Characterization of Berry Skin Phenolic Profiles in Dalmatian Grapevine Varieties

Abstract

Dalmatian vineyards host many autochthonous varieties. The phenolic profile, defined by the relative proportions of different phenolic compounds, is specific for each grape variety. The aim of this study was to determine and analyze the flavonoid compounds of twenty rare red grape varieties. Nineteen phenolic compounds, represented by anthocyanins, flavanols, and flavonols, were detected and quantified using HPLC in three consecutive vintages. The content of grape skin anthocyanins (10414.06 (Plavac mali crni)-19.58 (Trišnjavac) mg kg⁻¹ d.w. of grape skin), flavonols (1742.08 (Pošip crni)-215.56 (Crljenak viški) mg kg⁻¹ d.w. of grape skin), and flavan-3-ols (448.04 (Pošip crni)-87.88 (Glavinuša) mg kg⁻¹ d.w. of grape skin) showed significant differences in investigated varieties. According to the investigated phenolic compounds, Pošip crni, Ljutun, Zadarka, Dobričić, Plavac mali crni, and Trnjak differed from other investigated varieties. These local varieties can be perceived as an alternative to widespread varieties in Dalmatia. This was shown by one of the first studies on phenolic profiles of (mostly) rare autochthonous varieties.

1. Introduction

Dalmatia is a wine growing region in Croatia, with well-preserved native grapevine genetic resources. Out of more than 100 grapevine varieties native to Croatia, more than 50% are related to this region [1]. Even today they make a significant share in the overall grape and wine production of this region. Nevertheless, only a couple of these are widely cultivated, while others are related to limited and small cultivation areas and many of them are critically endangered [2]. However, a new trend of giving attention to autochthonous varieties has emerged due to consumers demands for new wines with special features, differing from the rest. Ultimately, a revitalization of minor grape varieties could lead to multiformity of the wine market but also the increase in grapevine genetic resources used in wine production, which is now extremely narrow.

Phenolic compounds play one of the major roles in the quality of red grapes and wines and they have many positive effects on human health [3]. Red grapes represent one of the most important sources of polyphenols for human diet in the form of fresh fruits, as well as red wine [4]. Research focused on the evaluation of specific grapevine genetic resources from certain geographic region [5,6,7] is of great interest since varietal differences in content of different polyphenolic compounds are of great importance for wine industry and wine market, breeding potential of this germplasm, as well as for growing industry of polyphenols.

Polyphenols can be classified in two groups: flavonoid and non-flavonoid. These compounds are responsible for color, bitterness, and astringency of red wine, as well as for different reactions with proteins, oxygen, and changes of wine during ageing [8]. Most flavonoids found in grapes are anthocyanins, flavanols, and flavonols, while non-flavonoids mainly include hydroxycinnamic and hydroxybenzoic acids [9]. Anthocyanins are red pigments, which are responsible for the color of red wines. The most important grape anthocyanins are 3-glucoside forms of cyanidin, peonidin, petunidin, delphinidin, and malvidin [10]. Flavanols, also known as flavan-3-ols, are mainly responsible for organoleptic properties of wine, such as astringency, bitterness, but also contribute to stabilizing wine color [11]. In grapes they can be found in the form of monomers or polymers called proanthocyanidins [3]. Flavonols, another group of flavonoids, are important in stabilizing wine color, because they act as copigments with anthocyanins [12].

The content and composition of phenolic compounds in grapes are influenced by grape variety, environmental, and viticultural practices [13]. However, some of the polyphenolic classes, such as anthocyanins and flavonols, are under strict genetic control. Meaning that the content of compounds is more influenced by environmental factors while composition is influenced by the genotype [14,15]. Furthermore, it was recently shown that geographic origin, combined with genetics, also influences the overall polyphenolic profiles [6]. Thus the aim of this research was to investigate the polyphenolic profiles of Dalmatian native varieties and to determine the differences based on the analyzed polyphenolic profiles.

2. Materials and Methods

2.1. Chemicals

Acetonitrile HPLC grade was purchased from J. T. Baker (Deventer, Netherlands). Acetic acid and 85% orthophosphoric acid were obtained from Fluka (Buchs, Switzerland). Ethanol was provided from Kemika (Zagreb, Croatia). Standards used for identification and quantification purposes were as follows: delphinidin 3-O-glucoside, cyanidin 3-O-glucoside, peonidin 3-O-glucoside, malvidin 3-O-glucoside, epigallocatechin, procyanidin B1, procyanidin B2, rutin, myricetin-3-O-glucoside, kaempferol-3-O-glucoside, and isorhamnetin-3-O-glucoside (Extrasynthese, Genay Cedex, France); (−)-epicatechin, (+)-catechin and epicatechin gallate (Sigma-Aldrich, St. Louis, MO, USA); quercetin-3-O-glucoside (Sigma, St. Louis, MO, USA).

