Morphological and Chemical Traits as Quality Determinants of Common Thyme ( Thymus vulgaris L.), on the Example of ‘Standard Winter’ Cultivar

: Common thyme is regarded as one of the most important culinary plants. The purpose of the work wasto determine the intraspeciﬁcvariability of commonthyme with respect to morphological and chemical characters including the content and composition of essential oil and phenolic compounds in the herb. The objects of the study were 12 clones, vegetatively multiplied from randomly selected individual plants of cultivar ‘Standard Winter’. The morphological observations and harvest of raw materials were carried out in the ﬁrst year plants’ vegetation. The highest di ﬀ erences between clones were on fresh and dry weight of herb (CV = 0.38 and 0.36, respectively), width of leaves (CV = 0.21), and density of glandular trichomes on the abaxial surface of leaves (CV = 0.29). Examined clones were also di ﬀ erentiated as to the chemical features. Essential oil content (performed by hydrodestillation) and composition (by GC-MS and GC-FID) were determined and they ranged from 2.10 to 4.38 g × 100 g − 1 DW. Here, thymol, γ -terpinen, and p -cymen were the dominant compounds. Clone no. 4 was distinctive as to the highest content of essential oil followed by the highest share of thymol (54.59%). The total content of phenolic acids and ﬂavonoids (determined according to PPh 6th) also di ﬀ ered among clones (CV = 0.38 and 0.36, respectively). Using a validated HPLC-DAD method, the following compounds were identiﬁed: ca ﬀ eic, rosmarinic, p -coumaric acids, luteolin 7- O -glucoside, naryngenin, and ( − )-epicatechin. Here, rosmarinic acid followed by luteolin 7- O -glucoside were present in the highest amounts (611.47–2675.59 and 46.77–325.11 mg × 100 g − 1 DW, respectively). The highest di ﬀ erences between clones were the contents of p -coumaric acid (CV = 0.59), luteolin 7- O -glucoside (CV = 0.50) and rosmarinic acid (CV = 0.40). Such a high range of variability can provide problems with raw material standardization. Nevertheless, it opens possibilities for breeders, whereas individual plants / clones may become valuable components for breeding.


Introduction
The genus Thymus, one of the most important taxon of the Lamiaceae family, is widely distributed in the Old World, with the center of origin located in the Mediterranean region. The genus is represented by 214 species and 36 subspecies, divided into eight sections: Micantes, Mastichina, Piperella, Teucrioides, Pseudothymbra, Thymus, Hyphodromi and Serpyllum [1]. Thymus species have been known for ages Given the use of common thyme variability in breeding programmes, many traits are taken into consideration, e.g., content and composition of essential oil, developmental features, yield of herb, resistance to biotic and abiotic stresses (especially frost), as well as the ability for regrowth. In Europe, a few ecotypes of common thyme that vary as to their morphological, physiological, and chemical traits, have been distinguished and cultivated. Here, 'German type' appeared as a result of selection managed by generations of farmers [26]. This type is frost resistance, cultivated as a perennial, characterized by erect type of growth and greyish-green color of leaves. 'French type', the one with narrow leaves, is considered as frost sensitive and cultivated as an annual plant. 'Mediterranean type', so called 'Valdonian type', was introduced into cultivation from a population growing wild in Italy. This type is distinguished by grayish-blue leaves, low frost resistance, lignified stems, and superior essential oil content and composition [26,36,37]. Another intraspecific division was mentioned by Melchior and Kastner [38] where two forms of common thyme were selected on the bases on the leaf blade shape and size. Here, T.vulgaris f. capitatus Willk. et. Lange is characterized by oblong-lanceolate leaves, 4-5 mm long and 1-2 mm wide, while T.vulgaris f. verticillatus Willk. et. Lange by elliptic and/or oblong-obovate leaves, 5-8 mm long and 3-4 mm wide. Both forms can cross with each other and give fertile progeny, being intermediate forms. Rumińska [37] claims that f. capitatus covers the 'French type' of common thyme, while f. verticillatus corresponds to 'German type'.
The aim of the study was to determine the intraspecific variability of common thyme 'Standard Winter' cultivar, with respect to selected morphological traits and the chemical composition of the herb, including the content and composition of essential oil, flavonoids and phenolic acids.

