Skip to main content
NCBI home page
Journal of Food Science and Technology logoLink to Journal of Food Science and Technology
. 2021 Mar 7;59(2):655–665. doi: 10.1007/s13197-021-05056-x

Influence of keeping the calyces during fermentation on physicochemical and sensory properties of Hibiscus sabdariffa wines

Ana Luisa Gutiérrez-Salomón 1, Jahir Antonio Barajas-Ramírez 2,✉, Victoria Guadalupe Aguilar-Raymundo 2, Araceli Castañeda-Ovando 3
PMCID: PMC8814298  PMID: 35185184

Abstract

Hibiscus (Hibiscus sabdariffa) wines are usually made through the fermentation of its calyces extracts. However, the exhausted calyces could still contain a considerable amount of extractable bioactive compounds. This work analyzed the effect of keeping the hibiscus calyces during fermentation on the physicochemical and sensory characteristics and acceptance of its wines. Hibiscus wines Filtered-Ch and Filtered-C were generated by fermentation of filtered musts of China and Colima varieties, respectively, while Unfiltered-Ch and Unfiltered-C were obtained by fermentation of musts from calyces that were kept for 120 days. Unfiltered-C and Unfiltered-Ch wines had higher contents of total monomeric anthocyanins (32.5 mg/L in Unfiltered-C and 48.5 mg/L in Unfiltered-Ch), condensed tannins (around 200 mgCE/L), total phenolic compounds (800 mgGAE/L), and antioxidant activity (8.4–8.8 mMTE/L), as well as a higher concentration of tartaric, citric and malic acids (0.26, 0.32 and 2.25 g/L, respectively) than Filtered-C and Filtered-Ch wines (P < 0.05). Moreover, Unfiltered-C and Unfiltered-Ch wines had darker red colors, evidenced by lower lightness, chroma and hue values, than those observed in Filtered-C and Filtered-Ch wines (P < 0.01). Furthermore, Unfiltered-Ch and Unfiltered-C wines were more appreciated, having a more intense taste and smell according to the descriptions obtained from consumers by applying the Check-All-That-Apply technique. Sensory attributes observed in hibiscus wines were statistically correlated with the physicochemical characteristics. Keeping hibiscus calyces during fermentation allows for the production of wines with acceptable sensory characteristics and a higher concentration of bioactive compounds than producing wines from filtered musts.

Keywords: Hibiscus sabdariffa, Fermentation, Physicochemical characteristics, Sensory analysis

Introduction

In general terms, wine is referred to as the fermented juice of grapes. However, wine can also be made from fruits and vegetables, and these beverages have shown such remarkable sensory and chemical properties that these wines can be classified as functional foods (Jackson 2017; Joshi et al. 2017). Hibiscus (Hibiscus sabdariffa L.) calyces have been used to produce an herbal wine with desirable sensory (Ifie et al. 2012; Tiwari et al. 2017) and physicochemical characteristics, as well as having a high bioactive compound content (Ifie et al. 2016).

The most reported process of producing wine from hibiscus calyces involves three steps: (1) The extraction of chemical compounds, (2) the separation of calyces from must, and (3) the fermentation step (Ifie et al. 2012; Tiwari et al. 2017). However, this process has a disadvantage, as it has been demonstrated that the separated calyces are still a good source of antioxidant fiber, which is linked to phenolic compounds with antioxidant activity (Sáyago-Ayerdi et al. 2014). The extraction step is fundamental to the wine production process. It has been reported that grape wine produced with a prolonged maceration time shows higher phenolic content than wines produced over a regular maceration time of up to 5 days (Alencar et al. 2018; Ivanova-Petropulos et al. 2016). The hibiscus wine has been produced keeping the calyces during fermentation (Ifie et al. 2016; Mounigan and Badrie 2007), which involves a longer time of maceration. This technique can provide important functional properties during the fermentation process as the presence of calyces during fermentation can facilitate the increased extraction of bioactive compounds in hibiscus wine as observed in grape wines. Several studies have explored sensory acceptance for hibiscus wines (Ifie et al. 2012; Tiwari et al. 2017), which can be considered as flavorful and desirable products for a segment of consumers who appreciate these characteristics (Jackson 2017). These beverages could be further enhanced by keeping the calyces during the fermentation process. The use of a novel and time-saving sensory technique such as the Check-All-That-Apply test could provide a reliable sensory description of hibiscus wines as well as determining the sensory attributes that are preferred by consumers (Jaeger et al. 2019).

To the best of our knowledge, there are no publications that compare the sensory and physicochemical characteristics of hibiscus wines obtained from the fermentation of filtered and unfiltered musts. Therefore, the aim of this work was to analyze the physicochemical characteristics and bioactive compound content of hibiscus wines made by fermentation of filtered and unfiltered musts as well as exploring the relationship with acceptance and sensory characteristics described by consumers.

Materials and methods

Materials

All solvents were of HPLC grade. Standards of organic acids, gallic acid, Folin-Ciocalteu’s reagent, 2,2-diphenyl-1-picrylhydrazyl (DPPH), 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox), sodium carbonate, catechin, methanol, vanillin, hydrochloric acid, sodium acetate and potassium chloride were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Dry hibiscus calyces of two varieties (China and Colima) were purchased from markets in Guadalajara, Jalisco, Mexico and Zamora, Michoacan, Mexico, respectively. The calyces of both varieties are characterized by their dark red color when fully mature. Moisture content in dry calyces was 8.0 ± 2.3 g/100 g in both varieties. The calyces were kept in plastic zip-lock bags at room temperature until the preparation of hibiscus wines.

