Sensory preferences for pomegranate arils in Italy: A comparison between different varieties and cultivation sites

Despite the growing worldwide interest on pomegranate due to its health benefits, little is known about the varieties produced and consumed in Italy. In this context, the aim of our study was to evaluate the factors affecting preferences of Italian consumers towards pomegranate arils of different varieties and cultivation sites. Three samples of pomegranate arils were analysed. Two retrieved from the local (Veneto) agricultural market: one of Turkish origin and one of Sicilian (Italy) origin. The third was retrieved from local producers in Veneto. Selected


| INTRODUC TI ON
The pomegranate is one of the oldest edible fruits known and there are more than 500 acknowledged varieties in the world, a few dozen of which are commonly grown. Thus, in the last few years, pomegranate has seen a great diffusion in various countries, especially those with a Mediterranean climate, such as Turkey, Tunisia, Egypt, Spain, Morocco and Italy, where it is mostly produced in Sicily, but also, in Iran, Afghanistan, India, the United States (California), China, Japan and Russia. 1 Due to its ability to grow at high temperatures, the potential for its expansion in arid or semi-arid areas is enormous, in particular where salinity and water deficiency are limiting for other crops. 2 It is harvested from September to early December, depending on the climate and the variety. Due to the growing market demand, it has become increasingly important to identify varieties and crosses of high quality and economic interest. 2 One of the reasons why pomegranate is gaining importance is that consumers are starting to focus on a healthier diet. Recently, this fruit has been promoted as one of the new 'superfoods', because of its bioactive phytochemical compounds based on ellagitannins, such as punicalagins and punicalins, anthocyanins, gallagic and ellagic acid, and on a distinct profile of fatty acids. All of them could have a positive effect on health, like prevention from serious chronic diseases, such as diabetes and cardiovascular diseases, and, even, from cancer. [3][4][5][6][7] Some scientific publication have shown that the pomegranate and its juice have anti-atherogenic, antioxidant and antihypertensive effects, generally associated with its content of polyphenols, as they contribute to the high antioxidant activity of the fruit. 8 On the other hand, however, the hydrolysable tannins (punicalagin, in particular) are responsible for the astringent sensation typical of the pomegranate. High levels of astringency can alter the overall sensory satisfaction of the consumer. 9 This attribute assumes a relevant role for pomegranate considering that the willingness of consumers to compromise taste for health properties is narrowing more and more. 10,11 Considering the consumers' acceptance for pomegranate, it should be recalled that red colour is one of the major quality attributes that affects consumers sensory acceptability, and it depends on the quantity of anthocyanins, the molecules responsible for the bright red colour of pomegranate juice. 12 Despite the increase in the commercial importance of pomegranate and despite its aromatic profile varies significantly among cultivars, as well as among the places where it is grown, relatively little is known about its genetic diversity related to its aromatic traits or consumer preferences.
Concerning sensory analysis and cultivar characterization, the literature shows that there are more researches and analyses for the pomegranate juice, compared to the fruit itself, perhaps for the greater industrial and economic interest that the juice finds, as it is easier to use and consume than the arils. The sensory profile of pomegranate juice has been studied with different analytical and sensorial measurements to define the differences among varieties, 13 post-harvest conditions, 14 climatic effects, 15 adulteration 16 and process. 17 The taste of the juice is usually measured by a combination of analytical instruments and sensory analysis. 18 Generally speaking, it emerges, from the related literature, that pomegranate juice is generally described with positive attributes such as sweet and sour taste, earthy and fruity smells, while astringency and bitterness have been classified by the tasters as negative attributes. Ferrara  When it comes to fresh fruit consumption, seed hardness is the key factor for consumers' acceptability. If seeds are too hard, consumer satisfaction will be drastically reduced. 24 Mayuoni-Kirshenbaum and Porat 13 demonstrated that sensory quality of pomegranate fruit greatly depends on the following factors: (a) cultivars; (b) degree of ripening; and (c) prolonged storage. The cultivars mainly determine the perception of sour taste, fruity odour and seed hardness mouthfeel. This is in line with evidences derived from the study of Chater et al, 25 in which six cultivars were utilized to determine consumer acceptance compared to the industry standard, 'Wonderful'. There were significant differences among cultivars for all traits assessed by the panellists: aril colour, sweetness, tartness, seed hardness, bitterness and overall desirability. There were also differences in acceptance among consumers for Wonderful cultivar depending on whether it was grown on the coast or in the inland. Therefore, this investigation indicated that site and cultivar have significant effects on fruit quality traits, which affected consumer acceptance of pomegranate fruit. This means that, in line with what Mayuoni-Kirshenbaum and Porat 13 said, the locus in which the plant grows also affects consumers' acceptability.
Despite the growing spread of pomegranate worldwide due to its health benefits and the stressed importance of characterize the cultivars traits preferred by consumers, little is known about the varieties produced and consumed in Italy. In Italy, the cultivation of pomegranate has at the moment a limited but growing diffusion. The Italian Institute of Statistics (ISTAT) estimate in 2019 the total area under pomegranate in 1234 ha (Table 1), while in 2013, the total area was only 133 ha. The main growing regions are Sicily and Apulia (respectively, 386 and 374 ha of total area), followed by Veneto, which is growing (210 ha).
In this context, the aim of our study is to assess consumers' preferences for pomegranate arils by comparing products produced in a newly grown area in Italy (namely in Veneto) with two products widely available on the local market: (a) Wonderful variety grown in Sicily and (b) Hicaz variety grown in Turkey. The study characterizes sensory and physicochemical characteristics of the products, and allows making a prediction on which attributes influence consumers overall liking for pomegranate in a growing market like Italy.

