Phytochemicals and Other Characteristics of Croatian Monovarietal Extra Virgin Olive Oils from Oblica, Lastovka and Levantinka Varieties

Virgin olive oils from the fruits of Croatian autochthonous varieties Oblica, Lastovka and Levantinka were characterized for the first time. Headspace volatiles were analyzed by HS-SPME/GC-FID/MS. The main volatiles were C6 compounds. The most abundant was (E)-hex-2-enal (62.60%–69.20%). (Z)-Hex-3-enal was not found in Lastovka oil, while Levantinka oil did not contain hexanal. Tocopherols, chlorophylls and carotenoids were determined by HPLC-FL. Levantinka oil was characterized by the highest α-tocopherol level (222.00 mg/kg). Total phenolic contents (TPs), as well as antioxidant activity (DPPH assay) of the oils hydrophilic fractions (HFs) were assessed by spectroscopic methods. The antioxidant activity of Oblica oil HF was the most pronounced (0.91 mmol TEAC/kg) and the HF contained the highest TPs amount (212.21 mg/kg). HFs phenolic composition was determined by HPLC-DAD. The main identified phenols were secoiridoids dominated in Oblica oil: decarboxymethyl ligstroside aglycone (p-HPEA-EDA up to 158.5 mg/kg), oleuropein aglycone (3,4-HPEA-EA up to 96.4 mg/kg) and decarboxymethyl oleuropein aglycon (3,4-DHPEA-EDA up to 93.5 mg/kg).


Introduction
Virgin olive oil (VOO) has been a valuable vegetable oil extracted from fresh and healthy olive fruits (Olea europeae L.) by mechanical and other physical methods (washing, decantation, centrifugation or filtration). Nowadays, it is well-known that regular dietary consumption of VOO manifests in health benefits associated with Mediterranean diet [1]. The nutritional value of VOO arises from high level of oleic acid and from minor components such as phytosterols, carotenoids, tocopherols and hydrophilic phenols. The major phenolic compounds are oleuropein derivatives, based on hydroxytyrosol which are strong antioxidants and radical scavengers [2]. The content of phenolic compounds is an important factor to be considered when evaluating the quality of VOO, since these compounds exhibit potent antioxidant activity and contribute significantly to the extraordinary oxidation stability of VOO [3]. Chain-breaking antioxidants, such as phenolic compounds, react with lipid radicals to form nonreactive radicals, interrupting the propagation chain. In fact, these compounds are able to donate an electron or a hydrogen atom to the lipid radical formed during the propagation phase of lipid oxidation [2]. Phenolic compounds are intimately associated with the positive attributes of bitterness and pungency which are typical sensory notes of VOO obtained from the olives that are green and turning color [4]. The polar phenolic compounds of VOO belong to different classes: phenolic acids, phenylethyl alcohols, hydroxy-isochromans, flavonoids, lignans and secoiridoids [2]. The amount of polar phenols and volatile compounds in the olive oil depend on many factors such as cultivar, agronomic practices, ripeness index, fruit pre-storage, extraction, procedure, storage conditions, etc. [4]. Color is one of the major attributes that affects consumer perception of VOO quality. The lipophilic nature of chloroplast pigments (chlorophyll and carotenoids) determines their affinity for the oily phase and the pigments are mainly responsible for the color of VOO ranging from yellow-green to greenish gold [5]. The presence of chlorophylls and carotenoids in olive oil depends on the fruits genetic factors (olive variety), the stage of fruits ripeness, environmental conditions, production year, extraction process and storage conditions. Carotenoids, together with polyphenols and tocopherols provide oxidative stability to the olive oils and exhibit synergistic antioxidant and anticarcinogenic action at physiological concentration [6]. In addition, extra VOO is appreciated worldwide for its taste and flavor that is colored by various volatile compounds: aldehydes, alcohols, esters, hydrocarbons, ketones, furans and others [7]. It has been stated [8] that C6 compounds, the major components of VOO headspace, mainly contribute to green odor notes. These compounds are produced by lipoxygenase-mediated oxidation of polyunsaturated fatty acids containing cis-cis-penta-1,4-diene structure during the crushing and malaxation steps of the oil production [9]. Variable amounts of hexanal, hexanol and hexyl acetate derive from degradation of linoleic acid, while (Z)-hex-3-enal, (E)-hex-2-enal, (E)-hex-2-enol, (Z)-hex-3-enol and (Z)-hex-3-enyl acetate result from the enzymatic degradation of linolenic acid [10].
VOOs exhibit positive effect on human health as well as specific and desirable sensory properties, which is why the demand for these oils is constantly growing with request to mark their geographical and varietal origin [4]. There are about 30 autochthonous cultivars of olives in Croatia. Oblica, Lastovka and Levantinka are the main autochthonous varieties in Dalmatia region (south Croatia) and Oblica is the most abundant [11]. The aim of present research is to perform detail chemical characterization of Oblica, Lastovka and Levantinka VOOs (first report including statistical analysis) by: (1) determination of basic characteristics (acidity, peroxide value, K232 and K270) according to EU regulations; (2) total phenols, chlorophylls and carotenoids evaluation by UV/VIS including targeted tocopherols analysis by HPLC-FL; (3) the headspace volatiles profiling (HS-SPME/GC-FID/MS); (4) HPLC-DAD targeted phenolics analysis of the hydrophilic oil fractions (HFs) and assessing the HFs antioxidant activity (DPPH assay).

