Postharvest Damage of Mangosteen and Quality Grading Using Mechanical and Optical Properties as Indicators

Transcription

1 International Journal of Food Properties ISSN: (Print) (Online) Journal homepage: Postharvest Damage of Mangosteen and Quality Grading Using Mechanical and Optical Properties as Indicators Bundit Jarimopas, Phanida Pushpariksha & S. Paul Singh To cite this article: Bundit Jarimopas, Phanida Pushpariksha & S. Paul Singh (2009) Postharvest Damage of Mangosteen and Quality Grading Using Mechanical and Optical Properties as Indicators, International Journal of Food Properties, 12:2, , DOI: / To link to this article: Copyright Taylor and Francis Group, LLC Published online: 26 Mar Submit your article to this journal Article views: 463 Citing articles: 2 View citing articles Full Terms & Conditions of access and use can be found at

2 International Journal of Food Properties, 12: , 2009 Copyright Taylor & Francis Group, LLC ISSN: print / online DOI: / POSTHARVEST DAMAGE OF MANGOSTEEN AND QUALITY GRADING USING MECHANICAL AND OPTICAL PROPERTIES AS INDICATORS Bundit Jarimopas 1, Phanida Pushpariksha 2, and S. Paul Singh 3 1 Department of Agricultural Engineering, Faculty of Engineering at Kamphaengsaen, Kasetsart University, Kamphaengsaen, Nakohnpathom, Thailand 2 Postharvest and Processing Technology Division, Faculty of Agricultural Technology, Rajamangala University of Technology Thanyaburee, Pratumtanee, Thailand 3 School of Packaging, Michigan State University, East Lansing, MI, USA Postharvest damage in fresh mangosteens at wholesale level in Thailand was investigated from April to October A total of 37.1% of the production yield was rendered inedible by damage during this period; damages included fruit cracking, hardened rinds, rough surfaces, translucent flesh, gummosis and decay. This study focused on a method of predicting damage based on the color of the skin of the affected mangosteen. As a first step, diameter, height, weight, and volume of large, medium, small, and undersize mangosteens were measured. The term, dimension ratio, was introduced as a sizing parameter identifying conventional trade size. The coefficient of static friction of the glossy- and rough-surface mangosteens on plexiglass, plywood, and galvanized steel sheet varied from 0.31 to The color of sound and defective fruits was measured in terms of their tristimulus values X, Y, and Z. The corresponding chromaticity coordinates of a mangosteen, x and z, depended on the maturity stage of the fruit while y depended on the type of fruit surface. A ratio was proposed to test the accuracy of predicting internal defects from the color variation between two spots on the surface of the same fruit. The highest percentage of correct prediction was 67.4% with a color ratio of X 1 (pink blush color on yellow ground color) to X 2 (pink color) that was greater than Keywords: Mangosteen, Grading, Color, Friction, Physical properties, Postharvest loss. INTRODUCTION Mangosteens (Garcinia mangostana L.) are one of the most popular tropical fruits in Thailand. The main planting areas cover 19,200 hectares of land in the eastern and southern regions of the country and annual production averages 108,000 metric tons. [1] There are two distinct harvesting seasons: April to June in the east, and July to September in the south. The mangosteen is a round, reddish-purple to dark purple fruit, which is capped by a prominent calyx at the stem end and also features 4 to 8 triangular flat remnants of the stigma in a rosette at the apex. The rind is 3.2 to 6.4 mm thick, red in cross-section, and Received 12 March 2007; accepted 3 December Address correspondence to Bundit Jarimopas, Department of Agricultural Engineering, Faculty of Engineering at Kamphaengsaen, Kasetsart University, Kamphaengsaen, Nakohnpathom, Thailand. jarimopas@ yahoo.com 414