| E XPERIMENTAL PROCEDURE S
The study has been performed between October and November 2019 in three steps: panel test, physicochemical analyses and consumer test.

| Sensory analysis
Sixteen students, five men and eleven women in the age range of 20-30 years, with a university training in sensory analysis, were recruited as panellists. During the panel selection, judges were subjected to three tests: a test for the discrimination of four tastes (salty, sweet, sour and bitter) in model solutions, a test for the recognition thresholds (sweet, acid and bitter tastes) and a test for the intensity of astringency.
For the discrimination of the tastes, each candidate had to try a set of 10 dilutions of reference substances, corresponding to the four specific tastes. In addition, in one of the cups, there was only water. The samples were identified by a three-digit code and randomized; the participants had to recognize and indicate the correct taste. For the second test, five cups containing different increasing concentrations of the reference substances for each taste (sweetness, sourness and bitterness) were presented to each judge in a random order, as can be seen in Table 2. Every judge had to indicate the sample at which he perceived every taste. For the last test, the candidates were given five cups containing water solutions of tannic acid at increasing concentration. They were asked to identify which cup had the same concentration of tannic acid of an additional cup with an unknown level of tannic acid.
As regards the score awarded, in the first test judges could score a maximum of 10 points. In the second one, they could score a maximum of nine points (three points per taste); the highest score (three points) was given to the judge who recognized the taste at the lowest concentration. In the third one, the judges obtained a maximum of one point if they indicated the correct concentration of tannic acid. To select the most suitable judges, the minimum score to be selected was 12 points out of 20, chosen by the panel leader.
The training required three sessions of 2 hours each. In the first session, judges were trained on two sensory aspects: texture (firmness, crunchiness, juiciness and seed intrusiveness), using a commercial pomegranate as standard, and taste, with standard taste solutions ( Table 2). In the second one, judges were trained on three sensory aspects: kinaesthetic characteristics (size and colour of the arils), taste and aroma. Young red wine, apple juice and red fruit juice were used to become familiar with typical aromas of pomegranate. 26 Judges were instructed on taste and aromas also in the last session.
Sensory evaluation was carried out in the Sensory Laboratory in compliance with the UNI-ISO standard 8589. 27 Samples were retrieved the day before the training of the panel and have been kept at refrigeration temperatures. They were randomly selected without any visible physical defects and with as similar shape and size as possible within the same cultivar. The studied samples were as follows: • WOND_V: Wonderful variety, produced by a local (Veneto, Italy) co-operative; • WOND_S: Wonderful variety produced in Sicily (Italy) bought from the local wholesales market, considered the standard commercial competitors; • HICAZ: Hicaz variety produced in Turkey origin bought from the local wholesales market.
The shelling of pomegranates and the preparation of samples took place in the sensory laboratory, 1 hour before the arrival of  Order and number assigned to each sample were randomized.
Unsalted crackers and water were used to clean palates between samples.
A descriptive sensory analysis was conducted by using 9-point structured scales, where the value one indicated the lower intensity of the attribute, while nine indicated the greater intensity; the value five represented the central term, identified as standard. A list of 13 descriptors was generated using previously published lexicons as guides. 20,28 The assessed sensory attributes were as follows: size and colour of the arils for the appearance; apple, red wine and red fruits for the aromas; sweetness, sourness, bitterness for the taste; firmness, crunchiness, juiciness, seed intrusiveness, astringency for mouthfeel sensations. The descriptors, their meanings and the reference standards used are presented in Table 2.

