Crop-Ecology and Post-harvest distillation method: The key regulators for yield and composition of essential oil of Valeriana

Valeriana jatamansi is an important temperate herb, which is used for pharmaceutical and essential oil industries. In India, this species is now at the verge of extinction due to the over-exploitation of its rhizomes from its natural habitat. It is hypothesized that the variations in bioactive compounds in its essential oil are very high among the wild populations as well as cultivated sources. Thus, the study was conducted to evaluate the chemical proling of essential oil of four wild populations (Rupena, Kugti, Garola, Khani) and two cultivated sources (CSIR-IHBT, Saloonie), which were distilled at three consecutive days. The variation in oil concentration in root/rhizome was found signicant (P ≤ 0.05), and the maximum value (0.35%) was registered with the population collected from Kugti and Khani. In essential oil, irrespective of population and distillation day, patchouli alcohol was the major compound, which ranged from 19 to 63.1%. The maximum value (63.1%) was recorded with the essential oil obtained from Garola's population and distilled on the rst day. The percentage of seychellene was abruptly increased with subsequent days of extraction in all the populations. The multivariate analysis revealed that the essential oil proles of Rupena, Kugti, Garola, and CSIR-IHBT populations were found similar during the rst day of distillation. However, during the second day Rupena, Kugti, Khani, and CSIR-IHBT come under the same ellipse of 0.95% coecient. The results suggest that the population of Kugti is superior in terms of oil concentration (0.35%) with a higher proportion of patchouli alcohol (63% on the rst day). Thus, repeated distillation is recommended for higher recovery of essential oil. Moreover, repeated distillation can be used to attain V. The results of the study yield and composition of essential oil of V. jatamansi is largely by the crop-ecology and distillation methods. Our results also conrm that different chemo-types exist in natural populations in Chamba (H.P., India). The total recovery of essential oil (%) was found signicant (P ≤ 0.05) variations among the source of materials, and the maximum value (0.35%) was recorded with the populations from Kugti and Khani. All the populations are characterized by the rich in patchouli alcohol. This study also conrmed that the repetitive distillation method can be used to obtain a higher quantity of essential oil of V. jatamansi with variance chemical proles. The concentrations of some compounds like seychellene, α-santalene, or β-elemene can be increased through repetitive distillation. It can also be concluded that the species V. jatamansi has wide adaptability, and it can be commercially cultivated with an altitudinal range of 1354–2140 m. It can also be grown in sandy loam to silty clay soil with a wide pH range of 5.59–8.85 and different nutritional levels. However, to discriminate the effects of population and crop-ecology and to identify the elite material, all the collected and/or characterized populations should be evaluated under the same conditions.


