**4. Discussion**

The identification of the *Thymus* species is extremely difficult because of the high levels of diversity within the genus. This genus contains several commercially important aromatic species. For this purpose, the relationship among the chemical composition of essential oils and molecular analysis was carried out for different *Thymus* species [20,23]. In this context, the essential oil composition and molecular analysis of nine thyme cultivars were investigated in this study, to distinguish between commercial thyme cultivars and Korean native thyme cultivars. In the morphological study, the *T. quinquecostatus* and *T. vulgaris* cultivars exhibited a significant level of variability in recorded parameters. In the qualitative traits, a considerable variability was observed in stem type, stem color, length and number of stem branches, leaf shape, and trichome position, among and within *T. quinquecostatus* and *T. vulgaris* cultivars.

The present study showed a high chemical diversity among nine thyme cultivars. Results revealed that essential oils from Korean cultivars (*T. quinquecostatus*) belonged to the geraniol, thymol, and linalool chemotypes. Essential oils from the commercial thyme cultivars (*T. vulgaris*) such as creeping, golden, and orange belonged to the geraniol chemotype and lemon, and the silver cultivars belonged to the thymol chemotype. Further, carpet cultivar belonged to the linalool chemotype. In particular, these essential oils were dominated by monoterpenes. 1-Octen-3-ol, γ-terpinene, linalool, borneol, α-terpineol, nerol, geraniol, thymol, β-cubebene, β-elemene, caryophyllene, β-bisabolene, butylated hydroxytoluene, β-sesquiphellandrene, and caryophyllene oxide were detected in all six essential oils from the commercial cultivars. With regards to the chemical composition of *T. vugaris* essential oils, seven different chemotypes such as thymol, carvacrol, linalool, geraniol, thujanol-4, terpineol, and 1,8-cineole were identified [15,16]. In the case of *T. quinquecostatus* essential oils, Shin and Kim [8] found that thymol (41.70%), γ-terpinene (16.00%), and *p*-cymene (13.00%) were the most prominent compounds. Similarly, thymol (30.54%), γ-terpinene (23.92%), and *p*-cymene (11.13%) were the major components in the essential oil obtained from the Odae cultivar. However, the major components in the essential oils obtained from Wolchul and Jiri cultivars of *T. quinquecostatus* were not identical. In the Wolchul cultivar, geraniol (42.94%) and geranyl acetate (26.49%) were detected as the

major components, whereas linalool (47.89%) and thymol (15.98%) were found to be abundant in the Jiri cultivar.

Hudaib and Aburjai [24] determined variations in the composition of essential oils from cultivated and wild-growing plants of *T. vulgaris* grown in Jordan. Higher oil yields were obtained in plants growing wild, when compared to the cultivated plants. Among the four different samples, thymol (0.8–63.8%) and carvacrol (6.9–86.1%) were the most abundant components in the *T. vulgaris* essential oils. A study indicated that the essential oil composition of *T. vulgaris* highly varied both qualitatively and quantitatively during the vegetative cycle [25]. The variations in the yield and composition of essential oils could be influenced by various factors, such as the geographical region of the plant, plant's maturity, cultivation practices, and weather parameters (temperature, humidity, sunlight duration, and rainfall) [26–28]. In addition, the genetic constitution of the cultivars also played a considerable role in the essential oil composition [1,25].

According to previous reports, it is difficult to distinguish *Thymus* species and cultivars by analyzing the essential oil profile alone. Hence, the combined analysis of chemical composition and molecular techniques was used for the correct identification of the different plant species. In recent decades, the correlation between the chemical composition and molecular analysis of different *Thymus* species were investigated by various researchers [1,13,20,21]. Previous studies showed that both essential oil composition and RAPD analysis could be used to distinguish different thyme cultivars, and especially, to determine their relationships [1]. In addition, RAPD analysis revealed high polymorphisms even when using closely related genotypes. Even though the essential oil composition of plants was different from one another, RAPD analysis clustered these plants together, owing to their similar genetic background [15].

