Diversity in Sensory Quality and Determining Factors Influencing Mandarin Flavor Liking Livnat Goldenberg, Yossi Yaniv, Tatiana Kaplunov, Adi Doron-Faigenboim, Nir Carmi, and Ron Porat
We evaluated the sensory quality of 42 mandarin varieties that belong to 7 different natural subgroups: common mandarin (Citrus reticulata Blanco), Clementine (Citrus clementina Hort. ex. Tan), Satsuma (Citrus unshiu Marcovitch), Mediterranean mandarin (Citrus deliciosa Tenore), King mandarin (Citrus nobilis Loureiro), and mandarin hybrids, such as tangor (Citrus reticulata × Citrus sinensis) and tangelo (Citrus reticulata × Citrus paradisi). Consumer flavor acceptance tests revealed wide diversity in flavor preferences among mandarin varieties and subgroups. Furthermore, descriptive flavor-analysis tests conducted with the aid of a trained sensory panel revealed that the 9 most preferred varieties had similar flavor profiles, characterized by high sweetness, moderate to low acidity levels, low bitterness and gumminess, strong fruity and mandarin flavor, and high juiciness. The average total soluble solids (TSS) and acidity levels among the highly preferred varieties were 13.1% and 1.1%, respectively. In contrast, the 8 least preferred varieties were either too sour or gummy or had low levels of sweetness, fruity, or mandarin flavor, and either high acidity levels (>1.4%) or low TSS levels ( 7.0, on a 1–9 scale) Ora 01. 16.13 12.88 0.84 15.37 12. 23.13 12.30 Murcott 02. 11.13 13.10 1.26 10.40 02. 18.14 15.38 Shani 02. 19.13 14.20 1.41 10.05 01. 20.14 16.13 Michal 11. 13.12 13.03 0.70 18.61 10. 27.13 12.35 Merav 12. 17.12 12.23 1.17 10.46 12. 9.13 12.68 Ponkan 12. 27.12 12.19 1.05 11.67 12. 9.13 12.95 Wilking 01. 13.13 11.03 0.84 13.21 01. 6.14 14.30 Nour 12. 9.12 14.95 0.88 17.16 11. 25.13 12.53 Fina 11. 19.12 11.9 0.95 12.53 11. 26.13 12.06 Moderate-preferred mandarin varieties (5.5 > flavor acceptance scores > 7.0, on a 1–9 scale) Yifat 01. 13.13 11.15 1.55 7.20 01. 6.14 13.75 Sigal 02. 11.13 15.58 0.89 17.47 01. 14.14 12.95 Dancy 12. 27.12 11.43 1.19 9.60 12. 17.13 11.28 Lee 12. 2.12 12.23 0.91 13.47 11. 4.13 12.58 Tami 11. 13.12 13.35 0.95 14.09 11. 13.13 11.68 Ortanique 03. 11.13 13.09 1.64 7.98 03. 4.14 13.27 Yafit 01. 16.13 13.25 1.05 12.82 12. 31.13 11.35 Afourer 01. 24.13 10.00 1.22 8.30 01. 20.14 12.10 Caffin 11. 19.12 13.81 0.95 14.54 11. 13.13 13.65 Oroval 11. 3.12 10.50 0.96 10.94 10. 27.13 10.40 Orit 02. 3.13 13.18 0.98 13.56 01. 22.14 12.78 King 02. 27.13 12.90 1.92 6.76 03. 4.14 13.68 Edit 12. 11.12 12.43 1.13 11.00 12. 1.13 12.33 Fairchild 12. 9.12 15.75 1.03 15.52 11. 25.13 14.30 Avana 01. 1.13 11.13 1.06 10.55 12. 1.13 11.05 Pazit 01. 24.13 10.28 0.95 10.82 01. 14.14 10.42 Vinola 01. 16.13 10.53 1.66 6.43 02. 9.14 14.80 Minneola 01. 1.13 10.30 1.35 7.68 12. 17.13 12.38 Niva 11. 26.12 13.63 1.27 10.78 11. 13.13 15.18 Temple 01. 1.13 13.15 1.97 6.72 12. 31.13 13.65 Odem 02. 3.13 11.60 0.50 23.20 01. 14.14 13.28 Tacle 01. 1.13 13.45 1.58 8.54 12. 9.13 13.15 Orlando 12. 27.12 10.58 1.27 8.31 11. 26.13 10.68 Satsuma Owari 10. 23.12 12.20 1.35 9.11 10. 13.13 10.65 Cami 01. 13.13 10.48 1.15 9.09 12. 23.13 9.45 Low-preferred mandarin varieties (flavor acceptance scores < 5.5, on a 1–9 scale) GP-377 10. 23.12 11.33 0.99 11.48 10. 15.13 10.88 Hadas 03. 17.13 12.05 1.83 6.61 03. 4.14 14.80 Rishon 10. 23.12 10.23 0.77 13.27 10. 3.13 10.65 Kiyomi 02. 4.13 11.20 1.69 6.68 01. 22.14 10.28 Satsuma Dubashi-Beni 11. 19.12 11.42 1.44 7.93 11. 13.13 11.28 Yusuf Efendi 12. 17.13 9.36 0.95 9.85 12. 9.13 9.44 Satsuma Okitsu 10. 16.12 9.58 1.20 7.98 10. 3.13 8.55 Fallglo 11. 14.12 13.38 1.80 7.44 11. 13.13 13.55