Introduction
Valeriana jatamansi Jones (Syn. V. wallichii DC.), commonly known as Indian valerian, belongs to the family of Valerianaceae; However, recently it has been categorized under Caprifoliaceae 1,2 . It is a small perennial medicinal herb, and naturally distributed in the temperate Himalaya between an altitude of 1000 and 3300 m. Asl 3,4 . In India and Nepal, V. jatamansi is widely used as a substitute for V. o cinalis, which is mostly used in Europe 5 . V. jatamansi is widely used in the Indian Ayurvedic and Unani systems of medicine since ancient times 6 . Now it becomes the most acceptable sedative over modern medicines 6,7,8 . V. jatamansi possesses a sedative property due to the presence of bioactive compounds like valepotriates in roots/rhizome 9,10 , which is a non-glycosidic iridoid esters 11 . It is also used in Chinese traditional systems of medicine for its tranquilizing hypnotic and antiviral activities 12 .
Importantly, the essential oil of V. jatamansi is now widely used in avor and fragrance industries, and about 30 pharmaceutical and essential oil-based products are commercially available 13 . Moreover, V. jatamansi has also been reported as a psychopharmacological agent 10,14,15 , and highly effective against leprosy 16 , epilepsy and hysteria 17 , nerve diseases 18 , scorpion bite 19 , cholera 20 , anxiolytic properties 21 and curing lewy body dementia 22 . However, the present study has been emphasized only on essential oil content and composition of different populations of V. jatamansi from the western Himalayan region in India.
In Himachal Pradesh, a Himalayan state of India, this plant is naturally grown on a moist slope of hills in the district of Chamba. However, a wide diversity of V. jatamansi exists in nature in this region 23 . Now in India, this species is at the edge of extinction due to the overexploitation of its roots/rhizome from its natural habitat to meet the burgeoning industrial demand. The industrial demand for this plant is increasing day by day for phytopharmaceutical as well as essential oil-based products. Since the major portion of the V. jatamansi is contributed through wild collections, there is a major concern about a consistent supply of quality raw material. Moreover, the collectors do not follow the standard protocol for harvesting to maintain the quality of essential oil. Essential oil of V. jatamansi is a complex mixture of volatile molecules including monoterpenes and sesquiterpenes. More than 290 compounds have been identi ed in the essential oil of V.
The variations in oil concentrations in roots/rhizome as well as the percentage share of different compounds are largely in uenced by seasonal variations 30 , populations 28 , plant parts 4,27,31 , geographical regions 32,33 , type of materials 29 and genetic makeup 4,33,34 . The variations in secondary metabolites of V. jatamansi due to habited and altitude differences have also been reported 35 . The differences between wild and cultivated source of V. jatamansi in terms of essential oil content and antioxidant activities have also been noticed 36 . The variations in chemical compositions of essential oil of Rosa damascena have been noticed due changes of ambient temperature and humidity during the time of owering 37 . Since V. jatamansi is widely found in nature in the Chamba region, there is a possibility of large variations in oil content and compositions among the populations. However, the information about the variety, chemotype, or population that occurs in this region is limited. Farmers collect/harvest roots/rhizomes from wild without having the information about chemical pro les. Farmers also use wild populations as planting material for cultivation in the under-utilized land or in orchards as an intercrop.
Another important issue with V. jatamansi is distillation methods, which largely in uence the oil recovery and quality of essential oil. Generally, farmers distil their harvested roots/rhizome in a commercial hydro-distillation unit after partial drying. Both dry and fresh roots/rhizome are used for oil extraction, and the variations in oil concentration between fresh and dry materials have been reported 29,36,38 .
Among the distillation conditions, time/duration is a key factor for the composition of essential oil like pine (Pinus ponderosa Dougl. ex Laws) and Sweet Sagewort (Artemisia annua L.) 39,40 , Japanese cornmint (Mentha canadensis L.) 41 , lemongrass (Cymbopogon exuosus Steud.), and palmarosa (Cymbopogon martini Roxb.) 42 . Moreover, the effect of distillation time on recovery of essential oil of V. jatamansi has been reported 29,43,44 . However, information regarding the effects of repetitive distillation on oil recovery and the quality of the oil is missing. It has been reported that essential oil fractions with speci c chemical pro les can be managed through altering the time of hydrodistillation 45 . In case of rose essential oil, higher methyl eugenol concentration is noticed with long-term fermented owers 46 .
We hypothesized that repetitive distillation would be effective to increase the oil recovery and modi cation of the composition of the oil. It is a fact that the pH level of water in uences the hydrolytic reactions during hydro-distillation. Seeing the above issues, there is a pressing need to understand the best population and effective distillation method for higher oil yield and desire quality. This experiment, therefore, conducted to (i) investigate the pro le of essential oil of different populations of V. jatamansi, and (ii) to understand the effects of repetitive distillation on recovery and quality of essential.

