Aroma Characterization of Cabernet Sauvignon Wine from the Plateau of Yunnan (China) with Different Altitudes Using SPME-GC/MS

Abstract

The aroma characteristics of Cabernet Sauvignon wines from five different altitudes (Xidang-2110, Dari-2249, Jiunongding-2330, Adong-2610, and Jiangpo-2788, respectively) in the plateau of Yunnan (China) were detected by headspace solid phase micro-extraction gas chromatography mass spectrometry. The results showed that the numbers of volatile compounds increased with altitude rising, while the concentration of the total volatiles decreased with the exception of Xidang-2110. Alcohol and esters accounted for more than 97% of the total volatile compounds. Ten out of thirty-six components (22%) identified and quantified in five wines were found at concentrations higher than their corresponding threshold values (OAVs > 1). Ethyl hexanoate, ethyl butyrate, and 1-octen-3-ol were the three most powerful odorants in wine from Xidang-2110; but odor activity values and relative odor contribution of 1-octen-3-ol were higher than ethyl hexanoate and ethyl butyrate in the wines from Dari-2249 and Adong-2610; meanwhile β-damascenone, ethyl hexanoate, and 1-octen-3-ol were the three most powerful odorants in wines from Jiunongding-2330 and Jiangpo-2788.

Keywords:

• Wine
• Aroma compounds
• Altitude
• HS-SPME
• GC/MS

INTRODUCTION

Aroma is one of the important sensory characteristics of wines, which influences the consumers’ overall acceptability.^[1−3] It has been reported that more than 800 volatile compounds have been found in grapes and wines, with a wide concentration range varying from hundreds of mg/L to the μg/L or ng/L level with different cultivars.^[4−6] Varietal, fermentative, and wine aging aromas are three main groups of wine flavors original.^[7] The concentration of these volatile compounds depends on grape variety, environmental conditions, cultural techniques, grape maturity state, winemaking, and aging techniques.^[1,⁸,^9]

Altitude can strongly affect the climatic conditions since it directly impacts temperature, humidity, and other environmental factors that affect grape maturation.^[10] It has been reported that the sunlight-exposed fruits were generally greater in soluble solids, anthocyanins, and phenolics and lower in titratable acidity, juice pH,^[11] while vine shading caused a decrease in the contents of glycosides of terpenols, phenols, and C13-norisoprenoids in berries.^[12] Low altitude was ideal for biosynthesis of catechin monomers, such as (+)-catechin, (-)-epicatechin, (-)-epicatechin gallate, procyanidin dimers, trimer C1, and total extractable proanthocyanidins;^[13] and higher altitude vineyard appeared to have larger amounts of total anthocyanidin monoglucosides in grapes and wines.^[10] Little literature has been reported on the influences of vineyard altitude on the monoterpenes and norisoprenoids.^[14,^15] Falcao et al.^[15] indicated that vineyard altitude had significant regression with 2-methoxy-3-isobutylpyrazine concentrations with no significant relation with α- and β-ionone and β-damascenone; at the same time, wines from higher altitudes had a “bell pepper” aroma, while wines from lower altitudes were correlated with “red fruits” and “jam” aromas.

The wines produced from many different climatic and geographically delineated areas of China should have a range of styles and a diverse expression of varietal characters. However, sensory data of Chinese wine are scarce, especially for those with a denomination of origin.^[6] So many areas are suitable for vine-growing in China, such as Qilian zone (at the edge of the desert), Helan Zone of Ningxia (beyond the Helan mountains), Yantai of Shandong and Changli of Hebei (on the seaside), etc. All these vine-growing regions are unique in ecological conditions, and allow the product to be typical, meaning that the product is representative of its terroir. Because specific climatic conditions (warm-arid climate; big temperature difference between daytime and nighttime; annual sunshine time: 1987 h; annual rainfall: 300–600 mm) were suitable for high quality grape growing, Cabernet Sauvignon was introduced into this area in the early 21st century. One hundred and sixty-five years ago, grape seeds, mainly rose honey and French wild, were brought to this area by missionaries from France.^[16] However, to date no research has reported about the sensory characteristics of chemical composition of Cabernet Sauvignon wines in this region. So the present work aimed to find the differences in the aroma compounds of wines from different cultivation altitudes of the plateau of Yunnan by headspace solid phase micro-extraction gas chromatography mass spectrometry (HS-SPME-GC/MS), so as to elucidate the aroma characteristics of wines from different altitudes and provide valuable information for producing high-quality wine in this area.