2.2. Grape Samples

Grape samples of twenty red grape varieties (Vitis vinifera L.) were used in this research: Babić (BA), Babica (BAB), Tribidrag (TRB), Crljenak viški (CV), Dobričić (DOB), Drnekuša vela (DV), Glavinuša (GLA), Gustopupica (GUS), Lasina (LAS), Ljutun (LJUT), Ninčuša (NIN), Soić (SO), Pošip crni (PC), Plavina (PLA), Plavac mali (PMC), Svrdlovina (SVR), Trišnjavac (TRI), Trnjak (TR), Vranac (VRA), and Zadarka (ZAD). Grape samples were collected in the germplasm collection at the Institute for Adriatic Crops and Karst Reclamation (Split, Dalmatia, Croatia) in three consecutive years (2011, 2012, and 2013). This germplasm collection is in the winegrowing region Central and South Dalmatia. All grape varieties in this collection are of the same age and pruning system. Grapevine is grafted onto 1013 Paulsen rootstock with middle head bilateral cordon training and planted 2.2 × 1.1 m. Each sample consisted of five bunches randomly picked from five different vines at full ripeness.

2.3. Grape Extraction Procedure

The grape skins were manually removed from the pulp and freeze-dried. Freeze-drying was performed at 0.1 mbar for 72 h. Dry skins were grinded and obtained powder (500 mg) was extracted by a 10 mL of 70% aqueous ethanol, containing 1% formic acid for one day in the dark at room temperature. The extract was centrifuged in LC-321 centrifuge (Tehtnica, Železnik, Slovenia) for 20 min at 5000 rpm at room temperature. Supernatant was collected, concentrated under vacuum to remove ethanol (40 °C) on rotary evaporator and brought to final volume of 10 mL with a mobile phase A. The extract was filtered with Phenex-PTFE 0.20 µm syringe filter (Phenomenex, Torrance, CA, USA) and analyzed by HPLC.

2.4. HPLC Analysis

Separation and identification of polyphenol compounds was performed to the method described by Tomaz and Maslov [16]. The analyses were performed on an HPLC Agilent 1100 (Agilent Technologies, Palo Alto, CA USA) comprising a binary pump, an autosampler, a diode array detector and Agilent 1200 fluorescence detector. Separation was performed on Luna Phenyl-Hexyl (Phenomenex, Torrence, CA USA) column (250 mm × 4.6 mm i.d., 5 µm particle size) with Phenyl guard column (4.0 × 3.0). Column was heated at 50 °C. The injection volume for all samples was 20 µL. Gradient consists of two phases: (A) water/phosphoric acid (99.5/0.5, v/v), and (B) acetonitrile/water/phosphoric acid (50/49.5/0.5, v/v/v). For detection and quantification of compounds, the chromatograms were recorded at 280, 360, and 518 nm by diode array detector and at excitation wavelength 225 nm and emission wavelengths at 320 nm by fluorescence detector. UV-VIS spectra were recorded in range of 200–700 nm. Quantification of non-commercial available standards of anthocyanins was made accordingly to the calibration curves of malvidin-3-O-glucoside. Samples were analyzed in triplicate.

2.5. HPLC-ESI-MS Analysis

For peak assignment, phenolic compounds were confirmed by HPLC-ESI-MS with Agilent 1200 Series System (Agilent Technologies, Waldbronn, Germany) coupled on-line to an Agilent model 6410 mass spectrometer fitted with ESI source. The same column was used as previously described. The mobile phase was fixed to 0.5 mL min ⁻¹. The solvents were (A) aqueous 0.1% formic acid and (B) acetonitrile containing 0.1% formic acid. Mass spectra were recorded from m/z = 100 to 1000 in a positive and negative ionization mode alternately. The electrospray ionization (ESI) parameters were: drying gas (N₂) flow and temperature 8 L min-1 and 300 °C; nebulizer pressure was 30 psi, capillary voltage 4500 V for negative ion mode or −4500 V for positive ion mode. Fragmentation voltage was 135 V.

2.6. Statistical Analysis

ANOVA was used to test the significance of the effects of grapevine variety and year of sampling, as well as their interaction for all phenolic compounds analyzed and for summarized compounds from three groups (anthocyanins, flavonols, and flavanols). In the case of significant results obtained by ANOVA, means were compared using Duncan’s multiple range test among varieties. To evaluate the total variability in phenolic profiles of 20 varieties, principal component analysis (PCA) was performed, and variables and observations scores for first two canonical factors were used to create scatter plots to explain multivariate differences among samples. All the analyses were carried out using XLSTAT software v.2020.3.1. (Addinsoft, New York, NY, USA).

