This study has been carried along the mountainous areas of Valencia region (which includes the administrative provinces of Castellón, Valencia and Alicante) and bordering ones of Aragón, Castilla-La Mancha and Murcia regions. The region, located at the East and Southeast of Spain, comprises a thin coastal strip with wetlands and a fertile plain giving way in a few kilometers to a craggy inland territory in the West, with some high peaks (Figure 1).

The ethnopharmacological field study was carried out between 2009 and 2012, with semi-structured interviews with male and female informants randomly selected. The study was carried out in accordance with the recommendations of the Code of Ethics of the International Society of Ethnobiology (2006). Ethics approval was not required as per the author's institutions guidelines and national regulations. Verbal informed consent was obtained from each informant prior the interview. In this process, the objectives of the research, the topics and procedures for the interview and the general outcomes expected were presented to the informant and both confidentiality and complete anonymity were guaranteed. The consent of the informant was audio recorded. Only non-sensitive personal data were collected. Although elders were the most interesting informants, a few younger people were also interviewed in order to determine the intergenerational transmission of herbal knowledge (Martínez-Francés et al., 2012). Literature search for additional ethnopharmacological data has also been performed.

Aerial parts of six populations of Salvia blancoana subsp. mariolensis, two of S. officinalis subsp. lavandulifolia, and one of S. x hegelmaieri, S. officinalis subsp. officinalis, S. x auriculata, and S. microphylla var. microphylla were collected during flowering stage in June-July 2010 in different localities from Valencia region (Spain). In addition, two more samples, one of S. blancoana subsp. mariolensis and another of S. officinalis subsp. lavandulifolia were collected in winter time, with the aim of detecting seasonal changes in their essential oil composition (Table 1, Figure 1). Plant material was air-dried at room temperature in a dark, dry place and later stored until used.

Identification of plant material was carried out using a stereomicroscope with the aid of literature data (Bolòs and Vigo, 1995; Laguna et al., 1998; López, 2001; Reales et al., 2004; Serra, 2007; Sáez, 2010; Mateo and Crespo, 2014). A voucher of each sample was deposited in the Herbarium BCF of the University of Barcelona (Barcelona, Spain) (Table 1).

Anisaldehyde (Ref.: A88107), diphenylboric acid aminoethyl ester (Ref.: 42810), PEG 400 (Ref.: P3265), and rutin (Ref.: 78095) were obtained from Sigma Chemical Company (St. Louis, MO, USA). Eupatorin was from our own laboratory collection (previously isolated from Salvia officinalis subsp. lavandulifolia leaves). Acetic acid, formic acid, sulfuric acid, dichloromethane, ethanol, ethyl acetate, methanol, and toluene were of analytical grade and purchased from Panreac Química (Barcelona, Spain). HPTLC glass plates coated with Silicagel 60F₂₅₄ (20 × 10 cm) were purchased from Merck (Darmstadt, Germany; Ref.: 1.05642.0001).

Hydroalcoholic and dichloromethane extracts were analyzed by HPTLC using a CAMAG (Muttenz, Switzerland) system constituted by a semiautomatic TLC Sampler Linomat 5 as application device, an Automatic Development Chamber ADC 2, a TLC Plate Heater III device, a Chromatogram Immersion Device III, Reprostar 3 (Digistore) as chromatogram documentation device, and the software WinCATS 1.4.6.

Test solutions were the hydroethanolic and dichlorometane extracts. Aliquots of 4 μL of test and reference solutions were applied on the plates as bands of 8 mm in length (distance from the left edge 15 mm and from the lower edge 8 mm). Plates were developed in ascending mode over a distance of 6 cm from the application position using a twin-through chamber fitted with filter paper previously saturated for 20 min with the mobile phase. Analyses were carried out at constant humidity (33%) at room temperature (23 ± 2°C). Two mobile phases were used depending on the test solutions analyzed: ethyl acetate, acetic acid, formic acid, water (100:10:10:25) for hydroethanolic extracts, and toluene, ethyl acetate, formic acid (70:30:1) for dichloromethane extracts.