Preparation of hibiscus wine

Hibiscus calyces (China and Colima varieties) were weighed (25 g each) and decocted by adding into 1 L of boiling water and kept for 10 min (Mercado-Mercado et al., 2015). Weighting and decoction were repeated twice for each hibiscus variety. From each variety, one decoction was filtered, while the other was used in its entirety to be used as must, which produced four treatments in total. Later, the four treatments were cooled at 40 °C, and the total soluble solids were adjusted to 21.7 ± 0.5°Bx by adding sucrose. Must fermentation was carried out by inoculation of active dried S. cerevisiae yeast (Tradi-Pan, Fermex, Mexico) 4 g/L−1 (Joshi et al. 2017) at 23 ± 2 °C in hermetically sealed 20 L plastic containers equipped with airlock valves. A volume of 18 L of must was used for fermentation in each container. After 4 months in fermentation containers, the wines obtained from filtered and unfiltered musts were filtered with muslin cloths, bottled in 750 mL dark bottles and stored at 4 °C. Fermentation was naturally stopped by the low sugar content in wine and by the low storage temperature (Joshi et al. 2017). A sample of 400 mL was taken from each container before and after the fermentation step for physicochemical analysis in order to determine the initial and final characteristics of products, respectively. For physicochemical measurements, all the samples were filtered (Whatman # 1), centrifuged (4090xg, for 10 min at 4 °C) and then stored at −20 °C until analysis.

Preparation of samples for sensory analysis

For the sensory analysis, sucrose was added to the bottles of the four treatments in order to reach 13°Bx or 16°Bx (low or high sweetness wines). In doing so, it was expected to distract to the participants from the hibiscus characteristics and obtain a wider differentiation between products. A total of eight samples were obtained through a combination of the three factors involved: variety (Colima and China), removing or keeping the calyces during fermentation (Filtered or Unfiltered) and low or high sweetness (LS or HS) to generate the treatments (Filtered-C.LS, Filtered-C.HS, Unfiltered-C.LS, Unfiltered-C.HS, Filtered-Ch.LS, Filtered-Ch.HS, Unfiltered-Ch.LS and Unfiltered-Ch.HS, respectively). The hibiscus wine bottles were stored at 4 °C until conducting the sensory tests, which occurred no later than 24 h after the bottles had been opened.

Physicochemical analysis

Color

Wine color was measured using the lightness, chroma and hue (L, C, h) values with a Minolta chroma meter CR-410 (Minolta, Japan). Samples of 15 mL were transferred into a glass Petri dish (7.4 cm diameter). Color was then measured by immersing the chroma meter’s measuring port at 1 cm into the liquid and pressing the measuring trigger (Juhari et al. 2018).

Total Soluble Solids (TSS)

For control purposes, TSS content was measured using a refractometer (Mod DR201-95, Krüss Optronic, Germany), calibrated with distilled water. TSS values were reported as Brix degrees (°Bx).

pH

pH measurements were carried out using a calibrated pH meter (Mod EA920, Orion, USA).

Alcohol content

Alcohol content was determined with a 100 mL sample from each product using the specific gravity method (AOAC 11.005).

Total anthocyanins content

The total anthocyanins content (TAC) was determined in each sample through the pH differential method at pH 1.0 and pH 4.5 (Lee et al. 2005). Two solutions were prepared at pH 1.0 (0.025 M KCl) and pH 4.5 (0.4 M CH3COONa) and were adjusted with HCl. The samples from extracts and wines were diluted to be within a measurable absorbance range.

Both samples (diluted to pH 1.0 and pH 4.5) were shaken and equilibrated under darkness for 20 min. Absorbance values were measured at 700 nm and 520 nm. TAC values in samples were expressed as mg cyanidin-3-glucoside (C3G) equivalents per liter using molecular weight (449.2 g ml−1) and the molar absorptivity coefficient (26,900 L mol−1 cm) for C3G.

Condensed tannin content

Condensed tannin content (CTC) was analyzed in musts and wines using the vanillin-HCl method (Broadhurst and Jones 1978), with slight modifications. A calibration curve was prepared with catechin (200–1000 µg/mL), while 500 µL of standard or sample (must or wine) was mixed with 3000 µL vanillin at 4% (in methanol) and 1500 µL of concentrated HCl (38%). Mixtures were kept in darkness for 20 min and the absorbance was measured at 500 nm. In each measurement, a blank was prepared using 500 µL of ethanol instead of standard or sample. The results were expressed as mg catechin equivalents (mgCE/L).

Total phenolic compound content

Total Phenolic Content (TPC) was measured using the Folin-Ciocalteu method reported by Song et al. (2010) with slight modifications. An aliquot of 150 µl (standard or sample) was taken and mixed with 600 µl of distilled water, 750 µl of Folin-Ciocalteu’s reagent (10%), and 750 µl sodium carbonate (7.5%, w/v). Reaction was kept to 30 min under darkness conditions. Absorbance was measured at 740 nm in a UV–Vis spectrophotometer (Genesys 10S, Thermo Scientific, United States). TPC were expressed as gallic acid equivalents (mgGAE/L) by using a calibration curve built within the range of 30 to 150 mg/L (r2 = 0.999).

Antioxidant activity

The DPPH assay was performed following Brand-William’s method (1995) with some modifications. A solution of DPPH (100 µM) was prepared and mixed with 1 mL of sample. The mixture was left to stand for 30 min under darkness conditions at 23 ± 2 °C. The absorbance was then measured at 515 nm. Trolox was used as a reference standard with results expressed as mMTE/L.

Determination of organic acids

Tartaric, malic, and ascorbic acid contents were determined by HPLC method. These analyses were performed using a Y9100 HPLC system (Young Lin, Korea) equipped with a quaternary pump (YL9110), degasser, thermostatted column compartment (YL-9130) and photodiode-array detector (YL-9160). All separations were performed using a SiliaChrom® RP-C18 column (4.6 × 250 mm, 5 µm). A 20 µL sample was manually injected after filtration through a 0.2 µm membrane filter (Millipore, USA). The mobile phase used for the separations was composed by NaH2PO4 50 mM in distilled at pH 2.8 (Park et al. 2017). The flow rate was kept at 0.5 mL/min, and the run time was 20 min.

Standard solutions were injected to obtain the retention time for each compound. Tartaric (prepared at concentrations of 0.01, 0.05, 0.1, 0.2, 0.4, 0.6, 0.8 and 1.0 g/L), malic (prepared at 0.1, 0.2, 0.4 and 0.5 g/L), and citric acid (prepared at concentrations of 0.15, 0.3, 0.6, 1.2, 1.8, 2.4, and 3.0 g/L) were used to obtain calibration curves at 210 nm. The quantification of individual organic acids was carried out by comparing peaks with their respective calibration curves of standard solutions.