MATERIALS AND METHODS

Vineyard Conditions and Vinification

The five vineyards are located in the Deqin County of Yunnan Province, China, a dry-hot valley zones at the altitude of 1900–4229 m. All the vines were cultivated in 2003, with bilateral cordon training and 2.0 × 1.0 m (row × vine) spacing. The experiment was carried out in 2011. All wines were fermented in 50 L stainless steel tanks. Grape berries of Cabernet Sauvignon were harvested manually and transported using ice box after harvesting. The detailed geographical information and sample time of each vineyard were presented in Table 1. Briefly, the grapes were crushed, destemmed, and placed separately in stainless steel tanks immediately after, the grape was transported to wine factory. Fifty mg/L sulfur dioxide and 25 mg/L pectinase (LALLZYME EX, Lallemand Company, France) were added to the musts and the contents were mixed by hand. After maceration of the musts for 24 h, 200 mg/L of activated dry yeast RC 212 (Lallemand Company, Denmark) was added to the musts. Alcoholic fermentation was carried out at 22 to 25℃ to dryness (reducing sugar <4 g/L). At the end of alcoholic fermentation the wines were separated from pomace, and added with SO₂ of 50 mg/L. Each wine was produced in three replicates. After fermentation, the wine samples were bottled and stored at 4℃ prior to analysis.

TABLE 1 The geographical information of five vineyards

Regions	Altitude (meter)	Latitude	Longitude	Sample time
Xidang	2110	28°35′35′′ N	99°10′40′′ E	2011.09.22
Dari	2249	28°29′20′′ N	98°49′21′′ E	2011.09.22
Jiunongding	2330	28°17′37′′ N	98°52′24′′ E	2011.10.02
Adong	2615	28°33′49′′ N	98°52′15′′ E	2011.10.13
Jiangpo	2788	28°40′06′′ N	98°43′47′′ E	2011.10.21

Analysis of General Composition of Wines

General composition of wines were performed according to the official methods established by the O.I.V.^[17] The parameters analyzed were residual sugar, total acidity, pH, alcohol degree, volatile acidity, and free SO₂.

Headspace Solid Phase Micro-Extraction (HS-SPME)

Aromatic compounds of the wine samples were extracted by HS-SPME and analyzed using gas chromatography/mass spectrometry (GC/MS) as described by Zhang et al.^[18] 5 mL of wine sample and 1 g NaCl were placed in a 15-mL sample vial. The vial was tightly capped with a PTFE-silicon septum and heated at 40 ℃ for 30 min on a heating platform agitation at 400 rpm. The SPME (50/30-μm DVB/Carboxen/PDMS, Supelco, Bellefonte, Pa., U.S.A.), preconditioned according to manufacturer’s instruction, was then inserted into the headspace, where extraction was allowed to occur for 30 min with continued heating and agitation by a magnetic stirrer. The fiber was subsequently desorbed in the GC injector for 30 min.

GC/MS Analysis

The GC/MS system used was an Agilent 6890 GC equipped with an Agilent 5975 mass spectrometer. The column was a 60 m × 0.25 mm HP-INNOWAX capillary with 0.25 μm film thickness (J & W Scientific, Folsom, CA, USA). The carrier gas was helium at a flow rate of 1 mL/min. Samples were injected by placing the SPME fiber at the GC inlet for 30 min in the splitless mode. The oven’s starting temperature was 50℃, held for 1 min, then raised to 220℃ at a rate of 3℃/min and held at 220℃ for 5 min. The mass spectrometer in the electron impact mode (MS/EI) at 70 eV was scanned in the range of m/z 20 to 450 U. The mass spectrometer was operated in the full scan and the selective ion mode (SIM) under autotune conditions at the same time. The area of each peak was determined by Chemstation software (Agilent Technologies). Analyses were carried out in triplicate.

Reagents

All standards were purchased from Aldrich (Milwaukee, WI, USA) and Fluka (Buchs, Switzerland). Purity of all standards was above 99%. Model solutions were prepared using the methods reported by Howard et al.^[19] 4-Methyl-2-pentanol was employed as the internal standard. For quantification, 8-point calibration curves for each compound were prepared using the method described by Ferreira et al.,^[20] which was also used as a reference to determine the concentration range of standard solutions. The regression coefficients of calibration curves were above 98%.

