Phytochemical Profile and Antioxidant Activity of Lavandula angustifolia and Lavandula x intermedia Cultivars Extracted with Different Methods

Lavender is a valuable perennial plant from the Lamiaceae family. It is grown mainly for its essential oil, but it also contains polar bioactive compounds such as polyphenols and coumarins. Their level depends on the species, cultivars, geographical origin, climatic conditions, harvest time and extraction method. The authors investigated the effect of several extraction procedures (maceration, decoction and ultrasound-assisted extraction) applied to three cultivars of Lavandula angustifolia (Betty’s Blue, Elizabeth, Hidcote) and two cultivars of Lavandula x intermedia (Grosso, Gros Bleu) on the yield of the polyphenolic compounds and antioxidant activity. HPLC analysis showed the presence of rosmarinic acid (2.52–10.82 mg/g), ferulic acid glucoside (2.94–8.67 mg/g), caffeic acid (1.70–3.10 mg/g), morin (1.02–13.63 mg/g), coumarin (1.01–5.97 mg/g) and herniarin (1.05–8.02 mg/g). The content of phenolic acids and flavonoids was higher in lavender, while the content of coumarins was higher in lavandin in all types of extracts. The antioxidant activity was determined by DPPH-EPR assay for antiradical properties (104.58–206.77 μmol Trolox/g) and FRAP assay for reducing properties (79.21–203.06 μmol Trolox/g). The obtained results showed that the cultivar is the dominant factor differentiating the samples. Still, the extraction method plays an important role in the final bioactive substances content and antioxidant properties of obtained extracts.

For the extraction of active compounds from plant materials, conventional and modern techniques are used. The conventional methods, including solid-liquid extraction, maceration [6,22], decoction [23], Soxhlet extraction or hydrodistillation, are applied generally for galenical preparations. They require large quantities of solvents, extended extraction times and may contribute to the degradation of thermolabile compounds. On the other hand, modern methods such as ultrasound-assisted extraction (UAE), microwave-assisted extraction (MAE), pressurised liquid extraction (PLE), accelerated solvent extraction (ASE) and supercritical fluid extraction (SFE) are more environmentally friendly and thus have been named "green techniques" [24,25]. The yield of active compounds depends on the extraction method, the fragmentation of raw material and the properties of the extraction solvent [26]. According to Costa et al. [14], water and the ethanol-water mixture are the most suitable solvents for the extraction of phenolic compounds. Other factors such as extraction steps, extraction time, temperature and the solvent to sample ratio may additionally affect the extractability of polyphenolics [24,27].
This study aimed to determine the impact of the extraction methods (maceration, decoction and ultrasound-assisted extraction) and cultivars of lavender (Lavandula angustifolia) and lavandin (Lavandula x intermedia) on the content of phenolic acids, flavonoids, coumarins and antioxidant activity of the extracts. Ferric reducing antioxidant power assay (FRAP) and free radical scavenging assay (DPPH-EPR-electron paramagnetic resonance test) were used to determine the antioxidant activity of plant extracts. The relationships between total polyphenols, total flavonoids and antioxidant activity were also analysed using Pearson's correlation coefficient and principal component analysis (PCA).

Plant Material
The inflorescences of different cultivars of lavender (Lavandula angustifolia Mill.) and lavandin (Lavandula x intermedia Emeric ex Loisel.) were harvested from the plantation "Przystanek Lawenda" (which possesses the status of National Collection of the genus Lavandula) located in Borowiczki-Pieńki near Płock (Mazovia Province) in the central part of Poland. The voucher specimens were deposited in the Herbarium Universitatis Varsoviensis (Warsaw, Poland) with voucher numbers WA0000111924 (Hidcote), WA0000111925 (Elizabeth), WA0000111926 (Betty's Blue), WA0000111927 (Gros Bleu), WA0000111928 Antioxidants 2022, 11, 711 3 of 16 (Grosso). Plant material was collected during the flowering season in June 2019. Lavender inflorescences were subjected to a freeze-drying process (Alpha 1-2 LDplus CHRIST, Donserv, Warsaw, Poland) at −25 • C, 0.63 mbar, for 50 h, and then ground in an electric milling machine. The powdered samples were stored in a dry, cool and dark place until further analysis.