Sensory evaluation

Generation of descriptors

Twenty consumers received all samples simultaneously and a list of the characteristics perceived was provided on a white sheet of paper. Subjects were asked to indicate the characteristics associated with appearance, taste, smell, and texture to help to distinguish samples, avoiding hedonic phrases or words. Responses from the participants were revised and synonyms grouped. The frequently mentioned descriptors (at least 25% of participants) were considered for the Check-All-That-Apply questionnaire.

Sensory characterization and acceptance

Ninety-three subjects (staff and students from the Universidad Politécnica de Pénjamo) participated in the study. Participants received all samples simultaneously. After tasting, the samples were ranked from the least to the most preferred, and the products were rated using a 15-point scale (only numbers, printed in a 150 × 10 cm cardboard), where 1 and 15 were considered as extremes of disliking and liking, respectively.

Once the first part of the sensory test was completed, subjects received a Check-All-That-Apply (CATA) questionnaire and were invited to taste the hibiscus wine samples and check all the attributes perceived in each one. Additionally, participants checked the characteristics required to be present in their ideal hibiscus wine. The order of presentation of sensory descriptors in CATA questionnaires was randomized to avoid bias by order of presentation (Jaeger et al. 2019).

In both experiments, one 30 mL sample of cold wine (4 °C) from each treatment was served in three-random-digit codified plastic cups.

Statistical analysis

A 22 factorial analysis was used (Colima or China varieties used for wine production and fermentation with or without hibiscus calyces in musts), with physicochemical results and sensory acceptance data analyzed by ANOVA and a Tukey’s test for means comparisons.

CATA question results were analyzed by Cochran’s Q to determine the difference in frequency of mentioned attributes among hibiscus wines. Correspondence analysis was performed on the frequency table to determine the spatial configuration and obtain a representation of the attributes and its relative distance to samples and liking data. Principal Component Analysis was performed to determine the relationship between CATA questions and physicochemical characteristics in hibiscus wine. All the statistical analysis was performed using XLSTAT 2014 software (Addinsoft, France).

Results and discussion

Physicochemical parameters

Table 1 shows physicochemical parameters observed in musts and wines. None of the musts contained alcohol because the fermentation process had not yet begun. pH values were around 2.7 in musts of all treatments. There were no differences between the color parameters of lightness, chroma, and hue for the treatments in musts. pH and color values in hibiscus extract’s agreed with values reported elsewhere (Juhari et al. 2018). Furthermore, adjusted TSS between musts did not show any significant differences. demonstrating that initial conditions for hibiscus wines were similar.

Table 1.

Physicochemical parameters in hibiscus musts and wines

Variety Sample Condition pH Color Alcohol (%v/v) TSS (°Bx)
L C h
China Must Filtered 2.6 24.8 9.4 356.2 ND 22.2
Unfiltered 2.8 31.0 10.8 1.2 ND 21.7
Wine Filtered 2.9 28.6a 15.3a 10.7a 9.3 8.8b
Unfiltered 2.9 26.2b 11.9b 2.6b 10.0 7.3c
Colima Must Filtered 2.7 24.8 10.1 358.9 ND 21.2
Unfiltered 2.7 38.1 11.7 10.6 ND 22.1
Wine Filtered 2.9 28.3a 13.3ab 6.9ab 1–1.0 8.8b
Unfiltered 3.0 25.6b 11.0b 3.4b 9.3 10.0a
Open in a new tab

Mean values n = 3. L: lightness, C: chroma, h: hue, TSS: total soluble solids; ND: not detected. Different letters between rows of the same sample (must or wine) indicate significant differences (Tukey, P ≤ 0.05)

The physicochemical parameters of all hibiscus wines obtained are shown in Table 1. Alcohol content in wines were 11.0 ± 1.0 g/100 mL, with TSS reaching values between 7.5 and 10.3, while pH values of wines increased slightly compared with their corresponding must, meaning that these parameters did not show significant differences. Similar pH values in hibiscus extracts were reported by Ifie et al. (2016), and they were reported to be stable during hibiscus must fermentation.

Color attributes (L, C, and h) were different between hibiscus wines obtained from the fermentation of filtered and unfiltered musts (P < 0.001, P < 0.01 and P < 0.001, respectively). Hibiscus wines Filtered-C and Filtered-Ch showed higher values for color attributes, meaning that these wines can be considered as lighter red beverages. In contrast, hibiscus wines obtained from the fermentation of unfiltered musts (Unfiltered-C and Unfiltered-Ch) showed darker and more intense red colors than Filtered-C and Filtered-Ch and all musts. This effect can be related to the longer maceration time, which could have contributed to a greater release of compounds from the hibiscus calyces causing a decrease in lightness, since it has been reported that they still contain anthocyanins and other phenolic compounds after decoction (Sáyago-Ayerdi et al. 2014).

Total anthocyanin content

Anthocyanins are phenolic compounds responsible for the purple and red colors in several plants, and are the major phenolic compounds found in varieties of hibiscus with dark and light red calyces (Ifie et al. 2016). The main anthocyanins found in hibiscus calyces are delphinidin 3-O-sambubioside and cyanidin 3-O-sambubioside (Ifie et al. 2016). It has been observed that Total Anthocyanin Content (TAC) decreases after 60 days of the hibiscus wine aging process, and this effect is faster when the beverage is stored at temperatures higher than 15 °C (Ifie et al. 2017).

According to the results (Table 2), musts showed slightly higher TAC than other reported values (18–27.3 mg of TAC/g of calyces) found in light red and dark red calyces of hibiscus varieties (Ifie et al. 2016). TAC values were not significantly different between varieties nor filtration conditions (P = NS).

Table 2.