3. Results and Discussion

In total, 19 compounds were analyzed belonging to the classes of anthocyanins, flavanols, and flavonols. In

Table 1. Anthocyanin profile of Croatian autochthonous grape varieties.

	Delphinidin-3-O-glucoside	Cyanidin-3-O-glucoside	Petunidin-3-O-glucoside	Peonidin-3-O-glucoside	Malvidin-3-O-glucoside	Total Anthocyanins
BA	238.12 ± 5.96 h*	11.68 ± 1.97 i	204.71 ± 2.33 l	33.15 ± 3.40 m	1279.73 ± 24.80 p	1767.39 ± 13.11 o
BAB	105.86 ± 10.41 m	1.12 ± 1.68 q	137.05 ± 1.06 m	21.38 ± 7.54 o	1789.12 ± 139.15 n	2054.53 ± 107.77 m
CV	63.41 ± 4.82 n	0.27 ± 0.41 s	85.30 ± 2.17 o	19.92 ± 1.55 p	1601.02 ± 8.48 o	1769.92 ± 14.82 o
DOB	1291.73 ± 64.41 a	79.13 ± 3.38 b	1255.22 ± 6.98 a	227.67 ± 14.94 c	5684.07 ± 27.11 d	8537.82 ± 97.11 d
DV	178.77 ± 19.78 j	35.51 ± 22.51 f	236.88 ± 28.11 j	84.21 ± 1.11 h	2330.07 ± 110.07 j	2865.44 ± 107.16 j
GLA	141.61 ± 3.75 l	14.81 ± 0.87 h	292.03 ± 3.68 h	62.34 ± 2.67 i	2049.75 ± 45.22 l	2560.54 ± 49.94 k
GUS	0.00 ± 0.00	1.78 ± 1.40 p	0.00 ± 0.00	26.96 ± 2.31 n	218.28 ± 7.65 r	247.02 ± 6.79 q
LAS	30.15 ± 5.48 p	70.11 ± 1.84 c	46.09 ± 0.84 q	102.48 ± 1.54 f	290.80 ± 8.72 q	539.62 ± 14.14 q
LJUT	433.48 ± 7.20 e	4.73 ± 1.07 k	443.35 ± 4.34 f	44.53 ± 5.95 l	3417.08 ± 25.63 g	4343.17 ± 21.20 f
NIN	223.547.82 i	2.28 ± 1.94 o	268.07 ± 22.31 i	44.12 ± 6.08 l	3447.36 ± 77.36 f	3985.37 ± 48.47 g
PC	0.00 ± 0.00	3.42 ± 1.07 n	0.00 ± 0.00	100.89 ± 2.39 g	49.54 ± 9.87 s	153.86 ± 9.04 r
PLA	37.57 ± 3.16 o	0.90 ± 0.78 r	72.74 ± 3.67 p	27.28 ± 2.08 n	1839.77 ± 12.43 m	1978.26 ± 11.91 n
PMC	1064.68 ± 70.91 b	44.44 ± 1.90 d	1002.76 ± 11.27 b	189.24 ± 5.75 d	8112.96 ± 131.43 b	10414.07 ± 219.92 a
SO	287.52 ± 8.68 g	8.77 ± 2.05 j	385.94 ± 22.54 g	54.98 ± 5.14 j	3009.15 ± 72.52 h	3746.36 ± 106.05 h
SVR	21.77 ± 1.37 q	0.00 ± 0.00	106.44 ± 1.14 n	14.87 ± 1.50 q	2199.25 ± 36.89 k	2342.33 ± 38.66 l
TR	598.96 ± 1.37 d	39.21 ± 2.14 e	774.01 ± 8.89 c	369.35 ± 18.58 a	8260.18 ± 102.12 a	10041.70 ± 116.64 b
TRB	145.28 ± 8.35 k	3.63 ± 4.29 m	208.77 ± 71.38 k	49.03 ± 11.27 k	2748.50 ± 24.94 i	3155.21 ± 109.78 i
TRI	0.00 ± 0.00	4.51 ± 0.56 l	0.00 ± 0.00	3.69 ± 0.49 r	11.39 ± 1.69 t	19.59 ± 2.75 s
VRA	634.03 ± 196.52 c	102.43 ± 20.92 a	718.00 ± 24.43 d	261.16 ± 9.57 b	4303.08 ± 152.99 e	6018.70 ± 326.08 e
ZAD	382.06 ± 6.93 f	15.54 ± 1.65 g	603.24 ± 2.61 e	162.23 ± 5.91 e	7620.85 ± 162.50 c	8783.92 ± 174.19 c
p value	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001

* Means with different superscript letters in the same row differ significantly (p ≤ 0.05).