Results and Discussion
To avoid the influence of other factors, olive trees were cultivated in the same orchard under identical agronomic (i.e., fertilization or irrigation) and pedoclimatic conditions and the olive fruits were picked at the same stage of ripeness and the oils were extracted with the same processing system (Section 3.1.). Therefore, it was possible to attribute the observed results exclusively to different cultivars.

Basic Characteristics of the Samples
The acidity, peroxide value (PV), K232 and K270 of investigated VOOs were assessed by EU method (Table 1).  [12]). From Table 1 it can be seen that Oblica variety exhibited the lowest acidity (0.12%), while the highest value was found in Lastovka oil (0.17%). Statistically significant differences were obtained in free acidity and peroxide value among the oils from the studied cultivars. Although significant differences were determined by Tukey's test, the absolute values of measured acidity of all oils are very similar. Peroxide value should amount less than 20 mEq O2/kg for extra VOO. Determined PVs ranged from 2.96 to 4.00 mEq O2/kg. The lowest PV showed Oblica oil (2.96 mEq O2/kg) and the highest Lastovka oil (5.32 mEq O2/kg). Determined PVs are significantly different in all tested oils, but in absolute values they are very close. K232 value demonstrates conjugated dienes and their oxidation products (absorbtion at λ = 232 nm) and K270 value indicates conjugated trienes and secondary oxidation products (carbonyl compounds; absorbtion at λ = 270 nm). K232 and K270 values ranged from 1.70 to 2.11 and from 0.12 to 0.19 (Table 1). Oblica oil showed the lowest K232 and K270 values while Lastovka oil showed the highest values. The investigated parameters (Table 1) were within the limits of EC Reg. 1989Reg. /2003Reg. (2003 [13] indicating the category of extra VOOs.

Headspace Composition
Olive oil, compared to other vegetable oils, is distinguished by a characteristic aroma. The sensory characteristics, together with high nutritional value are the main features that have resulted in the increase of VOO consumption in recent years [14]. Aroma is an important criterion for VOOs. Consequently, the identification of the compounds contributing to the aroma is considered as a key for quality and authentication control [15]. The headspace profile of tested monovarietal oils was dominated by C6 volatile organic compounds, mostly aldehydes ( Table 2). The C6 compounds responsible for green and fruity VOO perception are produced through lipoxygenase pathway during the olive fruit crushing and malaxation and incorporated into resulting oil. The most abundant was (E)-hex-2-enal (up to 63.8%, 69.8% and 63.6% respectively). (E)-Hex-2-enal was the most important positive contributor of lawn perception [9]. Considering other C6 compounds it can be seen that (Z)-hex-3-enal was not found in Lastovka oil, while Levantinka oil did not contain hexanal. Different values of identified C6 aldehydes in the samples could be due to different acyl hydrolase activity and consequently good or poor availability of free polyunsaturated fatty acids [16]. The results are in accordance with previous research [17] confirming that the cleavage by heterolytic hydroxydeperoxide lyase was the most important process (higher abundance of C6 compounds in comparison with C5 metabolites). The representatives of C5 compounds (Table 2) were pent-1-en-3-one, pentan-3-one and (E)-pent-2-enal. The percentage of pent-1-en-3-one is positively correlated with bitter taste of VOO [18]. In Lastovka oil the level of this compound was the highest (3.57%), while Oblica oil exhibited the lowest value (1.67%). In contrast to Oblica and Lastovka oils, only Levantinka oil contained (E)-pent-2-enal which can be useful potential biodiversity marker. Several hydrocarbons (such as α-copaene, trans-β-ocimene and two isomers of 3-ethyloct-1,5-diene) were found ( Table 2). trans-β-Ocimene was not identified only in Levantinka oil, whereas, α-copaene was identified in Oblica and Levantinka oils. 3-Ethyloct-1,5-diene isomers were detected in all the oils. Statistically significant differences were found almost among all identified compounds of Oblica, Lastovka and Levantika VOOs.
According to [19] the most common parameters that influence the composition of VOO volatiles are: variety, growing area, the degree of maturity of fruits, harvesting method, storage conditions, storage of olive fruits after harvest, processing of fruits into oil, others. The olive fruits of different varieties grown in the same area give VOOs of different volatiles composition, but similar results were found for VOOs of the same varieties grown in different geographic areas [19] Brkić Bubola et al. [20] investigated volatiles and sensory characteristics of the oils from three Istrian (Croatia) olive varieties (Buza, Črna and Rosinjola) and found that the volatiles generated depending on the varieties which indicated a close relation with the activity of a genetically determined enzyme.