Plant Material
The experiment was performed at the experimental field of the Department of Vegetable and Medicinal Plants, Warsaw University of Life Sciences (WULS-SGGW) (5210180 N; 2105234 E), on alluvial soil. Common thyme seeds (cultivar 'Standard Winter', commonly cultivated in Central Europe) were purchased from the Jelitto company (Schwarmstedt, Germany). Seeds were sown in the second week of February into multi-pots filled with a peat substrate, in a greenhouse. In the first week of May, 100 seedlings were randomly selected and planted. In August, 12 individual (maternal) plants were chosen and marked for vegetative multiplication. In December, the plants were transferred into the greenhouse. In February of the second year, green, soft cuttings were collected from each plant and rooted. Well rooted cuttings were planted out into the field, in a randomized block design (20 cuttings per plot; in 3 replications), with a spacing of 20 × 40 cm. Thus, the object of the study were 12 clones of common thyme, vegetatively multiplied from individual plants selected within the cultivar. The observations of morphological traits (Section 2.2) followed by harvest of herb was carried out on 1-year old plants (clones), at the beginning of plant blooming. The herb (upper, not woody parts of shoots) was cut from each plot, at a height of about 10 cm above ground. The fresh and dry weight of the herb was estimated (g per plant). After drying at 35 • C, leaves were separated from stems, then weighed. Leaves (so-called rubbed herb) were ground for chemical analysis (Section 2.3).

Morphological Observations
Morphological traits of the clones were estimated according to the List of Descriptors for Thymus vulgaris L. elaborated by the Medicinal and Aromatic Plants Working Group of European Cooperative Programme for Plant Genetic Recourses (MAPs WG ECP/GR) [39]. Observations were performed directly before harvest of raw material, on 10 plants per one clone. Plant growth habit, plant height (cm), plant diameter (mm), foliage density, leaf shape and color, leaf length (mm), and width Cooperative Programme for Plant Genetic Recourses (MAPs WG ECP/GR) [39]. Observations were performed directly before harvest of raw material, on 10 plants per one clone. Plant growth habit, plant height (cm), plant diameter (mm), foliage density, leaf shape and color, leaf length (mm), and width (mm), were determined. Moreover, observations concerning density of glandular trichomes on abaxial and adaxial surface of leaves were estimated. Samples of leaves (5 per plant) were collected, and directly subjected to microscopic observations. Stereoscope microscope Nikon SMZ 745 T (Tokyo, Japan) equipped with Camera Invenio 3SPixel CMOS and Coolview v.1.6 Programme (Precoptic Co, Warsaw, Poland) was used. The number of glandular trichomes was counted on both sides of the leaf, then expressed as number per 1 mm 2 . Photographic documentation was performed (Figures 1 and 2).

Chemical Analysis
2.3.1. Content of Essential Oil 30 g of air-dried raw material was used for hydrodistillation for 3 h using a Clevenger-type apparatus. The content of essential oil was expressed as g × 100 g −1 of dry weight (DW). Obtained essential oils were stored in amber vials, at 4 °C.

Analysis of Essential Oils by GC-MS and GC-FID
The qualitative and quantitative analysis was performed by usage an Agilent Technologies 7890A gas chromatograph equipped with a flame ionization detector (FID) and MS Agilent  Cooperative Programme for Plant Genetic Recourses (MAPs WG ECP/GR) [39]. Observations were performed directly before harvest of raw material, on 10 plants per one clone. Plant growth habit, plant height (cm), plant diameter (mm), foliage density, leaf shape and color, leaf length (mm), and width (mm), were determined. Moreover, observations concerning density of glandular trichomes on abaxial and adaxial surface of leaves were estimated. Samples of leaves (5 per plant) were collected, and directly subjected to microscopic observations. Stereoscope microscope Nikon SMZ 745 T (Tokyo, Japan) equipped with Camera Invenio 3SPixel CMOS and Coolview v.1.6 Programme (Precoptic Co, Warsaw, Poland) was used. The number of glandular trichomes was counted on both sides of the leaf, then expressed as number per 1 mm 2 . Photographic documentation was performed (Figures 1 and 2).

Chemical Analysis
2.3.1. Content of Essential Oil 30 g of air-dried raw material was used for hydrodistillation for 3 h using a Clevenger-type apparatus. The content of essential oil was expressed as g × 100 g −1 of dry weight (DW). Obtained essential oils were stored in amber vials, at 4 °C.