Odor Activity Values (OAVs) and Relative Odor Contribution (ROC)

The OAVs of each volatile compound was calculated to estimate the sensory contribution of these odorants to the general flavor grape. The OAVs were calculated by dividing the concentration of each compound in the sample by its odor threshold value of the compound in water/ethanol solution.^[21] The ROC of each aroma compound shown in parentheses and was calculated as the ratio of the OAVs of the respective compound to the total OAVs of each wine.

TABLE 2 The general composition of wines from five different altitudes

Composition	Xidang-2110	Dari-2249	Jiunongding-2330	Adong-2610	Jiangpo-2788
Residual sugar (g/L)	3.6 ± 0.4	2.0 ± 0.2	3.6 ± 0.3	2.2 ± 0.2	3.5 ± 0.3
Total acidity^a (g/L)	6.4 ± 0.04	9.5 ± 0.08	7.7 ± 0.06	9.7 ± 0.03	7.5 ± 0.04
pH	3.6 ± 0.04	3.1 ± 0.03	3.5 ± 0.03	3.0 ± 0.01	3.4 ± 0.02
Alcohol (%, v/v)	13.2 ± 0.08	12.4 ± 0.05	13.4 ± 0.04	13.4 ± 0.06	13.7 ± 0.04
Volatile acidity^b (mg/L)	0.30 ± 0.02	0.32 ± 0.03	0.31 ± 0.03	0.32 ± 0.01	0.31 ± 0.02
Free SO2 (mg/L)	18.26 ± 0.08	17.58 ± 0.06	18.34 ± 0.03	19.43 ± 0.03	18.55 ± 0.06

The data were mean values of triplicate samples (mean ± SD, n = 3);

^aexpressed as tartaric acid; ^bexpressed as acetic acid.

Statistical Analysis

Analysis of variance (ANOVA), principal component analysis (PCA) and cluster analysis were performed using the SPSS 16.0 (SPSS, Chicago, IL) for Windows with three replications of the same sample. Significant difference was calculated at 0.05 levels.

RESULTS AND DISCUSSION

General Composition of Wines from Different Altitudes

Table 2 shows the general composition of wines from different altitudes. Results showed there were no differences in residual sugar, alcohol, volatile acidity, and free SO₂ concentration of the young red wines from the five different altitudes. Wine of Xidang-2110 and Jiunongding-2330 had lower total acidity and higher pH than other areas.

Volatile Compounds of Wines from Different Altitudes

Table 3 lists the volatile compounds isolated from Cabernet Sauvignon wines made from the grapes in vineyards with different altitudes. In the wines from Xidang-2110, Dari-2249, Jiunongding-2330, Adong-2610 and Jiangpo-2788, 31, 28, 31, 33, and 35 free volatile compounds were identified and quantified, respectively. Like most studies,^[5,^6] the majority of volatiles were higher alcohols

TABLE 3 The concentration of volatile compounds in Cabernet Sauvignon wines from five different altitudes