Preparation of Macerates
Macerates were prepared according to the procedure described in the Polish Pharmacopoeia [28]. One gram of plant material was soaked without mixing in 20 mL of 50% ethanol for 30 min at room temperature. After maceration, the extracts were centrifuged at 3000 rpm for 10 min and filtered off. All samples were prepared in triplicate.

Preparation of Decoctions
Decoctions were prepared according to the procedure described in the Polish Pharmacopoeia [28]. Ten millilitres of MilliQ water was added to 1 g of plant material and boiled for 30 min. Water was systematically added during cooking, and finally, decoctions were filled to a 10 mL volume. Next, the extracts were centrifuged at 3000 rpm for 10 min and filtered off. All samples were prepared in triplicate.

Preparation of Ultrasound-Assisted Extracts (UAE)
Ultrasound-assisted extraction was performed according to the modified method described by Arceusz and Wesolowski [29]. One gram of plant material was mixed with 20 mL of 50% ethanol and sonicated for 30 min, in the first stage, with 10 mL and for 15 min in the second and third stages. During the whole sonication process, the temperature did not exceed 36 • C. The combined extracts were centrifuged at 3000 rpm for 10 min and filtered off. All samples were prepared in triplicate.
One part of the obtained supernatants (4 mL) was placed in Eppendorf tubes and stored at −32 • C until total polyphenol, total flavonoid and antioxidant analyses. The second part of the supernatants was evaporated to dryness using a rotary evaporator (Heidolph Instruments, Schwabach, Germany). Then, the dry samples were dissolved in 10 mL of MilliQ water and freeze-dried at −25 • C, 0.63 mbar for 50 h. Thus, obtained samples were used for HPLC analysis.

Total Phenolic Content (TPC)
The total phenolic content (TPC) was determined using the modified Folin-Ciocalteu colourimetric method, as previously described by Waterhouse [30]. Briefly, 15 µL of the extract solution or gallic acid solution was diluted with 1185 µL of MilliQ water and subsequently, 75 µL of Folin-Ciocalteu reagent was added and mixed on a vortex. After 2 min, 225 µL of 20% sodium carbonate was added, and the reaction mixture was thermostatted for 20 min at 40 • C. The absorbance was measured against the blank at 765 nm using an Evolution 60S UV-Visible spectrophotometer (Thermo Scientific, Waltham, MA, USA). All samples were analysed in triplicate. The results were expressed as milligrams of gallic acid equivalents (GAE) per gram of dry matter of plant material (mg GAE/g d.w.) based on the calibration curve: y = 0.0011x + 0.0016, R 2 = 0.999 (mg/L).