evaluation of various taste (sweetness, sourness, and bitterness), aroma (fruitiness and mandarin flavor), and mouth feel (juiciness and gumminess) attributes. We observed significant (P ࣘ 0.05) positive correlations between consumers’ flavor acceptance and perceptions of sweetness (R = 0.73), fruitiness (R = 0.67), and mandarin flavor (R = 0.75), and negative correlations between their flavor acceptance and perception of sourness (R = −0.44) and bitterness (R = −0.50; Figure 3). The 3 bitter varieties (bitter scores ࣙ 2) were “Rishon,” “Fallglo,” and “Orlando,” and among them “Rishon” and “Fallglo” were among the 8 leastpreferred varieties indicating the negative impact of bitterness on mandarin flavor likability (Figure 3). Hierarchical clustering and Pearson correlations among all chemical and sensory attributes tested revealed tight positive correlations of flavor acceptance with sensations of sweetness, fruitiness, and mandarin flavor, and negative correlations with acidity levels and sensations of sourness, bitterness, and gumminess (Figure 4). Clustering and Pearson
Means Ripening Ripening ratio TSS (%) Acid (%) ratio
1.01 1.49 1.39 0.68 1.21 0.81 1.48 0.87 0.91
12.40 10.49 11.60 19.33 10.50 16.04 9.68 14.36 13.25
12.59 14.24 15.16 12.69 12.45 12.57 12.66 13.74 11.98
0.93 1.37 1.40 0.69 1.19 0.93 1.16 0.88 0.93
13.89 10.45 10.83 18.97 10.48 13.86 11.44 15.76 12.94
1.47 0.68 0.92 1.01 0.89 1.74 1.37 1.10 0.89 0.88 1.35 1.55 1.07 1.50 1.01 1.01 1.96 2.05 1.42 2.05 0.64 1.67 1.31 1.28 0.89
9.50 19.02 12.23 12.49 13.21 7.62 8.31 11.22 15.34 11.81 9.48 8.89 11.59 9.54 11.01 10.32 7.58 6.05 10.71 6.69 21.27 7.90 8.20 8.34 10.77
12.45 14.26 11.35 12.40 12.51 13.18 12.30 11.05 13.73 10.45 12.98 13.29 12.38 15.03 11.09 0.35 12.66 11.34 14.40 13.40 12.44 13.30 10.63 11.43 9.96
1.51 0.79 1.06 0.96 0.92 1.69 1.21 1.16 0.92 0.92 1.16 1.73 1.10 1.27 1.03 0.98 1.81 1.70 1.34 2.01 0.57 1.62 1.29 1.32 1.02
8.35 18.24 10.91 12.98 13.65 7.84 10.57 9.76 14.89 11.43 11.52 7.82 11.30 12.53 10.78 10.61 7.01 6.86 10.74 6.71 22.24 8.22 8.26 8.73 9.93
0.98 2.33 0.70 1.50 1.36 0.99 1.17 1.81
11.12 6.39 15.41 6.86 8.29 9.54 7.43 7.49
11.10 13.43 10.44 10.74 11.35 9.40 9.06 13.46
0.99 2.08 0.74 1.59 1.40 0.97 1.18 1.81
11.30 6.50 14.34 6.77 8.14 9.71 7.70 7.46
correlations also revealed tight correlations between acidity levels and sourness, which were both negatively correlated with fruit ripening ratio (Figure 4). In previous studies of mandarin juices, it was suggested that flavor liking was associated with rich mandarin and fresh flavor attributes (Carbonell and others 2009).
Flavor profiles and TSS and acidity levels of most- and least-preferred varieties To evaluate the factors that affect liking of mandarin flavors, we compared the TSS and acidity levels in juices of all 42 mandarin varieties (Table 1). The results show that in the 9 most preferred varieties (flavor acceptance scores > 7.0, on a 1 to 9 scale) juice TSS levels were in the range of 12.0% to 15.2%, with an average of 13.1%; and juice acidity levels were in the range of 0.7% to 1.4%, with an average of 1.1% (Table 1). The average ripening ratio among the highly preferred mandarin varieties was 13.2 (Table 1). In contrast, we found that in the 8 least preferred varieties (flavor Vol. 80, Nr. 2, 2015 r Journal of Food Science S423
S: Sensory & Food Quality
2012/2013
Diversity in sensory quality of mandarins . . . acceptance scores < 5.5, on a 1 to 9 scale), either juice TSS levels were below the range of 12.0% to 15.2% detected in the highly preferred varieties, or juice acidity levels were above the range of 0.7% to 1.4% detected in the highly preferred varieties (Table 1). The low TSS levels and/or high acidity levels in the least preferred varieties resulted in fruits that were either less sweet or more sour
Fallglo 4.00 Juicy
Gummy
Satsuma Okitsu 4.66
Sweet 8 6 4 2 0
Mandarin f lavor
Sour
Juicy
Bitter
Gummy
Satsuma Dubashi-Beni 4.88 Juicy
Gummy Mandarin f lavor
Sour
Juicy
Bitter
Gummy Mandarin f lavor
Fruity
Bitter
Juicy
Gummy
S: Sensory & Food Quality
Mandarin f lavor
Gummy Mandarin f lavor
Fruity
Sweet 8 6 4 2 0
Sour
Figure 6–Flavor profiles of the 8 least preferred mandarin varieties. Flavor profiles were evaluated with the aid of a trained sensory panel, and data are means of scores given by 10 testers.