Main chemical composition of hibiscus musts and wines

Variety Sample Condition TAC (mg/L) CTC (mgCE/L) TPC (mgGAE/L) Antioxidant activity (mMTE /L) Tartaric acid (g/L) Malic acid (g/L) Citric acid (g/L)
China Must Filtered 139.3 292.0 248.4 0.64 0.29 0.22 1.70
Unfiltered 132.1 308.7 219.0 0.59 0.44 0.29 2.12
Wine Filtered 18.6c 85.5b 572.0b 3.15b 0.26 0.21ab 1.65ab
Unfiltered 48.5a 220.5a 833.0a 8.40a 0.26 0.32ab 2.25ab
Colima Must Filtered 89.5 254.7 259.8 0.73 0.16 0.29 1.57
Unfiltered 109.6 213.3 226.1 0.80 0.07 0.21 2.05
Wine Filtered 16.4c 70.5b 425.0b 3.14b 0.23 0.12b 0.85b
Unfiltered 32.5b 196.5a 802.0a 8.82a 0.29 0.39a 2.66a
Open in a new tab

Mean values n = 3. TAC: Total Anthocyanin Content (as mg cyanidin-3-glucoside equivalents/L); CTC: Condensed Tannins Content; TPC: Total Phenolic Content. Different letters between rows of the same sample (must or wine) indicate significant differences (Tukey, P ≤ 0.05)

Extraction conditions can lead to an increase in the release of several compounds from hibiscus calyces. Ifie and colleagues (2016) used water at 50 °C (30 min) for the hibiscus must preparation, while, in the present work, this process was carried out by decoction in boiling water for 10 min. Likewise, other studies have reported that boiling temperatures can lead to a higher recovery yield of chemical compounds from hibiscus calyces (Ramírez-Rodrígues et al. 2011).

TAC values decreased during fermentation in all wines (Table 2). Wines obtained from the fermentation of filtered musts of Colima and China varieties (Filtered-C and Filtered-Ch) were not significantly different. In contrast, the wine made with unfiltered must from the China variety (Unfiltered-Ch) showed higher TAC than the wine from the Colima variety made under the same conditions (Unfiltered-C).

The decrease in anthocyanin content was related to hue, as the initial must color was reddish-purple which then changed to a more intense red tone. Moreover, the reduction in TAC could be because monomeric anthocyanins can be modified to polymeric anthocyanins during fermentation, and these molecules are not sensitive to pH change, meaning that they cannot be measured by the pH differential method (Ifie et al. 2017).

Condensed tannin content

Tannins are phenolic compounds strongly associated with the sensory attribute of astringency (Jackson 2008). Condensed Tannins are derived from flavanols, and their reduction during winemaking has been previously reported (Lingua et al. 2016).

Tannin content decreased in all of the wines (Table 2) during the first 60 days of fermentation and there were no significant changes until the day that the wines were bottled (data not shown).

Hibiscus wines Filtered-C and Filtered-Ch showed lower CTC than Unfiltered-C and Unfiltered-Ch, which can be attributed to the musts’ properties. In the case of wines made from filtered musts, tannins are suspended in the extracts and cannot interact with the other calyx components, meaning that they can react with other compounds from the extracts and form more stable products.

In contrast, in unfiltered wines, the release of tannin compounds from the hibiscus calyces can continue during maceration. Consequently, these wines showed higher CTC values than those observed in filtered wines. This effect has also been reported in wines where grape skin and seeds were kept in maceration over nine days of fermentation (Ivanova-Petropulos et al. 2016), although the effects on sensory acceptance or sensory attributes were not evaluated in that study.

These results were compared with values from grape wines, where Filtered-C and Filtered-Ch showed values close to the Cabernet Sauvignon variety (from Argentina), with CTC values ranging from 76 to 86 mgCE/L (Lingua et al. 2016). Additionally, CTC observed in unfiltered wines was slightly lower than the reported values (232–300 mgCE/L) from some Chinese Merlot wines (Jiang and Zhang 2012).

Crucial factors that affect hibiscus winemaking and the resulting product’s properties are the contribution of environmental conditions, the winemaking technique, and aging. In fact, hibiscus wines obtained under specific conditions can have the equivalent content of bioactive compounds found in grape wine.

Total phenolic compounds

Total Phenolic Compounds (TPC) have important roles in wine and the production process, as they significantly contribute to their appearance, taste, and aroma (Ivanova-Petropulos et al. 2016; Setford et al. 2017) as along with helping to prevent some diseases (El Gharras 2009).

TPC content increased during the fermentation in all wines and reached a plateau on day 90 (data not shown). TPC content was significantly higher (P < 0.0001) in unfiltered wines than in filtered wines (Table 2). These results demonstrated that the hibiscus calyces maintained bioactive phenolic compounds after the decoction process and facilitated the release of compounds during fermentation.

The increase in TPC during hibiscus wine production was observed by Yokotsuka et al. (1997) after 40 days of fermentation, with TPC values of around 660 mgGAE/L. In this experiment, filtered wines had lower TPC values (425–572 mgGAE/L), while unfiltered wines showed higher TPC values (802–833 mgGAE/L).

Total phenolic compounds have also been determined in hibiscus aqueous extracts (1:20 m/V ratio) after the decoction process. In one study, TPC content was 66.1 mg/100 mL (Sáyago-Ayerdi et al. 2007), which was higher than the observed values for all hibiscus musts in this work (Table 2). However, the aforementioned study used twice the concentration of dried hibiscus, compared to the concentration used in the present work to obtain musts.

TPC content is an important wine parameter. Unfiltered hibiscus wines showed similar values to those found in other wines, such as a Merlot grape wine (860 mgGAE/L) produced in an arid area of northern China (Jiang and Zhang 2012), and the Snatušina variety from Macedonia (880 mgGAE/L) which was produced with must after 9 days of maceration (Ivanova-Petropulos et al. 2016). It has also been reported that phenolic acid content (mainly gallic and chlorogenic acids) in grape wines is considerably higher in Syrah wine after 30 days of maceration, although the common practice is to keep skins and seeds for only 5 days (Alencar et al. 2018).

Consequently, hibiscus calyces in must allow the continuous release of phenolic compounds, mainly anthocyanins and tannins, linked to dietary fiber (Sáyago-Ayerdi et al. 2007), and increase TPC in beverages through a diffusion effect, which has also been observed in the maceration process during grape winemaking (Setford et al. 2017).