Derivatization of the plates was carried out by dipping with either the Natural Products Reagent (NPR) and the PEG-400 solution (hydroethanolic extracts), or the anisaldehyde solution (dichloromethane extracts). In the case of hydroethanolic extracts, after elution, the plates were dried in cold air for 5 min, heated at 100°C for 3 min and dipped in the NPR solution (1 g of diphenylboric acid aminoethyl ester in 200 mL of ethyl acetate) while still hot. Once the plates were dried, they were subsequently dipped in the PEG-400 solution (10 g of PEG 400 in 200 mL of dichloromethane) and dried. Evaluation was performed under UV light at 366 nm. In the case of dichloromethane extracts, plates were immersed in anisaldehyde solution, prepared according to the European Pharmacopoeia (EDQM, 2016), and heated at 100°C for 5 min. They were examined immediately in daylight and eventually under UV light at 366 nm.

The ethnopharmacological study focused on traditional medicinal uses in Valencia region and closest mountainous areas by means of 87 interviews (2009–2012), 61 (33 men and 29 women) in the Valencia region, 19 (men) in Murcia and 6 (4 men and 2 women) in Aragón. The informants interviewed were aged between 24 and 83 years, being 61 the media in both sexes in Valencia region. In this area, nine Salvia species are used in folk medicine (Martínez-Francés et al., 2012). After a botanical identification of the different species used, some samples were collected for performing the chemical characterization (Table 1). The main focus of the present work is on S. blancoana subsp. mariolensis and S. officinalis subsp. lavandulifolia. For this reason, own and literature ethnopharmacological data of these two species is summarized in Table 2. Twenty nine medicinal uses were recorded for S. officinalis subsp. lavandulifolia, with a wider distribution area, from which 13 were also shared by S. blancoana subsp. mariolensis.

In Valencia region, S. blancoana subsp. mariolensis is often used to treat diseases of the circulatory, respiratory, digestive and genitourinary systems. Its aerial parts are commonly used as a tea, alone or mixed with other species. Informants believe that these preparations are hypotensive, antitussive or anti-cold, tonic-digestive, emmenagogue, febrifuge, sedative, and detoxifying. The local vernacular names are sàlvia, sàlvia de la Mariola, sàlvia de serra, sàlvia blanca, and savia (Pérez, 2011; Martínez-Francés et al., 2012) and it is considered the best sage for medicinal purposes.

Salvia officinalis subsp. lavandulifolia has similar uses in Valencia region, where it is known as sàlvia, sàlvia de serra, sàvia, sava, sèrvia, and sèlvia. Infrequent uses, are related to eye, skin, or the musculoskeletal or nervous system. In Castellón province, it is used in inhalation fumes for migraines, as ophthalmic antiinflamatory and for rheumatism. In the Aragonese borderlands, this sage is less frequent and less cited, although it is highly prized; it is known as sàlvia, sàlvia d'Aragó, estepera, and as salvia de Penyagolosa (mountain of Valencia region where it is collected). The main traditional uses in Aragon are as digestive and depurative, and it is considered very effective against menstrual problems and to eliminate the bad odor of the feet. In Murcia region, S. officinalis subsp. lavandulifolia is locally known as manisierva, manisielva, marisielva, mariselva, mariserva, marisierva, salvia, and savia. It is used for the same purposes cited above, but also as aperitif (Obón and Rivera, 1991) and as masticatory to clean the teeth. The last use is also found in Albacete (Castilla—La Mancha), where, macerated in alcohol, is additionally employed for rubbing back and chest to relieve fatigue. The decoction has been used as gargle for snoring and sore throats (Verde et al., 2008). In Albacete, bordering Valencia region, this species and the hybrid S. x hegelamieri, are indistinctly employed. For S. officinalis subsp. lavandulifolia, some popular names, marisielva, marisilva, salvia, savia, selva, sielva, and sierva, resemble to those collected in Murcia and Valencia regions.

The alcoholic preparation known as “salvieta” or “licor de salvia” (sage liquor), which is used as digestive, is especially made in S. blancoana subsp. mariolensis distribution area (Martínez-Francés et al., 2012). “Salvieta” is a homemade anise-flavored alcoholic maceration of the aerial parts of S. blancoana subsp. mariolensis, alone or, rarely, mixed with other sages, such as, S. officinalis subsp. officinalis, S. officinalis subsp. lavandulifolia, and S. microphylla var. microphylla. Nowadays this local product has not a great demand outside the zone of preparation, but some small industries of this area are already marketing the raw material for teas or as a kit for a homemade liquor elaboration. However, there is a potential market for this sage liquor, because its consumption in local festivities and annual fairs are consolidating its use among the youngest. To prevent adulterations, a correct identification of the raw material is necessary.