Targeted Tocopherols Analysis and Total Chlorophylls and Carotenoids
Tocopherols are considered as the most important lipid soluble natural antioxidants and they increase oxidation stability of the oils during storage [14]. The amount of their main component, α-tocopherol, varies by up to 300 mg/kg [21]. The concentration of β-, γ-and δ-tocopherols range from traces up to 25 mg/kg. A synergistic relationship between the antioxidant actions of some phenolics and tocopherols was demonstrated [14]. It is well accepted that tocopherols content seem to be reduced during the fruit's ripening, refining and hydrogenation process [22].
The levels of tocopherols, chlorophylls and carotenoids determined in the oils from all varieties are shown in Table 3. The results showed the dominance of α-tocopherol in all studied oils, followed by γ-tocopherol as expected for a typical VOO. Levantinka oil was characterized by the highest level of α-tocopherol (222.0 mg/kg), while significantly the lowest value was found in Lastovka oil (177.82 mg/kg). Considering γ-tocopherol, significantly the lowest value was measured in Lastovka oil, whereas Oblica and Levantinka oils contained higher values (33.19 and 31.84 mg/kg, respectively). The obtained results are in accordance with our previous paper [23] on Croatian varieties Krvavica and Mašnjača as well as by Ranalli et al. [24] on Italian varieties Leccino, Frantoio and Moraiolo and Douzane et al. [25] on several Algerian varieties. The results agree with Perrin [26] findings that tocopherols concentration is generally greater than 100 ppm in good quality oils with α-tocopherol representing about 95% of the total fraction. The color intensity of the oil is determined by changes in the source fruits' pigment content during ripening. Although the pigment concentration in the fruit differs greatly with the variety, it decreases with ripening when chlorophylls disappear faster than carotenoids [27]. Chlorophyll and carotenoid pigments greatly influences the color of VOOs ranging from green-yellow to golden, depending on the variety and the stage of maturity [28]. As shown in Table 3, significant differences among cultivars (p < 0.05) were observed in α-tocopherol and γ-tocopherol among Lastovka VOO in comparison with Oblica and Levantika VOOs. The results show significantly higher total pigment content (chlorophylls + carotenoids) in Lastovka oils, with mean value of 6.81 mg/kg. Oblica and Levantinka oils contained total pigments in the amount of 5.96 and 5.86 mg/kg respectively. Chlorophylls and carotenoids ranged, respectively from 3.86 to 4.75 mg/kg and from 1.89 to 2.06 mg/kg. These findings are in agreement with previous results [29]. Moreover, Giufrida et al. [30] reported that the presence of the pigment in the oil depends on several factors, such as the cultivar, soil and climatic conditions, fruit ripeness and the processing procedures. Data on the pigments composition of those Croatian olive oils could be used, along with other parameters, to guarantee the genuineness and authenticity of the products. These compounds also exhibit biological and health properties and occur in the oils at concentrations which usually correlate with those of phenols and volatiles [31].