Total Flavonoid Content (TFC)
The total flavonoid content (TFC) was determined using the modified colourimetric method described by Jindřiška and Neugebauerová [31]. The 1.4 mL of MilliQ water, 100 µL of an extract solution or a catechin solution, 60 µL of 5% sodium nitrite and 60 µL of 10% aluminium chloride were mixed. After 5 min of incubation at 25 • C, 0.4 mL of 1 M sodium hydroxide solution was added. The solution was mixed thoroughly, and the absorbance was measured against a blank solution at 510 nm using an Evolution 60S UV-Visible spectrophotometer (Thermo Scientific, Waltham, MA, USA). All samples were analysed in triplicate. The results were expressed as milligrams of catechin equivalents (CA) per gram of dry matter of plant material (mg CA/g d.w.) based on the calibration curve: y = 0.0016x + 0.0053, R 2 = 0.998 (mg/L). The stock solutions of chlorogenic acid, caffeic acid, ferulic acid, ellagic acid, rosmarinic acid, vanillin, isoquercitrin, morin, coumarin and herniarin, selected as the most common in plant material, were prepared by dissolving 1 mg of standard in 1 mL of methanol, then mixed together and filtered through a 0.45 µm syringe filter (Bionovo, Legnica, Poland). The concentrations of the most abundant compounds (caffeic acid, ferulic acid glucoside, rosmarinic acid, morin, coumarin and herniarin) were determined using an appropriate calibration curve. All measurements were performed in triplicate.
The linearity, detection limit and quantification limit were calculated from the curve to validate the method. The linearity was evaluated by the value of the coefficient of determination (R 2 ) of the calibration curve acquired for each standard. The limit of detection (LOD) and the limit of quantification (LOQ) were determined using the equations: LOD = 3.3 × σ/S and LOQ = 10 × σ/S, respectively (where σ is the standard deviation of the y-intercept and S is the slope of the calibration curve) ( Table S1). The precision of the method was determined by injecting three concentrations of standards: 40, 80 and 200 µg/mL for rosmarinic acid and herniarin, three times on the same day and the next day. The intraday and interday precision were assessed by calculating the relative standard deviation (RSD) ( Table S2). The coefficient of determination (R 2 ) for each compound was higher than 0.997, which indicates a high correlation of the data in the analysed concentration range. The LODs were from 0.008 to 0.021 mg/mL, while the LOQs were from 0.024 to 0.065 mg/mL (Table S1). The RSD values for the intraday and interday precision were below 1%, which indicated a good precision of the method (Table S2).

Extract Analysis
Qualitative and quantitative analyses of the lavender and lavandin extracts were performed using HPLC-DAD-UV/VIS. Separation of phenolic compounds and coumarins was achieved using the Hitachi Chromaster system (Tokyo, Japan) with a Purospher STAR RP-18e column (5 µm particle size, 250 mm × 4.6 mm, Merck) at 30 • C. The autosampler temperature was set at 20 • C, and the sample injection volume was 20 µL. The mobile phase consisted of 0.1% (v/v) formic acid in water (A) and 0.1% (v/v) formic acid in acetonitrile (B) at a flow rate of 1 mL/min. The gradient conditions were as follows: 10-20% B (0-35 min), 20-35% B (35-60 min), 35-10% B (60-60.1 min), 10% B (60.1-70 min). The freeze-dried extracts (40 mg) were dissolved in 2 mL of 50% ethanol (ultrasound-assisted extracts and macerates) or water (decoctions). The chromatograms were recorded at 250, 280 and 330 nm. The peaks were identified by comparing their retention times and UV-Vis spectra with standards.

MS Analysis
Mass spectrometry analyses were carried out using ultra-performance liquid chromatograph ACQUITY UPLC I-Class (Waters) coupled with Synapt G2-S mass spectrometer (Waters) equipped with the electrospray ion source and quadrupole-time-of-flight mass analyser. The resolving power of the TOF analyser was 40000 FWHM. The recorded data were processed using the MassLynx V4.1 software package (Waters). The chromatographic separations were performed on a Waters Acquity UPLC BEH C18 column (1.7 µm particle size, 2.1 mm × 100 mm). The mobile phase consisted of 0.1% (v/v) formic acid in water (A) and methanol (B) at a flow rate of 0.30 mL/min. The gradient conditions were as follows: 5% B (0-3 min), 5-100% B (3-30 min), 100-5% B (30-33 min), 5% B (33-35 min). The sample injection volume was 2 µL, and the wavelength of the UV detector was set at 254 nm and Antioxidants 2022, 11, 711 5 of 16 280 nm. The mass spectrometry measurement was achieved both in the positive and negative ion modes. The measurements were performed with capillary voltage set to 2.94 kV in positive ion mode and 3.05 kV in negative ion mode. The desolvation gas flow was 799 L/h and temperature 450 • C, and 774 L/h and temperature 350 • C, in positive ion mode and negative ion mode, respectively. The sampling cone voltage and source offset were set to 20 V in positive ion mode, while in negative ion mode, they were set to 32 V. The source temperature was 120 • C in both cases. The Leucine-Enkephaline solution was used as the Lock-Spray reference material. Commercial libraries such as MassBank (High-Quality Mass Spectral Database) and NIST Chemistry WebBook were used to identify compounds.