3 DAMAGE OF MANGOSTEEN AND QUALITY GRADING 415 contains a bitter yellow latex and purple staining juice. Its value derives from its 4 to 8 triangular segments of snow-white, juicy, soft flesh. Each segment may be seedless or else contain a developed seed that clings to the meat. Mangosteens were found to have a climacteric respiratory pattern. [2] This means that respiration and ethylene production rates increase during maturation and ripening, and these are accompanied by corresponding changes in fruit color from green to pink, red and purple. TISTR [3] recommends a skin color-based harvesting index for mangosteens as follows: stage 0 - yellowish white; stage 1 - greenish yellow; stage 2 - pink blush on yellow background; stage 3 - pink; stage 4 red; stage 5 - reddish purple; stage 6 dark purple. Fruit should be picked not earlier than stage 2. At this time, fruits can ripen and sweeten after harvesting. Stage 5 and 6 mangosteens are considered to be ripe. Grading is essential for marketing because pricing is tied to produce quality. Quality sorting criteria of fresh mangosteens include size, color and defects. Fruit size is graded differently by various traders. Fruit defects include fruit-cracking, rough surface, pericarphardening, translucent flesh, gummosis and decay. All defects are typically found at retail level and often reduce consumer demand for the fruit. Dangcham [4] reports symptoms of translucent flesh include flesh changes from white to translucent and textural changes from soft to firm and crisp. The translucency of flesh might occur because of excessive absorption of moisture via the root system, the leaves or the fruit. [5,6] Fruit defects cause a great deal of perceivable loss in mangosteen production and quality, although such post-harvest loss has not yet been reported quantitatively. One of the more common defects, rough surface, is inherent to the fruit before harvesting, but is considered to be undesirable because the fruit is unpleasant to the touch. However, rough surfaces can be polished off by mechanical brushing. Grading of fruits by size can nowadays be accomplished by various types of sizing machines. [7,8,9] Grading parameter often included firmness. [10] The degree of ripening of mangosteen is associated with the hue of the surface of the fruit. Bodria et al. [11] used light reflectance measurement technique to assess chlorophyll content in red-pigmented fruits to classify their ripeness. Raju et al. [12] applied acoustic resonance spectroscopy to nondestructively monitor mango ripeness. Jha et al. [13] evaluated the surface gloss of eggplants using a spectroradiometric system to quantify their freshness. Furthermore, Quevedo and Aguilera [14] determined the specular reflection of an incident light beam to quantify the roughness of a surface. Thus, to date, measuring light reflectance from fruits would employ sensor technique that best approximates the performance of human sorters. [15] This is also attested to by Chen, [16] who has indicated that one of the most practical and successful techniques for nondestructive quality evaluation and sorting of agricultural products is the electro-optical technique based on the optical properties of the product. Knowledge of the optical and mechanical properties of mangosteens thus is useful for the design and evaluation of handling and grading machines. Accordingly, the objectives of this research were (i) to investigate postharvest damages of mangosteen at the wholesale level; and (ii) to ascertain the mechanical and optical properties of mangosteens relate to quality grading. MATERIALS AND METHODS The mangosteens used in this study were harvested in 2004, from three commercial orchards in Janthaburee province and three commercial orchards in Chumpohn province, both of which are located in Thailand. Mangosteens of different maturity stages varying

4 416 JARIMOPAS, PUSHPARIKSHA, AND SINGH from pink blush on yellow background to dark-purple were carefully hand harvested to prevent drop damage. Fruits were placed into shoulder bags and then unloaded into 25-kg plastic baskets. Like other produce, the mangosteens harvested were sorted, which is the main post-harvest operation at the wholesale level. The sorting considerations included size, color, ripening degree, and obvious defects such as fruit cracking, hard rind, and rough surface. Determination of Post-harvest Damages at the Wholesale Level On the day of the harvest, the mangosteens were transported from the orchards in 25-kg plastic baskets to nearby wholesalers for grading and sizing. Mangosteen grading included large-glossy, medium-glossy, large-rough, small-glossy, medium-rough, smallrough, and undersize. For defect observation, 30 mangosteen samples were collected at random out of each grade for an occasion, a selected orchard in a selected province. As mentioned earlier, there were two provinces and three commercial orchards for each province. There were three separate occasions: at the beginning, at the middle and at the end of the harvesting season, which were April through June 2004, for Janthaburee, and July through September 2004, for Chumpohn. The percentage of defective fruit was evaluated based on the yield count. Determination of Physical Characteristics The diameter, height, weight, and volume of four sizes of mangosteens (large, medium, small, and undersize) were determined. The maximum diameter (D max ) and minimum diameter (D min ) of each of the fruits, the diameter of the minimum calyx circumscribing circle (d), the height of the fruit excluding the stem end (h) and including the stem end (H) (Figure 1) were measured by using a vernier caliper. The dimension ratio (D ratio ) was calculated by using the equation: D ratio = d/d max. Fruit weight was obtained by the use of an electronic balance (TANITA, Model: HFF521, Tanita Corporation, Tokyo). Fruit volume was determined by the water displacement technique. [17] Determination of Coefficient of Static Friction The mangosteens were stored at room temperature ((25 30 C) after harvest until they were fully ripe. Thirty replicates of glossy- and rough-surface fruits (Figure 2) were tested. The test was carried out once a month over a 3-month period of harvesting season, Figure 1 Dimensions of mangosteen fruit.