Total Phenolic Amounts and Targeted Phenolic Analysis
The phenolic composition of olive oil is very complex and the average concentration of these compounds depends on several factors including maturation stage, part of the fruit, variety, season, packaging, storage, climatologic conditions and the production technology [14]. VOO is well known for its high content of phenolics along with oleic acid and tocopherols that exhibit health-promoting properties [32]. Phenolic compounds influence the sensory properties (flavor, bitterness, etc.) of olives and the oil, and they protect against oxidative rancidity by acting as antioxidants, which are increasingly being recognized by playing a beneficial role in the diet [33].
Significant differences in total phenols (TPs) content among several HFs of the samples were observed (Table 4). Oblica oil exhibited the highest TPs amount (212.21 mg/kg), while Levantinka oil showed the lowest TPs content (144.60 mg/kg). Lastovka oil was quite similar to Oblica oil considering TPs amount (206.09 mg/kg). It is generally accepted that the TP's level varies in the oils obtained from different cultivars and areas. Discrimination among the olive oil samples with the same geographical origin and different cultivars was possible by comparing TPs [32]. Another authors [34] claim that phenolic composition was found to be not useful in discriminating the olive oil samples due to the fact that TPs content of the oils was affected not only by the olive cultivars, but also by the climatic and environmental conditions, agronomic practice and the technological process. TPs content ranged from 50 to 1.000 mg/kg, but the values usually amount from 100 to 300 mg/kg [35] that is similar to the values found in the investigated samples. TPs content of the samples in this study could be considered as medium-high levels in accordance with previous reports [23,32,34,36]. Olive oil hydrophilic extracts contain a large number of phenolic compounds including simple phenols, lignans, and secoiridoids, which exhibit antioxidant properties [37]. Among the phenolic compounds found in extra-virgin olive oils, o-dihydroxyphenolics are very potent antioxidants [38]. Free radicals are generated in the human body through aerobic respiration and exist in different forms, including superoxide, hydroxyl, hydroperoxyl, peroxyl and alkoxyl radicals. Natural antioxidant enzymes in healthy individuals remove these free radicals while dietary antioxidants assist the body in neutralizing free radicals. Therefore, it is important to consume foods with high contents of antioxidants, such as virgin olive oil, to reduce the harmful effects of oxidative stress [39]. Antioxidant activity measured by DPPH assay showed the highest value in hydrophilic fraction of Oblica oil (0.91 mmol TEAC/kg) and statistically significant lowest value in Levantinka oil (0.55 mmol TEAC/kg). The difference in antioxidant activity of tested oils may depend on the total phenol content in the varieties. Most of the previous studies reported strong correlation between total phenols and antioxidant capacity [23,37,39,40].
HPLC-DAD was used for targeted analysis of the polar compounds in Oblica, Lastovka and Levantinka VOOs. The most abundant secoiridoids of the samples (Table 5) were dialdehydic form of elenolic acid linked to hydroxytyrosol or tyrosol (p-HPEA) respectively assigned as 3,4-DHPEA-EDA, p-HPEA-EDA and 3,4-HPEA-EA. Statistically significant differences among phenolic levels were observed for the oils. Levantinka and Oblica VOOs were characterized statistically by highest amount of p-HPEA-EDA. Lastovka VOOs contained the highest amount of 3,4-HPEA-EA. 3,4-DHPEA-EDA, the dialdehydic form of decarboxymethylelenolic acid linked to hydroxytyrosol, was statistically different in all tested oils with the highest value in Oblica oil (85.0 mg/kg) and lowest value in Lastovka oil (34.4 mg/kg). Flavones, such as apigenin and luteolin, showed in total the highest value in Oblica oil, while the lowest value was measured in Levantinka oil. Lignans, such as pinoresinol, were estimated only in Oblica oil. These results are similar to those reported by several authors for other monovarietal VOOs [31,40,41].