Bitter Fruity
Rishon 5.34
Sweet 8 6 4 2 0
Juicy
Sour
Bitter
Gummy Mandarin f lavor
Fruity
Hadas 5.36
Sweet 8 6 4 2 0
Juicy
Kiyomi 5.28
Sweet 8 6 4 2 0
Sour
Mandarin f lavor
Fruity
Yusuf Efendi 4.75
Sweet 8 6 4 2 0
than those of the preferred varieties (Table 1). Thus, in 6 out of the 8 least preferred varieties, fruit ripening ratios were below the range detected in the most preferred varieties (Table 1). Worth notice is that for most mandarin varieties tested, juice TSS and acidity levels were similar among the 2012/2013 and 2013/2014 growing seasons, but for some varieties we detected notable
Sweet 8 6 4 2 0
Sour
Bitter Fruity
GP-377 5.40 Sour
Bitter Fruity
Juicy
Gummy Mandarin f lavor
Sweet 8 6 4 2 0
Sour
Bitter Fruity
Figure 7–Principle component analysis (PCA) of all mandarin varieties. The 9 most highly preferred varieties indicated in Figure 5 are encircled by an orange line, and the 8 least preferred varieties indicated in Figure 6 are encircled by a green line. The vector lines indicate TSS, acidity, and various flavor attributes.
S424 Journal of Food Science r Vol. 80, Nr. 2, 2015
Diversity in sensory quality of mandarins . . .
Principle component analysis PCA is routinely used in order to explore flavor diversity among varieties and to identify drivers of flavor liking in various fruits, such as peaches, nectarines, and figs (King and others 2012; Delgado and others 2013). To examine and define the relationships among flavor attributes, TSS and acidity levels and the genetic backgrounds of the mandarin varieties, we conducted a PCA (Figure 7). The PCA data clearly distinguished between the more and less preferred varieties, and demonstrated that high acceptance was tightly correlated with high TSS levels and perceptions of sweetness, fruitiness, and mandarin flavor, whereas the less acceptable varieties were rather correlated with high acidity levels, and perceptions of sourness, bitterness, and gumminess (Figure 7).
Conclusions Despite the increasing importance of easy-to-peel mandarins in global citrus markets, relatively little is yet known regarding their sensory quality or the factors that determine consumer flavor preferences among fresh fruits. In the present study, we conducted consumer flavor acceptance and descriptive sensory analysis tests of 42 different mandarin varieties belonging to 7 different natural subgroups, and observed wide variations in flavor preferences and sensory profiles (Figure 2, 5 and 6). We found that the flavors of common mandarins, Clementines and tangors (mandarin × orange hybrids) were somewhat preferred to those of Satsumas, Mediterranean mandarins, and tangelos (mandarin × grapefruit hybrids; Figure 2). The average TSS and acidity levels of highly preferred mandarin varieties were 13.1% and 1.1%, respectively, and these varieties had an average ripening ratio of 13.2 (Table 1). Moreover, we found that flavor acceptance showed significant positive correlations with sweetness, fruitiness, and mandarin flavor, and negative correlations with perceptions of sourness and bitterness (Figure 3). Finally, Pearson correlations and PCA revealed
that high TSS levels and perceptions of sweetness, fruitiness, and mandarin flavor were positive determinants, whereas high acidity levels and perceptions of sourness, bitterness, and gumminess were negative determinants of flavor liking of mandarins (Figure 6 and 7).
Acknowledgment This manuscript is contribution no. 695/14 from the Agricultural Research Organization, the Volcani Center, P.O. Box 6, Bet Dagan 50250, Israel.
References Aleza P, Juarez J, Hernandez M, Pina JA, Ollitrault P, Navarro L. 2009. Recovery and characterization of a Citrus clementina Hort. ex Tan. ‘Clemenules’ haploid plant selected to establish the reference whole Citrus genome sequence. BMC Plant Biol 9:110. Ares G, Barrios S, Lareo C, Lema P. 2009. Development of a sensory quality index for strawberries based on correlation between sensory data and consumer perception. Postharvest Biol Technol 52:97–102. Beltran F, Perez-Lopez AJ, Lopez-Nicolas JM, Carbonell-Brrachina AA. 2008. Effect of mandarin cultivar on quality of mandarin juice. Food Sci Technol Intl 14:307–13. Benjamin G, Tietel Z, Porat R. 2013. Effects of rootstock/scion combinations on the flavor of citrus fruit. J Agr Food Chem 61:11286–94. Bermejo A, Cano A. 2012. Analysis of nutritional constituents in twenty citrus cultivars from the Mediterranean area at different stages of ripening. Food Nutr Sci 3:639–50. Carbonell L, Izquierdo L, Carbonell I. 2007. Sensory analysis of Spanish mandarin juices. Selection of attributes and panel performance. Food Qual Prefer 18:329–41. Carbonell L, Bayarri S, Navarro JL, Carbonell I, Izquierdo L. 2009. Sensory profile and acceptability of juices from mandarin varieties and hybrids. Food Sci Technol Intl 15:375–85. Daillant-Spinnler B, MacFie HJH, Beyts PK, Hedderley D. 1996. Relationships between perceived sensory properties and major preference directions of 12 varieties of apples from the Southern hemisphere. Food Qual Prefer 7:113–26. Delgado C, Guinard J-X. 2011. How do consumer hedonic ratings for extra virgin olive oil relate to quality ratings by experts and descriptive analysis ratings? Food Qual Prefer 22:213–25. Delgado C, Cristoto GM, Heymann H, Cristoto CH. 2013. Determining the primary drivers of liking to predict consumers’ acceptance of fresh nectarines and peaches. J Food Sci 78:S605– 14. Goldenberg L, Yaniv Y, Kaplunov T, Doron-Faigenboim A, Porat R, Carmi N. 2014 .Genetic diversity among mandarins in fruit-quality traits. J Agric Food Chem. 62:4938–46. Gonzalez-Mas MC, Rambla JL, Alamar MC, Gutierrez A, Granell A. 2011. Comparative analysis of the volatile fraction of fruit juice from different citrus species. PLoS One 6(7): e22016. doi:10.1371/journal.pone.0022016. Goodner KL, Rouseff RL, Hofsommer HJ. 2001. Orange, mandarin, and hybrid classification using multivariate statistics based on carotenoid profiles. J Agr Food Chem 49:1146–50. Grierson W. 2006. Maturity and grade standards. In: Wardowsky WF, Miller WM, Hall DJ, Grierson G, editors. Fresh citrus fruits. 