DPPH Scavenging (Electron Paramagnetic Resonance Test)
The free radical scavenging activity of the extracts was measured in vitro by the 2,2diphenyl-1-picrylhydrazyl (DPPH) assay described by Sanna et al. [32]. The stock solution was prepared by dissolving 12.5 mg DPPH in 25 mL methanol and stored at 4 • C until required but no longer than 12 h. In the experiment, 20 µL of the sample or blank was mixed with 180 µL of DPPH solution and kept for 30 min in the dark. EPR spectra were recorded, and their intensity was assumed as the double integral of the spectra. All samples were analysed in triplicate. The results were expressed as micromoles of Trolox equivalents per gram of dry matter of plant material (µM Trolox/g d.w.). EPR measurements were performed using a MiniScope MS200 spectrometer (Magnettech GmbH, Berlin, Germany) with the following set of parameters: central field 330 mT, sweep range 9.9 mT, sweep time 20 s, microwave power 6 mW, modulation amplitude 0.15 mT.

The Ferric Reducing Antioxidant Power (FRAP Assay)
The ferric reducing antioxidant power (FRAP) assay was performed according to the modified procedure described by Benzie and Strain [33]. Briefly, 50 µL of the sample or Trolox solution was mixed with 1000 µL of working FRAP reagent. The reaction mixture was thermostatted for 4 min at 37 • C. The absorbance was measured against the blank solution at 593 nm using an Evolution 60S UV-Visible spectrophotometer (Thermo Scientific, Waltham, MA, USA). The working FRAP reagent was prepared by mixing 20 mM iron (III) chloride and 10 mM TPTZ solutions with 300 mM acetate buffer (pH 3.6) in a proportion of 1:1:10, respectively. All samples were analysed in triplicate. The results were expressed as micromoles of Trolox equivalents per gram of dry matter of plant material (µM Trolox/g d.w.) based on the calibration curve: y = 0.0021x − 0.0108, R 2 = 0.998 (µM/L).

Statistical Analysis
The results were expressed as mean ± SD (n = 3). One-way analysis of variance (ANOVA) coupled with the Tukey post-hoc test was used to compare the data and to identify means with significant differences (p < 0.05). Pearson's correlation coefficient (r) was used to assess the relationship between antioxidant activity and the phenolic compound content at a 5% probability level. All data were subjected to PCA analysis as well. All statistical analyses were carried out with the TIBCO Statistica 13.3 (StatSoft Poland).

Results and Discussion
The typical appearances of the inflorescences of studied cultivars of lavender (Lavandula angustifolia Mill.) and lavandin (Lavandula x intermedia Emeric ex Loisel.) are shown in Figure 1.
The typical appearances of the inflorescences of studied cultivars of lavender (Lavandula angustifolia Mill.) and lavandin (Lavandula x intermedia Emeric ex Loisel.) are shown in Figure 1.

Total phenolic and flavonoid content
In this study, the total phenolic content (TPC) varied from 14.88 to 32.82 mg GAE/g d.w. for Lavandula angustifolia extracts and from 15.55 to 30.79 mg GAE/g d.w. for Lavandula x intermedia ( Figure 2). The total flavonoid content (TFC), a subgroup of polyphenols, ranged between 8.51 and 23.70 mg CA/g d.w. for Lavandula angustifolia extracts and between 10.91 and 22.06 mg CA/g d.w. for Lavandula x intermedia ( Figure 3). Statistically significant differences were found between the extraction methods, especially between an ultrasound-assisted method and decoction, for the total phenolic content and total flavonoid content for both species. An ultrasound-assisted method (UAE) gave the highest values of the total phenolic content (23.11-32.82 and 25.07-30.79 mg GAE/g d.w., for Lavandula angustifolia and Lavandula x intermedia extracts, respectively), followed by maceration (16.73-28.