| Sample preparation
Morpho-pomological measurements of fruits, arils and seeds characteristics and chemical analyses were carried out on samples of 20 mature fruits per type. The selected fruits were initially washed with cold distilled water, drained and then cut. The arils were hand-separated from the skin and pith and collected to form a homogeneous mixture for each product, then thoroughly mixed.
The following arils characteristics were analysed: size, colour, texture, total soluble solids, titratable acidity, total polyphenols and tannins and chromatographic analysis.

| Size
For the size attribute, the maximum diameter and length (mm) of the arils were measured by a digital caliper (Mitutoyo) with a 0.01 mm accuracy.

| Colour
The colour was measured using the CIE L*, a*, b* coordinates with a calibrated Minolta CM-600d, with a reflectance spectrum in the range from 400 to 700 nm.
The total anthocyanins concentration was determined with the Ribéreau-Gayon method 29

| Texture analysis
For the seed intrusiveness, a woody portion index was measured as seed weight/aril weight ratio (mg/mg); weight of arils and seeds was determined with a precision weighing device (AND HR-120 METTLER TOLEDO) with an accuracy of 0.0001 g. For the analysis, 30 replications were performed, because of the heterogeneity of the arils, as confirmed by Szychowski et al. 24 Texture profile analysis (TPA) and puncture test (PT) were conducted for each product, using a TEXTURE ANALYSER TA.TX plus (Stable Micro System), with a load cell capacity of 50 N. The TPA was performed according to Rosenthal. 31 Two compression cycles were made, and the following texture profile attributes were evaluated: hardness (N), cohesiveness (mm), springiness (mm) and chewiness (N).
All these measurements were associated with the attributes used to represent the texture of an aril in the evaluation by the trained panel: firmness and crunchiness.

| Total soluble solids and titratable acidity
For the sweetness, the sugar content or total soluble solid (TSS, °Brix) from pomegranate juice was measured using a digital refractometer, calibrated with distilled water at 20°C.
For the sourness, titratable acidity (TA) and pH were detected.

| Polyphenols and tannins
The total phenolic (TP) concentration was determined in triplicate using Folin-Ciocalteu method as described by Makkar et al. 32 The value obtained provided the total polyphenols concentration in mg/L, expressed as gallic acid equivalents.
The total tannins concentration was determined, using the Bate- follows: 5% B for 2 minutes, followed by 5%-12% B in 5 minutes, 12%-55% B in 31 minutes, and 55%-100% B in 1 minute. After washing for 2 minutes with solvent B, the column was re-equilibrated with 95% solvent A. The flow rate was 1.0 mL/min. The injection volume was 20 µL, and the column temperature was set to 37°C.
Identification and quantification were performed at 520 nm for anthocyanins and 350 nm for ellagitannins. Before injection, the juices, manually extracted from the three sample (see Section 2.2.1 for more details), were centrifuged in an Eppendorf tube (10 minutes at 14 000 g) and the centrifuged supernatant was passed through a 0.45 μm cellulose acetate filter (Advantec, CA).
Quantification for each compound was obtained comparing each peak area against the standard curve for the reference solutions: cyanidin chloride (Cy) for the anthocyanins and ellagic acid for the ellagitannins.