It is a common criterion to consider S. officinalis subsp. officinalis, traditionally cultivated in Spain for centuries, as an introduced species from the Eastern Mediterranean along with other species, such as, S. fruticosa and S. x auriculata Mill., an hybrid of both (Reales et al., 2004). Concerning S. officinalis subsp. lavandulifolia, its independence (as S. lavandulifolia) was recognized by several authors. Nevertheless, Bolòs and Vigo (1995) and, more recently, Reales et al. (2004) subordinated it to S. officinalis, mainly due to the high coincidence of the indumentum of the calyx. This criterion is nowadays internationally accepted (TPL, 2017) and has been followed in the present work. S. microphylla var. microphylla, a foreign species coming from Mexico (Valdés et al., 1995), cultivated in this area for more than 150 years, is also used in Valencian folk medicine together with S. officinalis subsp. officinalis, S. officinalis subsp. lavandulifolia, S. blancoana subsp. mariolensis, and S. x auriculata. All of them have been collected for morphological and chemical characterization.

As it has been exposed above, discrepancies between Spanish botanists on the classification of Salvia taxa greatly complicate the identification of the raw material for the liquor and herbal markets, in which an unambiguous identification should be a priority before commercialization. In order to avoid misunderstanding, the taxonomic treatment proposed by Reales et al. (2004) and Sáez (2010) is followed here. Salvia officinalis related sages present in the Iberian Peninsula are organized in seven taxa: Salvia fruticosa Mill., S. officinalis L. subsp. officinalis, S. officinalis L. subsp. lavandulifolia (Vahl) Gram, S. officinalis L. subsp. oxyodon (Webb & Heldr.) Reales, D.Rivera & Obón, S. blancoana Webb and Heldr. subsp. blancoana, S. blancoana Webb & Heldr. subsp. vellerea (Cuatrec.) W. Lippert, S. blancoana Webb and Heldr. subsp. mariolensis Figuerola, and two of their hybrids: S. x auriculata Mill. (S. officinalis subsp. officinalis x S. fruticosa) and S. x hegelmaieri Porta Rigo (S. officinalis subsp. lavandulifolia x S. blancoana subsp. mariolensis).

Macro and micromorphological characters are useful for the identification of the different Iberian Salvia subgenus Salvia taxa related to S. officinalis. As it is shown in Table 3, leaves of S. blancoana subsp. mariolensis are oblanceolate, while those of S. officinalis subsp. lavandulifolia are linear-lanceolate and those of S. officinalis subsp. officinalis and S. fruticosa are ovate. Salvia blancoana subsp. mariolensis present persistent upper floral bracts in contrast with the rest. The presence of cup-shaped glandular trichomes and long patent eglandular trichomes can separate it from S. officinalis subsp. lavandulifolia. Salvia blancoana subsp. mariolensis and S. blancoana subsp. vellerea, with different distribution areas, exhibit different type of leaves, the second lacking sessile glands. Salvia microphylla var. microphylla (subgenus Calosphace), is not included in Table 3 due to its distinct morphology, especially its delta-shaped leaves and red flowers, which clearly differentiates it from the other sages. The raw herbal material from all these taxa is frequently dried and cut before marketed, making even more difficult its identification. Therefore, other techniques, such as, chemical profiling, are useful.

In order to contribute to the chemical characterization of S. blancoana subsp. mariolensis, the essential oil contents of the air-dried aerial parts of seven samples of this taxon have been determined and compared with those obtained for samples of other sages used in the region (Table 1). The essential oil content (% v/w) of S. blancoana subsp. mariolensis was in general higher than 2% (1.8–4.5%), even for a sample collected in winter (SB7), which yielded 3.8% of oil. Only one sample (SB4) gave a lower yield (1.8%). The essential oil contents of the three samples of S. officinalis subsp. lavandulifolia were always below 2% (1.2–1.3%) and lower than those shown by S. blancoana subsp. mariolensis. Finally, their hybrid, S. x hegelmaieri, gave a content of 2.5%, which is in between of those shown by the parent taxa.

It is the first time that the detailed composition of the essential oil of S. blancoana subsp. mariolensis is described. Only some preliminary data were previously published by us (Martínez-Francés et al., 2012). In the present work, the essential oils distilled from the air-dried aerial parts of seven samples of this taxon (SB1–SB7), collected in different locations and including wild and cultivated plant material, were submitted to exhaustive analysis by GC-FID and GC-MS. The results, which are shown in Table 4 (main constituents) and Supplementary Table 1 (complete list of constituents), were compared with those given by the oils of three samples of S. officinalis subsp. lavandulifolia (SL1–SL3). In addition, the composition of the essential oil of the aerial parts of S. x hegelmaieri, the hybrid of the two former taxa growing in the same geographical area, is reported for the first time (Table 4, Supplementary Table 1).

