Phytochemical Variability of Essential Oils of Two Balkan Endemic Species: Satureja pilosa Velen. and S. kitaibelii Wierzb. ex Heuff. (Lamiaceae)

Satureja pilosa and S. kitaibelii (Lamiaceae) are Balkan endemic plant species, and the composition of their essential oil (EO) is highly variable. The aim of the present study was to establish: (1) the EO variability in two populations of S. pilosa (the intrapopulation), and (2) the EO variation in S. kitaibelii between nine populations (interpopulation) from Bulgaria and two from Serbia. The EOs of two Satureja species were obtained from aboveground plant parts by hydrodistillation and were analyzed by GC/MS/FID. Overall, the EO yield on the intrapopulation level of S. pilosa varied from 0.54% to 2.15%, while the EO of S. kitaibelii varied from 0.04% to 0.43% (interpopulation). The EO of S. pilosa was found to contain thymol and carvacrol as the main constituents, with other major constituents being p-cymene and γ-terpinene. S. pilosa samples in both studied populations formed six chemical groups. The major constituents (p-cymene, terpinen-4-ol, bornyl acetate, γ-muurolene, endo-borneol, cis-β-ocimene, trans-β-ocimene, carvacrol, α-pinene, thymoquinone, geranial, geranyl acetate, spathulenol, and caryophyllene oxide) of S. kitaibelii EO were considered for grouping the populations into ten chemotypes. The present study is the first report on the interpopulation diversity of S. kitaibelii EOs in Bulgaria. It demonstrated variability of the EOs between and within the populations of S. kitaibelii from Bulgaria. This study identified promising genetic material that could be further propagated and developed into cultivars for commercial production of S. kitaibelii and S. pilosa, thereby reducing the impact of collection on wild populations.


Introduction
Bulgarian flora is characterized by a diversity of plant species, including Bulgarian and Balkan endemics [1]. This is due to the specific geographical location of Bulgaria and its geological history [2]. Satureja pilosa and S. kitaibelii (Lamiaceae) are Balkan endemic species and both have a limited distribution [3]. Satureja kitaibelii (S. montana ssp. kitaibelii) is spread on stony poor soils in the northern part of Bulgaria (Znepole region; Sofia region; the Northern part of the Balkan Mountains), while S. pilosa is spread on rocky habitats in the Central and Eastern Balkan Mountains and the Eastern Rhodope Mountains [4]. Satureja species are well known as culinary and medicinal herbs in Bulgaria [5]. Satureja products are used as natural preservatives for food and in many other industries (perfume, cosmetic, Table 1. The main compounds of essential oils of this study and literature reports on S. kitaibelii.