				Concentration (μg/L)
Compounds	Retention time	Threshold (μg/L)	Odor description	Xidang-2110	Dari-2249	Jiunongding-2330	Adong-2610	Jiangpo-2788
Alcohols
1-propanol	9.41	306,000^b	Fresh, alcohol	7093.78 ± 821.218^ab	—	6539.90 ± 998.96^b	7842.98 ± 83.60^a	7167.49 ± 309.86^ab
Isobutanol	11.05	40,000^c	Fusel, alcohol	20,180.99 ± 950.77^ab	26,707.39 ± 763.28^a	16,374.69 ± 432.58^b	17,187.89 ± 18.63^ab	21,106.17 ± 563.28^ab
1-butanol	12.82	150,000^d	Medicinal, alcohol	5715.46 ± 564.91^ab	5518.39 ± 66.23ab	4750.73 ± 383.11^b	5646.75 ± 162.80^ab	6164.33 ± 11.29^a
Isoamyl alcohol	14.89	30,000^c	Cheese	193,538.98 ± 4003.29^d	446,662.77 ± 3777.17^a	271,292.73 ± 2765.86^b	272,105.44 ± 3447.73^b	238,934.03 ± 3147.08 c
1-pentanol	16.48	80,000^a	Fruity, balsamic	14,252.98 ± 2.06^b	14,389.80 ± 24.76^ab	14,282.10 ± 45.21^ab	14,526.59 ± 208.56^a	14,351.36 ± 1.13^ab
4-methyl-1-pentanol	19.04	50,000^e	−	59.27 ± 8.05^c	152.42 ± 18.33^a	89.05 ± 11.24^b	109.96 ± 2.02^b	95.17 ± 0.64^b
2-heptanol	19.20	70^f	Fruity, moldy, musty	12.04 ± 2.16^a	—	—	—	16.65 ± 3.42^a
1-hexanol	20.62	8000^c	Green, grass	1176.30 ± 103.09^d	2144.75 ± 158.84^b	1666.48 ± 57.27^c	2681.69 ± 102.44^a	1799.98 ± 172.81^bc
(E)-3-hexen-1-ol	21.16	400^c	Green, floral	trace	trace	trace	trace	245.75 ± 2.42
1-octen-3-ol	24.83	1^g	Mushroom	20.15 ± 3.21^c	106.01 ± 12.44^b	44.64 ± 4.46^c	259.12 ± 10.16^a	36.21 ± 3.01^c
1-octanol	29.12	120^c	Intense citrus, roses	231.36 ± 1.48^b	121.98 ± 3.63^c	—	333.23 ± 41.33^a	221.66 ± 0.92^b
Nonanol	33.01	58^f	Floral, green, fatty	—	26.81 ± 2.13^b	—	41.14 ± 3.62^a	—
Benzyl alcohol	41.31	10,000^g	Sweet, floral	233.32 ± 14.89^c	415.02 ± 39.41^bc	685.79 ± 27.80^b	1016.29 ± 259.64^a	542.36 ± 52.95^bc
2-phenylethanol	42.60	14,000^c	Rose, honey	177,482.08 ± 4079.41^a	64,313.19 ± 1323.09^b	28,524.58 ± 1053.99^b	26,060.12 ± 994.42^b	15,604.60 ± 983.03^b
Subtotal				419,996.71	560,558.53	344,250.69	347,811.20	306,285.76
Proportion (%)				84.37	91.63	86.55	86.77	84.42
Esters
Ethyl acetate	6.33	7500^c	Fruity, sweet	69,786.18 ± 289.76^a	36,511.28 ± 898.64^b	38,453.62 ± 688.32^b	41,562.79 ± 948.47^b	42,578.02 ± 689.44^b
Ethyl butyrate	9.11	20^h	Fruity, sweet	460.06 ± 30.24^a	485.20 ± 29.61^a	522.48 ± 48.36^a	530.97 ± 35.12^a	551.41 ± 44.37^a
Isoamyl acetate	11.61	30^c	Banana	289.57 ± 175.97^b	678.73 ± 284.26^ab	988.98 ± 170.10^a	508.50 ± 41.77^ab	602.59 ± 214.24^ab
Ethyl hexanoate	15.97	5^c	Fruity, anise	250.35 ± 58.15^a	330.38 ± 7.93^a	303.32 ± 59.64^a	377.32 ± 33.58^a	318.78 ± 68.00^a
Ethyl oenanthate	19.78	—	—	—	—	—	3.94 ± 0.65	trace
Ethyl lactate	20.46	154,636^d	Lactic, raspberry	2969.56 ± 32.43^c	4976.98 ± 243.14^a	3639.03 ± 199.26^bc	3408.21 ± 5.98^bc	3795.15 ± 584.36^b
Ethyl nonanoate	28.36	—	—	trace	17.60 ± 1,03	trace	—	trace
Ethyl decanoate	32.84	200^c	Fruity, fatty, pleasant	44.48 ± 5.54^c	—	147.35 ± 0.05^a	74.33 ± 0.91^b	71.64 ± 4.83^b
Diethyl succinate	33.93	500,000^d	Light fruity	trace	64.77 ± 1.80^c	318.82 ± 2.14^b	1752.59 ± 5.67^a	trace
Ethyl dodecanoate	40.23	1500^a	Flowery, fruity	trace	trace	8.00	trace	trace
Phenethyl acetate	39.24	250^c	Floral, honey	22.10 ± 2.53^c	48.07 ± 0.67^a	28.69 ± 1.50^bc	31.87 ± 3.13^b	25.55 ± 2.00^c
Subtotal				73,822.30	43,113.01	44,410.29	48,250.52	47,943.14
Proportion (%)				14.83	7.05	11.17	12.04	13.21
Acids
Isobutyric acid	29.70	200,000^c	Fatty	843.87 ± 112.08^b	3588.63 ± 126.35^ab	5435.96 ± 311.25^a	—	5201.42 ± 433.76^a
Isovaleric acid	33.66	—	Rancid, acidic	843.87 ± 87.08^b	2158.70 ± 7.14^a	929.77 ± 16.71^b	1883.52 ± 140.71^ab	1002.61 ± 228.17^b
Hexanoic acid	40.08	3000^c	Cheese, rancid, fatty	871.45 ± 59.42^c	1271.76 ± 83.27^a	1024.33 ± 25.84^bc	1196.27 ± 121.03^ab	1005.07 ± 27.35^bc
Octanoic acid	47.30	500^c	Rancid, harsh, cheese, fatty acid	400.90 ± 18.54^b	790.77 ± 60.14^a	770.65 ± 51.06^a	698.13 ± 90.63^ab	544.67 ± 40.69^ab
Decanoic acid	53.80	15,000^c	Fatty, unpleasant	127.06 ± 9.38^c	227.66 ± 10.88^a	191.73 ± 4.15^b	189.73 ± 21.75^b	162.32 ± 3.84^b
Subtotal				3087.15	8037.52	8352.44	3967.65	7916.09
Proportion (%)				0.62	1.31	2.1	0.99	2.18
Aldehydes and ketones
Furfural	25.70	14,100^c	Pungent	776.48 ± 21.23^a	—	735.04 ± 13.68^a	710.92 ± 9.85^a	685.04 ± 23.36
Benzaldehyde	28.14	2000^b	Fruity, berry	137.07 ± 3.14	—	trace	trace	trace
Subtotal				913.55	0	735.04	710.92	685.04
Proportion (%)				0.18	0	0.18	0.18	0.19
Terpenes and Norisoprenoids
4-terpineol	30.97	250^h	Sweet, herbaceous	—	—	—	53.69 ± 1.23^a	46.69 ± 2.45^a
α-terpineol	34.71	250^h	Floral, flower, peach	—	46.36 ± 0.03^a	—	55.05 ± 7.51^a	—
β-damascenone	39.26	0.05^c	Sweet, floral	—	—	4.28 ± 0.36^a	—	3.90 ± 0.21^a
Geranylacetone	39.49	60^a	Floral	—	—	2.32 ± 0.83^a	5.21 ± 1.21^a	—
Subtotal				0	46.36	6.6	113.95	50.59
Proportion (%)				0	0.008	0.002	0.028	0.014
Total				497,819.71	611,755.42	397,755.06	400,854.24	362,808.98