Organic acids

Wines contain a large number of compounds, including organic acids, which are partially responsible for their taste balance and chemical stability, as well as providing antioxidant properties and health benefits. The main organic acids contained in extracts of hibiscus calyces are tartaric, malic and lactic acids (Ifie et al. 2018), which significantly influence sensory characteristics (sourness and astringency) and the formation of esters that contribute to the wine aroma (Jackson 2008).

Tartaric, malic, and citric acids were analyzed in this work. Figure 1a shows the chromatograms for hibiscus wine samples. Citric acid was most abundant in both musts and wines (Table 2), while there was no difference in the concentration of tartaric acid between the wines (P = NS). Results show that citric and malic acid contents were affected by the hibiscus calyces (both P < 0.05), with citric and malic acid contents significantly lower in wines made from filtered musts, especially from the Colima variety.

Fig. 1.

Fig. 1

Open in a new tab

a Chromatograms of the reference curve for organic acid analysis and chromatograms of hibiscus wines obtained by fermentation of the Colima and China varieties with unfiltered and filtered musts. b Images of hibiscus wines. From left to right: Colima and China wines made with filtered musts, Colima and China wines made with unfiltered musts

Dried hibiscus calyces are considered as an excellent source of tartaric and citric acids (Juhari et al. 2018). Differences in these organic acids and volatile compounds have been attributed to drying conditions and hibiscus varieties (Juhari et al. 2018; Suliman et al. 2011).

The organic acid content was comparable with that found in other previous works. For example, citric acid content has been found within a range of 0.11 mg/g (Suliman et al. 2011) to 4.0 mg/g (Jabeur et al. 2019) in hibiscus calyces extracts. Malic acid content has been reported in the range of 0.12 mg/g (Ifie et al. 2018) to 63 mg/g in calyces (Jabeur et al. 2019), which is higher than citric and tartaric acids in hibiscus wines. Finally, tartaric acid levels have been reported to be in the range of 0.1 mg/g (Ifie et al. 2016) to 1.7 mg/g in calyces (Wong et al. 2002). Malic and citric acids have been regarded as the main organic acids in hibiscus calyces depending on the season (Ifie et al. 2016, 2018), and there is a wide range of reported organic acid levels in dried hibiscus calyces. Additionally, other organic acids, such as succinic acid (Wong et al. 2002) and quinic acid (Jabeur et al. 2019) have been considered as the main organic acids in hibiscus calyces.

Differences in organic acid content in hibiscus calyces can be influenced by several factors including climatic and growing conditions, genotypic and phenotypic varieties, and drying conditions used as postharvest treatment (Ifie et al. 2016, 2018; Jabeur et al. 2019).

Antioxidant activity

Antioxidant activity is usually related to the concentration of phenolic compounds such as tannins, anthocyanins and another simple (monomers and oligomers) or complex (polymeric) compounds (Jiang and Zhang 2012) which have proven health benefits for consumers (El Gharras 2009). In this experiment, the recorded antioxidant activity was higher in unfiltered wines than in filtered wines (P < 0.05). This result was related to the higher content in TPC, TAC, CTC and organic acids in unfiltered wines. It was also observed that antioxidant activity was positively correlated with total phenolic content (r = 0.553; P < 0.0001) and with the content of citric and malic acids in hibiscus wines (r = 0.309; P < 0.05 and r = 0.436; P < 0.001, respectively) which are in accordance with the results reported for other hibiscus wines by other authors (Tiwari et al. 2017).

Antioxidant values observed in musts obtained by a decoction of 25 g of hibiscus calyces/L of water were slightly higher than expected compared to the values reported by other authors who analyzed hibiscus samples from several regions and varieties in Mexico (Borrás-Linares et al. 2015). However, agroclimatic conditions as well as agronomic and postharvest practices may well have contributed to differences in the observed values (Ifie et al. 2018). The antioxidant activity values found in Unfiltered-C and Unfiltered-Ch hibiscus wines were similar to the values reported by some authors for three out of 23 samples of Cabernet Sauvignon wines from the Mudgee region in Australia which ranged from 7.76 to 8.57 mMTE/L (Yoo et al. 2011). These values were also higher than those reported for Cabernet Sauvignon and Merlot wines from some Chinese regions, which ranged from 4.0 to 6.5 mMTE/L (Jiang and Zhang 2012). On the other hand, antioxidant activity in wines obtained in this experiment was lower than the values reported for Syrah, Merlot, and Cabernet Sauvignon wines from Argentina, which ranged from 10.5 to 12.8 mMTE/L (Lingua et al. 2016). As previously mentioned, the differences between regions and studies in antioxidant activity and other variables can be attributed to several factors. However, one important result obtained from this study is that hibiscus wines made under appropriate conditions could be considered as functional beverages according to their bioactive compound content.

Sensory characterization and acceptance

Figure 2 shows the overall liking degree for hibiscus wines. The acceptance was not different between the wines obtained by fermentation of musts from China and Colima varieties (P = NS). The differences in the level of sucrose added to wines just before they were bottled did not affect their overall liking degree either (P = NS). On the other hand, keeping the calyces of hibiscus during fermentation affected the acceptance of products (P < 0.01). Therefore, Unfiltered-C.LS, Unfiltered-C.HS, Unfiltered-Ch.LS, and Unfiltered-Ch.HS wines (all of which kept the calyces) obtained the highest overall liking values among the hibiscus wines analyzed. It is worth mentioning that the acceptance values given by consumers for the products were in the upper middle-range of the scale, which is a positive zone and indicates that the products are perceived as being pleasant. However, this also indicates that the products are not perceived as outstanding and this could be related to certain extraordinary characteristics that are lacking, as well as the fact that participants in this study were not used to drinking wines, as 62% of the participants were university students with beer as their most consumed alcoholic product. More research in this area with the participation of wine consumers or the use of a commercial product as a reference for comparison could produce different outcomes and conclusions.

Fig. 2.