Analysis of Essential Oils by GC-MS and GC-FID
The qualitative and quantitative analysis was performed by usage an Agilent Technologies 7890A gas chromatograph equipped with a flame ionization detector (FID) and MS Agilent

Chemical Analysis
2.3.1. Content of Essential Oil 30 g of air-dried raw material was used for hydrodistillation for 3 h using a Clevenger-type apparatus. The content of essential oil was expressed as g × 100 g −1 of dry weight (DW). Obtained essential oils were stored in amber vials, at 4 • C.

Total Content of Phenolic Acids and Flavonoids
Both groups of compounds were determined according to the Polish Pharmacopoeia 6th edition [41]. Phenolic acids content was determined by Arnov's method, and expressed as caffeic acid equivalent (g × 100 g −1 DW), while flavonoids were determined by the aluminum chloride colorimetric method and expressed as quercetin equivalent (g × 100 g −1 DW). A detailed description of these methods has already been given by Kosakowska et al. [42].

Validation
The standards were purchased from Merck (Darmstadt, Germany) and ChromaDex ® (Irvine, CA, USA) and separately dissolved with MeOH in a 25 mL volumetric flask according to the ChromaDex's Tech Tip 0003: Reference Standard Recovery and Dilution and used as standard stock solutions [43]. Working solutions were prepared by diluting 10 µL and 100 µL of standard stock solutions with methanol in 10 mL volumetric flasks, 500 µL and 1000 µL in 5 mL volumetric flasks as well as 1000 µL in 2 mL volumetric flasks. The working solutions and undiluted stock solutions were injected (1 µL) on a column in six replicates (n = 6) using SIL-20AC HT. Six-point calibration curves were plotted according to the external standard method by correlating concentration with peak area. Curves parameters were calculated with Microsoft Excel 14. Signal-to-noise ratio approach were used to determine LOD (S/N of 3:1) and LOQ (S/N of 10:1) ( Table 1). The peak table and UV-spectra library (190-450 nm) of individual compounds were also created. Table 1. Validation parameters of the HPLC-DAD analysis (n = 6).

Sample Preparation
Raw material (1.000 g) was homogenized and extracted with 100 mL of methanol in Extraction System B-811 (Bűchi Labortechnik AG, Flawil, Switzerland). Soxhlet hot extraction with twenty-five extraction cycles, flushing and drying was used. After evaporation of solvent, the residue was dissolved in 5 mL of methanol. The obtained extracts were filtered with ISO-Disc™ Filters PTFE-25-2, diameter 25 mm, pore size 0.20 µm (Supelco Analytical™, Bellefonte, PA, USA) and subjected to HPLC-DAD.

Parameters of Integration
Peak identification was carried out by comparison of retention time as well UV-spectra with standards. The content of the determined compounds was calculated in mg × 100 g −1 DW.

Statistical Analysis
Data were subjected to statistical analysis using Statistica software. The mean values were compared by using the one way analysis of variance (ANOVA) followed by Tukey's multiple range test. The differences between individual means were deemed to be significant at p < 0.05. Standard deviation (±SD) and coefficient of variation (CV) was estimated.