2nd ed. Longboat Key, FL: Florida Science Source, Inc. p 43–8. Hodgson RW. 1967. Horticultural varieties of citrus. In: Reuther W, Webber HJ, Batchler LD, editors. The citrus industry, Vol. 1. Berkeley, CA: University of California. p 431–588. King ES, Hopfer H, Haug MT, Orsi JD, Heymann H, Cristoto GM, Cristoto CH. 2012. Describing the appearance and flavor profiles of fresh fig (Ficus carica L.) cultivars. J Food Sci 77:S419–29. Lawless HT, Heymann H. 1999. Sensory evaluation of food. New York, NY: Kluwer Academic Publishers. Malundo TMM, Shewfelt RL, Ware GO, Baldwin AEA. 2001. An alternative method for relating consumer and descriptive data used to identify critical flavor properties of mango (Mangifera indica l.). J Sens Stud 16:199–214. Mayuoni-Kirshinbaum L, Bar-Ya’akov I, Hatib K, Holland D, Porat R. 2013. Genetic diversity and sensory preference in pomegranate fruits. Fruits 68:517–24. Miyazaki T, Plotto A, Goodner K, Gmitter FG. 2011. Distribution of aroma volatile compounds in tangerine hybrids and proposed inheritance. J Sci Food Agric 91:449–60. Miyazaki T, Plotto A, Baldwin EA, Reyes-De-Corcuera JI, Gmitter FG. 2012. Aroma characterization of tangerine hybrids by gas-chromatography-olfactometry and sensory evaluation. J Sci Food Agric 92:727–35. Obenland D, Collin S, Mackey B, Sievert J, Fjeld K, Arpaia ML. 2009. Determinants of flavor acceptability during the maturation of navel oranges. Postharv Biol Technol 52:156–63. Puri M, Marwaha SS, Kothari RM, Kennedy JF. 1996. Biochemical basis of bitterness in citrus fruit juice and biotech approaches for debittering. Crit Rev Biotechnol 16:145–55. Ramana KVR, Govindarajan VS, Raganna S. 1981. Citrus fruits—varieties, chemistry, technology, and quality evaluation. Part I: Varieties, production, handling and storage. Crit Rev Food Sci Nutr 15:383–431. Shamir R, Sharan R. 2002. Algorithmic approaches to clustering gene expression data. In: Jiang T, Smith T, Xu Y, Zhang MQ, editors. Current topics in computational biology. Cambridge, MA: MIT Press. p 269–99. Tietel Z, Bar E, Lewinsohn E, Feldmesser E, Fallik E, Porat R. 2010. Effects of wax coatings and postharvest storage on sensory quality and aroma volatiles composition of ’Mor’ mandarins. J Sci Food Agric 90:995–1007. Tietel Z, Bar E, Lewinsohn E, Fallik E, Porat R. 2011a. Sensory and aroma profiling of fresh and stored ’Or’ mandarins. Acta Hortic 892:373–82. Tietel Z, Plotto A, Fallik E, Lewinsohn E, Porat R. 2011b. Taste and aroma of fresh and stored mandarins. J Sci Food Agric 91:14–23. USDA (United States Department of Agriculture). 2014. Citrus: World markets and trade. Available at: http://apps.fas.usda.gov/psdonline/circulars/citrus.pdf. Accessed May 2014.
Vol. 80, Nr. 2, 2015 r Journal of Food Science S425
S: Sensory & Food Quality
differences among seasons (Table 1). To the best of our knowledge, this is the first reported finding that the optimal TSS and acidity levels for high flavor liking in mandarins are approximately 13.1% and approximately 1.1%, respectively (Table 1). In a previous study, conducted in California, it was reported that flavor liking of “Navel” oranges increased significantly when juice TSS levels increased from 8% to 10% during ripening (Obenland and others 2009). It is also known that high acidity of early-season mandarin varieties impairs flavor quality (Grierson 2006). In the present study, descriptive sensory analysis tests revealed that all 9 most preferred mandarin varieties exhibited more or less similar sensory profiles (Figure 5A), which were characterized by high sweetness, moderate to low sourness, low bitterness and gumminess, strong fruity and mandarin flavor, and high juiciness (Figure 5B). In contrast, the flavor profiles of the least preferred varieties were distinguished from those of the most preferred varieties according to one or more parameters (Figure 6). For example: “Fallglo” and “Hadas” were sourer; “Fallglo,” Satsuma “Okitso, ” Satsuma “Dubashi-Beni,” “Kiomi,” “Rishon,” “Hadas,” and “GP-377” had lower sweetness scores; and “Fallglo,” Satsuma “Okitso,” “Kiomi,” and “Rishon” had weak fruity and mandarin flavor (Figure 6). Interestingly, we previously obtained similar results for pomegranate fruits, in which out of 18 distinct varieties all most preferred varieties had more or less similar sensory profiles, whereas the least preferred varieties differed from each other and were considered to be either too sour or bitter, to have weak pomegranate and fruity flavor, or to have hard seeds (MayuoniKirshinbaum and others 2013).
