Changes in orange juice color by addition of mandarin juice

Eur Food Res Technol (2006) 222: 516 520 DOI 10.1007/s00217-005-0099-6 ORIGINAL PAPER Antonio J. Pérez-López Felipe Beltran Marta Serrano-Megías Domingo Saura López Ángel A. Carbonell-Barrachina Changes in orange juice color by addition of mandarin juice Received: 4 May 2005 / Revised: 20 June 2005 / Accepted: 21 June 2005 / Published online: 1 December 2005 C Springer-Verlag 2005 Abstract The most natural way of improving the color of orange juices is by adding other juices, which provide a more intense coloration. The US legislation allows the addition of up to 10% of mandarin juice to the orange juice to improve its color. The first objective of this study was to compare the color characteristics of juices from 11 mandarin cultivars, currently being grown in Spain. Experimental results proved that only the green-red coordinate, a, of the orange juices can be improved by adding mandarin juice. The mandarin cultivar that provided a juice with the highest values of a was Clemenules. Once this selection was made, the effects of adding mandarin juice at different ratios, up to 10%, on the color characteristics of the orange juice were studied. Values of the a coordinate went from 5.50 for the pure orange juice up to 6.29 for the mixture of 90% orange juice plus 10% of mandarin juice. Finally, hedonic tests proved that regular juice consumers preferred the color of mandarin juice to that of orange juice and that they liked better the color of the juice mixture containing 10% mandarin juice than that containing 3% mandarin juice. A. J. Pérez-López M. Serrano-Megías Departamento de Ciencia y Tecnología de los Alimentos, San Antonio Catholic University, Campus de los Jerónimos, s/n., 30107 Guadalupe, Murcia, Spain F. Beltran J. García Carrión S.A., Jumilla, Murcia, Spain D. S. López Unidad Mixta IBMC, FMC FoodTech. Instituto de Biología molecular y Celular, Universidad Miguel Hernández, Carretera de Beniel, km 3.2, 03312 Orihuela, Alicante, Spain Á. A. Carbonell-Barrachina ( ) Departamento Tecnología Agroalimentaria, Universidad Miguel Hernández, Carretera de Beniel, km 3.2, 03312 Orihuela, Alicante, Spain e-mail: angel.carbonell@umh.es Tel.: +34-966749754 Fax: +34-966749677 Keywords Citrus. Color. CIELab. Juice mixture Introduction Consumers purchase a product based mainly on their perception of the product quality and the relationship price/quality [1]. The importance of color as a quality parameter in citrus products has been demonstrated in several studies [2]. The natural bright colors of the citrus juices have been considered traditionally as one of their main advantages over other juices [3]. The main carotenoides responsible for the orange color of the orange and mandarin juices are α-carotene and β-carotene, zeta-antheraxanthin (yellow), violaxanthin (yellow), β-citraurine (reddish orange) and β-cryptoxanthin (orange) [4, 5]. Large quantities of β-cryptoxanthin in orange and mandarin juices produce a highly desirable bright orange color [2]. It is easy to understand why color is an important attribute to the food industry. Consumers frequently judge food quality based on color. In the marketplace, it is rare that consumers are allowed to taste food products prior to purchasing them; however, they can look at the product. Therefore, they finally decide whether to buy a particular product largely based on the overall appearance, including color. Frequently color and flavor are directly related. However, food processors are often limited in their capability to adjust color in the final product. Because of this limitation, manufacturers pay special attention to the color of raw materials/ingredients and to the changes that occur during each step of the production chain [6]. The US Department of Agriculture (USDA) recognizes the importance of the color in the quality of the commercial citrus juices and it awards the same importance to both color and flavor. However, the USDA for their classification of orange juice color uses a series of six standard plastic tubes but makes no reference to CIELAB coordinates, which is the most widely used color space. In Mediterranean climate, dry and fresh, as that of Spain, the pigmentation of the fruits is developed fully,