The Samples and Preparing of the Hydrophilic Fractions
All the olive oils were produced from the fruits from the Dalmatia region (Zadar hinterland, Croatia) in 2014. The olive trees were cultivated under same agronomic and agrotechnical conditions. 12-year-old olive trees were planted in squares (7 × 5 m spacing) in the same orchard (ca. 2 ha). There was no irrigation. Five batches of the healthy olive fruits (replicates; each with 200 kg of the fruits) were handpicked from each olive variety at the same maturity index (MI = 4.3). MI was calculated as a subjective evaluation of the skin color and flesh as proposed by Uceda and Frias [42]. The fruits of each variety were separately processed in the extraction plant Molinova TG (Gruppo Pieralisi, Pieralisi S.p.A. Jesi, Italy) within 24 h of collection. Before the extraction of each batch, the plant was cleaned. The fruits were crushed with a hammer crusher and olive paste was malaxed for 35 min at 26 ± 1 °C in the mixer. The olive oil was separated by centrifugation through two phase decanter (without addition of warm water). All oils were filtered through 25 mm GD/X 0.45 μm cellulose acetate filters (Whatman, Milan, Italy) and thereafter stored in dark glass bottles at 4 °C until the analyses.
Hydrophilic fractions (HFs) of the oils were prepared by adding 5 mL of CH3OH-H2O (80:20 v/v) mixture in 3 g of the oil placed in 20 mL screw cap test-tube. The mixture was blended in an ultrasonic bath for 15 min at 30 °C (the emulsion was allowed to separate). The hydrophilic layer was placed in a round flask. The oil extraction was repeated two times; the hydrophilic extracts were combined and evaporated on a rotary vacuum evaporator at 30 °C. The residue was dissolved up to 5 mL with the 80:20 CH3OH-H2O solution and filtered through a Whatman 13 mm GD/X 0.2 μm cellulose acetate syringe filter (Whatman, Milan, Italy).

Determination of the Oils Basic Characteristics
Free acidity (% of oleic acid (%18:1)), peroxide value (mEq O2/kg of the oil) and UV absorption characteristics (K232 and K270) were determined according to the European Union Commission Regulations EC 1989 as in our previous paper [23]. K232 and K270 were determined with an UV spectrophotometer (Specord 200, Analytik Jena AG, Jena, Germany) at 232 and 270 nm using 1% solution of the oil in cyclohexane and a path length of 1 cm. All parameters were determined in triplicate for each sample.

Headspace Solid-Phase Microextraction (HS-SPME)
HS-SPME was performed on SPME fiber (divinylbenzene/carboxen/polydimethylsiloxane; DVB/CAR/PDMS) purchased from Supelco Co (Bellefonte, PA, USA). The fiber was conditioned according to Supelco Co instructions before the extraction. VOO (5 g) was placed in 15 mL glass vial and sealed with PTFE/silicone septa. The vial was placed in a water bath at 40 °C for equilibration (15 min) and extraction (40 min) under constant stirring with a magnetic stirrer (1000 rpm). After sampling, the fiber was inserted into the injector (250 °C) of GC-FID/MS for 6 min for thermal desorption of the volatiles into the GC column.

Gas Chromatography and Mass Spectrometry (GC-FID/MS)
An Agilent Technologies (Palo Alto, CA, USA) gas chromatograph (7890A) with flame ionization detector, mass selective detector (5975C) and HP-5MS capillary column (5%-phenyl)-methylpolysiloxane Agilent J & W GC column; 30 m, 0.25 mm i.d., coating 0.25 μm) was utilized. Helium was applied as carrier (1.5 mL/min). The injector operated in split mode (2:1 split ratio) at 260 °C. The column was heated at 40 °C for 3 min, thereafter to 100 °C (5 °C/min) and later to 260 °C (3 °C/min) and then held to 260 °C for 3 min. The MS conditions were: source temperature 230 °C; quadrupole at 150 °C; transfer line at 270 °C; EI 70 eV and m/z 29-350. The peaks were identified by comparison of the retention indices with authentic samples (relative to C9-C25 n-alkanes) and literature as well as by comparing their mass spectra with Wiley 9 MS library (Wiley, New York, NY, USA) and NIST08 (Gaithersburg, MD, USA) database. The percentage composition was computed from the peak areas using the normalization method (without correction factors). The component percentages ( Table 2) were calculated as mean values from duplicate GC-FID analyses of each sample.