Essential oil content (%)
Analyzed data revealed that essential oil of V.jatamansi was obtained up to the third day of distillation; however, in the case of Rupana and Garola, oil was not obtained (Fig. 1a-e) on the third day. Irrespective of sources of roots/rhizome, the recovery rates (%) were higher on the rst day of distillation, and thereafter the trends were declined. The total recovery (addition of 2/3 days' samples) of essential oil (%) was found signi cant (P ≤ 0.05) variations among the natural populations and cultivated source (Fig. 1f). The populations collected from Kugti and Khani registered signi cant (P ≤ 0.05) higher oil content (0.35%) compared with other populations, a cultivated source at CSIR-IHBT, and oil obtained from the traditional extraction method. The lowest oil content (0.17%) was recorded with the population of Garola.
Composition of essential oil GC and GC-MS analysis data revealed that a maximum of 18 volatile compounds was identi ed from the essential oil obtained from hydrodistillation of V. jatamansi roots/rhizome of different populations and cultivated sources (Fig. 2). These compounds contributed up to 81.8% of the total volume ( Fig. 2) in this study. The highest number (16) of volatile compounds were identi ed with the cultivated source from CSIR-IHBT during third-day distillation (Fig. 2e) whereas the lowest number (9) of compounds were identi ed with the population collected from Garola on Day-one distillation (Fig. 2c). However, the maximum (81.8%) and minimum (62.1%) sharing by the identi ed compounds in a total volume of essential oil was observed with the day-one distillation of the Kugti population and day-two distillation of Garola population, respectively (Fig. 2b, c). In all populations, the numbers of identi ed compounds were higher in the essential oil extracted on the second day compared to the rst day even these values were higher than the Saloonie sample (traditional practice).
The chemical pro les of the volatile oils from different natural and cultivated sources of V. jatamansi have been presented in Table 2. The data on compositions of essential oils revealed a great variation among the different sources of material and distillation process. Irrespective of the source of materials and distillation day, the major compounds detected in the essential oil were patchouli alcohol, δguaiene, selinene < 7-epi-alpha->, α-selinene, α-patchoulene, seychellene, and β-patchoulene, which have been illustrated in the representative chromatograms of essential oil in Fig. 3. Patchouli alcohol was found the most abundant compound in all the essential oil obtained from different populations and distilled on different days ( Table 2). Patchouli alcohol ranged from 19 to 63.1%. The maximum value (63.1%) was recorded with the essential oil obtained from Garola's population and distilled the rst day whereas the lowest value (19%) was recorded with the population collected from Khani and distilled third day ( Table 2). The essential oil of V. jatamansi extracted through the traditional method (Saloonie sample) registered maximum percentage of β-patchoulene (10.2%), α-guaiene (5.5%), α-humulene (5.0%), and δ-guaiene (7.3%) compared with the rest of the samples in this experiment. Humulene epoxide II was found only in three populations (Rupena, Garola, and Khani) during second-day distillation, and the maximum value (2.5%) was recorded with Garola's population. Seychellene concentration was found maximum (13.3%) during the third day of distillation of the population collected from Kugti. Similarly, β-caryophyllene concentration was found maximum (3.8%) with Khani and CSIR-IHBT populations during the third day of distillation (Table 2).   The dynamic of important compounds in the essential oil of V. jatamansi populations has been in uenced by the distillation days (Fig. 4).
The results indicated that maximum percentages of patchouli alcohol, irrespective of the source of materials, were registered with the oils obtained during the rst day of distillation (Fig. 4a). The patchouli alcohol concentrations in the oils obtained during the rst day of distillation, irrespective of populations, were found substantially higher than the oil obtained from the traditional distillation method (Saloonie). However, the concentrations of patchouli alcohol were sharply declined during subsequent days of extraction (Fig. 4a). In contrast, the percentage of seychellene was abruptly increased with subsequent days of extraction in all the populations (Fig. 4b). The concentration of δ-guaiene in the essential oil of V. jatamansi was increased with second and third-day samples of Kugti, Garola, Khani, and CSIR-IHBT populations (Fig. 4c). However, the trend of α-patchoulene concentration was not found consistent across the populations (Fig. 4d).