517 but it is known that in warmer and/or more humid regions as in Florida (USA), the coloration is less intense. For this reason, the optimization of juice color plays a more important role in the Florida industries as compared to the Spanish ones. The only way to improve the juice color is by mixture with other juices, which provide a more intense coloration. The US Federal legislation allows the addition of mandarin juice (Citrus reticulata) and its hybrids, up to 10%, to the orange juice to improve its color; even with this addition, the mixture is still legally called orange juice [7]. Mandarin fruits are prized for their delicious flavor, but relatively scarce information is available on their color characteristics when compared to those of other citrus fruits, such as orange or lemon [8]. Spain, the second world producer of mandarin (1,779,800 t in 2003), it is also the main supplier of the international market, especially toward northern European countries and the United States of America. The most grown and exported mandarin cultivars are Clementine. Spain concentrates the production of mandarin on the areas of Valencia and Murcia and presents an annual per capita consumption of 5kg[9]. On the other hand, the selling prices of mandarin in eastern Spain sometimes do not cover even the irrigation, pesticides and picking expenses. Thus, there is an imperative need for finding new uses for mandarin in the food industry because otherwise this agricultural and industrial sector is going to collapse with all the implications that this breakdown can have in the agriculture, industry and culture of eastern Spain. Currently, the main two items elaborated with mandarin fruits are canned slices and juice, and they only represent about 7% of the total trade of this fruit [10]. Since addition of mandarin juice to orange juice is allowed by the US legislation, the main objective of this study was to compare the color properties of 11 mandarin cultivars, being currently grown in Spain. Once, a specific cultivar is selected for juice making, based on its color characteristics, the effect of its addition, in different ratios (up to 10%), to orange juice (cultivar Valencia Late) will be evaluated. Materials and methods Fruit material Mandarins, Citrus reticulata (Nova, Ortanique, Marisol, Fortuna, Ellendale, Clemenules, Oronules, Clementpons, Orogrande, and Hernandina) and Citrus unshiu (Satsuma Owari), and oranges, Citrus sinensis (Valencia Late) were all grown under identical conditions of soil, irrigation, and illumination in eastern Spain (Librilla, Murcia). Fruits were collected in winter (first week of December 2004). Fruits from these cultivars were selected based on their diameter, ranging from 48.30±0.02 mm (Oronules) and 82.36±0.12 mm (Clemenules), and maturity index, 12.0±0.3. Sample preparation Mandarin and orange juices were processed in a commercial plant (Murcia, Spain). The juices were obtained by using a Premium Juice Extractor (FMC Corporation, Florida, USA) [12]. This type of machinery leads to a juice with a low content of essential oils [11]. Freshly squeezed samples are referred as juices; samples were kept in refrigeration (2 C) until analyses and were not subjected to any thermal treatment. Color measurement Color determinations were made, at 25±1 C, using a Hunterlab Colorflex (Hunterlab, Reston, Virginia, USA). This spectrophotometer uses an illuminant D 65 anda10 observer as references. A sample cup for reflectance measurements was used (5.9 cm internal diameter 3.8 cm height) with a path length of light of 10 mm. Blank measurements were made with the cup filled with distilled water against a reference white background. Color data are provided as CIELAB coordinates [13], which define the color in a three-dimensional space. L indicates lightness and a and b are the chromaticity coordinates, green-red and blue-yellow coordinates, respectively. L is an approximate measurement of luminosity, which is the property according to which each color can be considered as equivalent to a member of the gray scale, between black and white, taking values within the range 0 100; a takes positives values for reddish colors and negative values for the greenish ones, whereas b takes positive values for yellowish colors and negative values for the bluish ones. C is chroma [C = (a 2 ) + (b 2 )], and is 0 at the center of a color sphere and increases according to the distance from the center. Finally, h ab is the hue angle [h ab = arc tg(b /a )], which is defined as starting at the +a axis and is expressed in degrees; 0 would be +a (red), 90 would be +b (yellow), 180 would be a (green), and 270 would be b (blue) [14]. Sensorial analysis Sensory evaluation (hedonic tests) was used to determine the color liking for orange and mandarin juices. Stone and Sidel [15] recommended using at least 50 consumers for hedonic studies. One hundred consumers were recruited with a small advertisement in Alicante and Murcia (35.1% men and 64.9% women, all of them between 18 and 40 years of age). The principal selection criterion was that subjects had to be regular consumers of juice at least twice a week. Commercial orange and mandarin juices samples, fresh squeezed, were given to consumers. The juices mixtures were prepared by adding to the orange juice 3, 6, 9, and 10% of mandarin juice. Samples, 30 ml, were served in 50 ml transparent glass and were marked with randomized