5 DAMAGE OF MANGOSTEEN AND QUALITY GRADING 417 Fruit cracking. Rough and Glossy surface. Figure 2 Mangosteens with rough, glossy and cracked surfaces. resulting in a total of 180 fruit samples. The coefficient of static friction (μ) could be estimated from tangent of the angle of the inclined plane, [17,18,19] upon which the mangosteen rested. The test was performed by placing a mangosteen on the inclined plane and slowly raising the plane so that the mangosteen started to slide at the inclination angle (θ). The corresponding tangent of θ was μ. The test was run with the variation of three control factors; a) two types of fruit surfaces (glossy and rough surfaces), b) three types of the inclined surfaces (plexiglass, plywood, and galvanized steel sheet), and c) two sliding directions of fruit (the longitudinal and the transverse axis of the mangosteen). The selected material for the inclined plane was the construction material of conveyor in the packing line. The data of μ were analyzed by standard analysis of variance and Duncan s multiple range tests. Determination of Color A JC801-colorimeter (Juki, Japan) was used to measure fruit color in terms of the tristimulus values X, Y, and Z. The chromaticity coordinates x, y, and z were consequently calculated by using the following expressions: x = X / (X+Y+Z), y = Y / (X+Y+Z) and z = Z / (X+Y+Z). [20] Color measurements were carried out with fruits at the following different developed skin color stages: stage 2 - pink blush color on yellow ground color; stage 3 - pink color; stage 4 - red color; stage 5 - reddish purple color; and stage 6 - dark-purple color. For glossy-surface fruit with uniform color development, a sector on the equator of the fruit was taken for color measurement. The same experiment was done for rough-surface fruits. For non-uniform color development (Figure 3), which implied internal defects according to grower s experience, two sectors of the fruit surface having different color Normal color Previous stage color Figure 3 Non-uniform color development in mangosteen.

6 418 JARIMOPAS, PUSHPARIKSHA, AND SINGH stages were measured and analyzed for the color ratio in order to determine the accuracy of the defect symptoms. Color ratio was defined as X 1 /X 2, Y 1 /Y 2, and Z 1 /Z 2, where the subscript 1 represented the first sector in younger stage (stage 2) and 2, the second sector in normal stage (stage 3). The younger stage refers to the maturity stage prior to that of the normal stage. RESULTS AND DISCUSSION Postharvest Damages at the Wholesale Level Figure 4 shows the typical manual postharvest quality sorting system of fresh mangosteen in Thailand. The number of sorting considerations varied from place to place and time to time, and according to the inclination of the middlemen. Cracked and hard rind fruits were culled out of the harvested fruits and fully ripe fruits were then separated at the culling output stage. Finally, after sizing had been completed, the mangosteen were graded into six groups; (a) large-glossy, (b) medium-glossy, (c) large-rough, (d) small-glossy, (e) medium-rough, and small-rough and (f) undersize. The fully ripe fruits were delivered to the local market. Mangosteens of grades a, b and c were sold to exporters while the other graded groups were distributed to the domestic market. Table 1 shows the percentages of fresh mangosteens of each grade in Janthaburee and Chumphon provinces. Only 33.7% of production in Janthaburee could be exported. Meanwhile, 47.0% of the fruits from Janthaburee and 23.8% from Chumphon were considered to be inferior grade fruits, which yielded lower prices and grower income. Hand harvesting Rejected Rind hardening Fruit cracking Culling Fully ripening Local market Domestic market Sizing and Grading Grade a, b and c Grade d Undersize, Under grade Exporter Domestic market Domestic market Figure 4 A block diagram of the sorting system of fresh mangosteens in Thailand. Table 1 The percentages of fresh mangosteens of each grade. Grade % Janthaburee % Chumpohn a) Large-glossy 14.5 ± ± 15.3 b) Medium-glossy 10.8 ± c) Large-rough 8.4 ± ± 12.2 d) Small-glossy 9.5 ± ± 11.8 e) Medium-rough and Small-rough 47.0 ± ± 21.6 f) Undersize 9.8 ± ± 6.1 Total number (fruits) 11,010 7,871