Total Phenolic and Flavonoid Content
In this study, the total phenolic content (TPC) varied from 14.88 to 32.82 mg GAE/g d.w. for Lavandula angustifolia extracts and from 15.55 to 30.79 mg GAE/g d.w. for Lavandula x intermedia ( Figure 2). The total flavonoid content (TFC), a subgroup of polyphenols, ranged between 8.51 and 23.70 mg CA/g d.w. for Lavandula angustifolia extracts and between 10.91 and 22.06 mg CA/g d.w. for Lavandula x intermedia (Figure 3). Statistically significant differences were found between the extraction methods, especially between an ultrasoundassisted method and decoction, for the total phenolic content and total flavonoid content for both species. An ultrasound-assisted method (UAE) gave the highest values of the total phenolic content (23.11-32.82 and 25.07-30.79 mg GAE/g d.w., for Lavandula angustifolia and Lavandula x intermedia extracts, respectively), followed by maceration (16.73-28. (Figure 3). The values of total phenolic and total flavonoid contents obtained for Lavandula x intermedia for the Grosso cultivar, with higher results for decoction (22.28 mg GAE/g d.w. and 14.57 mg CA/g d.w.) than macerate (19.69 mg GAE/g d.w. and 13.83 mg CA/g d.w.), were the exception (Figure 2, Figure 3). Similar to the total phenolic content, the aqueous-ethanolic extracts of both species showed a higher total flavonoid content than the aqueous extracts. The highest total flavonoid content was observed for the extracts obtained by the ultrasound-assisted method    Our results are in agreement with those reported by Tušek et al. [34], where the total phenolic content of Lavandula x intermedia was 17.83 mg GAE/g d.w. and 26.21 mg GAE/g d.w. when the extraction was performed with water at 40 • C and 60 • C, respectively. Likewise, Jiménez-Zamora et al. [11] obtained a result equal to 322 mg GAE/L, which corresponds to 24.15 mg GAE/g d.w., for the lavender infusion. Yet, these results are higher than those reported by Kontogiorgis [23]. In their study, the total phenolic content for decoction prepared from Lavandula angustifolia was 130.59 mg GAE/L, which corresponds to 3.92 mg GAE/g d.w. This could be due to a shorter extraction time of 5 min as compared with the 30 min extraction time in this work. Moreover, research on methanolic extracts from 30 different populations of Lavandula x intermedia indicated the presence of phenolic compounds in the range of 0.31-1.05 mg GAE/g d.w., with the flavonoid content of 0.28-0.72 mg QE/g d.w. [12]. On the other hand, in the study of Spiridon et al. [35], the extraction of lavender with methanol afforded a higher total phenolic content equal to 50.6 mg GAE/g as well as a higher total flavonoid content equal to 27.6 mg rutin/g. In addition, our results were lower than those obtained by Hawrył et al. [13], where the total Antioxidants 2022, 11, 711 8 of 16 phenolic content of methanolic extracts was 1.46 mg GAE/mL (which could be recalculated to 72.87 mg GAE/g d.w. for plant material) for Lavandula angustifolia Atropurpurea and 1.98 mg GAE/mL (98.88 mg GAE/g d.w.) for Lavandula angustifolia Rosea. The above values are much higher than those reported in our study, probably because other extraction procedures and different varieties of lavender were used.
In general, the highest values of the total phenolic content (TPC) and total flavonoid content (TFC) were obtained for the Betty's Blue and Elizabeth cultivar extracts, whereas the lowest ones were obtained for the Hidcote cultivar extracts. The order of other cultivars depended on the extraction method. In the case of extracts obtained by the ultrasoundassisted method (UAE), there were no significant differences between the Betty's Blue, Elizabeth and Gros Bleu cultivars (TPC: 30.25; 32.82; 30.79 mg GAE/g d.w. and TFC: 21.45; 23.70; 22.06 mg CA/g d.w.). Nevertheless, the above cultivars had significantly higher total phenolic and total flavonoid contents than the Grosso and Hidcote cultivars (TPC: 25.07; 23.11 mg GAE/g d.w. and TFC: 18.57; 14.12 mg CA/g d.w., respectively) (Figures 2 and 3).