518 three digital codes. A complete block design was made and the juice were presented one by one following a Williams latin squared design balanced for order and first-order carry over effects to avoid the position error [16]. Measurements were performed in individual booths with controlled illumination and temperature [17]. Consumers participated in three different hedonic tests; each session lasted about half hour. First, consumers participated in a paired comparison to identify, whether consumers were able to distinguish between orange and mandarin juices based exclusively in their color. Only those consumers being able to distinguish the two types of juices were allowed to participate in the second test, in which they were ask which juice, orange or mandarin, they preferred, according to their color. Finally, consumers participated in a ranking test in which mixtures of orange and mandarin juices should be sorted out according to their preference for their color (subjects were instructed to assign rank 1 to the sample with a more appealing color and 4 to the sample with a less appealing color). This third test was run in triplicate [17]. Statistical analysis Instrumental color measurements were run in 18 replications. All data were subjected to analysis of variance (ANOVA) and the Tukey least significant difference multicomparison test to determine significant differences among mandarin cultivars [18]. Significance of differences was represented as p 0.001. The statistical analyses were done using SPSS 12.0 (SPSS Science, Chicago, USA) and figures using Sigma Plot 8.0 (SPSS Science, Chicago, USA). Principal component analysis (PCA) examined the relationship among color coordinates and cultivars. Analysis without axis rotation provided the most informative interpretation of the relationships; eigenvalues greater than 1 were used as a selection criterion for the number of components [19]. Data from paired comparison test was analyzed by the chi-square test (ISO 5495:1983) [20]. Data for preference test was described by a frequency table. Finally, ranking data was analyzed by a Friedman test (ISO 8587:1988) [21]. Results and discussion The pulp content has a direct effect on the measurement of reflectance and to avoid this effect all samples must have similar pulp content [22]. The fiber contents of all mandarin and orange juice samples were equilibrated by using a FMC FoodTech Quick Fiber device (FMC Corporation, Florida, USA); in this way, juice samples presented a mean value for the centrifugable pulp content of 15.0±1.1 g l 1 (no significant differences were found among juice samples). Besides, the mean value for the soluble solids contents of the mandarin juices was 11.5±0.5 Brix and no adjustment was needed due to the close values of all studied samples. The main reason for selecting Valencia Late as the only orange cultivar is that it is the most industrially processed orange cultivar for juice production in Spain [23, 24]. Besides, Valencia Late oranges are worldwide appreciated due to their deep orange color [2]. The main color characteristics of the orange juice analyzed in this study were: L =52.99±0.02; a =5.50±0.01; b =33.83±0.02. One of the peculiarities of the ultrafrozen orange juices analyzed by Meléndez-Martínez et al. [2]was their deep orange color (L =63.23±1.27; a =16.18±0.56; b =64.64±3.78), because the juice was subjected neither to high temperatures nor to concentration process during the production. On the other hand, the orange juice samples analyzed in this study have been subjected to thermal processing (T=98 C for 20 s). The chroma values agreed with less vivid orange colors in the present orange juice samples (C =34.27±0.01) as compared with those of Meléndez- Martínez et al. [2] (C =66.64±3.66). Selection of a mandarin cultivar The statistical study of experimental color data showed significant differences among mandarin cultivars at p<0.001 (Table 1), for all five color parameters: L, a, b, C, and h ab. Relationships among cultivars were established by the Tukey s multiple-range test. According to the L data, the lightest samples were those from Clemenules mandarins (53.02±0.02) and Valencia Late oranges (52.99±0.02). On the other hand, the darkest samples came from Ortanique (50.45±0.02), Ellendale (50.51±0.01), Nova (50.53±0.01), Fortuna (50.55±0.02), and Satsuma owari (50.59±0.01) mandarins. As can be seen, less than three units (2.57) of difference were detected in lightness between the lightest and the darkest mandarin