Bulgaria
Slavkovska et al. [19] α-pinene (6.0); camphene (3.7); myrcene (1.0); p-cymene (20.9); γ-terpinene (2.0); limonene (16.0); borneol (9.8); carvacrol (0.9). Yugoslavia Chalchat et al. [25] α-pinene (1.  Table 2 reveal the presence of considerable variability of the EO yield between samples at the intrapopulation level (each plant is one sample, a total of 12 plants). In the first population from the Balkan Mountains (the village of Selce), the oil yield between the different plants (samples) varied from 1.27% to 2.15%, while in the second population (the village of Samokitka in the Rhodope Mountains), the EO yield varied from 0.54% to 1.89%. In some of the studied samples the amount of EO was high (1.89-2.15%), and the results obtained by us were close to the yield obtained by Bezić et al. [26] for other types of Satureja (S. montana L. 2.8%, S. cuneifolia Ten. 2.6%, S. subspicata Vis. 2.0%, S. visianii Šilić 2.4%). Similar variability in the S. pilosa EO yield was detected in previous research of the populations in the territory of Bulgaria [12,17], Turkey [14], and Greece [13]. Overall, the data of the S. pilosa EO yield on the intrapopulation level cannot be compared with literature data because our study reported the yield of 24 individual plants (samples) within two geographically isolated populations. Very often, researchers take and analyze a compound sample of a given population. Mostly, genetic features affect the yield of S. pilosa because: (1) the samples were collected at the same time; (2) the samples were at the same phenological stage; (3) the plants within a population were growing under the same climatic conditions; and (4) the EO was extracted using the same method. Generally, the EO yield of S. pilosa varied from 0.54% to 2.15% between the two populations ( Table 2). The S. kitaibelii EO yield from eleven locations varied from 0.04% (Kostenkovci 1) to 0.43% (Gradec 1), while the EO yield of cultivated S. montana was 0.70% (Table 3). A similar variation of EO yield of S. kitaibelii was previously reported in one population each in Bulgaria (0.19%) [17] and Serbia (0.11% to 0.27%) [20].  Table 2). The same results were found in a previous study on S. pilosa samples collected from 33 locations [12]. The most abundant compounds of class monoterpenes were thymol, carvacrol, p-cymene, and γ-terpinene (Supplementary Table S1).
The analysis of variance results shown in Table 4, followed by the multiple means comparison results shown in Tables 5 and 6, reveal that the EO profile of S. pilosa was specific to each population studied, and the EO had a clearly differentiated profile, which supports our working hypothesis. For example, carvacrol (38.99-77.09%) was the main compound of EO in 10 samples collected from the Balkan Mountains (village of Selce), and thymol (19.34% and 1.55%, respectively) was found in only two samples from this population (Table 5). For the samples from the Eastern Rhodope Mountains (Samokitka) in the EO composition, thymol predominated (52.52-72.20%), and in none of the samples was the presence of carvacrol detected. Bezić et al. [26] showed that the production of phenolic compounds in the EO was stimulated by hot and dry conditions of the environment. The studied populations of S. pilosa were found in petrophytic cenoses of the xerophytic mountain range. The plants were located on open, rocky terrain with southern exposure and high solar radiation, which only partly supports the assumption of Bezić et al. [26]. In this study, the population in the Balkan Mountains (Selce) was under the influence of a temperate continental climate zone and lower temperatures, while the population in the Eastern Rhodope Mountains (Samokitka) was under the influence of a Mediterranean climate characterized by higher temperatures [27]. Additionally, Bezić et al. [26] showed that the yield of carvacrol varied during ontogenesis and peaked during blossoming and high summer temperatures. In this study, the results from the population in the Eastern Rhodope Mountains (Samokitka) do not support the report of Bezic et al. [26]; the plants were grown at higher temperatures but carvacrol was not identified in them. Therefore, one may assume that temperature during vegetation may be less significant on carvacrol accumulation compared with that of plant genetics. Apparently, the EO profile of S. pilosa is a function of mostly genetic traits.