As concerns macerates, the highest total phenolic and total flavonoid contents were obtained again for the Betty's Blue and Elizabeth cultivars (TPC: 28
The peaks were identified by comparing their retention times and UV spectra with standards and by the presence of a characteristic fragmentation pattern (Table S3, Figure S2). Only the identification of compound four was based solely on a fragmentation pattern analysis of UPLC-MS results since no standard was available. It was identified as ferulic acid glucoside based on fragment ions released at m/z 193.05, 149.06 and 134.04 ( Figure S3). Previously, this compound has been found in Lavandula angustifolia aqueous-ethanolic extract [35], Lavandula dentata hydromethanolic extract [37] and Lavandula x intermedia waste [41] obtained after the steam distillation of essential oils.
On the other hand, Hawrył et al. [13] reported the presence of ferulic acid only in La-331 vandula viridis methanolic extract, while Adaszyńska-Skwirzyńska and Dzięcioł [40] [35], Lavandula dentata hydromethanolic extract [37] and Lavandula x intermedia 346 waste [41] obtained after the steam distillation of essential oils. All the chromatograms showed a similar profile of phenolic compounds and coumarins, as shown in Figure 4. However, differences were found in the content of individual phenolic compounds between the cultivars and the extraction method. The contents of phenolic compounds: caffeic acid (no. 2; CA), ferulic acid glucoside (no. 4; FA), rosmarinic acid (no. 9; RA), morin (no. 10; M) and coumarins: coumarin (no. 8; C) and herniarin (no. 11; H), as the most abundant ones, were analysed ( Figure 5). The obtained results were calculated as mg/g of freeze-dried extract. The content of ferulic acid glucoside was expressed as ferulic acid. Amounts of rosmarinic acid (2.52-7.18 mg/g), ferulic acid glucoside (2.94-5.97 mg/g) and morin (1.02-6.14 mg/g) were lower in decoctions than in the ethanolic extracts: UAE (3.31-10.82 mg RA/g, 3.79-7.64 mg FA/g, 3.42-12.70 mg M/g) and macerates (3.91-9.17 mg RA/g, 3.85-8.67 mg FA/g, 3.91-13.63 mg M/g). On the other hand, there was no significant difference in the content of caffeic acid. The highest values for caffeic acid were obtained for decoctions (2.07-3.10 mg/g), followed by maceration (1.83-2.88 mg/g) and UAE (1.70-2.47 mg/g). In turn, significant differences were found between the extraction methods, especially between macerates (2.63-5.97 mg C/g and 3.06-8.02 mg H/g) and decoction (1.01-2.72 mg C/g and 1.05-2.23 mg H/g), for the content of coumarin and herniarin, respectively. Thus, it indicates that the best extraction method should be chosen considering the specific compounds and not the whole groups.  The content of rosmarinic acid, ferulic acid glucoside, morin and caffeic acid was higher in lavender (Betty's Blue, Elizabeth, Hidcote), while the content of coumarin and herniarin was higher in lavandin (Grosso, Gros Bleu) in all types of extracts. The exception was the content of ferulic acid glucoside obtained for the Grosso cultivar, with a higher result for maceration (6.85 mg/g), UAE (6.21 mg/g) and decoction (4.15 mg/g) than for the Betty's Blue cultivar, 3.85 mg/g, 4.01 mg/g and 3.46 mg/g, respectively ( Figure 5).
The highest variance among cultivars of the same species was observed for morin, though only for lavender species, ferulic acid glucoside and herniarin. On the other hand, the rosmarinic acid and coumarin content differed between species but not between cultivars of the same species. Thus, it would indicate that the path of synthesis of these compounds is species-specific, not cultivar-specific.