More than 98% of each essential oil was identified, which, considering all the samples analyzed, involved a total of 174 constituents. In all cases, the main group of constituents were oxygenated monoterpenes (46.5–71.2%), followed by monoterpene hydrocarbons (21.3–41.1%). The essential oil of S. blancoana subsp. mariolensis is characterized by 1,8-cineole (13.7–45.7%) and camphor (12.1–28.6%) as major constituents, followed by myrcene (3.0–5.6%), α-pinene (3.0–7.4%), camphene (4.2–8.5%), β-pinene (5.0–7.6%), limonene (1.7–5.0%), and borneol (0.8–7.4%). Among sesquiterpenes (2.6–8.5%), β-caryophyllene (0.2–3.5%), β-caryophyllene oxide (0.3–1.2%), and spathulenol (0.2–1.3%) reached the higher percentages. The essential oil of S. x hegelmaieri showed a similar pattern, with 1,8-cineole (40.0%), camphor (15.9%), α-pinene (6.0%), camphene (5.8%), β-pinene (5.8%), myrcene (4.3%), borneol (3.4%), and limonene (3.3%) as the main constituents.

As it can be seen in Table 4, the essential oils of the three taxa analyzed showed a similar profile, mainly concerning their major constituents. They are, in general, in accordance with those previously reported for the essential oils from S. officinalis subsp. lavandulifolia (Lawrence et al., 1970; Marcos Sanz et al., 1988; Martínez-Francés et al., 2012; Usano-Alemany, 2012), with 1,8-cineole and camphor as major constituents. However, some minor distinguishing features could be recognized. In fact, p-cymene was found in the three samples of S. officinalis subsp. lavandulifolia in a range from 2.0 to 3.3%, while it did not reach 1% either in S. blancoana subsp. mariolensis or in S. x hegelmaieri. Furthermore, the phenylpropanoid estragole was detected in traces in only one sample from S. blancoana subsp. mariolensis, whereas it was found in S. x hegelmaieri (0.8%) and two samples of S. officinalis subsp. lavandulifolia (0.8 and 2.2%).

In order to investigate seasonal variations in the essential oil composition of S. blancoana subsp. mariolensis and S. officinalis subsp. lavandulifolia, two samples were collected at winter time (SB7 and SL3, respectively). In general, the chromatographic profile of the latter was inside the range found for spring-early summer samples of S. officinalis subsp. lavandulifolia. Nevertheless, in the case of S. blancoana subsp. mariolensis, the essential oil of the winter sample (SB7) showed a significantly higher percentage of 1,8-cineole and a lower percentage of camphor, the two major constituents, when compared to the essential oils from samples collected in late spring-early summer (SB1–SB6).

In addition to the similarity of the composition of the essential oils of the three taxa investigated, it is noteworthy to remark that they show chromatographic profiles that are mostly in accordance with those prescribed in the current standards for “Spanish sage oil” in the European Pharmacopoeia (EDQM, 2016) (monograph 1849), for “oil of sage, Spanish” by the International Organization for Standardization (standard ISO 3526) (ISO, 2005) and for “aceite esencial de salvia de España” by the Spanish Association for Standardization (AENOR, standard UNE 84310) (AENOR, 2001; Table 5). Minor differences can be observed, that could be due to the fact that the essential oils studied here were hydrodistilled in the laboratory from small amounts of dried plant material, whereas standards are established for essential oils obtained by steam distillation at industrial scale. The species accepted for the preparation of the essential oil in these standards is S. lavandulifolia (referred here as S. officinalis subsp. lavandulifolia). Nevertheless, according to our results, also S. blancoana subsp. mariolensis should be considered as a possible source of this oil. In fact, this endemic Levantine sage appears in a small mountainous area between northern Alicante and southern Valencia provinces, immersed within the wide distribution area of S. officinalis subsp. lavandulifolia. Therefore, when wildgrowing, these two sages could be collected without any distinction.

The essential oil of S. officinalis subsp. lavandulifolia, after a single oral dose of 50 μL, has shown to improve cognitive function in healthy humans compared to placebo, specially the performance of secondary memory and attention tasks (Kennedy et al., 2011). This effect has been mainly related to the inhibition of acetylcholinesterase by the oil, 1,8-cineol followed by α-pinene being the most active constituents (Perry et al., 2003; Kennedy et al., 2011). Taking in account the similarity in composition with the essential oil of S. blancoana subsp. mariolensis, which is particularly rich in 1,8-cineol, the effect on cognitive function could also be expected from the oil of this species.