The data were mean values of triplicate samples (mean ± SD, n = 3). “–”, not determined. ^a[24]; ^b[25]; ^c[26]; ^d[27]; ^e[28]; ^f[29]; ^g[30]; ^h[20]; ⁱ[31].

and esters, which accounted for more than 97% of the total volatile compounds. Fatty acids, aldehydes, ketones, terpenes, norisoprenoids, and volatile phenols were identified as minor compounds. According to the quantitative data, concentrations of the total volatiles from the five samples ranged from 362,808.98 (Jiangpo-2788) to 611,755.42 μg/L (Dari-2249). Xidang-2110 had the highest concentrations of esters, aldehydes, and ketones and Dari-2249 had the highest alcohols, but Jiunongding-2330 had the highest concentration of acid. It proved that the numbers of volatile compounds increased with altitude rising, while concentrations of the total volatiles decreased (except for Xidang-2110). The factors that influence aroma containing grape variety, management practices, winemaking techniques, and yeast were same in this study. With altitude rising, temperature generally decreased and light intensity increased.^[13,^22] The research of Olusola Lamikanra et al.^[23] on the cantaloupe melon showed that, with fruit exposed to UV light, the content of most of the aliphatic esters decreased by over 60% of the amounts present in the corresponding fresh cut fruit. Therefore, the influence of vineyard altitude on wine aroma can be attributed to the effects of local climate on berry composition.