,

Table 2. Flavonols profile of Croatian autochthonous grape varieties.

	Myricetin-3-O-glucoside	Rutin	Hyperoxide	Quercetin-3-O-glucoside	Kaempferol-3-O-glucoside	Isorhamnetin-3-O-glucoside	Total Flavonols
BA	53.09 ± 3.05 l*	13.05 ± 0.53 mn	10.00 ± 0.21 hi	111.07 ± 2.19 s	6.09 ± 0.94 j	26.21 ± 0.56 k	219.49 ± 3.15 q
BAB	58.74 ± 2.88 i	14.57 ± 2.29 l	11.22 ± 2.02 h	176.07 ± 3.31 n	2.25 ± 3.37 k	37.13 ± 5.01 h	299.98 ± 18.74 o
CV	37.72 ± 6.56 n	15.79 ± 2.83 k	8.69 ± 6.66 i	120.77 ± 10.56 k	0.00 ± 0.00	32.58 ± 9.53 i	215.56 ± 26.12 q
DOB	142.06 ± 12.23 c	13.50 ± 1.89 m	0.00 ± 0.00	169.19 ± 1.44 o	0.00 ± 0.00	41.66 ± 4.06 g	366.42 ± 17.27 j
DV	46.97 ± 5.34 m	27.73 ± 3.76 f	10.83 ± 0.84 hi	277.00 ± 3.77 f	10.66 ± 1.13 f	25.40 ± 0.78 k	398.58 ± 7.79 h
GLA	81.80 ± 1.01 h	19.03 ± 0.85 i	20.91 ± 1.18 e	227.70 ± 25.25 k	15.53 ± 0.46 e	26.13 ± 2.52 k	391.11 ± 28.75 i
GUS	31.95 ± 1.82 o	10.62 ± 0.73 p	20.75 ± 2.57 e	244.81 ± 88.22 j	21.58 ± 0.47 c	29.60 ± 2.48 j	359.32 ± 88.25 k
LAS	32.29 ± 0.26 o	0.00 ± 0.00	22.55 ± 2.42 de	249.69 ± 1.71 e	15.02 ± 0.57 e	14.35 ± 1.65 m	333.89 ± 2.86 lm
LJUT	87.41 ± 2.17 f	44.94 ± 0.83 b	23.70 ± 3.40 d	272.79 ± 3.72 g	18.53 ± 1.05 d	37.30 ± 3.08 h	484.67 ± 5.81 g
NIN	56.25 ± 7.46 j	24.76 ± 0.44 g	14.46 ± 1.83 f	151.04 ± 3.37 p	0.00 ± 0.00	49.00 ± 7.01 f	295.51 ± 16.09 o
PC	21.52 ± 1.96 r	21.29 ± 1.66 h	113.14 ± 2.45 b	1358.82 ± 10.11 a	213.22 ± 4.27 a	14.09 ± 1.10 m	1742.08 ± 16.23 a
PLA	54.76 ± 1.33 k	16.42 ± 0.46 j	13.56 ± 1.01 fg	116.44 ± 1.68 r	7.11 ± 0.89 ij	41.27 ± 3.13 g	249.56 ± 3.91 p
PMC	329.63 ± 32.46 a	99.18 ± 2.62 a	439.96 ± 34.59 a	446.35 ± 12.05 d	114.82 ± 50.53 b	96.40 ± 9.15 c	1526.32 ± 91.61 b
SO	59.11 ± 1.6 1 i	24.41 ± 3.01 g	12.02 ± 2.82 gh	207.31 ± 4.06 l	8.15 ± 6.22 hi	19.42 ± 10.22 l	330.42 ± 11.94 m
SVR	90.38 ± 11.50 e	0.00 ± 0.00	10.52 ± 0.65 hi	85.54 ± 0.88 t	8.88 ± 0.62 gh	125.75 ± 10.05 b	321.07 ± 20.15 n
TR	184.01 ± 7.18 b	12.32 ± 2.22 o	21.73 ± 0.68 de	265.85 ± 9.95 h	10.23 ± 0.36 f	129.99 ± 4.38 a	624.13 ± 16.9 d
TRB	97.76 ± 1.55 d	34.10 ± 3.49 d	13.62 ± 3.07 fg	279.99 ± 14.63 e	0.00 ± 0.00	74.12 ± 3.28 e	499.59 ± 21.24 f
TRI	27.17 ± 3.31 p	28.52 ± 2.64 e	29.35 ± 0.98 c	500.64 ± 4.33 b	10.13 ± 0.21 fg	0.00 ± 0.00	595.80 ± 5.98 e
VRA	83.29 ± 4.96 g	12.70 ± 0.46 no	12.26 ± 0.78 fgh	192.02 ± 9.37 m	0.00 ± 0.00	38.16 ± 0.99 h	338.43 ± 14.45 l
ZAD	26.26 ± 8.40 q	36.58 ± 8.87 c	28.49 ± 2.65 c	470.79 ± 28.99 c	19.43 ± 0.81 d	79.92 ± 3.32 d	661.46 ± 31.57 c
p value	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001

* Means with different superscript letters in the same row differ significantly (p ≤ 0.05).

and

Table 3. Flavanols profile of Croatian autochthonous grape varieties.