samples. Data on Table 1 showed that the mandarin juice with the highest values of the green-red coordinate, a,was Table 1 Color coordinates for mandarin and orange freshly squeezed juices Cultivar L a b C h ab Mandarin Clemenules 53.02 a 8.57 a 29.40 b 30.63 b 73.75 c Oronules 51.05 b 8.17 b 29.17 c 30.29 c 74.35 b Marisol 50.90 c 8.11 b 29.04 d 30.15 c 74.40 b Hernandina 50.81 c 7.87 c 28.77 e 29.82 d 74.70 b Orogrande 50.75 cd 7.79 d 28.42 f 29.47 e 74.67 b Clementpons 50.70 d 7.73 d 28.19 g 29.23 f 74.67 b Satsuma owari 50.59 e 7.72 d 28.12 g 29.16 f 74.65 b Fortuna 50.55 e 7.71 d 28.39 f 29.42 e 74.81 b Ellendale 50.51 ef 7.70 d 28.05 h 29.09 g 74.65 b Nova 50.53 e 7.71 d 28.10 g 29.14 f 74.66 b Ortanique 50.45 f 7.58 e 27.82 i 28.83 h 74.76 b Orange Valencia Late 52.99 a 5.50 f 33.83 a 34.27 a 80.77 a Note. Mandarin and orange cultivars with the same letters were not significantly different at p<0.001 for the attribute evaluated (Tukey multiple range test)

519 Clemenules (8.57±0.01) and the lowest values belonged to samples of the cultivar Ortanique (7.58±0.01). An important remark is the low value of a in the Valencia Late oranges (5.50±0.01) as compared to the mandarin cultivars, which have a mean value of 7.90±0.09. The Valencia Late oranges provided a juice with the highest values of the blue-yellow coordinate, b, 33.83±0.01, with a mean value of 28.50±0.01 for all mandarin juices. In this way, the addition of mandarin juice will not increase the b value of the initial orange juice. From all studied mandarin cultivars, Clemenules had the highest values of b, with a value of 29.40±0.01. Again, the mandarin cultivar which presented the highest values of the chroma parameter was Clemenules (30.63±0.01). Oronules and Marisol also provided juices with high values of chroma, 30.29±0.02 and 30.15±0.04, respectively. However, none of the mandarin juices had C values as high as those from the Valencia Late oranges, 34.27±0.01. The other color parameter establishing a relationship between a and b is the Hue angle. The only mandarin juice with statistically different values of h ab was Clemenules, 73.75±0.05, compared to a mean value of 74.63±0.05 for the other 10 mandarin cultivars and to a value of 80.77±0.04 for the orange juice. Therefore and considering that the experimental results proved that only the green-red coordinate, a, of the Valencia Late oranges can be improved by addition of mandarin juices from any of the studied cultivars, Clemenules mandarins were selected for the next experiment. However, the cultivars Oronules and Marisol could also be a good option if Clemenules mandarins are not available for any reason. In summary, it can be concluded that Clemenules mandarins provide a juice with the highest values of a, b, and C, thus providing a high intensity of the orange color. Besides, juices from the Clemenules mandarins are lighter than any other mandarin juices studied. Orange and mandarin juice mixtures Samples of the orange and mandarin mixtures were prepared by adding to Valencia Late juice 3, 6, 9, and 10% of mandarin juice, cultivar Clemenules. After the mixtures were prepared, instrumental and sensory measurements were immediately carried out. Data on Table 2 proves that addition of mandarin juice to orange juice increases the values of L and a but decreases the values of b, C, and h ab of the initial orange juice. The juice consisting of orange (90%) and mandarin (10%) was more pale ( L =0.04) and dull ( C = 0.50) than the initial orange juice. On the other hand, the green-red coordinate was significantly increased from 5.50±0.01 on pure orange juice to 6.29±0.01 on the 90% orange 10% mandarin mixture. Another instrumental measurement was carried out to prove that addition of mandarin juice to orange juice improves the orange color of the final product. Reflection spectra of (1) pure orange juice, cultivar Valencia Late, (2) 90% orange juice, cultivar Valencia Late+10% man- Table 2 Color coordinates for pure mandarin, pure orange (VL, Valencia Late), and mixtures of orange and mandarin freshly squeezed juices Sample L a b C h ab Valencia Late 52.99 b 5.50 d 33.83 a 34.27 a 80.77 a Clemenules 53.02 a 8.57 a 29.40 d 30.63 e 73.75 d VL+3% Clemenules 52.47 c 5.55 d 32.66 c 33.13 d 80.36 ab VL+6% Clemenules 52.98 b 5.85 c 33.10 b 33.67 c 79.98 ab VL+9% Clemenules 53.06 a 6.18 b 33.10 b 33.67 c 79.42 bc VL+10% Clemenules 53.03 a 6.29 b 33.18 b 33.77 b 79.27 c Note. Samples with the same letters were not significantly different at p<0.001 for the attribute evaluated (Tukey multiple range test) darin juice, cultivar Clemenules, (3) pure mandarin juice, cultivar Clemenules, were measured in order to determine color values in the spectral wavelength range from 595 to 