Principal component analysis (PCA)
The multivariate analysis in terms of PCA was performed to explore the relationship among the volatile compounds in the essential oil of V. jatamansi populations and to understand how these populations were different from each other. The compounds of the essential oil were treated as dependent variables. The separate PCAs were constructed for all three days of distillation revealed that the rst two components, PC1, and PC2, jointly explained 79.86, 63.06, and 86.03% of the total variations for rst, second, and third-day distilled samples, respectively ( Fig. 5a-f). The eigenvalues of these two PCs are 9.53 and 2.45, respectively. The relationships among the compounds of essential oil the magnitude of their contributions in the space of PC1 and PC2 have been presented in Fig. 5a, c, and e for the rst, second and third day, respectively. For rst day sample, PC1 has positive coe cients with β-patchoulene, β-elemene, α-guaiene, seychellene, α-humulene, αpatchoulene, δ-guaiene, selinene < 7-epi-alpha->, and spathulenol; the loading values of these compounds were quite high (> 0.83). Hence, it is clear that these eight compounds are highly correlated with each other (Fig. 5a). However, PC1 has a negative coe cient with patchouli alcohol with a loading value of − 0.94. The PCA bi-plot of the rst-day sample indicated that there were three distinct clusterings among the populations and cultivated sources. The populations from Kugti, Garola, Rupena, cultivated source of CSIR-IHBT were placed in the negative coordinate of both PCs, and all four samples came under the same ellipse of 0.95% coe cient (Fig. 5b). The population of Khani and Saloonie created independent groups.
During the second day of distillation, β-patchoulene, α-guaiene, α-humulene, δ-guaiene were situated in the positive coordinate of PC1with loading value of 0.81, 0.56, 0.81, and 0.87, respectively (Fig. 5c). The compound like β-elemene, santalene, β-selinene, α-selinene, and humulene epoxide II, were also separated by the PC1 and placed in the negative coordinate with loading value of 0.78, 0.75, 0.78, 0.87, and 0.71, respectively. On the third day of distillation, the compounds were separated by both PCs, but there was no distinct cluster among the compounds (Fig. 5e). The PCA bi-plots of second and third days of distillation also indicated three distinct groups among the natural populations and cultivated sources (Fig. 5d, f). For the second-day sample, Kugti, Khani, and CSIR-IHBT fall under the same group, whereas CSIR-IHBT is separated from Kugti and Khani for the third-day sample. The oil from Saloonie did not form any group with the rest of the populations in this study.

Physico-chemical properties of soil of naturally distributed locations of Valeriana jatamansi
The pysico-chemical properties of the soil of different sample collecting locations have been presented in Table 2. The soil texture of the different locations ranged from sandy loam to silty clay. Sandy loam soil was found in Rupena and Saloonie, whereas loamy soil was observed at Khani and Kugti. The soil of Palampur was heavy (silty clay). The signi cant (P ≤ 0.05) differences in terms of chemical properties (pH, EC, OC, available N, P, and K) of the soil were found among the locations. The highest pH (8.85) of soil was registered at Kugti that was signi cantly (P ≤ 0.05) different from the rest of the locations. The lowest pH value (5.59) was registered with the soil collected from Palampur. The EC values of soils from different locations were also found signi cant (P ≤ 0.05) difference, and the highest and lowest values were recorded in Garola and Palampur, respectively. Soil OC ranged from 0.71% at Garola to 3.34% at Saloonie ( Table 2). The bulk densities of soil collected from Kugti, Khani, Garola, and Palampur were found almost similar value (~ 1.48 g cc-1), but this value was signi cantly (P ≤ 0.05) higher than the values recorded from Rupena (1.25 g cc-1) and Saloonie (1.34 g cc-1). Signi cant (P ≤ 0.05) variations in available N, P, and K content in soil were found among the locations. Maximum N and K contents were registered in the soil of Rupena, whereas maximum P was recorded with soil from Saloonie (Table 3). Thus, it is evidenced from the result that the V. jatamansi is grown in a wide range of soil types with different nutritional statuses. Analysis of variance outcomes for determining the effect of location on oil concentration in different populations of V. jatamansi and on soil properties have been presented in Table 4. The values of mean squares with corresponding level of signi cant are presented.