	Epicatechin Gallate	Gallocatechin	Procyanidin B1	Epigallocatechin	Procyanidin B3	Catechin	Procyanidin B4	Procyanidin B2	Total Flavanols
BA	6.37 ± 1.36 f*	2.02 ± 0.09 m	11.25 ± 1.37 k	43.39 ± 3.93 h	4.67 ± 0.50 g	16.30 ± 1.45 k	13.67 ± 0.79 gh	11.56 ± 1.69 hi	109.22 ± 6.25 kl
BAB	4.71 ± 0.65 i	4.81 ± 0.74 l	55.70 ± 1.65 e	69.60 ± 2.41 d	1.42 ± 2.13 k	73.55 ± 2.97 b	30.77 ± 6.07 b	40.32 ± 3.43 b	280.88 ± 12.09 d
CV	2.74 ± 0.48 l	4.72 ± 0.36 l	28.81 ± 1.31 g	37.83 ± 2.00 ij	0.99 ± 1.49 l	16.02 ± 1.89 k	12.52 ± 1.05 hi	12.11 ± 1.16 hi	115.74 ± 4.81 k
DOB	14.85 ± 1.23 a	7.69 ± 3.42 i	72.81 ± 4.54 b	65.07 ± 3.09 e	2.37 ± 0.48 i	33.67 ± 2.86 f	18.44 ± 3.96 e	26.66 ± 3.56 c	241.56 ± 20.39 f
DV	6.54 ± 1.05 f	15.27 ± 1.81 d	23.88 ± 0.86 i	25.30 ± 1.64 m	0.00 ± 0.00	10.62 ± 0.94 mn	11.82 ± 1.48 i	12.28 ± 0.98 hi	105.71 ± 3.56 l
GLA	5.98 ± 0.50 g	8.05 ± 0.81 i	0.00 ± 0.00	31.03 ± 2.25 l	3.57 ± 0.50 h	16.85 ± 2.17 k	9.91 ± 1.51 j	12.49 ± 0.98 h	87.88 ± 3.45 m
GUS	0.57 ± 0.85 n	16.06 ± 1.62 c	25.50 ± 1.95 h	48.95 ± 2.23 g	6.60 ± 0.55 f	30.97 ± 1.66 g	18.47 ± 1.19 e	24.94 ± 3.51 d	172.07 ± 6.03 h
LAS	0.00 ± 0.00	16.30 ± 1.16 c	24.09 ± 4.44 hi	40.16 ± 7.07 i	0.00 ± 0.00	30.02 ± 1.61 g	30.22 ± 1.29 bc	39.51 ± 2.11 b	180.29 ± 8.81 h
LJUT	5.34 ± 0.24 h	10.89 ± 2.10 g	66.49 ± 4.79 cd	100.99 ± 2.19 b	10.47 ± 0.76 d	63.81 ± 2.68 d	68.49 ± 4.24 a	19.83 ± 1.43 f	346.32 ± 16.62 c
NIN	11.24 ± 1.58 e	14.59 ± 2.76 e	16.68 ± 1.48 j	26.48 ± 2.19 m	3.39 ± 0.81 h	9.97 ± 2.04 n	9.90 ± 1.07 j	11.30 ± 0.99 i	103.54 ± 8.64 l
PC	0.00 ± 0.00	46.33 ± 2.57 b	8.47 ± 2.86 l	232.86 ± 7.96 a	28.07 ± 2.55 a	96.55 ± 4.78 a	24.60 ± 1.81 d	11.17 ± 1.21 i	448.04 ± 14.62 a
PLA	3.02 ± 0.48 k	7.07 ± 1.15 j	15.71 ± 1.60 j	35.07 ± 3.32 k	2.70 ± 0.50 i	42.03 ± 1.7 1 e	14.28 ± 1.06 fg	20.84 ± 2.07 ef	140.72 ± 5.52 i
PMC	11.12 ± 1.18 e	74.20 ± 4.96 a	41.62 ± 2.31 f	97.09 ± 8.03 c	21.90 ± 1.51 c	66.38 ± 4.29 c	29.16 ± 2.65 c	51.71 ± 2.85 a	393.16 ± 25.05 b
SO	13.16 ± 1.67 d	6.51 ± 1.29 jk	29.44 ± 3.11 g	19.00 ± 1.28 n	27.46 ± 1.46 b	19.58 ± 0.82 i	10.48 ± 1.04 j	11.99 ± 0.58 hi	137.61 ± 4.60 i
SVR	0.93 ± 0.85 m	8.97 ± 0.34 h	41.02 ± 1.80 f	35.66 ± 1.54 jk	3.64 ± 0.77 h	13.45 ± 0.91 l	12.43 ± 0.74 i	12.53 ± 0.67 h	128.63 ± 5.75 j
TR	14.17 ± 1.44 b	11.84 ± 3.23 f	97.54 ± 2.87 a	54.36 ± 1.75 f	3.32 ± 0.50 h	18.79 ± 0.94 ij	25.30 ± 1.59 d	24.62 ± 1.31 d	249.94 ± 9.42 e
TRB	3.98 ± 2.57 j	7.05 ± 1.87 j	67.30 ± 8.02 c	71.26 ± 6.39 d	7.55 ± 0.54 e	26.62 ± 2.80 h	15.31 ± 1.12 f	17.36 ± 1.29 g	216.44 ± 13.82 g
TRI	0.00 ± 0.00	6.04 ± 0.42 k	2.10 ± 0.21 m	39.41 ± 1.20 i	3.48 ± 0.38 h	17.25 ± 0.72 jk	11.90 ± 0.51 i	12.19 ± 0.52 hi	92.36 ± 3.44 m
VRA	6.60 ± 1.58 f	12.29 ± 1.39 f	73.08 ± 2.49 b	17.73 ± 1.63 n	3.36 ± 0.52 h	12.28 ± 4.46 lm	0.00 ± 0.00	17.52 ± 1.11 g	142.85 ± 7.31 i
ZAD	13.52 ± 1.13 c	6.33 ± 0.34 k	65.05 ± 3.55 d	37.63 ± 1.49 ij	1.88 ± 0.95 j	19.37 ± 2.93 i	9.40 ± 0.91 j	21.32 ± 1.18 e	174.50 ± 6.30 h
p value	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001