7 DAMAGE OF MANGOSTEEN AND QUALITY GRADING 419 Table 2 The percentages of damages of fresh mangosteens. Damage % Janthaburee % Chumpohn average Fruit cracking 1.2 ± ± ± 1.7 Hard rind 0.5 ± ± ± 2.7 External defect - Rough surface 37.2 ± ± ± 20.4 Internal defects - Translucent flesh 14.5 ± ± ± Gummosis 10.8 ± ± ± Translucent flesh & Gummosis 4.3 ± ± ± Decay 1.4 ± ± ± 1.9 Nearly 10% of fruits from both regions were undersized, and therefore, unable to be sold at market. It seemed clear therefore that it would be beneficial to develop pre-harvest techniques to minimize the proportion of inferior grade fruits. After storing the fruit at room temperature until all samples were fully ripe, the mangosteens (excluding those that had cracked) were cut open to observe their internal condition. Table 2 shows the percentages of fresh mangosteens from Janthaburee and Chumphon provinces that were spoilt. Damages included fruit cracking, hard rinds and other external and internal defects. Fruits that had cracked amounted to 1.3% of the yield count. This category of fruits, along with that which had translucent flesh, could be sold for a very low price in Janthaburee but not at all in Chumphon. Other damages included hard rinds (1.9%) and rough surfaces (48.6%), both of which could be detected visually. Internal defects included translucent flesh, gummosis and decay, which made up an average 33.9% of the yield. Only 62.9% of production yield was healthy edible fruits. A small percentage of hard rind-fruits were found at wholesale level due to careful harvesting. The fruits whose pericarp had hardened outwards from inside the pericarp often also showed decay from pathogens. The sum of damaged fruit and healthy fruit exceeded 100% because external or internal defects could appear singly or in combination in one fruit. Most of the mangosteen defects at the wholesale level were likely result of physiological and pathological disorders during storage. Physical Characteristics Mangosteens are normally classified into four sizes based on its mass large, medium, small, and undersize. A total of 500 fruits at stage 1 were collected from each size-group for this experiment. Physical characteristics determined included diameter, height, weight and volume of the fruits. Table 3 shows the physical characteristics of these mangosteens. The lowest values for the physical characteristics were found in the undersize fruits. The large fruits showed the greatest values in all characteristics except the dimension ratio; for this indicator, the undersize fruits displayed the greatest value, which was higher than one. This indicated that the diameter of minimum calyx circumscribing circle of the undersize fruits was bigger than the maximum diameter. Hence, the dimension ratio, with a value greater than 1, could be used as grading threshold for the undersize fruits. Besides, it was also found that the weight of the mangosteens decreased considerably ( 14.5%) with time after harvest (2 weeks), while the maximum diameter of the fruit