Liquid Chromatography with Diode Array Detector (HPLC-DAD)
HF phenolic compounds were detected and quantified with the HPLC-DAD method described by Tuberoso et al. [43]. A ProStar HPLC system (Varian Inc., Walnut Creek, CA, USA) was employed equipped with a pump module 230, an autosampler module 410, a ThermoSeparation diode array detector SpectroSystem UV 6000lp (Thermo Separation, San Jose, CA, USA) and a Gemini C18 column (150 × 4.60 mm, 3 μm, Phenomenex, Casalecchio di Reno, BO, Italy). The mobile phase was 0.2 M H3PO4 (solvent A) and CH3CN (solvent B) at a constant flow rate (1.0 mL/min), mixed (linear gradients) as follows: at 0 min A:B ratio 85:15 (v/v), reaching 60:40 (v/v) in 30 min, then 40:60 (v/v) in 10 min and finally at 100% B until 50 min. Prior to each injection the system was 10 min stabilized with A:B ratio 85:15 (v/v). The injection volume was 10 μL. The phenols analysis was carried out at 280 nm (hydroxytyrosol, tyrosol, vanillic acid, pinoresinol, oleuropein, and ligstroside derivatives) and 360 nm (luteolin and apigenin). Oleuropein and ligstroside derivatives were tentatively identified according to the literature data [44,45]. The obtained chromatograms and spectra were elaborated with a ChromQuest V. 2.51 data system (ThermoQuest, Rodano, Milan, Italy). Stock standard solutions were prepared in CH3OH and working solutions in CH3OH-H2O (80:20, v/v). The method was validated according to the International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) guidance note [46] as previously reported [23]. All compounds were dosed using the calibration curve constructed with corresponding standard, except oleuropein and ligstroside derivatives that were dosed using the oleuropein calibration curve. The correlation values were comprised between 0.9992 and 0.9999.

Determination of Total Chlorophylls and Carotenoids
The oil solutions of 5% (w/w) in CH3COCH3 were prepared and absorbances were measured (at 464 nm for carotenoids and at 669 nm for chlorophylls) on a 10 mm quartz cuvette utilizing a Varian Cary 50 UV-Visible spectrophotometer (Varian, Leini, TO, Italy). Chlorophyll a and β-carotene stock standard solutions were prepared in CH3COCH3, as well as working solutions that were prepared with proper dilutions (0.1-2.0 mg/kg, r = 0.9997 and 0.02-0.50 mg/kg, r = 0.9994 for chlorophyll a and β-carotene, respectively).

Folin-Ciocalteu Assay
HF total phenolic content was determined spectrophotometrically with a modified Folin-Ciocalteu method [47]. Shortly, HF (100 μL) was added to Folin-Ciocalteu phenol reagent (500 μL). After 5 min, 3 mL of 10% Na2CO3 (w/v) was added, the mixture was shaken, and thereafter diluted with H2O to a final volume of 10 mL. After a 90 min incubation period at room temperature, the absorbance was read at 725 nm (against a blank) on a 10 mm optical polystyrene cuvette (Kartell 01937, Kartell Spa Noviglio, Mi, Italy) utilizing a Varian Cary 50 spectrophotometer (Varian, Leini, TO, Italy). The total polyphenol content expressed as mg/kg of gallic acid equivalent (GAE) were obtained using a calibration curve of a freshly prepared gallic acid standard solution (5-100 mg/kg, r = 0.9999).

DPPH Assay
HF antiradical activity was assessed with the DPPH spectrophotometric method and the obtained data were expressed as Trolox equivalent antioxidant capacity (TEAC) [47]. HF (50 μL) was dissolved in 2 mL of 0.04 mmol/L DPPH in CH3OH. Spectrophotometric readings were carried out at 517 nm with a Varian Cary 50 spectrophotometer using 10 mm optical polystyrene cuvette after an incubation period of 60 min in dark at room temperature. A Trolox calibration curve (0.02-1.00 mM) was prepared (r = 0.9997) and data were expressed in TEAC (mmol/kg).

Statistical Analysis
Statistical analysis by SPSS 17.0 statistical software (SPSS Inc., Chicago, IL, USA) was applied to datasets to perform descriptive multivariate statistical study. The mean values were compared using the Tukey's honestly significant difference test (p < 0.05).

Conclusions
In continuation of our research on monovarietal extra VOOs from Croatian autochthonous olive varieties, it was established that Oblica, Lastovka and Levantika VOOs exhibited all characteristics within the limits for the extra VOO category. Among the headspace volatiles, the most abundant was (E)-hex-2-enal. (Z)-hex-3-enal was not found in Lastovka oil and was different abundant in Levantika and Oblica oils and therefore useful for their differentiation. Levantinka oil did not contain hexanal. The results showed α-tocopherol dominance in all samples followed by γ-tocopherol. The lowest value of γ-tocopherol was measured in Lastovka oil. Higher total pigment content (clorophylls + carotenoids) was found in Lastovka oil. The most abundant secoiridoids were the dialdehydic forms of elenolic acid linked to hydroxytyrosol or tyrosol, particularly 3,4-DHPEA-EDA, p-HPEA-EDA and 3,4-HPEA-EA, which dominated Oblica oil, and 3,4-HPEA-EA in Lastovka oil. Different abundance of these compounds could be useful to distinguish olive oils from different cultivars.