Alcohols

Alcohols, accounting for 84.37–91.63% of the aromatic content of these wines, were the largest group of free volatile compounds in Cabernet Sauvignon wines made from grapes in vineyards with different altitudes (Table 3). This is consistent with previous literature.^[5,³²,^33] In the wine from Xidang-2110, Dari-2249, Jiunongding-2330, Adong-2610, and Jiangpo-2788, 13, 12, 11, 13, and 13 alcohols were identified and quantified, respectively. The subtotal concentration of alcohols ranged from 306,285.76 (Jiangpo-2788) to 560,558.53 μg/L (Dari-2249). The alcohols, produced during alcoholic fermentation, play an important role in the flavor of wines, depending on the types of compounds and their concentrations.^[34−36] Isoamyl alcohol, isobutanol, 1-pentanol were among the higher concentrations in the five wines from different altitudes, in accordance with previous studies.^[6,^32] These compounds, recognized by their strong and pungent smell and taste, are related to herbaceous notes.^[5] At concentrations below 300 mg/L, they certainly contributed to the desirable complexity of the wine, but when their concentration goes beyond 400 mg/L, the higher alcohol concentrations shows a negative effect on quality.^[5,^35] In this study, all the alcohols concentrations, except isoamyl alcohol, were below 300 mg/L. In addition, 2-phenylethanol was a high concentration benzenic compound, as shown in Table 3. This compound is formed principally by the metabolism of yeast and provides the aroma of flowers with scents of rose.^[5,^37] The concentration of 2-phenylethanol decreased with altitude rising.

Esters and Acids

Esters have long been considered important contributors to wine aroma, because they exhibit fruity odors similar to those often used to describe wines.^[5] As can be seen in Table 3, the wines from the five different altitudes contained esters ranging from 43,113.01 (Dari-2249) to 73,822.30 μg/L (Xidang-2110), which accounted for 7.05–14.83% of the total aroma compounds detected. Like previous studies in Shaanxi, Shanxi, Hebei of China,^[6,^32] ethyl acetate, ethyl lactate, isoamyl acetate, and ethyl butanoate are the four esters of higher concentrations, with Xidang-2110 having the highest concentrations of ethyl acetate, and Dari-2249 having the highest ethyl lactate and phenethyl acetate, while Jiunongding-2330 having the highest isoamyl acetate, ethyl decanoate, and diethyl succinate. Although the contents of ethyl nonanoate, ethyl dodecanoate were trace quantities in some wines, they caused wine aroma more abundant and complex.^[24]

Fatty acids have been described with fruity, cheesy, fatty, and rancid notes.^[38,^39] The contents of fatty acid are determined by the initial composition of the must and fermentation conditions.^[32,^40] In this study, five acids, including isobutyric acid, isovaleric acid, hexanoic acid, octanoic acid, and decanoic acid, were identified in wine samples. The concentration of fatty acids detected in the sample wines were between 3087.15 and 8352.44 μg/L, accounting for 0.62–2.10% of total aroma compounds. Except for the fact that the isobutyric acid in Jiunongding-2230 was of the highest content, other acids were more in Dari-2249 wine than those of other wines sampled. All the concentrations of C6 to C10 fatty acids were lower than 4 mg/L, giving positive effect on the global aroma quality. According to Shinohara’s study,^[41] the C6 to C10 fatty acids at concentrations of 4 to 10 mg/L impart mild and pleasant aroma to wine; however, at levels beyond 20 mg/L, their impact on wine becomes negative. Because the concentrations of C6 to C10 fatty acids were all far below 4 mg/L in the current study, they have not a significant impact on the aroma of the five wines examined.

Terpenes and Norisoprenoids, Aldehydes and Ketones, and Others

Terpene compounds have a low olfactory threshold and are generally associated with floral and citric aromas;^[26] they are not affected by yeast metabolism during fermentation and hence a good indicator for the variety and quality of grape.^[42] 4-terpineol was detected in wines from Adong-2610 and Jiangpo-2788, and this compound was not detected in other regions of China.^[6,⁷,^43] α-terpineol was detected in wines from Dari-2249 and Adong-2610. These compounds were all found in low concentrations and not beyond their odor thresholds, as expected for a neutral variety, and were not present in the free fraction of wines.^[5] Norisoprenoids are volatile compounds that may come from the direct degradation of carotenoid molecules, such as β-carotene, lutein, neoxanthin, and violaxanthin.^[5] β-damascenone was detected in wines from Jiunongding-2330 and Jiangpo-2788; geranylacetone was found in wines from Jiunongding-2330 and Adong-2610. Although these compounds have a very low odor threshold and low concentration was found in all the wines detected, β-damascenone (0.05 μg/L) exceeded its odor threshold in wines from Jiunongding-2330 and Jiangpo-2788. This compound is considered a positive contributor to wine aroma.^[44] As shown in Table 3, no aldehydes and ketones were detected in Dari-2249. Two aldehydes and ketones detected from four other wines were furfural and benzaldehyde. The concentration of furfural was higher than benzaldehyde, and no significant difference was among four wines.