* Means with different superscript letters in the same row differ significantly (p ≤ 0.05).

, the mean values of analyzed polyphenolic compounds from three consecutive years are presented, while in Supplementary Table S1, the mean values of analyzed polyphenolic compounds for each year can be found.

3.1. Anthocyanins

The total anthocyanin concentration in berry skin and its relative profile during three-year research, together with the comparison of mean values by Duncan’s multiple range test, is presented in Table 1.

Significant differences between varieties in the content of all investigated monoglucoside anthocyanins were detected. Plavac mali crni (10,414.06 mg kg⁻¹ d.w. of grape skin), Trnjak (10,041.69 mg kg⁻¹), Zadarka (8783.91 mg kg⁻¹), and Dobričić (8537.81 mg kg⁻¹) showed the highest content of total anthocyanins among investigated varieties. These values are very similar to the values obtained for Merlot [17]. A very low content of total anthocyanins has been detected in varieties Trišnjavac, Pošip crni, Gustopupica, and Lasina, 19.58 mg kg⁻¹, 153.85 mg kg⁻¹, 247.02 mg kg⁻¹, and 539.61 mg kg⁻¹, respectively. Such content of total anthocyanins in these varieties is expected, as Trišnjavac, Pošip crni and Gustopupica have a pink color of berry skin. Pinot gris is one of the most cultivated pink color grape variety. Ferreira et al. [18] determined a very low content of total anthocyanins in the above-mentioned grape variety. This amount is very close to the amount determined in Croatian pink color grape varieties. Overall, the obtained results suggest that the genotype has the major impact on the relative content of individual anthocyanins, which is in a close agreement with previous research [19,20].

The most abundant anthocyanin in investigated varieties is malvidin-3-O-glucoside, which is represented from 32.20% (Pošip crni) to 93.89% (Svrdlovina) of total anthocyanins (Table 1.). These results are in line with the previous findings [20]. In general, malvidin derivative forms are stabile molecules that give stability to the wine color during winemaking due to these molecules resistance to oxidation [21]. The distribution of other anthocyanins is not uniform and it depends on a number of factors, such as genotype, environmental conditions, climate, soil type, etc. [22]. Petunidin-3-O-glucoside was the second abundant anthocyanin in most of the investigated varieties. However, in Babić, Dobričić and Plavac mali, delphinidin-3-O-glucoside was the second most abundant anthocyanin, 13.47%, 15.73%, and 10.22%, respectively. Similar findings were observed earlier with Merlot (18.68%), Zinfandel (12.40%) [17], Cabernet Sauvignon (32.64%) [23,24], and Syrah (16%) [25]. Petunidin- and delphinidin-3-O-glucoside percentages were similar within the same varieties and varied among other varieties. Lasina and Trišnjavac are varieties rich with cyanidin-3-O-glucoside and peonidin-3-O-glucoside. Cyanidin-3-O-glucoside, as a minor anthocyanidin, varies from Svrdlovina (0.00%) to Lasina with 12.99%. It was expected to find low shares of cyanidin-3-O-glucoside, since this anthocyanin is a precursor of all others [20,26].