A total of 18 compounds were identified in the EO of the species from both studied populations, as p-cymene (9.53-24.54%), γ-terpinene (1.23-10.75%), and terpinen-4-ol (0.36-3.87%) were present in all analyzed samples (Tables 4-7). This is not surprising, because p-cymene and γ-terpinene are precursors in the biosynthesis of thymol and carvacrol [28]. Poulose and Croteau [28] noted that thymol is biosynthesized by the aromatization of γ-terpinene to p-cymene followed by hydroxylation of p-cymene. Thymol and carvacrol have important biological activities such as antibacterial activity, antioxidant activity, anticancer potency, and antiseptic activity [29][30][31]. Previous studies reported thymol, carvacrol, γ-terpinene, and p-cymene as the main compounds of Satureja species EO [13,[32][33][34][35]. Furthermore, the cited authors reported quite a few chemotypes in Satureja species [36][37][38]. Apparently, there is high variability of EO in Satureja species, and chemotypes are common in the genus Satureja.   Sesquiterpenes was the second established class in the samples from both studied populations in minimal quantities (0.89-5.34%) ( Table 3). From these, trans-caryophyllene and caryophyllene oxide were present in the EO of all analyzed samples (Table 6). Overall, this is the first comprehensive study on the endemic plant S. pilosa that assessed the variability of EO on intrapopulation level. Based on the % ratio of the major compounds of the monoterpenes that this intrapopulation study, six chemotypes from two populations of S. pilosa were identified. In the two studied populations, three different chemical groups emerged. For a population from the Balkan Mountains (village of Selce), the following distribution was found: in seven samples there was predominance of (1) the carvacrol, p-cymene type; in four samples (2) the carvacrol, p-cymene and γ-terpinene type; and in one sample (3) the carvacrol, p-cymene, thymol, and γ-terpinene type, respectively. EO chemical differentiation was also found in the samples collected from the Eastern Rhodope Mountains (Samokitka). In three of the analyzed samples from this population, the main EO compound groups were found to be (4) the p-cymene and thymol type; in one sample (5) the p-cymene, thymol, cis-β-ocimene, and γ-terpinene type; and in eight samples, the main EO compounds were (6) p-cymene, thymol, and γ-terpinene. The lowest values (lowest percent composition limits) for the respective compounds in the above groups were: 19.34% for thymol, 9.53% for p-cymene, 5% for γ-terpinene, and 38.99% for carvacrol. Similar variability in the EO content of S. pilosa was found in our previous study in 33 locations in Bulgaria, where we established five different chemotypes [12]. Similar intrapopulation and interpopulation variability in EO content and composition were observed on S. rechingeri Jamzad from populations in Iran [33]. We should note that in a previous study we found thymol and carvacrol in samples both from the Balkan Mountains and from the Eastern Rhodope Mountains [12]. Both of the studied populations are geographically isolated (allopatric), and there is a small possibility of genetic exchange between them. Due to the fact that S. pilosa is a cross-pollinated plant, genetic material is exchanged within populations. As a result of the genetic drift within populations, genes are combined with large effects and tight control of the biosynthesis [39]. Apparently, distinct patterns in compositional variability of EO were the result of a combination of genes; as a result, the S. pilosa EO composition is quite variable [12]. Due to ongoing genetic processes in the populations of S. pilosa, morphological differences in the species have been found [4,13]. It is on the basis of morphological differences that some authors have proposed new taxonomic ranks for S. pilosa. Anchev [4] proposed under rank S. pilosa subsp. pilosa, S. pilosa var. slavjankae Anchev for the territory of Bulgaria, while Dardioti et al. [13] proposed under rank S. pilosa subsp. origanita Dardioti & Kokkini for the territory of Greece. Obviously, populations of S. pilosa are not genetically balanced, and there are ecotypes of the species.
In general, in order to select plants of the EO species rich in thymol, it is necessary to collect material from the population of the Eastern Rhodope Mountains (Samokitka) and carvacrol from the population of the Balkan Mountains (village of Selce).