We evaluated the sensory quality of 42 mandarin varieties that belong to 7 different natural subgroups: common mandarin (Citrus reticulata Blanco), Clementine (Citrus clementina Hort. ex. Tan), Satsuma (Citrus unshiu Marcovitch), Mediterranean mandarin (Citrus deliciosa Tenore), King mandarin (Citrus nobilis Loureiro), and mandarin hybrids, such as tangor (Citrus reticulata × Citrus sinensis) and tangelo (Citrus reticulata × Citrus paradisi). Consumer flavor acceptance tests revealed wide diversity in flavor preferences among mandarin varieties and subgroups. Furthermore, descriptive flavor-analysis tests conducted with the aid of a trained sensory panel revealed that the 9 most preferred varieties had similar flavor profiles, characterized by high sweetness, moderate to low acidity levels, low bitterness and gumminess, strong fruity and mandarin flavor, and high juiciness. The average total soluble solids (TSS) and acidity levels among the highly preferred varieties were 13.1% and 1.1%, respectively. In contrast, the 8 least preferred varieties were either too sour or gummy or had low levels of sweetness, fruity, or mandarin flavor, and either high acidity levels (>1.4%) or low TSS levels ( 7.0, on a 1–9 scale) Ora 01. 16.13 12.88 0.84 15.37 12. 23.13 12.30 Murcott 02. 11.13 13.10 1.26 10.40 02. 18.14 15.38 Shani 02. 19.13 14.20 1.41 10.05 01. 20.14 16.13 Michal 11. 13.12 13.03 0.70 18.61 10. 27.13 12.35 Merav 12. 17.12 12.23 1.17 10.46 12. 9.13 12.68 Ponkan 12. 27.12 12.19 1.05 11.67 12. 9.13 12.95 Wilking 01. 13.13 11.03 0.84 13.21 01. 6.14 14.30 Nour 12. 9.12 14.95 0.88 17.16 11. 25.13 12.53 Fina 11. 19.12 11.9 0.95 12.53 11. 26.13 12.06 Moderate-preferred mandarin varieties (5.5 > flavor acceptance scores > 7.0, on a 1–9 scale) Yifat 01. 13.13 11.15 1.55 7.20 01. 6.14 13.75 Sigal 02. 11.13 15.58 0.89 17.47 01. 14.14 12.95 Dancy 12. 27.12 11.43 1.19 9.60 12. 17.13 11.28 Lee 12. 2.12 12.23 0.91 13.47 11. 4.13 12.58 Tami 11. 13.12 13.35 0.95 14.09 11. 13.13 11.68 Ortanique 03. 11.13 13.09 1.64 7.98 03. 4.14 13.27 Yafit 01. 16.13 13.25 1.05 12.82 12. 31.13 11.35 Afourer 01. 24.13 10.00 1.22 8.30 01. 20.14 12.10 Caffin 11. 19.12 13.81 0.95 14.54 11. 13.13 13.65 Oroval 11. 3.12 10.50 0.96 10.94 10. 27.13 10.40 Orit 02. 3.13 13.18 0.98 13.56 01. 22.14 12.78 King 02. 27.13 12.90 1.92 6.76 03. 4.14 13.68 Edit 12. 11.12 12.43 1.13 11.00 12. 1.13 12.33 Fairchild 12. 9.12 15.75 1.03 15.52 11. 25.13 14.30 Avana 01. 1.13 11.13 1.06 10.55 12. 1.13 11.05 Pazit 01. 24.13 10.28 0.95 10.82 01. 14.14 10.42 Vinola 01. 16.13 10.53 1.66 6.43 02. 9.14 14.80 Minneola 01. 1.13 10.30 1.35 7.68 12. 17.13 12.38 Niva 11. 26.12 13.63 1.27 10.78 11. 13.13 15.18 Temple 01. 1.13 13.15 1.97 6.72 12. 31.13 13.65 Odem 02. 3.13 11.60 0.50 23.20 01. 14.14 13.28 Tacle 01. 1.13 13.45 1.58 8.54 12. 9.13 13.15 Orlando 12. 27.12 10.58 1.27 8.31 11. 26.13 10.68 Satsuma Owari 10. 23.12 12.20 1.35 9.11 10. 13.13 10.65 Cami 01. 13.13 10.48 1.15 9.09 12. 23.13 9.45 Low-preferred mandarin varieties (flavor acceptance scores < 5.5, on a 1–9 scale) GP-377 10. 23.12 11.33 0.99 11.48 10. 15.13 10.88 Hadas 03. 17.13 12.05 1.83 6.61 03. 4.14 14.80 Rishon 10. 23.12 10.23 0.77 13.27 10. 3.13 10.65 Kiyomi 02. 4.13 11.20 1.69 6.68 01. 22.14 10.28 Satsuma Dubashi-Beni 11. 19.12 11.42 1.44 7.93 11. 13.13 11.28 Yusuf Efendi 12. 17.13 9.36 0.95 9.85 12. 9.13 9.44 Satsuma Okitsu 10. 16.12 9.58 1.20 7.98 10. 3.13 8.55 Fallglo 11. 14.12 13.38 1.80 7.44 11. 13.13 13.55
