCC = sample under UV light in a colorless flask.

Figure 1 .
Figure 1.Lethal concentration of Piper aduncum essential oil subjected to different storage conditions by residual contact on Spodoptera frugiperda caterpillars.Symbols are the mean of four replicates and bars the 95% CI obtained by Probit analysis.CONT = sample stored in a domestic refrigerator, amber glass flask covered with aluminium foil; REFA = sample kept in a refrigerator at 6 ºC in an amber flask; REFC = sample kept in a refrigerator at 6 ºC in a colorless flask; SUNA = sample kept under sunlight exposition (approximately 12 hours per day) in an amber flask; SUNC = sample kept under sunlight exposition (approximately 12 hours per day) in a colorless flask; INAF= sample kept indoors at ambient temperature in an amber flask; INCF = sample kept indoors at ambient temperature in a colorless flask; UVCA = sample under UV light in an amber flask; UVCC = sample under UV light in a colorless flask.

Figure 3 .
Figure 3. Toxicological effect by residual contact of fractions of Piper aduncum essential oil containing different dillapiole concentrations on Spodoptera frugiperda larvae.Points are the mean of four replicates and bars the 95% CI obtained by Probit analysis.R 2 = coefficient of determination.
LRI calc = linear retention indices calculated according toVan den Dool and Kratz (1963) for a DB-5 stationary phase.LRI lit = linear retention indices from the literature(Adams 2007; Joulain and Koenig 1998).UVCA = sample under UV light in an amber flask; UVCC = sample under UV light in a colorless flask; INAF = sample kept indoors at ambient temperature in an amber flask; INCF = sample kept indoors at ambient temperature in a colorless flask; REFA = sample kept in a refrigerator at 6 ºC in an amber flask; REFC = sample kept in a refrigerator at 6 ºC in a colorless flask.SUNA = sample kept under sunlight exposition (6 hours per day) in an amber flask; SUNC = sample kept under sunlight exposition (6 hours per day) in a colorless flask; CONT = sample stored in a domestic refrigerator, amber glass flask covered with aluminium foil and tr = trace (< 0.1 in área %).

Figure 4 .
Figure 4. Toxicological effect by topical contact of fractions of Piper aduncum essential oil containing different dillapiole concentrations on Spodoptera frugiperda larvae.Points are the mean of four replicates and bars the 95% CI obtained by Probit analysis.R 2 = coefficient of determination.
LRI calc = linear retention indices calculated according toVan den Dool and Kratz (1963) for a DB-5 stationary phase.LRI lit = linear retention indices from the literature(Adams 2007; Joulain and Koenig 1998).UVCA = sample under UV light in an amber flask; UVCC = sample under UV light in a colorless flask; INAF = sample kept indoors at ambient temperature in an amber flask; INCF = sample kept indoors at ambient temperature in a colorless flask; REFA = sample kept in a refrigerator at 6 ºC in an amber flask; REFC = sample kept in a refrigerator at 6 ºC in a colorless flask.SUNA = sample kept under sunlight exposition (6 hours per day) in an amber flask; SUNC = sample kept under sunlight exposition (6 hours per day) in a colorless flask; CONT = sample stored in a domestic refrigerator, amber glass flask covered with aluminium foil and tr = trace (< 0.1 in área %).
LRI calc = linear retention indices calculated according toVan den Dool and Kratz (1963) for a DB-5 stationary phase.LRI lit = linear retention indices from the literature(Adams 2007; Joulain and Koenig 1998).UVCA = sample under UV light in an amber flask; UVCC = sample under UV light in a colorless flask; INAF = sample kept indoors at ambient temperature in an amber flask; INCF = sample kept indoors at ambient temperature in a colorless flask; REFA = sample kept in a refrigerator at 6 ºC in an amber flask; REFC = sample kept in a refrigerator at 6 ºC in a colorless flask.SUNA = sample kept under sunlight exposition (6 hours per day) in an amber flask; SUNC = sample kept under sunlight exposition (6 hours per day) in a colorless flask; CONT = sample stored in a domestic refrigerator, amber glass flask covered with aluminium foil and tr = trace (< 0.1 in área %).
LRI calc = linear retention indices calculated according toVan den Dool and Kratz (1963) for a DB-5 stationary phase.LRI lit = linear retention indices from the literature(Adams 2007; Joulain and Koenig 1998).UVCA = sample under UV light in an amber flask; UVCC = sample under UV light in a colorless flask; INAF = sample kept indoors at ambient temperature in an amber flask; INCF = sample kept indoors at ambient temperature in a colorless flask; REFA = sample kept in a refrigerator at 6 ºC in an amber flask; REFC = sample kept in a refrigerator at 6 ºC in a colorless flask.SUNA = sample kept under sunlight exposition (6 hours per day) in an amber flask; SUNC = sample kept under sunlight exposition (6 hours per day) in a colorless flask; CONT = sample stored in a domestic refrigerator, amber glass flask covered with aluminium foil and tr = trace (< 0.1 in área %).

Table 1 .
Sample codes for the experimental treatments of four-year degradation of Piper aduncum essential oil.
SUNC Control (refrigerator, amber flask covered with aluminium foil) CONT

Table 2 .
Composition (area %) of a sample of essential oil of Piper aduncum from Acre state, Brazil during four years of storage in an amber flask wrapped in aluminium foil in a domestic refrigerator (control sample).

Table 3 .
Constituents of the essential oil of Piper aduncum from Acre state (Brazil) after four years of storage under different conditions (in area %).

Table S1 .
Constituents of the essential oil of Piper aduncum from Acre state (Brazil) after one year of storage under different conditions (in area %).

Table S2 .
Constituents of the essential oil of Piper aduncum from Acre state (Brazil) after two years of storage under different conditions (in area %).

Table S3 .
Constituents of the essential oil of Piper aduncum from Acre state (Brazil) after three years of storage under different conditions (in area %).