Antioxidant Properties
The antioxidant properties were determined by two assays: DPPH for antiradical properties and FRAP for reducing properties. Both methods gave similar results (104.58-206.77 µmol Trolox/g d.w. and 79.21-203.06 µmol Trolox/g d.w. for DPPH and FRAP assay, respectively). The values obtained for macerates and decoctions by DPPH assay were significantly higher than those obtained by FRAP assay, whereas the extracts obtained in the ultrasound-assisted process (UAE) of both methods gave similar results ( Figure 6).

Antioxidant Properties
The antioxidant properties were determined by two assays: DPPH for antiradical properties and FRAP for reducing properties. Both methods gave similar results (104.58-206.77 μmol Trolox/g d.w. and 79.21-203.06 μmol Trolox/g d.w. for DPPH and FRAP assay, respectively). The values obtained for macerates and decoctions by DPPH assay were significantly higher than those obtained by FRAP assay, whereas the extracts obtained in the ultrasound-assisted process (UAE) of both methods gave similar results ( Figure 6). Generally, the highest radical scavenging and reducing properties were observed for the extracts obtained in the ultrasound-assisted process (104.58-187.04 μmol Trolox/g d.w. and 142.69-203.06 μmol Trolox/g d.w. for DPPH and FRAP assay, respectively). The only exceptions were the values obtained in the DPPH assay for the Hidcote cultivar, with the highest result for macerate (156.75 μmol Trolox/g d.w. versus 104.58 μmol Trolox/g d.w. for UAE extract and 116.98 μmol Trolox/g d.w. for decoction). The differences were more prominent in FRAP assay results, where, for all cultivars, the reducing properties were significantly stronger for the extracts obtained in the ultrasound-assisted process than for Generally, the highest radical scavenging and reducing properties were observed for the extracts obtained in the ultrasound-assisted process (104.58-187.04 µmol Trolox/g d.w. and 142.69-203.06 µmol Trolox/g d.w. for DPPH and FRAP assay, respectively). The only exceptions were the values obtained in the DPPH assay for the Hidcote cultivar, with the highest result for macerate (156.75 µmol Trolox/g d.w. versus 104.58 µmol Trolox/g d.w. for UAE extract and 116.98 µmol Trolox/g d.w. for decoction). The differences were more prominent in FRAP assay results, where, for all cultivars, the reducing properties were significantly stronger for the extracts obtained in the ultrasound-assisted process than for both macerates and decoctions. The differences between FRAP results for the latter two extraction procedures were not significant in most cases, except for the Elizabeth cultivar ( Figure 6).
Jiménez-Zamora et al. [11] reported a value of 3.43 mmol Trolox/L for lavender infusion in DPPH assay, which corresponds to 257.25 µmol Trolox/g d.w. for plant material. This value is higher than most values obtained in this work, which could be due to the use of a different extraction method. On the other hand, Ferreira et al. [42], in a DPPH assay, obtained results equal to 9.2 mg DPPH/g d.w. (which roughly corresponds to 10 µmol Trolox/g d.w.) for Lavandula angustifolia and to 28 mg DPPH/g d.w. (circa 30 µmol Trolox/g d.w.) for Lavandula pedunculata decoctions. The above values are much lower than those obtained in our study for decoctions, as well as for other extraction procedures.
As for decoctions, the sample order observed for both assays was the same. The highest result was obtained for Betty's Blue cultivar (174.93 and 131.20 µmol Trolox/g d.w. for DPPH and FRAP assay, respectively), followed by Grosso (162. 34  The variations in the order of studied cultivars depending on the assay could be due to different mechanisms of the assays. In the FRAP assay, there is only the electron transfer mechanism present, whereas, in the DPPH assay, there is usually a combination of hydrogen transfer and electron transfer mechanisms. That could also be due to a different reaction medium, namely the acidic aqueous solution in FRAP assay and methanol in DPPH assay, which could affect the activity of phenolic compounds present in the extracts.