Results and Discussion
Investigated clones differed in respect of morphological traits. They were characterized by the sub-erect (7 clones) or erect (5 clones) type of growth habit ( Table 2). The plants' heights ranged from 19.75 (clone no. 4) to 28.00 cm (clone no. 6), with the CV at the level of 0.08, while plants' diameters ranged from 306.7 (clone no. 12) to 456.7 cm (clone no. 10) ( Table 3). Dalir and Safarnejad [34] claim that the minimum diameter of the common thyme plant is 19.42 cm, while the maximum is 29.26 cm. Listed features, especially type of growth habit and plant's height, are important when suitability for mechanical harvesting is taken into consideration. However, a high and stable yield seems to be the most desirable trait from the practice point of view. In the present study, fresh weight of herb ranged from 48.33 to 129.33 g per plant, while dry weight ranged from 9.27 to 31.23 g per plant. Clones no. 5, 7 and 10 gave the highest fresh and dry weight of harvested herb. These features differentiated the clones at a relatively high degree (CV = 0.38 for fresh weight; CV = 0.36 for dry weight) ( Table 3). This corresponds to the results obtained by Mewes et al. [33], where dry weight of common thyme herb varied from 20 to 47 g per plant, and a high diversity (CV = 0.44) among investigated accessions was observed. When given common thyme, the mass of leaves is usually estimated, since dry leaves themselves (so-called rubbed herb) are a commercial and traded product. In our work, the weight of leaves ranged from 5.80 (clone no. 12) to 19.83 g per plant (clone no. 10), CV = 0.33. This trait, related to foliar density, is one of the most important descriptors recommended by ECP/GR for common thyme evaluation [39]. Here, clones no. 7, 8, 9 were characterized by sparse, clones no. 1, 4, 5, 10, 12 by dense, while clones no. 2, 3, 6 and 11-medium foliar density (Table 2).  In present work, leaves of investigated clones were subjected to observations concerning colour and shape of the leaf blade, as well as its length and width. Most were characterized by green leaves (6 clones), however there also were dark green (2 clones), pale green (2 clones), greyish-green (1 clone) and grayish-blue (1 clone) (Table 2, Figure 1a,b). The shape was described as oblong-obovate (clones no. 1 and 3), obovate (clones no. 5,8,11), oblong-lanceolate (clones no. 4, 7, 10), lanceolate (clones no. 2 and 9) and elliptic (clones 6 and 12) ( Table 2). The length of the leaf blade varied from 5.29 (clone no. 12) to 7.91 mm (clone no. 1) (CV = 0.14), while its width from 2.38 (clone no. 7) to 4.79 (clone no. 1) (CV = 0.21). On the bases of these results, clones no. 1, 3, 5, 8, 11 can be classified as 'broad-leaved', clones no. 2, 4, 7, 9, 12 as 'narrow-leaved', while clones no. 6 and 10 can be considered as intermediate forms (Table 4, Figure 2a). These results correspond to divisions mentioned by Melchior and Kastner [38]. As it was mentioned before, the examined clones originate from individual plants selected from the cultivar 'Standard Winter'. Thus, differences between clones may represent a range of variability within this cultivar. It seems that morphological traits, typical European ecotypes and forms of common thyme [37,38], appear within the investigated cultivar as a result of phenotypical plasticity of the species, caused probably by its allogamous way of reproduction.
With regard to aromatic plants, including common thyme, the issue concerning the content and composition of essential oil is one of the most important, since this substance determines both the sensory value and pharmacological activity of raw material. In common thyme (as well as in other Lamiaceae), essential oil is stored in multicellular epidermal glands, the glandular trichomes, situated on the surface on both sides of the leaves. The glandular trichomes are composed of the epidermal and stalk cells. The head of the gland is formed by 10-14 secreted cells, where essential oil is produced before being transferred and stored in sub-cuticular space. Glandular trichomes are subdivided into two groups: peltate and capitate [44,45]. There is a direct correlation between glands number and essential oil production. Moreover, the distribution of glandular trichomes have valuable taxonomical significance at the species level [46]. Our results show that the examined common thyme clones differed with respect to number of glandular trichomes per 1 mm 2 of a leaf. The density of these glands on the adaxial surface of the leaf varied from 13.08 to 18.36 (CV = 0.12), while on the abaxial surface from 3.96 to 9.36 (CV = 0.29) per 1 mm 2 (Table 4, Figure 2b). Trichome density was related to essential oil content and its was observed that clones no. 4 and 11, characterized by the highest number of glandular trichomes on leaves, were also distinguished by the highest essential oil content (4.38 and 4.15 g × 100 g −1 DW, respectively) ( Tables 4 and 5). In general, the content of essential oil ranged from 2.10 to 4.38 g × 100 g −1 DW, and corresponds with results obtained by others [10,27,28,33]. The content of essential oils in aromatic plants, including common thyme, depends on various factors, both internal (genetic, ontogenetic, morphogenetic) and external (environmental such as temperature, day length, solar radiation, and postharvest treatment) [30,[47][48][49]. Table 5. The total content (g × 100 g −1 ) and gas chromatographic composition (% peak area) of essential oil samples. In the present work, 26 compounds were detected in essential oils, comprising 97.84-99.95% of the total identified fraction. Monoterpenes were present in the highest amounts. This class of compounds was subdivided into three groups: phenolic monoterpenes with the domination of thymol (from 36.74 to 54.59%), monoterpene hydrocarbons (up to 53.08%) where γ-terpinen and p-cymen were present in the highest amounts, and oxygenated monoterpenes (up to 3.83%) (Table 5, Figure 3). The fraction of sesquiterpenes was represented by sesquiterpene hydrocarbons (e.g., β-caryophyllene, α-cadinene, β-copaene), and trace amounts of oxygenated sesquiterpenes. Aliphatic alcohol 1-octen-3-ol was classified as the 'other' compound ( Table 5). The predominance of thymol led to qualifying the examined essential oils as 'thymol' chemotype. Here, clones no. 4 and 7 were characterized by the highest amount of this compound (54.59, 54.02%, respectively). Due to its high pharmacological activity, mainly antimicrobial and antioxidant, thymol is considered as the most important constituent of thyme herb [21,22]. It is also responsible for the sensory profile of thyme herb, conditioning its pleasant, herbal and spicy aroma [50]. The fluctuation of thymol content in common thyme essential oil was found to depend on various factors, and the chemotype seems to be one of the most important [4,[6][7][8][9][51][52][53][54][55].