Discussion
In this study, signi cant variations were found in the essential oil concentrations, the main attribute of V. jatamansi, among the naturally distributed populations in Chamba region and cultivated sources. The highest oil concentration (0.35%) was registered with the population collected from Kugti and Khani. These differences among the populations were probably due to variations in micro-climatic conditions, soil characteristics, and genetic make-up. The altitude of these two locations is higher. Although the physicochemical properties of the soils were found differences among the locations, no de nite correlation was found between soil characteristics and essential oil concentration.
Despite the higher altitude, the lowest concentration of oil (0.17%) was registered with the population from Garola. This location is characterized by higher EC (1.07 dS m − 1 ) and relatively low OC (0.71%) and available N (61.12 kg ha − 1 ) that might be another cause for the low concentration of essential oil in roots/rhizome. Crop-ecological factors such as light, ambient temperature and soil nutrients availability control the biosynthesis of essential oil 47  Pradesh, and the major chemical constituents and essential oil content are negatively correlated with altitude 33 . The low percentage of patchouli alcohol with the population from Khani during rst cycle of distillation was noticed probably due to the edaphic factor, particularly soil available K content. Patchouli alcohol is synthesized in various stages through cis-farsenylpyrophosphate 51 .The variations in essential oil composition of V. jatamansi due to growing conditions and location have also been reported 44,52 . Thus, the population having higher amount of patchouli alcohol could be used for mass multiplication for commercial cultivation. It is also suggested that variation in oil quality can be reduced bycollecting the raw material from the cultivated sources.
Despite 6 h distillation on the rst day, some quantities of essential oil of V. jatamansi were obtained up to the third day of distillation for the populations collected from Kugti and Khani and cultivated sources (CSIR-IHBT). The populations of Rupena and Garola also produced oil on the second day. These results were because some compounds of essential oil of V. jatamansi like β-elemene, santalene, and seychellene were not removed from the hypodermal layer in the root cortex region. The outcome is noteworthy, and the result can be used to maximize the recovery of essential oil of V. jatamansi. The effects of duration of distillation time on recovery of essential oil have been reported in many aromatic plants 39,42,45,53−58 .
In this study maximum of 18 volatile compounds were identi ed, which contributed up to 81.8% of the total volume (Fig. 2), and the maximum number (16) of volatile compounds were identi ed with the cultivated source from CSIR-IHBT during third-day distillation (Fig. 2e). This result could be due to the fact that the plants at CSIR-IHBT were grown with proper agronomic practices, which ultimately facilitated better conditions for the synthesis of a large number of compounds compared with other naturally distributed populations. The variations in genetic makeup among the populations may be another reason for changing the chemical pro le of essential oil. On the other hand, the minimum number of volatile compounds were identi ed in the oil obtained at rst-day distillation from the Garola population (Fig. 2c). The variations in the number of compounds in the essential oil of V. jatamansi have been reported in the literature 4,25,30,59 . Thus, it is established from the present study that, irrespective of populations, the number of compounds has been increased either at rst-day or on the second day of distillation (Fig. 2a-f).
The variations in compositions of essential oils were found noticeable due to populations and distillation methods ( Table 2). The differences in crop-ecology may be another cause of variations in compositions of essential oil among the populations. In this study, variations in oil characteristics and altitudinal were observed (Table 1, 2). The variations in essential oil composition from the different parts of India have been reported in V. jatamansi 31,36,43,52 . Patchouli alcohol, the most abundant compound in all the populations, was found noticeable differences, and the maximum concentrations were recorded on the rst day of distillation, irrespective of populations ( Table 2). The essential oil with different in fractions was probably due to boiling point, degree of solubility, molecular weight, and polarity of the compounds. The extraction of polar compounds during hydro-distillation is easy over that of terpene hydrocarbons 60 . The concentration of patchouli alcohol in Saloonie (distilled in traditional method) was substantially lower compared with the rest of the populations distilled on the rst day. This result could be due to the fact that patchouli alcohol was lost during the drying and storage or post-harvest practices. Low concentrations of patchouli alcohol were also found in the second and third days' oil for all populations. This result could be due to the fact that hydrolytic reactions occurred in the remaining hydrolat, which leads to the breakage of the functional group in patchouli alcohol.
However, the higher concentrations of β-patchoulene, α-guaiene, α-humulene, and δ-guaiene were found in the Saloonie sample probable due to the generation of more heat during the traditional distillation method. Disproportionation, as well as the cyclization process, occurs in monoterpenes at elevated temperatures 61 . Moreover, at high temperatures compounds are formed through termination reactions 62 . Moreover, the higher concentration of humulene epoxide II (2.5% on the second day) makes difference to Garola from other populations.
Interestingly, the concentrations of seychellene were abruptly increased with subsequent days of extraction for all the populations. The use of seychellene compound as a non-selective candidate for inhibitor cyclooxygenase on pre-osteoblast cells has been reported 63 . The αsantalene, generally used in cosmetic, perfumery, and aromatherapy industries 64 , was found only in second and third days. The α-santalene is the precursor of α-santalol, which is the main component of sandalwood oil from eastern India 65 . Similarly, β-elemene was found in the second and third days' oil. β-elemene is a novel anticancer agent and also reported for antitumor activity 66 . The improvement and/or generation of the new compound during the second and third days was probably due to oxidation, thermal degradation, chemical degradation, eventually a chemical conversion. This nding con rmed that the distillation method can be used to obtain the higher quantity of essential oil of V. jatamansi with variance chemical pro les. For example, if a high-patchouli alcohol oil is desirable, 6 h distillation is required for a single day. Similarly, if high concentrations of seychellene, α-santalene, or β-elemene are desirable, roots/rhizome of V. jatamansi need to be distilled for second and/or third times. The effects of cyclic/repeated distillation on the dynamics of compounds in the essential oil of V. jatamansi are explained in the form of a schematic diagram (Fig. 6).
The results of this study have been proven the hypothesis that crop-ecology, source of materials, and distillation method determine the yield and composition of the essential oil of V. jatamansi. The results of the present study also demonstrate that the species V. jatamansi has wide adaptability, and it can be commercially grown with an altitudinal range of 1354-2140 m. It can also be grown in sandy loam to silty clay soil with a wide pH range of 5.59-8.85 and different nutritional levels.