Finally, it is important to remark that α-thujone and β-thujone are absent or only found in traces in the essential oils of S. blancoana subsp. mariolensis, S. officinalis subsp. lavandulifolia, and S. x hegelmaieri. Due to its activity on the central nervous system, thujone is considered toxic, especially the α isomer which is the most potent, and the control of its content in foods and medicines is necessary (Emmert et al., 2004). The low content/absence of thujone in the essential oils analyzed here is important in order to avoid any safety concern for this reason in relation to the herbal products and, specially, the liquors prepared thereof, since they will follow the recommendations of the European Medicines Agency for herbal medicinal products (EMA, 2011), and the regulations for alcoholic and non-alcoholic beverages established in the European Directive 1334/2008/EC (European Parliament and Council, 2008).

With the aim of setting a simple test allowing an easy distinction of plant material coming from different species, hydroalcoholic and dichloromethane extracts of the dried aerial parts of the taxa were analyzed by HPTLC according to the methods described in the experimental section. Two HPTLC systems with optimized mobile phases were developed for the analysis of each type of extract.

Figure 2 shows the HPTLC chromatograms of the hydroalcoholic extracts examined under UV light at 366 nm after derivatization with NPR and PEG-400 solution. With this system, the chromatographic profiles of S. blancoana subsp. mariolensis, S. officinalis subsp. lavandulifolia, and S. x hegelmaieri were very similar to each other with only slight differences concerning the intensities of certain zones. The only remarkable difference is a fluorescent blue zone at Rf 0.4 present in the case of S. officinalis subsp. lavandulifolia, which is not observed in S. blancoana subsp. mariolensis nor in S. x hegelmaieri. High similarity is also observed between S. officinalis subsp. officinalis and S. x auriculata, which, nevertheless, could be distinguished from the former three thanks to the presence of one additional orange fluorescent zone in the lower third of the chromatogram and the much higher intensity of the lower of the two orange fluorescent zones present in the middle third. Finally, the hydroalcoholic extract of S. microphylla var. microphylla showed a clearly different chromatographic profile from all other taxa analyzed, mainly due to the presence, in the middle third, of one intense blue fluorescent zone (Rf 0.45) and a yellow fluorescent zone (Rf 0.49), which are absent in the other taxa. Thus, the fingerprints obtained with the hydroalcoholic extracts are not useful enough for the distinction of S. blancoana subsp. mariolensis from S. officinalis subsp. lavandulifolia. Although these two taxa have partially similar uses, they do not have overlapping distribution and the former constitutes a local resource for inland highland areas of Alicante province. For a potential commercialization and valorization of the traditional beverage “salvieta,” used only in this area, it is worth to set up a suitable HPTLC system for differentiating these two species.

The HPTLC analysis of dichloromethane extracts derivatized with anisaldehyde reagent and examined under daylight or UV 366 nm, provided characteristic fingerprints for most of the taxa analyzed, as it is shown in Figure 3. All samples of S. blancoana subsp. mariolensis (SB1-SB5) showed a homogeneous fingerprint, which was similar to the one obtained with S. x hegelmaieri, a hybrid of the former, botanically very close. These two taxa were clearly differentiated from the others by two zones observed at daylight, appearing in the lower part of the chromatogram (Figure 3A), at Rf 0.12 (red-brown) and Rf 0.19 (gray-blue). Furthermore, examination of the plate under UV light at 366 nm (Figure 3B) showed three blue fluorescent bands at Rf 0.12, Rf 0.14 and Rf 0.50 which were absent in the other taxa. Therefore, all these zones might be considered as specific markers for S. blancoana subsp. mariolensis and S. x hegelmaieri. In the case of S. officinalis subsp. lavandulifolia, a blue-green zone in the middle third (Rf 0.45) of the chromatogram observed under white light (Figure 3A) appears to be characteristic of this taxon. This zone is also seen under UV light at 366 nm as a greenish gray zone (Figure 3B). The fingerprints of the cultivated species S. officinalis subsp. officinalis, S. x auriculata, and S. microphylla var. microphylla showed visible differences after derivatization of the plates with anisaldehyde and observation under white light (Figure 3A).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer RM and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review.