TABLE 4 The OAVs and ROC of the volatile compounds in wines from five different altitudes

				OAVs (ROC)
Compounds	Retention time	Threshold (μg/L)	Odor description	Xidang-2110	Dari-2249	Jiunongding-2330	Adong-2610	Jiangpo-2788
Ethyl acetate	6.33	7500^c	Fruity, sweet	9.30 (6.94%)	4.87 (1.98%)	5.13 (1.91%)	5.54 (1.39%)	5.68 (2.33%)
Ethyl butyrate	9.11	20^h	Fruity, sweet	23.00 (17.16%)	24.26 (9.86%)	26.12 (9.76%)	26.55 (6.66%)	27.57 (11.33%)
Isoamyl acetate	11.61	3^c	Banana	9.65 (7.20%)	22.62 (9.20%)	32.97 (12.31%)	16.95 (4.25%)	20.09 (8.26%)
Isoamyl alcohol	14.89	30,000^c	Cheese	6.45 (4.81%)	14.89 (6.05%)	9.04 (3.38%)	9.07 (2.27%)	7.96 (3.27%)
Ethyl hexanoate	15.97	5^c	Fruity, anise	50.07 (37.35%)	66.08 (26.87%)	60.66 (22.66%)	75.46 (18.93%)	63.76 (26.20%)
1-octen-3-ol	24.83	1^g	Mushroom	20.15 (15.03%)	106.01 (43.11%)	44.64 (16.67%)	259.12 (64.99%)	36.21 (14.88%)
1-octanol	29.12	120^c	Intense citrus, roses	1.93 (1.44%)	1.02 (0.41%)	−	2.78 (0.70%)	1.85 (0.76%)
β-damascenone	39.26	0.05^c	Sweet, floral	−	−	85.60 (31.97%)	−	78.00 (32.06%)
2-phenylethanol	42.60	14,000^c	Rose, honey	12.68 (9.46%)	4.59 (1.87%)	2.04 (0.76%)	1.86 (0.47)	1.11 (0.46%)
Octanoic acid	47.30	500^c	Rancid, harsh, cheese	0.80 (0.60%)	1.58 (0.64%)	1.54 (0.58%)	1.40 (0.35%)	1.09 (0.45%)

^c[26]; ^g[30]; ^h[20].

OAVs

The computation of OAVs and ROC are useful indexes for determining the important aromatic components in a complex system.^[43] In this study, the odor threshold values were taken from information available in the literature, as shown in Tables 3 and 4. According to Guth,^[38] those present at concentrations higher than their odor threshold (OAVs > 1) are mainly considered as active odorants. As can be seen in Table 4, 10 out of 36 components (22%) identified and quantified in five wines were found at concentrations higher than their corresponding threshold values (OAVs > 1). Therefore, only a few volatile compounds were potentially active odorants. The most important active odorants of five wines were 1-octen-3-ol, ethyl buyrate, isoamyl acetate, ethyl hexanoate, β-damascenone. The OAVs and ROC values of aroma compounds in five wines were different with different altitudes. They can be divided into three groups. The first was Xidang-2110, whose three most powerful odorants were ethyl hexanoate (37.35%), ethyl butyrate (17.16%), and 1-octen-3-ol (15.03%). They gave the wine fruity, anise, sweet, and mushroom odors. The second group was Dari-2249 and Adong-2610, 1-octen-3-ol (31.97 and 32.06%), ethyl hexanoate (26.87 and 18.93%), and ethyl butyrate (9.86 and 6.66%) were three most powerful odorants. β-damascenone (31.97 and 32.06%), ethyl hexanoate (22.66 and 26.20%), and 1-octen-3-ol (16.67 and 14.88%) were the three most powerful odorants (sweet, floral, fruity, anise, and mushroom) in wines from Jiunongding-2330 and Jiangpo-2788. In general, norisoprenoids detected in negligible amounts or not found in the wine free fraction, were relatively abundant in the bound fraction, but β-damascenone detected in the free fraction in wines was not detected in the bound form.^[5] 1-octanol, 2-phenylethanol (except for Dari-2249), octanoic acid each accounted for less than 1% of the global aroma perceptions of wines from Dari-2249, Jiunongding-2330, Adong-2610, and Jiangpo-2788. According to Escudero and others,^[45] even if they were present at a concentration higher than their threshold values, compounds such as fusel alcohols, acids, esters, and volatile phenols were not able to affect individually the flavor of the wines. The OAVs and ROC of 1-octen-3-ol in wines from Dari-2249 and Adong-2610 were dominated; this is inconsistent with the results in previous studies ^[5,³²,^43] showing that ethyl octanoate, ethyl hexanoate, and isoamyl acetate were the predominant odorants.