Conclusions
The results of the present study elucidate that yield and composition of essential oil of V. jatamansi is largely governed by the crop-ecology and distillation methods. Our results also con rm that different chemo-types exist in natural populations in Chamba (H.P., India). The total recovery of essential oil (%) was found signi cant (P ≤ 0.05) variations among the source of materials, and the maximum value (0.35%) was recorded with the populations from Kugti and Khani. All the populations are characterized by the rich in patchouli alcohol. This study also con rmed that the repetitive distillation method can be used to obtain a higher quantity of essential oil of V. jatamansi with variance chemical pro les. The concentrations of some compounds like seychellene, α-santalene, or β-elemene can be increased through repetitive distillation. It can also be concluded that the species V. jatamansi has wide adaptability, and it can be commercially cultivated with an altitudinal range of 1354-2140 m. It can also be grown in sandy loam to silty clay soil with a wide pH range of 5.59-8.85 and different nutritional levels. However, to discriminate the effects of population and crop-ecology and to identify the elite material, all the collected and/or characterized populations should be evaluated under the same conditions.

Study material
The The voucher specimen of this material has also been deposited in our Institutional herbarium (a recognized and publicly available herbarium). The voucher specimen number has also been received (Accession # 11701).
Harvested rhizomes and brous roots were washed with running tap water, and excess water was removed with the help of blotting paper. After recording the fresh weight of roots/rhizome, samples were placed in air-tight polyethylene bags and brought to the laboratory for extraction of essential oil. The same procedures were followed for the cultivated sample from the Institutional farm. Besides, one oil sample was collected in three replicate from the farmer, which was extracted through a commercial hydro-distillation unit after partial drying of the roots/rhizome in the Saloonie region (Chamba). This is the common practice, which is being followed by the growers. This sample is also considering as a cultivated source. The collected oil sample was placed in an ice bag for further analysis. Geophysical positioning of the sampling locations was recorded with the help of the Garmin-eTrex 30x GPS, and the details have been presented in Table 1.

Soil collection and analysis
The soil samples were collected from all locations, where fresh roots/rhizome were collected. The soil samples were collected up to a depth of 15 cm after removing the litter from the surface. The soil samples were also collected in three replicates from the plant sampling area. Then these soil samples were brought to the lab for further analysis. The collected soil samples were dried under shade and then grind with pestle and mortar passing through the sieve of spacing 2 mm. Prepared samples were used for the analysis of pH, electrical conductivity (EC), soil organic carbon (SOC), available nitrogen (N), available phosphorus (P), and available potassium (K). For determination of pH and EC value of the collected sample, the soil water suspension in the ratio of 1:2 was prepared as per standard method described by Jackson 67 . Then the pH and EC of soil water suspension (1:2) were determined with the help of pH meter (model Eutech Instruments pH 510) and EC meter (model Century), respectively. The soil OC was analyzed by the wet oxidation method 68 . Available N, P, and K were estimated as per standard methods reported in the literature 69,70 .