Conclusions
The results presented in this study indicate high level variability in common thyme. The examined clones, being representative of single plants within the cultivar 'Standard Winter' appeared to be strongly diverse, both in morphological and chemical traits. Here, even though a quite significant homogeneity was noticed with regard to type of growth habit and height of plants, there was a strong differentiation in the frame of weight of the herb, shape, colour, length and width of leaf blade, as well as density of glandular trichomes, which influenced variability in the essential oil content. The analyzed clones represent 'thymol' chemotype, however a relatively high diversity within content of this constituent in the essential oil was noticed. Among phenolic acids and flavonoids, rosmarinic acid, followed by luteolin 7-O-glucoside, were the dominant compounds. However, the highest differences between investigated clones concerned p-coumaric acid (CV = 0.59) content. The obtained results indicate that except from association between number of glandular trichomes and essential oil content, there was no clear relationship between other investigated characters, not morphological nor chemical.
Results of our work suggest that 'Standard Winter' cultivar is not homogenous in respect of traits important from the industrial point of view. It is worth noting that in practice, besides cultivars, many varieties, landraces, and even populations of common thyme are cultivated by farmers. It can be suspected that in these cases, the level of heterogeneity may be much higher. Such a high range of variability may create problems with raw material standardization. From another side, it opens a huge possibilities for breeders, where single plants/clones may become valuable components for breeding. Moreover, clones distinguished by a desirable traits (e.g., a high content of thymol and/or rosmarinic acid), can be vegetatively multiplied on a smaller scale and directly consumed as a fresh culinary herb.  In general, the creation of chemotypes is not limited only to terpenes. Such phenomenon covers non-volatile phenolic compounds, as well. Although these two groups of compounds are derived from distinct metabolic pathways, they serve complementary roles in mediating plant interactions with the environment. Here, both terpenes and phenolics play a crucial role in plant physiology, since they are considered to take part in mechanisms of protection, adaptation, competition, signaling, and defense [30,47,48,56]. On the other hand, terpenes and phenolics reveal a wide range of pharmacological activities, which make them one of the most important secondary metabolites. In the present work, investigated common thyme clones were compared as to the total content and composition of non-volatile phenolic compounds, namely phenolic acids and flavonoids. The total content of phenolic acids ranged from 0.59 to 1.65 g × 100 g −1 DW (CV = 0.38) ( Table 6). Among this group, three phenolic acids were identified: caffeic, rosmarinic and p-coumaric acid. Here, rosmarinic acid appeared to be a clear dominant, and its content ranged from 611.47 to 2675.59 mg × 100 g −1 DW (CV = 0.40). It is worth noting that clones no. 3 and 12 were distinguished by the highest content of this compound. Caffeic and p-coumaric acids were present in smaller quantities, ranging from 9.71-37.53 and 0.16-4.13 mg × 100 g −1 DW, respectively (Table 6, Figure 3). It was observed that p-coumaric acid differentiated investigated clones at the highest degree (CV = 0.59). The listed phenolic acids represent cinnamic acids derivatives, where caffeic and p-coumaric acids are free compounds, while rosmarinic acid (ester of caffeic and 3, 4-dihydroxyphenyl lactic acids) belongs to depsides. The presence of these compounds in common thyme herb were reported earlier by other authors [18,[57][58][59]. Common thyme herb contains other caffeic and rosmarinic acids derivatives, i.e.: salvianolic acid (trimeric form of caffeic acid), rosmarinic acid glucoside, methyl rosmarinate. Rosmarinic acid is common in plants and is a dominant compound in the Lamiaceae species, including the Thymus genus, where it is reported to account for about 70% of total polyphenols [18]. However, its content is quite variable: in common thyme it oscillated between 3.4 and 22 mg × 1 g −1 [18,60]. According to Lukas et al. [61], the content of rosmarinic acid in wild growing populations of common oregano (Origanum vulgare L. ssp. vulgare) varied from 0.6 to 37.2 mg × 1 g −1 . Such a wide range of content suggests that this compound may be used as a marker of intraspecific variability. Moreover, when taking into account its extremely high antioxidant activity, a considerable amount of rosmarinic acid in plant material appears to be a desirable trait from a pharmacological point of view [62]. The investigated common thyme clones differed as to the total content and composition of flavonoids. Total content of these substances ranged from 0.19 to 0.47 g × 100 g −1 DW (CV = 36). Three compounds were identified within the group, namely luteolin 7-O-glucoside, naryngenin, and (−)-epicatechin. Luteolin 7-O-glucoside was present in the highest amount (from 46.77 to 325.11 mg × 100 g −1 DW (CV = 0.50)), followed by naryngenin (from 59.22 to 185.12 mg × 100 g −1 DW (CV = 0.38)). In turn, content of (−)-epicatechin was lower, and ranged from 32.70 to 78.32 mg × 100 g −1 DW (CV = 0.26). Among examined clones, clone no. 3 was characterized by the highest content of luteolin7-O-glucoside and (−)-epicatechin (Table 7, Figure 3). The presence of listed flavonoids in common thyme herb were demonstrated earlier [19,59,63,64]. According to literature data, other flavones and methylated flavones have also been noticed in this raw material, e.g., luteolin and its derivatives (luteolin O-diglucoside, luteolin acetyl-O-glycoside), apigenin, apigenin 7-glucuronide, apigenin 7-O-rutinoside, cirsilinol, eriodictyol, thymonin, etc. [16]. In general, flavones are regarded as a health promoting agents. Here, luteolin and its derivatives play a remarkable role in preventing a wide range of chronic diseases. Due to its antioxidant, anti-inflammatory, antitumor and antimicrobial activities, it prevents and reduces incidences of common diseases like: cancer, heart failure, and arteriosclerosis [65][66][67]. Thus, flavones determine the pharmacological activity of common thyme herb and, due to its physiological and chemical stability, it may be a good chemotaxonomic marker within the species [16].