Correlation between Antioxidant Activity and Phenolic Content
According to previous studies [11][12][13]43], antioxidant activity is associated with phenolic compounds. In our study, a highly positive correlation was observed between FRAP assay and total flavonoid contents. The highest Pearson's correlation coefficient (r = 0.99) was obtained for the macerates, followed by the ultrasound-assisted method (UAE) (r = 0.97) and decoctions (r = 0.85). In addition, a significant positive correlation was found for the FRAP assay with the total phenolic content for macerates (r = 0.95) and the ultrasound-assisted method (r = 0.89). Our study also showed that the coefficient (r = 0.97) of correlation between the total phenolic content and the total flavonoid content for aqueous-ethanolic extracts was higher than for the aqueous extracts (r = 0.90). The correlation circle of the PCA (Figure 7) was also used to show correlations between the total content of polyphenols and flavonoids and antioxidant assays for all extraction methods. A positive correlation between flavonoids, polyphenols and FRAP was observed, while between flavonoids, polyphenols and DPPH, there was only a weak positive correlation. These variations could be related to the different mechanisms of action of antioxidant assays. The PCA was also a useful statistical tool to observe the separation between groups. The results are presented graphically as a score plot (Figure 8). PCA analysis showed that the first two principal components were able to explain 97.58% of the variability in the data.
The first principal component (PC 1) described 88.07%, while the second (PC 2) was 9.51% of the total variance. The PCA allowed for confirming our previous results that the total content of polyphenolic compounds and antioxidant activity of extracts of lavender and lavandin cultivars depends dominantly on the cultivar and secondary on the extraction method. The most prominent effect was observed for ultrasound-assisted extracts, which were distinctly separated from macerates and decoctions.

Conclusions
The authors of this research investigated the effect of different extraction procedures on the recovery of the polyphenolic compounds of several cultivars of lavender (Lavandula angustifolia Mill.) and lavandin (Lavandula x intermedia Emeric ex Loisel.). HPLC analysis showed the presence of phenolic acids (rosmarinic acid, ferulic acid glucoside, caffeic acid, ellagic acid), flavonoids (morin, isoquercitrin, vanillin) and coumarins (herniarin, coumarin). In our study, the content of phenolic acids and flavonoids was higher in lavender (Betty's Blue, Elizabeth, Hidcote), while the content of coumarins was higher in lavandin (Grosso, Gros Bleu) in all types of extracts. The extractability of bioactive compounds, both polyphenols and coumarins, depended on the extraction method as well. This effect was especially noticeable in the case of morin, rosmarinic acid, coumarin and herniarin. It could be due to such factors as the number of extraction steps, extraction time, solvent used and temperature. Too high temperature may contribute to the degradation of thermolabile compounds, which results in low extraction yields. In our study, aqueousethanolic extracts (UAE and macerates) showed a higher amount of bioactive compounds and antioxidant activity than aqueous extracts (decoctions). Our results indicated that the ultrasound-assisted extraction was the best method for extraction of polyphenolic compounds, while the macerate seems to be the optimum method for coumarin extraction. Furthermore, a good correlation was found between the total phenolic and total flavonoid contents and antioxidant activity (FRAP) of the lavender and lavandin extracts. To sum up, the results showed that the cultivar is the main factor differentiating the samples. Still, the extraction method significantly influences the composition and antioxidant properties of obtained extracts, even when the same solvent is used.