Essential oil extraction
The fresh samples were chopped into small pieces and placed in 5L asks. The samples size for all population was 400 g, and the samples included only roots and rhizomes. The size of the rhizome was upto 15 cm. The ratio between water and fresh roots/rhizome was 2:1 (v/w) during distillation in laboratory conditions. The oil was extracted through the hydro-distillation process in a Clevenger-type apparatus for 6 hours. After the collection of oil, the Clevenger apparatus was left as it is without removing the water and roots/rhizome for re-distillation. The next day, the respective Clevenger apparatus were run again without changing the water for 6 hours, and the oil recovery of recorded and collected separately. The same procedure was repeated in 3rd day. The amount of oil extracted was recorded daily, and the percentages of oil recovery (v/w) for each cycle were calculated based on the plant material used at rst day. The total oil yield and oil percentage were calculated by adding all three days' recovery. Then the extracted oil was dehydrated over anhydrous sodium sulfate (Merck), and stored in sealed glass vials at 4 ˚C for further studies.

GC-MS analysis
The GC-MS analysis of essential oil of V. jatamansi was done by a QP2010 (Shimadzu Corp., Tokyo, Japan) GC-MS system, which was tted out with an AOC 5000 Auto-injector and a ZB-5 (SGE International, Ringwood, VIC, Australia) silica capillary column with a dimension of 30 m × 0.25 mm i.d., and 0.25 µm lm thickness. The temperature of the oven was programmed in such a way that started from 70°C (for 4 min) to 220°C with a slope of 4°C min-1 and detention time of 5 min. The xed temperatures for injector and interface were 240°C and 250°C, respectively. The ow rate of helium, a carrier gas, was 1.05 mL min − 1, and the voltage for ionization was 70 eV. GC analysis GC analyses of essential oil of V. jatamansi were carried out by a Shimadzu GC-2010 gas chromatograph (Shimadzu, Tokyo, Japan) that was attached with a ame ionization detector (FID) and a ZB-5 MS capillary column (30 m 9 0.25 mm, fused silica, and lm thickness 0.25 lm). The temperature of the oven was programmed at 70°C for 3 min and gradually increased up to 220°C with a slope of 4°C min-1 and detention time of 5 min. The temperatures for injector and interface were 220°C and 250°C, respectively. Then the individual compounds were quanti ed based on the peak-area percentage of the chromatogram.

Identi cation of components
The different compounds of essential oil of V. jatamansi were identi ed through using retention indices (RI) that were computed with reference to homologous series of n-alkanes (C8-C24). The compounds were con rmed through a comparison between RI and mass spectra with the National Institute of Standards and Technology mass spectral (NIST-MS) database 71 . The GC/ GC-MS analyses were conducted for all three replications and oil obtained at all rounds of extraction. The total percentage shares of compounds were computed by addition of the individual percentage of all identi ed compounds.

Experimental design and statistical analysis
The data obtained from the four populations from wild and one cultivated samples were subjected to analysis of variance (ANOVA). Differences among the populations in terms of oil concentration were assessed with the Fishers least signi cant differences (LSD) test values only when the F-test in ANOVA was found signi cant (P = 0.05). In case of soil parameters, Fishers LSD post-hoc test was applied to compare the means. Principal component analysis (PCA) was performed to categorize the populations based on chemical compositions of essential oil. The PCA was conducted separately for all three types of oil samples, which were obtained from three different distillation days. However, a sample received from traditional practice was included in all the PCA for comparative studies. All the ANOVA and PCA were performed with the help of statistical software (Statistica 7 software; Stat. Soft Inc., Tulsa, Oklahoma, USA). The data on compounds of essential oil of this study were presented as mean ± standard error (SE) since many compounds were not detected during different distillation cycles. Figure 1 Concentration of essential oil in the different populations of V. jatamansi under repetitive distillation method  The dynamic of patchouli alcohol (a) seychellene (b) δ-guaiene (c) and α-patchoulene (d) under repetitive distillation method for the different populations