| Consumer evaluation
The consumer test was used to assess consumer liking of the About 20 arils, manually extracted 1 hour before the test, were served during the test into odour-free plastic cups with 80 mL capacity, coded with three-digit random numbers. The survey was administered to 203 students aged between 18 and 35 years. Since in Italy pomegranate is a product still not very widespread and little consumed in general and since young people are generally more inclined to try new products than older consumers, 35 we considered young people as target consumers, selecting students and university employees for the test.
Participants were instructed on how to use the scale; they were also allowed to re-taste and change their previous scores, if needed. Water and unsalted crackers were provided to consumers to rinse their palate before and between tasting. Lastly, consumers answered questions related to their sociodemographic profile (age, gender, education level, income and number of family members), as well as questions measuring their consumption patterns and frequency purchasing about pomegranate and fruits, in general.

| Statistical analysis
The analysis of differences on physicochemical values started veri-

| Consumer evaluation
Descriptive statistics on consumers' habits are presented in Table 3.
The 44.3% of respondents affirmed that they consume pomegranate several times a year but less than once a month. The favourite form of purchased product was the whole fruit, rather than the ready-toeat arils (63.5%) preferring the consumption of fresh arils (74.9%).  Table 5 shows mean value and standard deviation of the physical and chemical analyses of the three samples of pomegranate.

| Physicochemical analysis
As regards the visual appearance, results underlined that, when it comes to the size of the arils, a significance difference was observed among WOND_S and WOND_V and HICAZ. The first one had, also, the higher length (9.98 mm) and the higher width Similar results have been recorded through chromatographic analysis. In particular, as shown in Table 6, the content of Pg3.
In the supporting information, a typical HPLC chromatogram showing the anthocyanins peak for WOND_V sample can be seen ( Figure S1).
Recovering the results of the consumers, we can observe that the favourite one considering the colour has been the WOND_S, this could mean that the consumers prefer in the arils a more brilliant red colour rather than a dark red one, tending to the violet.

TA B L E 6
Mean scores, standard deviation and P-values of anthocyanins and ellagitannins peak area (mg/L) based on sample effect for bitterness in the consumer test. There was a significant difference between HICAZ and WOND_S. To measure astringency, the total tannins content was analysed using the Bate-Smith method.
It emerges that there is a significant difference between WOND_S and the other two products. WOND_S has the highest content in tannins (0.88 g/L) and WOND_V the lowest one (0.57 g/L). Despite the higher tannin content, WOND_S achieved the highest rating in the consumer test. This could be partly explained by the fact that, although the Bate-Smith method was used to detect the total tannins content, this method measures only proanthocyanidins. On the other hand, it is well known that in pomegranate juice a share of the tannins belong to the ellagitannin family, which is not detected by the Bate-Smith method. For this reason, ellagitannins analysis in HPLC was also performed. As reported in Table 6, it emerged that the content of punicalagin B was significantly lower in WOND_S As can be seen in Figure 2, the two different varieties, Wonderful and Hicaz, differ in the content of the ellagitannins, since the ellagic acid prevails in the samples WOND_S and WOND_M and the punicalagin B in HICAZ sample, while there is no significant difference in the content of punicalagin A between the samples.
In the supporting information, a typical HPLC chromatogram showing the ellagitannins peak can be seen ( Figure S2).