evaluation of various taste (sweetness, sourness, and bitterness), aroma (fruitiness and mandarin flavor), and mouth feel (juiciness and gumminess) attributes. We observed significant (P ࣘ 0.05) positive correlations between consumers’ flavor acceptance and perceptions of sweetness (R = 0.73), fruitiness (R = 0.67), and mandarin flavor (R = 0.75), and negative correlations between their flavor acceptance and perception of sourness (R = −0.44) and bitterness (R = −0.50; Figure 3). The 3 bitter varieties (bitter scores ࣙ 2) were “Rishon,” “Fallglo,” and “Orlando,” and among them “Rishon” and “Fallglo” were among the 8 leastpreferred varieties indicating the negative impact of bitterness on mandarin flavor likability (Figure 3). Hierarchical clustering and Pearson correlations among all chemical and sensory attributes tested revealed tight positive correlations of flavor acceptance with sensations of sweetness, fruitiness, and mandarin flavor, and negative correlations with acidity levels and sensations of sourness, bitterness, and gumminess (Figure 4). Clustering and Pearson
Means Ripening Ripening ratio TSS (%) Acid (%) ratio
1.01 1.49 1.39 0.68 1.21 0.81 1.48 0.87 0.91
12.40 10.49 11.60 19.33 10.50 16.04 9.68 14.36 13.25
12.59 14.24 15.16 12.69 12.45 12.57 12.66 13.74 11.98
0.93 1.37 1.40 0.69 1.19 0.93 1.16 0.88 0.93
13.89 10.45 10.83 18.97 10.48 13.86 11.44 15.76 12.94
1.47 0.68 0.92 1.01 0.89 1.74 1.37 1.10 0.89 0.88 1.35 1.55 1.07 1.50 1.01 1.01 1.96 2.05 1.42 2.05 0.64 1.67 1.31 1.28 0.89
9.50 19.02 12.23 12.49 13.21 7.62 8.31 11.22 15.34 11.81 9.48 8.89 11.59 9.54 11.01 10.32 7.58 6.05 10.71 6.69 21.27 7.90 8.20 8.34 10.77
12.45 14.26 11.35 12.40 12.51 13.18 12.30 11.05 13.73 10.45 12.98 13.29 12.38 15.03 11.09 0.35 12.66 11.34 14.40 13.40 12.44 13.30 10.63 11.43 9.96
1.51 0.79 1.06 0.96 0.92 1.69 1.21 1.16 0.92 0.92 1.16 1.73 1.10 1.27 1.03 0.98 1.81 1.70 1.34 2.01 0.57 1.62 1.29 1.32 1.02
8.35 18.24 10.91 12.98 13.65 7.84 10.57 9.76 14.89 11.43 11.52 7.82 11.30 12.53 10.78 10.61 7.01 6.86 10.74 6.71 22.24 8.22 8.26 8.73 9.93
0.98 2.33 0.70 1.50 1.36 0.99 1.17 1.81
11.12 6.39 15.41 6.86 8.29 9.54 7.43 7.49
11.10 13.43 10.44 10.74 11.35 9.40 9.06 13.46
0.99 2.08 0.74 1.59 1.40 0.97 1.18 1.81
11.30 6.50 14.34 6.77 8.14 9.71 7.70 7.46
correlations also revealed tight correlations between acidity levels and sourness, which were both negatively correlated with fruit ripening ratio (Figure 4). In previous studies of mandarin juices, it was suggested that flavor liking was associated with rich mandarin and fresh flavor attributes (Carbonell and others 2009).
Flavor profiles and TSS and acidity levels of most- and least-preferred varieties To evaluate the factors that affect liking of mandarin flavors, we compared the TSS and acidity levels in juices of all 42 mandarin varieties (Table 1). The results show that in the 9 most preferred varieties (flavor acceptance scores > 7.0, on a 1 to 9 scale) juice TSS levels were in the range of 12.0% to 15.2%, with an average of 13.1%; and juice acidity levels were in the range of 0.7% to 1.4%, with an average of 1.1% (Table 1). The average ripening ratio among the highly preferred mandarin varieties was 13.2 (Table 1). In contrast, we found that in the 8 least preferred varieties (flavor Vol. 80, Nr. 2, 2015 r Journal of Food Science S423
S: Sensory & Food Quality
2012/2013
Diversity in sensory quality of mandarins . . . acceptance scores < 5.5, on a 1 to 9 scale), either juice TSS levels were below the range of 12.0% to 15.2% detected in the highly preferred varieties, or juice acidity levels were above the range of 0.7% to 1.4% detected in the highly preferred varieties (Table 1). The low TSS levels and/or high acidity levels in the least preferred varieties resulted in fruits that were either less sweet or more sour
Fallglo 4.00 Juicy
Gummy
Satsuma Okitsu 4.66
Sweet 8 6 4 2 0
Mandarin f lavor
Sour
Juicy
Bitter
Gummy
Satsuma Dubashi-Beni 4.88 Juicy
Gummy Mandarin f lavor
Sour
Juicy
Bitter
Gummy Mandarin f lavor
Fruity
Bitter
Juicy
Gummy
S: Sensory & Food Quality
Mandarin f lavor
Gummy Mandarin f lavor
Fruity
Sweet 8 6 4 2 0
Sour
Figure 6–Flavor profiles of the 8 least preferred mandarin varieties. Flavor profiles were evaluated with the aid of a trained sensory panel, and data are means of scores given by 10 testers.