Fig. 2

Open in a new tab

Overall liking for hibiscus wines. C. Colima; Ch. China; LS, HS. Low and high sweetness (13 and 16°Bx), respectively, Filtered and Unfiltered. Fermented without and with calyces, respectively

Sensory characteristics of hibiscus wines

The responses from the participants allowed to identify 11 sensory descriptors that were useful for discriminating between hibiscus wines (Table 3). These sensory descriptors were used by consumers in the CATA questionnaires to characterize the samples.

Table 3.

Frequency of mentions of attributes by hibiscus wines

Attributes Hibiscus wines
Colima variety China variety Ideal
Filtered Unfiltered Filtered Unfiltered
LS HS LS HS LS HS LS HS
Taste
 Bitter**** 47ab 40abc 22c 23c 49a 40abc 31abc 28bc 26c
 Alcohol ns 39 41 40 37 44 50 40 42 50
 Sweet**** 22c 26bc 51a 46ab 29bc 33abc 27bc 37ab 46ab
 Intense** 38ab 32ab 30ab 25b 44ab 49a 31ab 37ab 28b
 Hibiscus** 22b 24ab 42a 35ab 31ab 29ab 28ab 32ab 39ab
Smell
 Alcohol** 39ab 31ab 28b 29ab 49a 48ab 37ab 45ab 40ab
 Intense*** 24b 21b 24b 20b 32ab 47a 29ab 31ab 28ab
Texture
 Smooth**** 28abc 15c 40ab 44a 24bc 31abc 34ab 33abc 42ab
Appearance
 Intense red color**** 36abc 28bc 23c 33ab 51a 52a 49a 51a 48ab
 Light red color**** 60a 55a 54a 53a 9c 11bc 13bc 12bc 30b
 Presence of bubbles*** 34abc 47a 37abc 25abc 38abc 39ab 19c 25bc 34abc
Open in a new tab

Filtered. Fermented without calyces Unfiltered. Fermented with calyces. LS Added with sucrose to reach 13°Bx HS. Added with sucrose to reach 16°Bx Significant at **** P < 0.0001; *** P < 0.001; **P < 0.01; *P < 0.05; ns = no significant differences according to Cochran’s Q test

Significant differences (at least P < 0.01) were found in the frequencies of 10 out of 11 terms of the CATA questions used to describe the hibiscus wines (Table 3), suggesting that consumers perceived differences between wines in terms of sensory characteristics. From the five descriptors related to taste, only the “alcohol taste” option recorded no differences between samples, which could be associated with the similar levels of alcohol content as determined instrumentally. The values of the organic acids contained in musts and wines is generally low, which could be related to the sensory characteristics perceived in wines, where their respective acids and astringents were not mentioned as important.

The wines most appreciated by consumers and closer to their ideal product were perceived as sweeter, less bitter, smooth, and with the presence of a hibiscus taste (Fig. 3a). These attributes were found more frequently in Unfiltered wines. It is noteworthy that this study found a relationship between certain sensory characteristics and the acceptance for hibiscus wines. This could be explained by some factors such as (1) the sensory descriptors were generated by consumers, meaning that they could be easy to understand for non-experts (Jaeger et al. 2019); (2) when the terms used in Check-All-That-Apply questions are understood by participants, they can contribute to identifying the drivers of liking (Jaeger et al. 2019); and (3) a big enough sample size was used. This particular point is relevant since other authors could not find a relationship between descriptive sensory scores and hedonic values when hibiscus wines were compared by a sample of 50 hibiscus wine consumers (Mounigan and Badrie 2007).

Fig. 3.

Fig. 3

Open in a new tab

a Sensory map of hibiscus wines obtained by applying CATA questions. b Principal Components for the sensory and physicochemical characteristics of hibiscus wines. Open circles are characteristics and close circles correspond to treatments. Filtered. Hibiscus wines obtained from fermentation of filtered musts; Unfiltered. Hibiscus wines obtained from fermentation of unfiltered musts. C. Colima variety; Ch. China variety; LS. Wines added with sucrose to reach 13°Bx. HS. Wines added with sucrose to reach 16°Bx

The variety of hibiscus used determined the appearance of the wines (Fig. 1b). The products generated from the Colima variety exhibited a light red coloration, while those originating from the China variety showed an intense dark red coloration. These attributes were related to higher hue and chroma values in the case of wines obtained from the Colima variety. The hue and chroma values were particularly low under the must-filtered conditions, corresponding to the treatments with the lowest anthocyanin and total condensed tannin contents.

Figure 3b shows a representation of the sensory attributes and hibiscus wine samples obtained by Principal Components Analysis. The first two principal components explained 77.62% of the data’s variability. The first principal component was associated with the acceptance of hibiscus wines by consumers, with a preference expressed for the Unfiltered-C.LS, Unfiltered-C.HS, Unfiltered-Ch.LS and Unfiltered-Ch.HS wines. These wines were frequently associated with the sensory attributes of sweetness, smoothness, and hibiscus taste, as they showed higher contents of total phenolic compounds, condensed tannins, organic acids, as well as antioxidant activity. In contrast, the bitter attributes and higher values of luminosity, chroma, and hue angle corresponding to more luminous tones were associated with lower acceptance for Filtered-C.LS, Filtered-C.HS, Filtered-Ch.LS and Filtered-Ch.HS wines. The principal component 2 facilitates discrimination between varieties. The wines obtained from the Colima variety showed a lighter red tone, while those generated from the China variety were distinguished by the presence of an intense smell and red color. It was also observed that the adjustment of sugar in wines did not influence the perception of these wines as being different, as expected. This effect could be explained by the stronger presence of attributes such as hibiscus taste and bitterness, which predominated over sweetness, and conditioned the consumers’ preferences.

The citric acid content in wines made from unfiltered musts was higher than that observed in filtered musts, contributing to the acidic taste in the former, which the participants could have associated with a hibiscus taste. Tartaric and malic acid content in the wines analyzed in this experiment were low in comparison to the values observed in red wines, providing a modest contribution to sourness and astringency. Given that astringency sensation is frequently confused with bitterness (Jackson 2008), it was expected that this attribute would be reported more frequently in wines made from unfiltered musts, although the opposite effect was observed. Due to the fact that bitterness was highly related to luminosity, chroma and hue angle attributes, it is reasonable to think that appearance is partially responsible for these results. The effect of appearance has also been reported by Wang and Spence (2019), who found that novice and expert wine consumers can be influenced by the color of wine with regard to their perception of aroma and flavor.