3.2. Flavonols

Grape berries contain almost exclusively flavonol glycosides [27]. Flavonols detected in the present study include myricetin-3-O-glucosde, rutin, quercetin-3-O-galactoside, quercetin-3-O-glucoside, kaempferol-3-O-glucoside, and isorhamnetin-3-O-glucoside. The total flavonol content in berry skin and its relative profile during three-year research, together with the comparison of mean values by Duncan’s multiple range test, is represented in Table 2.

Significant differences between varieties in content of all investigated flavonols glycosides were detected. Pošip crni (1742.08 mg kg⁻¹), Plavac mali (1526.32 mg kg⁻¹), Zadarka (661.46 mg kg⁻¹), and Trnjak (624.13 mg kg⁻¹) showed the highest amount of total flavonols among investigated varieties. A very low amount of total flavonols has been detected in varieties Crljenak viški, Babić, and Plavina, 219.49 mg kg⁻¹, and 249.56 mg kg⁻¹, respectively. These observations were in close agreement with previously findings. The determined total content of flavonol glycosides is in the range of 232.2 mg kg⁻¹ for Vinhao [28] up to 1542.83 mg kg⁻¹ for Syrah [25].

Looking at the pattern of flavonols, the main aglycon in the investigated varieties by far was quercetin (quercetin-3-O-galactoside, quercetin-3-O-glucoside, rutin). The portion of quercetin derivatives in the total flavonols content determined in Croatian varieties was in the range from 29.92% (Svrdlovina) up to 93.70% (Trišnjavac). These results are in close agreement with previous findings. During analysis of 64 red grape varieties, Mattivi, et al. [29] determined a range of quercetin derivatives portion from 12.34% up to 87.76%. All Croatian grape varieties that were analyzed contained quercetin-3-O-glucoside as a dominant flavonol glycoside. Quercetin-3-O-galactoside was not identified in Crljenak viški, Dobričić, Ninčuša, Tribidrag, and Vranac. Many studies concerning the content and composition of phenolics in grapes did not find rutin in high concentrations. Nicoletti, et al.’s [30] investigation of content of flavonols in ten grape varieties showed that two of them did not contain rutin (Malvasia Nera and Alphonse Lavalle). The second most abundant flavonol in the most of the examined grape varieties was myricetin-3-O-glucoside (Dobričić 38.77%, Trnjak 29.48%, Svrdlovina 28.15%). The small percentage of this flavonol was observed in pink color grape varieties, Trisnjevac (4.56%) and Pošip crni (1.24%). When it comes to red grape varieties, myricetin derivatives can be present in range from 2.35% up to 81.61% [29]. Third most abundant flavonol was isorhametin 3-O-glucoside. Levels ranged from 14.35 mg kg⁻¹ (Lasina) to 129.98 mg kg⁻¹ (Trnjak). These values were like those obtained in Verdelho (21 mg kg⁻¹) and Cabernet Sauvignon (261.25 mg kg⁻¹) [17,28]. Isorhamnetin 3-O-glucoside was not identified in Trisnjavac. Moreover, low levels of isorhametin 3-O-glucoside were detected in other varieties with pink color of berries, such as Pošip crni and Gustopupica. Some authors have showed that myricetin and isorhamnetin glycosides are specific to the red grape varieties [31], and the obtained results for pink color grape varieties agreed with their genotype. The levels of kapmpferol-3-O-glucoside ranged from 2.25 mg kg⁻¹ (Babica) to 114.82 mg kg⁻¹ (Plavac mali crni) with the average percentage of 4.52%. The highest portion was determined in Pošip crni (12.24%). This flavonol was not detected identified in Crljenak viški, Dobričić, Lasina, Tribidrag, and Svrdlovina. Jin, et al. [17] did not identify kaempferol-3-O-glucoside in Zinfandel. The obtained results agree with earlier studies, which determined that kaempferol derivatives are minor flavonols with a percentage range from 0% to 17.52% [29].

3.3. Flavanols

Flavanols detected in this study included monomers epicatechin-3-O-gallate, galocatechin, epigallocatechin, catechin, as well as oligomeric procyanidins B1, B2, B3, and B4. The total flavanols concentration in berry skin and its relative profile during the three-year research, together with the comparison of mean values by Duncan’s multiple range test, is presented in Table 3.