Bitter Fruity
Rishon 5.34
Sweet 8 6 4 2 0
Juicy
Sour
Bitter
Gummy Mandarin f lavor
Fruity
Hadas 5.36
Sweet 8 6 4 2 0
Juicy
Kiyomi 5.28
Sweet 8 6 4 2 0
Sour
Mandarin f lavor
Fruity
Yusuf Efendi 4.75
Sweet 8 6 4 2 0
than those of the preferred varieties (Table 1). Thus, in 6 out of the 8 least preferred varieties, fruit ripening ratios were below the range detected in the most preferred varieties (Table 1). Worth notice is that for most mandarin varieties tested, juice TSS and acidity levels were similar among the 2012/2013 and 2013/2014 growing seasons, but for some varieties we detected notable
Sweet 8 6 4 2 0
Sour
Bitter Fruity
GP-377 5.40 Sour
Bitter Fruity
Juicy
Gummy Mandarin f lavor
Sweet 8 6 4 2 0
Sour
Bitter Fruity
Figure 7–Principle component analysis (PCA) of all mandarin varieties. The 9 most highly preferred varieties indicated in Figure 5 are encircled by an orange line, and the 8 least preferred varieties indicated in Figure 6 are encircled by a green line. The vector lines indicate TSS, acidity, and various flavor attributes.
S424 Journal of Food Science r Vol. 80, Nr. 2, 2015
Diversity in sensory quality of mandarins . . .
Principle component analysis PCA is routinely used in order to explore flavor diversity among varieties and to identify drivers of flavor liking in various fruits, such as peaches, nectarines, and figs (King and others 2012; Delgado and others 2013). To examine and define the relationships among flavor attributes, TSS and acidity levels and the genetic backgrounds of the mandarin varieties, we conducted a PCA (Figure 7). The PCA data clearly distinguished between the more and less preferred varieties, and demonstrated that high acceptance was tightly correlated with high TSS levels and perceptions of sweetness, fruitiness, and mandarin flavor, whereas the less acceptable varieties were rather correlated with high acidity levels, and perceptions of sourness, bitterness, and gumminess (Figure 7).
Conclusions Despite the increasing importance of easy-to-peel mandarins in global citrus markets, relatively little is yet known regarding their sensory quality or the factors that determine consumer flavor preferences among fresh fruits. In the present study, we conducted consumer flavor acceptance and descriptive sensory analysis tests of 42 different mandarin varieties belonging to 7 different natural subgroups, and observed wide variations in flavor preferences and sensory profiles (Figure 2, 5 and 6). We found that the flavors of common mandarins, Clementines and tangors (mandarin × orange hybrids) were somewhat preferred to those of Satsumas, Mediterranean mandarins, and tangelos (mandarin × grapefruit hybrids; Figure 2). The average TSS and acidity levels of highly preferred mandarin varieties were 13.1% and 1.1%, respectively, and these varieties had an average ripening ratio of 13.2 (Table 1). Moreover, we found that flavor acceptance showed significant positive correlations with sweetness, fruitiness, and mandarin flavor, and negative correlations with perceptions of sourness and bitterness (Figure 3). Finally, Pearson correlations and PCA revealed
that high TSS levels and perceptions of sweetness, fruitiness, and mandarin flavor were positive determinants, whereas high acidity levels and perceptions of sourness, bitterness, and gumminess were negative determinants of flavor liking of mandarins (Figure 6 and 7).
Acknowledgment This manuscript is contribution no. 695/14 from the Agricultural Research Organization, the Volcani Center, P.O. Box 6, Bet Dagan 50250, Israel.
References Aleza P, Juarez J, Hernandez M, Pina JA, Ollitrault P, Navarro L. 2009. Recovery and characterization of a Citrus clementina Hort. ex Tan. ‘Clemenules’ haploid plant selected to establish the reference whole Citrus genome sequence. BMC Plant Biol 9:110. Ares G, Barrios S, Lareo C, Lema P. 2009. Development of a sensory quality index for strawberries based on correlation between sensory data and consumer perception. Postharvest Biol Technol 52:97–102. Beltran F, Perez-Lopez AJ, Lopez-Nicolas JM, Carbonell-Brrachina AA. 2008. Effect of mandarin cultivar on quality of mandarin juice. Food Sci Technol Intl 14:307–13. Benjamin G, Tietel Z, Porat R. 2013. Effects of rootstock/scion combinations on the flavor of citrus fruit. J Agr Food Chem 61:11286–94. Bermejo A, Cano A. 2012. Analysis of nutritional constituents in twenty citrus cultivars from the Mediterranean area at different stages of ripening. Food Nutr Sci 3:639–50. Carbonell L, Izquierdo L, Carbonell I. 2007. Sensory analysis of Spanish mandarin juices. Selection of attributes and panel performance. Food Qual Prefer 18:329–41. Carbonell L, Bayarri S, Navarro JL, Carbonell I, Izquierdo L. 2009. Sensory profile and acceptability of juices from mandarin varieties and hybrids. Food Sci Technol Intl 15:375–85. Daillant-Spinnler B, MacFie HJH, Beyts PK, Hedderley D. 1996. Relationships between perceived sensory properties and major preference directions of 12 varieties of apples from the Southern hemisphere. Food Qual Prefer 7:113–26. Delgado C, Guinard J-X. 2011. How do consumer hedonic ratings for extra virgin olive oil relate to quality ratings by experts and descriptive analysis ratings? Food Qual Prefer 22:213–25. Delgado C, Cristoto GM, Heymann H, Cristoto CH. 2013. Determining the primary drivers of liking to predict consumers’ acceptance of fresh nectarines and peaches. J Food Sci 78:S605– 14. Goldenberg L, Yaniv Y, Kaplunov T, Doron-Faigenboim A, Porat R, Carmi N. 2014 .Genetic diversity among mandarins in fruit-quality traits. J Agric Food Chem. 62:4938–46. Gonzalez-Mas MC, Rambla JL, Alamar MC, Gutierrez A, Granell A. 2011. Comparative analysis of the volatile fraction of fruit juice from different citrus species. PLoS One 6(7): e22016. doi:10.1371/journal.pone.0022016. Goodner KL, Rouseff RL, Hofsommer HJ. 2001. Orange, mandarin, and hybrid classification using multivariate statistics based on carotenoid profiles. J Agr Food Chem 49:1146–50. Grierson W. 2006. Maturity and grade standards. In: Wardowsky WF, Miller WM, Hall DJ, Grierson G, editors. Fresh citrus fruits. 