700 nm (580 595: yellow; 595 605: orange; 605 750: red) [25]. Figure 1 shows that sample 2, orange and mandarin juice presented higher reflectance values from about 575 to 700 nm (yellow, orange, and red colors) than pure orange juice, sample 1. However, the mix showed lower reflectance values than pure mandarin juice, sample 3. Once this point is reached, the next question is clear: is this instrumentally detected improvement of the orange juice color, caused by the addition of mandarin juice, detected by the regular consumer of this type of products? Sensorial analysis of the orange and mandarin mixtures The only way of getting a realistic answer to the previous question is carrying out hedonic tests with a consumer panel. The first question made to the consumer panel, consisting on 100 regular consumers of citrus juices, was: which of REFLECTANCE (%) 50 40 30 20 10 0 400 450 500 550 600 650 700 λ(nm) Fig. 1 Absorption spectra of the following juice samples: (1) pure orange juice, cv Valencia Late orange juice; (2) mixture of orange (90%) and mandarin (10%) juices; (3) pure mandarin juice, cv Clemenules 3 2 1

520 the following two samples (pure orange juice or mandarin juice) do you prefer according to their color? The panel answer was clear and had statistical significance (χ 2 =288; df=1; p=0.001), 100% of the consumers preferred the mandarin juice over the orange juice. After knowing that consumers, participating in these hedonic tests, liked the color of mandarin juice better than that orange juice, the next step was logical: are these consumers able to distinguish among the assayed percentages of mandarin juice (3 10%) being added to the orange juice? Consumers were asked to arrange the four samples they received (orange juice plus 3, 6, 9, and 10% mandarin juice) according to their liking of the samples color. Experimental results demonstrated that consumers were able to significantly differentiate among 3 6, 9, and 10% according to the Friedman s test. References 1. Thai CN, Shewfelt RL (1991) Modelling sensory colour quality of tomato and peach: neural networks and statistical regression. ASAE/CSAE Annual International Meeting, Paper N, 90 6038, Pretoria, South Africa 2. Meléndez-Martínez AJ, Vicario IM, Heredia FJ (2003) J Agric Food Chem 51:7266 7270 3. Barron RW, Maraulja MD, Huggart RL (1967) Instrumental and visual methods for measuring orange juice color. Proceeding of the Florida State Horticultural Society 80:308 4. Stewart I (1980) Color as related to quality in citrus. In: Nagy S, Attaway J (eds) Citrus nutrition and quality. American Chemical Society, Washington, DC, pp 129 150 5. Lin S (1995) J Food Biochem 18:273 283 6. Gunasekaran S (1990) Crit Rev Food Sci Nutr 29:19 34 7. Terblanche E (1999) Effect of temperature on the colour of citrus during degreening. ASAE/CSAE Annual International Meeting, Paper N, 996120, Toronto, Ontario, Canada 8. Shaw P (1991) Fruits II. In: Maarse V (ed) Volatile compounds in foods and beverages. Marcel Dekker, Inc., New York, pp 312 316 9. MAPA (Ministerio Agricultura, Pesca y Alimentación) (2003) La alimentación en España. MAPA, Madrid, Spain 10. FAO (Food and Agriculture Organization of the United Nations) (2005) http://www.fao.org 11. Kimball DA (2002) Procesado de cítricos.ed.acribia,zaragoza, Spain 12. FMC (FMC FoodTech) (2004) http://www.fmcfoodtech.com 13. McGuire R (1992) HortScience 27:1254 1255 14. Minolta (1994) Precise color communication. Color control from feeling to instrumentation. Minolta, Co. Ltd., Osaka, Japan 15. Stone H, Sidel JL (1993) Sensory evaluation practices, 2nd edn. Academic Press, San Diego, California 16. MacFie HJ, Bratchell N, Greenhoff K, Vallis LV (1989) J Sensory Stud 4:129 148 17. Meilgaard M, Civille GV, Carr BT (1999) Selection and training of panel members. In sensory evaluation techniques, 3rd edn. CRC Press, Boca Raton, Florida, pp 133 159 18. Genard M (1992) Scientia Hort 52:37 51 19. Bower JA, Baxter IA (2000) Br Food J 102:821 837 20. ISO (International Organization of Standardization) (1983) Sensory Analysis. Methodology. Paired Comparison. Ref. No. ISO 5495:1983 (E). ISO, Geneva, Switzerland 21. ISO (International Organization of Standardization) (1988) Sensory Analysis. Methodology. Ranking. Ref. No. ISO 8587:1988 (E). ISO, Geneva, Switzerland 22. Arena E, Fallico B., Maccarone E (2000) J Food Sci 65:458 460 23. Sarooshi RA (1995) Report on a citrus study tour to Spain and attendance at Mandarin Symposium, Corsica, Italy. NSW Plant Industries Report No., p. 16 24. Offner N, Warren T, Davidson P, Simpson J, Burford J, Braniff J (1998) National Citrus Industry Mission to Europe. February, CMDG 25. MacDougall DB (2002) Colour in food. Improving quality. CRC Press, Cambridge, UK