Significant differences between varieties in content of all investigated flavanols were detected. Pošip crni (448.04 mg kg⁻¹), Ljutun (346.32 mg kg⁻¹) and Babica (280.88 mg kg⁻¹) showed the highest amount of total flavanols among investigated varieties. Very low amounts of total flavanols have been detected in varieties Glavinuša, Trišnjavac, and Ninčuša, with 87.88 mg kg⁻¹, 92.36 mg kg⁻¹, and 103.54 mg kg⁻¹, respectively.

The most abundant monomeric flavanol in investigated varieties is epigallocatechin, which is represented from 51.97% (Pošip crni) to 12.41% (Vranac) of total flavanols. In other studies, the high portion of this compound was observed. Mattivi et al. [32] determined that the epigallocatechin in Marzemino represents 47.27% of total flavanols content, while in Albarino it represents 6.86% of total flavanol content [33]. The highest content of epigallocatechin was found in Pošip crni (232.86 mg kg⁻¹) and the lowest Vranac (17.73 mg kg⁻¹). Catechin was the second most abundant monomeric flavanol in investigated varieties represented from 7.52% (Trnjak) to 26.19% (Babica) of total flavanols. In general, catechin is one of the most abundant flavonol in grape skins. Its content strongly depends on genotype and environmental conditions, climate, soil type, etc. The portion of catechin can be from 18.34% [34] up to 80% [35]. Pošip crni had the highest content of catechin (96.55 mg kg⁻¹) as opposed to Ninčuša with the lowest content (9.97 mg kg⁻¹). Other monomeric flavanols epicatechin-3-O-gallate and gallocatechin accounted for 3.30% and 7.04% of total flavanols, respectively. These two monomeric flavanols are, in general, minor flavanols [28,36]. Procyanidin B1 was the most abundant oligomeric flavanol in investigated varieties representing from 51.97% (Pošip crni) to 12.41% (Vranac) of total flavanols. These results are in line with previous findings [37,38]. Trnjak has the highest content of procyanidin B1 (97.54 mg kg⁻¹). Procyanidin B1 has not been detected in Glavinuša. Di Lecce, et al. [33] did not observe procyanidin B1 in Albarino. Procyanidins B2, B3, and B4 accounted 10.36%, 3.57%, and 9.66%, respectively, of total flavanols in all investigated varieties. These findings are in close agreement with the previously published results [39,40].

3.4. PCA Analysis

To evaluate total variability in phenolic profiles of 20 varieties, PCA was performed. The results of PCA are shown in Figure 1, which presents the distribution of the varieties in the space defined with the first two canonical factors (explaining 64.80% of variability). The figure also shows the vectors, which explain the direction of variables in the same space. Most of the varieties were located near x axis in the first and third quadrant. The varieties that separated from the rest were Plavac mali, Pošip crni, Ljutun, Dobričić, Trnjak, Vranac, and Zadarka. Varieties Plavac mali, Pošip crni, and Ljutun appear to be separated by higher content of flavanols and flavonols, especially compounds quercetin-3-O-glucoside, epigallocatechin, kaempferol-3-O-glucoside, procyanidins B2, B3, and B4. On the other hand, varieties Dobričić, Trnjak, Zadarka, and Vranac were separated by higher content of anthocyanins malvidin-, petunidin-, peonidin-, cyanidin-, and delphinidin-3-O-glucoside. The results suggest that all analyzed classes of polyphenolic compounds contribute to the differentiation of varieties. This corresponds to the findings of Makris et al. [41] who reported that flavanols and anthocyanins have a major influence on the differentiation of varieties. However, the authors report that flavonols have a rather minor impact, which is not in accordance with our results. Moreover, the biosynthesis of flavonols is also under strict genetic control due to its close relationship to the biosynthesis of anthocyanins [42]. Thus, the genotype has an influence on the flavonol profile, which also contributed to the differentiation of Dalmatian varieties.

4. Conclusions

This study determined the phenolic profiles of anthocyanins, flavonols, and flavanols of twenty red wine grape varieties grown in Dalmatia. Significant differences were found in the analyzed compounds verifying their importance in the varietal classification and characterization. The phenolic profiles of Pošip crni, Ljutun, Zadarka, Dobričić, Plavac mali crni and Trnjak differed from the other investigated varieties, as they have their own and unique phenolic fingerprint. This study accentuates the importance of valuing and recovering minor grape varieties that are in danger of extinction. In addition, recognizing the importance of these varieties could prove beneficial from an economic standpoint by enabling competitiveness within the fast-evolving wine industry by adding something new to the market. This ought to happen through continuous efforts in researching and further winemaking process of these autochthonous varieties.