2nd ed. Longboat Key, FL: Florida Science Source, Inc. p 43–8. Hodgson RW. 1967. Horticultural varieties of citrus. In: Reuther W, Webber HJ, Batchler LD, editors. The citrus industry, Vol. 1. Berkeley, CA: University of California. p 431–588. King ES, Hopfer H, Haug MT, Orsi JD, Heymann H, Cristoto GM, Cristoto CH. 2012. Describing the appearance and flavor profiles of fresh fig (Ficus carica L.) cultivars. J Food Sci 77:S419–29. Lawless HT, Heymann H. 1999. Sensory evaluation of food. New York, NY: Kluwer Academic Publishers. Malundo TMM, Shewfelt RL, Ware GO, Baldwin AEA. 2001. An alternative method for relating consumer and descriptive data used to identify critical flavor properties of mango (Mangifera indica l.). J Sens Stud 16:199–214. Mayuoni-Kirshinbaum L, Bar-Ya’akov I, Hatib K, Holland D, Porat R. 2013. Genetic diversity and sensory preference in pomegranate fruits. Fruits 68:517–24. Miyazaki T, Plotto A, Goodner K, Gmitter FG. 2011. Distribution of aroma volatile compounds in tangerine hybrids and proposed inheritance. J Sci Food Agric 91:449–60. Miyazaki T, Plotto A, Baldwin EA, Reyes-De-Corcuera JI, Gmitter FG. 2012. Aroma characterization of tangerine hybrids by gas-chromatography-olfactometry and sensory evaluation. J Sci Food Agric 92:727–35. Obenland D, Collin S, Mackey B, Sievert J, Fjeld K, Arpaia ML. 2009. Determinants of flavor acceptability during the maturation of navel oranges. Postharv Biol Technol 52:156–63. Puri M, Marwaha SS, Kothari RM, Kennedy JF. 1996. Biochemical basis of bitterness in citrus fruit juice and biotech approaches for debittering. Crit Rev Biotechnol 16:145–55. Ramana KVR, Govindarajan VS, Raganna S. 1981. Citrus fruits—varieties, chemistry, technology, and quality evaluation. Part I: Varieties, production, handling and storage. Crit Rev Food Sci Nutr 15:383–431. Shamir R, Sharan R. 2002. Algorithmic approaches to clustering gene expression data. In: Jiang T, Smith T, Xu Y, Zhang MQ, editors. Current topics in computational biology. Cambridge, MA: MIT Press. p 269–99. Tietel Z, Bar E, Lewinsohn E, Feldmesser E, Fallik E, Porat R. 2010. Effects of wax coatings and postharvest storage on sensory quality and aroma volatiles composition of ’Mor’ mandarins. J Sci Food Agric 90:995–1007. Tietel Z, Bar E, Lewinsohn E, Fallik E, Porat R. 2011a. Sensory and aroma profiling of fresh and stored ’Or’ mandarins. Acta Hortic 892:373–82. Tietel Z, Plotto A, Fallik E, Lewinsohn E, Porat R. 2011b. Taste and aroma of fresh and stored mandarins. J Sci Food Agric 91:14–23. USDA (United States Department of Agriculture). 2014. Citrus: World markets and trade. Available at: http://apps.fas.usda.gov/psdonline/circulars/citrus.pdf. Accessed May 2014.
Vol. 80, Nr. 2, 2015 r Journal of Food Science S425
S: Sensory & Food Quality
differences among seasons (Table 1). To the best of our knowledge, this is the first reported finding that the optimal TSS and acidity levels for high flavor liking in mandarins are approximately 13.1% and approximately 1.1%, respectively (Table 1). In a previous study, conducted in California, it was reported that flavor liking of “Navel” oranges increased significantly when juice TSS levels increased from 8% to 10% during ripening (Obenland and others 2009). It is also known that high acidity of early-season mandarin varieties impairs flavor quality (Grierson 2006). In the present study, descriptive sensory analysis tests revealed that all 9 most preferred mandarin varieties exhibited more or less similar sensory profiles (Figure 5A), which were characterized by high sweetness, moderate to low sourness, low bitterness and gumminess, strong fruity and mandarin flavor, and high juiciness (Figure 5B). In contrast, the flavor profiles of the least preferred varieties were distinguished from those of the most preferred varieties according to one or more parameters (Figure 6). For example: “Fallglo” and “Hadas” were sourer; “Fallglo,” Satsuma “Okitso, ” Satsuma “Dubashi-Beni,” “Kiomi,” “Rishon,” “Hadas,” and “GP-377” had lower sweetness scores; and “Fallglo,” Satsuma “Okitso,” “Kiomi,” and “Rishon” had weak fruity and mandarin flavor (Figure 6). Interestingly, we previously obtained similar results for pomegranate fruits, in which out of 18 distinct varieties all most preferred varieties had more or less similar sensory profiles, whereas the least preferred varieties differed from each other and were considered to be either too sour or bitter, to have weak pomegranate and fruity flavor, or to have hard seeds (MayuoniKirshinbaum and others 2013).