Satureja kitaibelii
The results of the EO composition of S. kitaibelii spread in nine Bulgarian locations and two Serbian locations very close to the Bulgarian border are shown in Tables 8 and 9, as well as in Supplementary Table S2. The main class of compounds in 10 samples of the 11 studied samples of S. kitaibelii was the monoterpenes (39.40-81.23%), with the exception of one sample from Serbia, where monoterpenes were 17.75% and the unknown compounds reached 76.28%. For S. montana ssp. montana (cultivated), the amount of monoterpenes reached 92.87% (Table 8). In our study, we identified 33 substances in EO compositions that varied from sample to sample. EO compounds of the monoterpene class (39.40-81.23%) were α-pinene, p-cymene, γ-muurolene, endo-borneol, terpinen-4-ol, thymoquinone, geraniol, geranial, bornyl acetate, carvacrol, and geranyl acetate. Cluster analysis results of S. kitaibelii and S. montana ssp. montana that showed the similarity of location (11 levels) and the concentrations of 11 compounds are shown in Figures 1 and 2. EO gas chromatographic analysis showed that samples collected from the same location have different EO compositions, although the plants developed under the same climatic conditions. For example, samples collected from the same location (Gradec 1 and Gradec 2) contained p-cymene amounts of 20.38% and 19.39%, respectively, and γ-muurolene amounts of 7% and 11.34%, respectively (Table 9). However, we should also note that in the first sample from this location (Gradec 1), terpinen-4-ol (14.95%) and bornyl acetate (5.77%) were found, which are compounds not present in the EO composition of the other sample from this location (Gradec 2). Variability in the EO composition of species has also been found for samples from other habitats. For example, the main EO compounds of the plants from Buchin prohod were endo-borneol (12.20%), p-cymene (5.56%), and γ-muurolene (18.72%), as well as carvacrol (5.06%) (Supplementary Table S2); while plant samples from Glozhene monastery contained geraniol (34.78%) and geranyl acetate (18.5%); and plants from Beldie han contained endo-borneol (11.98%), p-cymene (12.10%), γ-muurolene (6.68%), and αpinene (4.78%) as major constituents (Table 8). In a previous study of samples from the same location in Bulgaria (Beldie han), Konakchiev and Tsankova [17] found high levels of limonene (15.7%) and p-cymene (13.1%). In this study, limonene was not found as a major constituent of the EO in plants from this population at Beldie han. Moreover, we did not find limonene in any of our analyzed samples.
It is evident that the EO of S. kitaibelii varies considerably both between population and within the populations themselves. A similar conclusion was reached by authors wh analyzed samples of the species in Serbia [15,20,24]. The cited authors showed that ther  Table 8.
It is evident that the EO of S. kitaibelii varies considerably both between populations and within the populations themselves. A similar conclusion was reached by authors who analyzed samples of the species in Serbia [15,20,24]. The cited authors showed that there is EO variability both between different plant parts and between different populations of the species in Serbia [15,20,24]. Samples tested in Serbia showed high concentrations of geranyl acetate, geraniol, linalool, 1,8-cineole, limonene, bornyl acetate, and camphene [15,19,25].
The habitats of S. kitaibelii in the populations of Gradec, Beldie han, Buchin prochod, and Petrohan prochod are located in the western part of the Balkan Mountains in Bulgaria. The plants grow on limestone-based rocks in the north Bulgarian climate area under the influence of a temperate-continental climate zone [27]. According to Kopralev et al. [27], the Balkan Mountain range is divided into three parts-Northern, Central, and Eastern Balkan Mountains. The samples from the Western Balkan Mountains (locations of Gradec, Beldie han, Buchin prochod, and Petrohan prochod) have different phytochemical EO profiles than the samples obtained from the Central Balkan Mountains (Kostenkovci) and the samples from the locations that border the Western and Central Balkan Mountains (Glozhene monastery) (geraniol, geranyl acetate) ( Table 7). For example, the samples from the Western Balkan Mountains are predominated by p-cymene, γ-muurolene, endo-borneol, bornyl acetate, and carvacrol, while thymoquinone, geranial, geranyl acetate, spathulenol, and caryophyllene oxide were the main compounds of samples from the Central Balkan Mountains (Kostenkovci). Different EO compositions (geraniol, geranyl acetate) were found in the samples from the locations that border the Western and Central Balkan Mountains (Glozhene monastery) ( Table 7).
Geraniol and geranyl acetate were found only in two samples from the Glozhene monastery (34.78% and 18.15%, respectively) and Kostenkovci (39.27% and 33.04%, respectively), while one of the samples from Serbia contained high concentrations only of geraniol (47.54%). The location of S. kitaibelii was spread around the Glozhene monastery and Kostenkovci (Central Balkan Mountains), and those from Serbia were geographically separated (allopatric). As is well-known, the Balkan Mountains comprise the longest mountain range in Bulgaria and Serbia [27]. Our studied locations of S. kitaibelii in the Balkan Mountains (Bulgaria and Serbia) represent the general geographical population of the species. Currently, S. kitaibelii populations are fragmented with a mosaic structure. Probably in the past, the studied locations in the geographical population were interconnected. The species had much larger areas of distribution and numbers, which is why there was a free exchange of genetic materials. This is probably one of the reasons why samples that are geographically separated have similar EO phytochemical compositions. Likely, the geographical location of the population of S. kitaibelii does not significantly affect the composition of the essential oil, but it is mainly genetic features that do so.
In general, we can note that of the studied locations of S. kitaibelii with the most specific and different EO compositions were the samples collected from the Central Balkan Mountains (Kostenkovci). The main EO compounds for these samples (Kostenkovci) were thymoquinone (31.65%), geranial (5.68%), geranyl acetate (28.87%), spathulenol (6.09%), and caryophyllene oxide (8.61%) (Supplementary Table S2). In EO compositions of another sample from the same location were predominant thymoquinone (29.78%), geranyl acetate (30.73%), and γ-muurolene (7.25%) (Supplementary Table S2). We should emphasize that high concentrations of thymoquinone (31.65% and 29.8%, respectively) in the EO of S. kitaibelii were reported for the first time. Thymoquinone is one of the abundant compounds in Nigella sativa L. oil and is responsible for many pharmacological activities (inflammatory) and degenerative diseases, including cancer [40]. Geraniol is a commercially important compound and is very important for the flavor and fragrance industries [41]. Geraniol has a wide range of biological activities as an antimicrobial, antioxidant, and anti-inflammatory, but also as a chemoprevention agent for cancer [41].