Conclusion

The production of hibiscus wines from unfiltered musts of the China and Colima varieties led to products with differentiated physicochemical characteristics and bioactive compound content compared to those obtained from filtered musts. Hibiscus wines made from unfiltered musts were generally more accepted than those made from filtered musts. Sensorial characteristics concerning taste and texture as well as physicochemical characteristics associated with organic acid and bioactive compound content were related and responsible for the acceptance of hibiscus wines.

A limitation of this study is that the sensory tests were carried out with participants not used to consuming wine, meaning that the results, especially those for acceptance, could be different compared to other segments of consumers. Therefore, it would be interesting to conduct another study only with wine consumers to determine their perceptions of hibiscus wines.

Author’s contributions

JABR, ALGS and VGAR conceived and planned the experiments. JABR carried out the experiments. JABR, ALGS, VGAR and ACO contributed to analysis and interpretation of the results. BRJA took the lead in writing the manuscript. All authors provided critical feedback and helped shape the research, analysis and manuscript.

Funding

No funding, grants or other support was received.

Availability of data and material

The data related to this article may be requested through the email jabarajas@uppenjamo.edu.mx, as I will make available the spreadsheets containing the results of the study on reasonable request.

Declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Ethics approval

This article does not contain any studies with human or animal subjects.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. Alencar NMM, Corrêa CCBB, Maróstica JMR, Biasoto ACT, Behrens JH. Influence of maceration time on phenolic compounds and antioxidant activity of the Syrah must and wine. J Food Biochem. 2018;42(2):e12471. doi: 10.1111/jfbc.12471. [DOI] [Google Scholar]
  2. Borrás-Linares I, Fernández-Arroyo S, Arráez-Roman D, Palmeros-Suárez PA, Del Val-Díaz R, Andrade-Gonzáles I, Fernández-Gutiérrez A, Gómez-Leyva JF, Segura-Carretero A. Characterization of phenolic compounds, anthocyanidin, antioxidant and antimicrobial activity of 25 varieties of Mexican Roselle (Hibiscus sabdariffa) Ind Crops Prod. 2015;69:358–394. doi: 10.1016/j.indcrop.2015.02.053. [DOI] [Google Scholar]
  3. Brand-Williams W, Cuvelier ME, Berset C. Use of a free radical method to evaluate antioxidant activity. LWT Food Sci Technol. 1995;28(1):25–30. doi: 10.1016/S0023-6438(95)80008-5. [DOI] [Google Scholar]
  4. Broadhurst RB, Jones WT. Analysis of condensed tannins using acidified vanillin. J Sci Food Agric. 1978;29(9):788–794. doi: 10.1002/jsfa.2740290908. [DOI] [Google Scholar]
  5. El Gharras H. Polyphenols: food sources, properties and applications—a review. Int J Food Sci Technol. 2009;44:2512–2518. doi: 10.1111/j.1365-2621.2009.02077.x. [DOI] [Google Scholar]
  6. Ifie I, Abrankó L, Villa-Rodríguez JA, Papp N, Ho P, Williamson G, Marshall LJ. The effect of ageing temperature on the physicochemical properties, phytochemical profile and α-glucosidase inhibition of Hibiscus sabdariffa (roselle) wine. Food Chem. 2017;267:263–270. doi: 10.1016/jfoodchem.2017.05.044. [DOI] [PubMed] [Google Scholar]
  7. Ifie I, Ifie BE, Ibitoye DO, Marshall LJ, Williamson G. Seasonal variation in Hibiscus sabdariffa (Roselle) calyx phytochemical profile, soluble solids and α-glucosidase inhibition. Food Chem. 2018;261:164–168. doi: 10.1016/j.foodchem.2018.04.052. [DOI] [PubMed] [Google Scholar]
  8. Ifie I, Marshall LJ, Ho P, Williamson G. Hibiscus sabdariffa (Roselle) extracts and wine: phytochemical profile, physicochemical properties, and carbohydrase inhibition. J Agric Food Chem. 2016;64(24):4921–4931. doi: 10.1021/acs.jafc.6b01246. [DOI] [PubMed] [Google Scholar]
  9. Ifie I, Olurin TO, Aina JO. Production and quality attributes of vegetable wine from Hibiscus sabdariffa Linn. Afr J Food Sci. 2012;6(7):212–215. doi: 10.5897/AJFS12.036. [DOI] [Google Scholar]
  10. Ivanova-Petropulos V, Durakova S, Ricci A, Parpinello GP, Versari A. Extraction and evaluation of natural occurring bioactive compounds and change in antioxidant activity during red winemaking. J Food Sci Technol. 2016;53(6):2634–2643. doi: 10.1007/s13197-016-2235-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Jabeur I, Pereira E, Caleja C, Calhelha RC, Sokovic M, Catarino L, Barros L, Ferreira ICFR. Exploring the chemical and bioactive properties of Hibiscus sabdariffa L. calyces from Guinea-Bissau (West Africa) J Food Sci Technol. 2019;10:2234–2243. doi: 10.1039/C9FO00287A. [DOI] [PubMed] [Google Scholar]
  12. Jackson RS (2008) Sensory perception and wine assessment. In: Wine science. Principles and applications, 3rd edn. Associated Press, Amsterdam, pp 641–685
  13. Jackson RS. Innovations in winemaking. In: Kosseva MR, Joshi VK, Panesar PS, editors. Science and technology of fruit wine production. Amsterdam, Netherland: Academic Press; 2017. pp. 617–662. [Google Scholar]
  14. Jaeger SR, Hunter DC, Vidal L, Chheang SL, Ares G, Harker FR (2019) Sensory product characterization by consumers using check-all-that-apply questions: Investigations linked to term development using kiwifruit as a case study. J Sens Stud e12490. 10.1111/joss.12490