PCA

In order to reveal the influence of altitudes on aromatic constituents, PCA was performed. Figure 1 shows the loading plot on the plane defined by the first and second principal components and the corresponding projection of samples. The first principal component (PC1) describes 43.54% of the variability in the data and the second (PC2) describes 27.23%. PC1 correlates highly with α-terpineol, hexanoic acid, isovaleric acid, octanoic acid, decanoic acid, 4-methyl-1-pentanol, while PC2 correlates highly with ethyl acetate, geranylacetone, diethyl succinate, 4-terpineol. The five regions with different altitudes could be clearly divided into four groups according to their relationships between scores and their aromatic constituents. The first group was composed of Xidang-2110 with the higher negative value of PC2, since its higher values of ethyl acetate and 2-phenylethanol. The second group was Jiangpo-2788 and Jiunongding-2330, which could be characterized by 1-butanol and β-damascenone. The third group was composed of Adong-2610 with the higher positive value of PC2. This group was characterized by geranylacetone, diethyl succinate and 4-terpineol. However, Dari-2249 was located at the highest values of positive PC1 and negative PC2 owing its higher values of phenethyl acetate and isoamyl alcohol.

FIGURE 1 Distribution of wine samples and loadings of volatile compounds in the first two PCs of volatile compounds.

Cluster Analysis

To better elucidate aroma characteristics of the wines from different altitudes, cluster analysis using Ward’s method was carried out on these identified volatile compounds. As shown in Fig. 2, the five wines can be divided into four groups: the first group was Xidang-2110, the second group contained Jiunongding-2330 and Jiangpo-2788, the third was Adong-2610, and the last was Dari-2249. The result of cluster analysis made the same group with PCA (Fig. 1a). It suggested the vineyards in the same group had similar aroma characterizations.

FIGURE 2 Cluster analyses of five wines from different altitudes.

CONCLUSIONS

Altitudes had an obvious influence on aroma compounds of wine. In the wines from Xidang-2110, Dari-2249, Jiunongding-2330, Adong-2610 and Jiangpo-2788, 31, 28, 31, 33, and 35 free volatile compounds were identified and quantified, respectively. The majority of volatiles were higher alcohols, esters. Ten out of 36 components (22%) identified and quantified in five wines were found at concentrations higher than their corresponding threshold values (OAVs > 1). Ethyl hexanoate, ethyl butyrate, and 1-octen-3-ol gave the wine of Xidang-2110 fruity, anise, sweet, mushroom odors. 1-octen-3-ol, ethyl hexanoate, and ethyl butyrate (with mushroom, fruity, anise, and sweet odors) were three most powerful odorants of wine from Dari-2249 and Adong-2610. β-damascenone, ethyl hexanoate, and 1-octen-3-ol were three most powerful odorants (sweet, floral, fruity, anise, and mushroom) in wines from Jiunongding-2330 and Jiangpo-2788. It showed that the numbers of volatile compounds increased with altitude rising, while concentrations of the total volatiles were decreased (except for Xidang-2110).

ACKNOWLEDGMENTS

The author’s would like to acknowledge the assistance from Ke-Xu Cui, Jian-Hong Cao, and Jia-Kui Wang of Shangri-La Winery Company Limited for providing the grape samples. The authors thank the Center for Viticulture and Enology, China Agricultural University for technical assistance in the completion of the GC–MS experiments. The authors thank Miss Hui-Fang Du for her critical reading of the manuscript.

FUNDING

This research was supported by China Agriculture Research System for Grape Industry (CARS-30-zp-9).

Additional information

Funding

This research was supported by China Agriculture Research System for Grape Industry (CARS-30-zp-9).