In the present study, 16 primers were used to amplify segments of DNA of the genome of three Korean thyme cultivars and six commercial thyme cultivars, to investigate the genetic variations. A total of 133 bands were obtained and the average percentage of the polymorphic bands was 93.23%. Based on the RAPD data, the similarity of the cultivars, estimated by the Jaccard's coefficient, is depicted in Figure 5. The nine cultivars of thyme fell into two clusters. Cluster 1 was formed by six cultivars (lemon, golden, creeping, silver, carpet, and Jiri) and cluster 2 by three cultivars (orange, Wolchul, and Odae). This emphasized the obvious variation between the Korean cultivars (except Jiri cultivar) and the commercial cultivars. The dendrogram indicated a clear separation of *T. quinquecostatus* from *T. vulgaris*, with the exception of the Jiri cultivar. According to the RAPD similarity matrix, it was observed that the Wolchul and Odae cultivars were closely related. Nevertheless, there was no significant relationship between the essential oil composition and RAPD data. The ability to discriminate all studied cultivars using RAPD bands indicated that RAPD analysis can provide a rapid and inexpensive technique to identify phenotypically similar thyme cultivars.

Based on previous reports, a high correlation between genetic and chemical relationships was attained in several plants. These data indicated that the composition of the essential oil is regulated by a number of genes that are extensively distributed throughout the plant genome [1,29,30]. Khalil et al. [31] used RAPD analysis to determine the genetic relationship between *T. vulgaris* populations collected in Syria. In their study, 13 individuals were analyzed using 27 primers, which generated 180 polymorphic bands from 198 bands. The authors found a significant correlation between *T. vulgaris* populations and their geographic areas. The present study also proved that the geographic distribution had a significant influence on genetic variation. Comparing the groups formed by the cluster analysis based on RAPD data (Figure 5) and chemotype, based on essential oil composition, we can observe that the groups formed in both cases were not identical.

In another study, the composition of essential oils and genetic relationships between six commercial cultivars of *T. vulgaris* were analyzed. A total of 104 were polymorphic RAPD bands (63.8%) were obtained using 15 primers. Among 15 primers, the highest percentage of polymorphism was obtained by the OPA-05 primer (90.9%). Similar to the essential oil composition, the six *T. vulgaris* cultivars fell into two major clusters, according to the RAPD patterns, with a correlation coefficient of −0.779 [1]. The chemical and genetic variations of 20 taxa from four Hungarian *Thymus* species (*T. glabrescens*, *T. pannonicus*, *T. praecox*, and *T. pulegioides*) were studied by Pluhár et al. [23]. In the molecular analysis, 114 polymorphic RAPD bands (80.8%) were obtained using 13 primers. The results revealed that partial correlation was found between the essential oil and RAPD analyses. The essential oil composition and genetic variation in six micropropagated genotypes (in vitro and in vivo) of *T. saturejoides* were investigated by Nordine et al. [32]. RAPD results and the essential oil composition grouped these six genotypes into three clusters exhibiting significant intraspecific chemical and genetic di fferences. Furthermore, a significant correlation was observed between RAPD and essential oil composition obtained from the in vitro genotypes.

Similar to our report, several studies also reported that the combined use of RAPD and essential oil analyses were not significantly correlated. For example, the genetic and chemical relationships among 31 individuals of *T. caespititius* collected from the islands of Pico, Sao Jorge, and Terceira (Azores) were determined. In the RAPD analysis, 187 polymorphic bands were obtained using 17 primers. However, there was no close relationship between the collection site, the essential oil composition, and RAPD analysis [15]. Rustaiee et al. [20] also studied the essential oil composition and genetic variability between some *Thymus* species such as *T. daenensis* (two populations), *T. fallax*, *T. fedtschenkoi*, *T. migricus*, and *T. vulgaris*, using GC-MS and RAPD. Although the RAPD markers allowed a perfect distinction among di fferent *Thymus* species according to their characteristic genetic background, there was no identical clustering with the essential oil composition. In addition, Masi et al. [33] found that the essential oil compositions did not match with the results achieved from agronomic and genetic analyses in *Ocimum basilicum*. In another study, there was no correlation between RAPD and the essential oil obtained from the in vivo genotypes of *T. saturejoides* [32]. Based on the previous and present studies, marker-assisted RAPD technique had a high advantage for the assessment of the genetic di fferences of plant species without prior molecular knowledge.

Results of the present study revealed that there was a significant correlation between the genetic and geographic distances of the Korean thyme cultivars (Wolchul and Odae cultivars), compared to the commercial thyme cultivars. However, the chemical polymorphism of these thyme cultivars is not well-understood. Hence, other molecular techniques should be investigated in order to understand this question in *T. quinquecostatus* and other *Thymus* cultivars.