  15. Jiang B, Zhang Z-W. Comparison of phenolic compounds and antioxidant properties of Cabernet Sauvignon and Merlot wines from four grape-growing regions in China. Molecules. 2012;17:8804–8821. doi: 10.3390/molecules.17088804. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Joshi VK, Panesar PS, Rana VS, Kaur S. Science and technology of fruit wines: an overview. In: Kosseva MR, Joshi VK, Panesar PS, editors. Science and technology of fruit wine production. Amsterdam: Academic Press; 2017. pp. 1–73. [Google Scholar]
  17. Juhari NH, Bredie WLP, Toldam-Andersen TB, Petersen MA. Characterization of roselle calyx from different geographical origins. Food Res Int. 2018;112:378–389. doi: 10.1016/j.foodres.2018.06.049. [DOI] [PubMed] [Google Scholar]
  18. Lee J, Durst RW, Wrolstad RE. Determination of total monomeric anthocyanin pigment content of different juices, beverages, natural colorants, and wines by the pH differential method: collaborative study. J AOAC Int. 2005;88(5):1269–1278. doi: 10.1093/jaoac/88.5.1269. [DOI] [PubMed] [Google Scholar]
  19. Lingua MS, Fabani MP, Wunderlin DA, Baroni MV. From grape to wine: changes in phenolic composition and its influence on antioxidant activity. Food Chem. 2016;208:228–238. doi: 10.1016/j.foodchem.2016.04.009. [DOI] [PubMed] [Google Scholar]
  20. Mercado-Mercado G, Blancas-Benitez FZ, Velderrain-Rodríguez GR, Montalvo-González E, González-Aguilar GA, Alvarez-Parrilla E, Sáyago-Ayerdi SG. Bioaccesibility of polyphenols released and associated to dietary fibre in calyces and decoction residues of Roselle (Hibiscus sabdariffa L.) J Funct Foods. 2015;18:171–181. doi: 10.1016/j.jff.2015.07.001. [DOI] [Google Scholar]
  21. Mounigan P, Badrie N. Physicochemical and sensory quality of wines from red sorrel/roselle (Hibiscus sabdariffa L.) calyces: effects of pretreatments of pectolase and temperature/time. Int J Food Sci Technol. 2007;42:569–475. doi: 10.1111/j.1365-2621.2007.01270.x. [DOI] [Google Scholar]
  22. Park J-M, Shin J-A, Lee JH, Lee K-T. Development of a quantitative method for organic acid in wine and beer using high performance liquid chromatography. Food Sci Biotechnol. 2017;26(2):349–355. doi: 10.1007/s10068-017-0047-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Ramírez-Rodrígues MM, Balaban MO, Marshall MR, Rouseff RL. Hot and cold water infusion aroma profiles of Hibiscus sabdariffa: Fresh compared with dried. J Food Sci. 2011;76(2):212–217. doi: 10.1111/j.1750-3841.2010.01989.x. [DOI] [PubMed] [Google Scholar]
  24. Sáyago-Ayerdi SG, Arranz S, Serrano J, Goñi I. Dietary fiber content, and associated antioxidant compounds in roselle flower (Hibiscus sabdariffa L.) beverage. J Agric Food Chem. 2007;55:7886–7890. doi: 10.1021/jf070486b. [DOI] [PubMed] [Google Scholar]
  25. Sáyago-Ayerdi SG, Velázquez-López C, Montalvo-González E, Goñi I. By-product from decoction process of Hibiscus sabdariffa L. calyces as a source of polyphenols and dietary fiber. J Sci Food Agric. 2014;94:898–904. doi: 10.1002/jsfa.6333. [DOI] [PubMed] [Google Scholar]
  26. Setford PC, Jeffery DW, Grbin PR, Muhlack RA. Factors affecting extraction and evolution of phenolic compounds during red wine maceration and the role of process modelling. Trends Food Sci Technol. 2017;69:106–117. doi: 10.1016/j.tifs.2017.09.005. [DOI] [Google Scholar]
  27. Song F-L, Gan R-Y, Zhang Y, Xiao Q, Kuang L, Li H-B. Total phenolic contents and antioxidant capacities of selected Chinese medicinal plants. Int J Mol Sci. 2010;11:2362–2372. doi: 10.3390/ijms11062362. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Suliman AMA, Ali AO, Idriss SEAA, Abdualrahman MAY. A comparative study on red and white karkade (Hibiscus sabdariffa L.) calyces, extracts and their products. Pak J Nutr. 2011;10(7):680–683. doi: 10.3923/pjn.2011.680.683. [DOI] [Google Scholar]
  29. Tiwari S, Shukla S, Kishor K. Production, optimization, characterization and evaluation of antimicrobial activities in Hibiscus rosa-sinensis wine. J Pharmacogn Phytochem. 2017;6(3):19–26. [Google Scholar]
  30. Wang QJ, Spence C. Drinking through rosé-coloured glasses: Influence of wine colour on the perception of aroma and flavour in wine experts and novices. Food Res Int. 2019;126:108678. doi: 10.1016/j.foodres.2019.108678. [DOI] [PubMed] [Google Scholar]
  31. Wong P-K, Yusof S, Ghazali HM, Che-Man YB. Physico-chemical characteristics of roselle (Hibiscus sabdariffa L.) Nutr Food Sci. 2002;32(2):68–73. doi: 10.1108/00346650210416994. [DOI] [Google Scholar]
  32. Yokotsuka K, Yajima M, Seki T, Matthews M. Changes in color parameters during fermentation and storage of red wine using Thai roselle under different pHs and SO2 concentrations. Food Sci Technol Int. 1997;3(2):105–109. [Google Scholar]
  33. Yoo YJ, Prenzler PD, Saliba AJ, Ryan D. Assessment of some Australian red wines for price, phenolic content, antioxidant activity, and vintage in relation to functional food prospects. J Food Sci. 2011;76(9):C1355–C1364. doi: 10.1111/j.1750-3841.2011.02429.x. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The data related to this article may be requested through the email jabarajas@uppenjamo.edu.mx, as I will make available the spreadsheets containing the results of the study on reasonable request.

Articles from Journal of Food Science and Technology are provided here courtesy of Springer

RESOURCES