Drying of the Samples
All Satureja samples were air dried in a shady area at temperatures below 30 • C to avoid oil losses and changes in the EO profile prior to the extraction.

Preparation of Samples for the EO Isolation
Subsamples of S. kitaibelii were generated randomly from each air-dried sample. The whole aboveground plant parts were used in this study. Each of the twelve collected plants per population (total 24) of S. pilosa represented an individual sample.

Essential Oil (EO) Isolation of the Satureja Biomass Samples
The EO of the aboveground plant parts was extracted via hydrodistillation in 2 L distillation units (Laborbio Ltd. Sofia, Bulgaria, www.laborbio.com, accessed on 1 May 2022) at the Research Institute for Roses and Medicinal Plants in Kazanluk, Bulgaria. Each extraction was performed in two replicates. Samples of dried aboveground plant parts (Supplementary Table S3) plus 0.800 L of water were placed in a Clevenger distillation system. After isolation of each subsample, the EO volume and weight were measured, and the EO samples were stored in a freezer until analyses. Here, we report the oil content (yield) based on weight in dried matter.

Gas Chromatography Mass Spectrometry Flame Ionization Detection (GC-MS-FID) Essential Oil Analysis
The extracted EO from all Satureja samples in two replications were analyzed for chemical profile by gas chromatography (GC)-mass spectroscopy (MS)-flame ionization detection (FID) techniques. Using a micropipette, 50 µL of oil (weight measured on a tared balance) from each sample was transferred into a 10 mL volumetric flask. Samples were brought to volume with CHCl 3 . A 1 mL aliquot of each diluted oil sample was placed by glass pipet into a GC vial for analysis.
Oil samples were analyzed by GC-MS-FID on an Agilent (Santa Clara, CA, USA) 7890A GC system coupled to an Agilent 5975C inert XL MSD. Chemical standards and oils were analyzed using a DB-5 column (30 m × 0.25 mm fused silica capillary column, film thickness of 0.25 µm) operated using an injector temp of 240 • C, column temperature of 60 to 240 • C at 3 • C/min and held at 240 • C for 5 min, helium as the carrier gas, an injection volume of 1 µL (split ratio 25:1), and an MS mass range from 50 to 550. FID temperature was 300 • C. Post-column splitting was performed so that 50% of outlet flow proceeded to FID and 50% to mass spectrometry (MS) detection.
Compounds were quantified by performing area percentage calculations based on the total combined FID area. For example, the area for each reported peak was divided by total integrated area from the FID chromatogram from all reported peaks and multiplied by 100 to arrive at a percentage. The percentage of a peak is a percentage relative to all other constituents integrated in the FID chromatogram.
3.6. Statistical Analyses 3.6.1. Satureja pilosa The effect of location (Selce and Samokitka) and sample (695-697 and 728-736 in Selce, 715-727 in Samokitka) nested in location on 15 constituents (α-pinene, α-thujene, camphene, myrcene, p-cymene, α-terpinene, γ-terpinene, unknown, terpinen-4-ol, trans-caryophyllene, caryophyllene oxide, cis-β-ocimene, thymoquinone, thymol, and carvacrol) was determined using a nested design with the two effects in the model being location and sample (location). The analysis of variance (ANOVA) was done using Proc Mixed of SAS [45]. Since there were 12 samples from the Selce location and 12 samples from the Samokitka location, the multiple means comparison of the samples nested within location was done on 24 means, which is large. Therefore, to minimize the potential over inflation of Type II experiment-wise error rate, LSD at 1% significance level was used for letter grouping.
For each response variable, normal distribution and constant variance assumptions on the error terms were validated as described in Montgomery [46], and an appropriate transformation was applied on some of the response variables where the assumption was violated; however, the results in the tables are presented after back transforming them to the original scale.

Satureja kitaibelii and S. montana (Cultivated)
The significance of the location (11 levels) effect on EO yield, and the concentrations of 11 constituents (α-pinene, camphene, myrcene, p-cymene, cis-β-ocimene, γ-terpinene, endo-borneol, trans-caryophyllene, γ-muurolene, spathulenol, and caryophyllene oxide) were determined using ANOVA, and the data were analyzed using Proc Mixed of SAS. Model assumptions were also verified as described in Montgomery [46]. Since location had significant (p-value < 0.05) effects on EO yield and the concentrations of all constituents, Tukey's multiple range test at α = 5% was used to generate letter groupings.
To determine the similarity level of the locations in terms of all 11 constituents, and the constituents in terms of all 11 locations, cluster analysis (complete linkage clustering) was conducted to generate dendrograms for the locations and the constituents as described in Johnson and Wichern [47].

Conclusions
This is the first comprehensive study on the EO chemical variability of the Balkan endemic S. pilosa on the intra-population level and the variability of EO of S. kitaibelii collected from wild populations in Bulgaria. The quantitative and qualitative EO compositions of S. pilosa in both of the studied populations varied widely. Expression of the S. pilosa EO variations are the established chemical types that are most probably due to genetic specificity. In the first population (Balkan Mountains) S. pilosa chemical groups were: (1) the carvacrol and p-cymene type; (2) the carvacrol, p-cymene, and γ-terpinene type; and (3) the carvacrol, p-cymene, thymol, and γ-terpinene type; while for the second population (the Eastern Rhodope Mountains), the chemotypes included (1) the p-cymene and thymol type; (2) the p-cymene, thymol, cis-β-ocimene, and γ-terpinene type; and (3) the p-cymene, thymol, and γ-terpinene type.
The interpopulation variability in the Bulgarian populations of S. kitaibelii EO was studied for the first time. Significant variability in EO quantity and quality was found. Thus, 10 chemical types were identified. The presence of thymoquinone in the EO of S. kitaibelii from the population in Kostenkovci is reported for the first time. If the chemical profile of the plant population at Kostenkovci is of industrial interest, then plants can be selected for introduction into a crop culture.