Conclusions
The results presented in this study indicate high level variability in common thyme. The examined clones, being representative of single plants within the cultivar 'Standard Winter' appeared to be strongly diverse, both in morphological and chemical traits. Here, even though a quite significant homogeneity was noticed with regard to type of growth habit and height of plants, there was a strong differentiation in the frame of weight of the herb, shape, colour, length and width of leaf blade, as well as density of glandular trichomes, which influenced variability in the essential oil content. The analyzed clones represent 'thymol' chemotype, however a relatively high diversity within content of this constituent in the essential oil was noticed. Among phenolic acids and flavonoids, rosmarinic acid, followed by luteolin 7-O-glucoside, were the dominant compounds. However, the highest differences between investigated clones concerned p-coumaric acid (CV = 0.59) content. The obtained results indicate that except from association between number of glandular trichomes and essential oil content, there was no clear relationship between other investigated characters, not morphological nor chemical.
Results of our work suggest that 'Standard Winter' cultivar is not homogenous in respect of traits important from the industrial point of view. It is worth noting that in practice, besides cultivars, many varieties, landraces, and even populations of common thyme are cultivated by farmers. It can be suspected that in these cases, the level of heterogeneity may be much higher. Such a high range of variability may create problems with raw material standardization. From another side, it opens a huge possibilities for breeders, where single plants/clones may become valuable components for breeding. Moreover, clones distinguished by a desirable traits (e.g., a high content of thymol and/or rosmarinic acid), can be vegetatively multiplied on a smaller scale and directly consumed as a fresh culinary herb.