| Sensory analysis
Thirteen attributes describing appearance (size and colour of the arils), aromas (apple, red wine and red fruits), taste (sweetness, sourness and bitterness), texture (firmness, crunchiness, juiciness and seed intrusiveness) and mouthfeel sensations (astringency) were generated to characterize the sensory profile of the three types of pomegranate, as it can be seen in Table 7 However, from chemical analysis, WOND_S had the highest content in total polyphenols, the molecules responsible, among others, of bitterness. Although the consumer test did not reveal a significant difference for the bitterness, the most appreciated sample for this attribute was WOND_V; this discrepancy in the results could be due to interactions with other tastes, as stated by Keast and Breslin. 38 The values attached to juiciness are found to be higher in WOND_S (5.31) and lower in WOND_V (4.12); comparing these results with consumer reviews, we observe that the highest score was assigned to HICAZ, immediately followed with a slight difference from WOND_S. This would seem to indicate that consumers still prefer a fruit with greater juiciness.
Observing the seed intrusiveness, a negative attribute, the panel indicated WOND_V as the sample with the most intrusive seed (7.25) while WOND_S has the least intrusive one (6.25 suggest that the highest score in colour intensity is given to WOND_V (7.44) while HICAZ had the lightest colour arils (6.73); this is in line with chemical analysis, observing the content of total anthocyanins, that is a colour indicator, highest in WOND_V and lowest in HICAZ. In fact, the consumers have assigned the lowest rating to HICAZ; therefore, they seem to prefer a darker colour in the arils.

| Factors affecting preferences
The importance of pomegranate and its consumption is growing in Italy, as in other parts of the world, thanks to its important nutritional con- liking and acceptance for the juices obtained from separated arils are significantly higher than those for juice extracted pressing the whole fruit. These results suggest that consumers' acceptability, subordinated to the negative product features such as sourness or astringency, could be different according to the nature of the product.
Polyphenols (which include also tannins) and anthocyanins are not only responsible of the astringency sensation, but they are closely linked to the renowned health properties of pomegranate and the red colour of husk and arils. In this context, it should be recalled that the success of pomegranate lies properly in its health benefit properties and it is often used as an additional product in juice blends not only because of its sensory profile but also because of its antioxidant properties, which ensure better preservation, intended as technological features, 42 and greater palatability by consumers.
In this sense, due to the growing interest in healthy diets, it may be interesting to study the product WOND_V more in depth with regard to potential health properties and antioxidant activity. Indeed, our results highlight the highest content in total anthocyanin (1445 mg/L). These data are comparable with the study of Adiletta et al, 43  In fact, anthocyanins are more easily degraded at higher temperatures, as for example in Sicily (region in which WOND_S was cultivated), rather than in Veneto (region of WOND_V). This is confirmed by Schwartz et al, 15 who described differences in the chemical composition of the arils of 11 cultivars grown in Mediterranean and desert climates in Israel.
The level of total anthocyanins was significantly higher in fruits obtained from the Mediterranean area compared to those from the desert area.
However, although WOND_V is the richest variety among those analysed in anthocyanin, it should be stressed that explaining these characteristics on the label may not be enough to guide consumers' purchasing choices towards this product, considering that the willingness of consumers to compromise taste for health is narrowing more and more. 10,11

| CON CLUS IONS
In the present study, we looked for the factors affecting preferences of Italian consumers towards pomegranate arils of different varieties and cultivation sites. This has been possible thanks to three different types of analysis: sensory characterization through a trained panel, physicochemical analyses and a questionnaire for consumer preferences.
To sum up, we found that site and cultivation area affect consumers' perception and acceptance. Wonderful pomegranate from Sicily (WOND_S) was the most preferred sample, with the highest overall liking (6.65). Juiciness, arils size, colour and firmness are the attributes that affect the most the overall liking. Sourness, bitterness and astringency are the product features that are negatively correlated to consumers' acceptance. However, even products that do not fully reflect the needs of consumers as fresh products, as in the case of Wonderful pomegranate from Veneto (WOND_V), could be used in the food chain as processed products (such as juices). Therefore, our results highlight that even others less-known varieties grown in new production areas, such as Veneto, can have commercial potential. Further studies could be carried out on other local varieties in Veneto, such as Parfianka, S. Pietro, Mondrone, Dente di Cavallo, Mollar and Valenciana, to have a more appropriate overview of pomegranate cultivation and its consumers' acceptance.