8 420 JARIMOPAS, PUSHPARIKSHA, AND SINGH Table 3 Physical characteristics of large, medium, small, and undersize mangosteen fruits. Physical characteristic Large Medium Small Undersize Weight (g) 102 ± ± 6 65 ± 8 44 ± 7 Volume (cm 3 ) 105 ± ± 7 67 ± 6 45 ± 6 Fruit maximum diameter (mm) 60 ± 4 56 ± 1 51 ± 3 44 ± 3 Fruit minimum diameter (mm) 58 ± 3 54 ± 2 49 ± 3 43 ± 3 Diameter of minimum calyx 52 ± 3 47 ± 2 49 ± 4 47 ± 4 circumscribing circle (mm) Height of fruit with stem end (mm) 66 ± 5 65 ± 3 59 ± 5 55 ± 5 Height of fruit without stem end (mm) 50 ± 4 48 ± 2 42 ± 4 37 ± 3 Dimension ratio 0.86 ± ± ± ± 0.03 decreased only slightly ( 0.4%). Hence, the sizing of fruit was rather inconsistent if only weight was the sizing parameter. Coefficient of Static Friction The comparison of the coefficient of static friction between the glossy- and the rough-surface fruits was as shown in Table 4. The analysis of variance (ANOVA) showed that the sliding direction significantly affected the static friction coefficient of the roughsurface fruit (P < 0.05). However, the effect of the sliding direction on the static friction coefficient of glossy-surface fruits was insignificant. Generally speaking, the mangosteen whose longitudinal axis was perpendicular to the sliding direction gave higher static friction coefficient than that was parallel to the sliding direction. Furthermore, the inclined surface showed significant influence on the coefficient of static friction of glossy- and rough-surface fruits. Significant differences were found between the means of static friction coefficient for all surfaces, except plywood and galvanized steel sheet on roughsurface fruits in the same sliding direction. Fruit surface also significantly affected the static friction coefficient even though the values for the coefficients of glossy- and roughsurface fruits were close to each other for plywood in the same sliding direction. The static friction coefficient obtained between the plexiglass and the rough-surface fruits was low ( 0.32) while that between the plexiglass and glossy-surface fruits was high ( 0.46). In the case of glossy-surface fruits sliding on plexiglass, an extremely smooth surface, the high coefficient was probably due to the wax on the fruits an effect similar to that Table 4 Coefficients of static friction of the glossy- and rough-surface mangosteen fruits. Description Plexiglass* Plywood* Galvanized steel sheet* Glossy-surface fruits - Longitudinal axis 0.46 ± 0.06f 0.37 ± 0.04c 0.33 ± 0.03b - Transverse axis 0.46 ± 0.07f 0.38 ± 0.04cd 0.34 ± 0.03b Rough-surface fruits - Longitudinal axis 0.31 ± 0.04a 0.38 ± 0.06cde 0.38 ± 0.08cd - Transverse axis 0.33 ± 0.04b 0.39 ± 0.06de 0.40 ± 0.07e *Means followed by the same letter are not different at a = 0.05 significance level according to Duncan s multiple range test.

9 DAMAGE OF MANGOSTEEN AND QUALITY GRADING 421 reported for oranges by Chen and Squire. [21] This could be because the wax layer on the surface of the fruits created a very high viscous friction between two smooth surfaces, resulting in an increase in adhesion and cohesion. Color Color and gloss-rough surface relationship. The tristimulus values X, Y, and Z of mangosteens decreased with the increase in fruit maturity. The most prominent decrease for the entire range was found from stage 2 to stage 3 for both glossy- and roughsurface fruits. The values for rough-surface fruits were smaller than those for glossysurface fruits in stages 2 and 3 but larger in the remaining stages of 4, 5, and 6 (Table 5). Figure 5 presents the chromaticity coordinates x, y, and z for mangosteens from stages 2 to 6. The value of y was not dependent on maturity stage and was nearly constant. The y value for rough-surface fruits was generally higher than that for glossy-surface fruits. The average y value for rough-surface fruits was 0.389, varying from , whereas the y value for glossy-surface fruits varied from , with the mean recorded at It could also be observed that some rough-surface fruits showed chromaticity coordinates close or equal to color of glossy-surface fruits in the lower range. However, the cluster of the chromaticity coordinates for rough-surface fruits was not clearly separated from that for glossy-surface fruits in stage 2 [Fig. 6(a)]. When the mangosteens became more mature, their x values decreased, while the z values increased with respect to time (Fig. 5). The change in chromaticity coordinates described the color change from red to purple. The cluster of the chromaticity coordinates for glossy-surface fruits was well separated from that for rough-surface fruits in this final stage [Fig. 6(b)]. The threshold values of x or z, separating fully ripe mangosteens between those with glossy and those with rough surfaces was characterized by x ( 0.335) and z ( 0.280). Color and internal defect relationship. A relationship was found between the amount of discoloration on the mangosteen rind to the amount of translucent fruit flesh and secretion of gum on the inside. As a result, non-destructive evaluation of color of the outer rind could be used to predict fruit quality. The following section discusses that relationship. Ninety-five non-uniform color development mangosteens were used to study the potential of using visible light excitation to detect internal defects. Sixty-one percent of these fruits showed internal defects and the remaining 39% were healthy fruit. Generally speaking, a normal, healthy mangosteen exhibits uniformly distributed color over its skin at every stage while mangosteens with internal defects shows two sectors of different colors (Fig. 3). Table 5 Tristimulus values of mangosteens at different stages of ripening X Y Z Stage glossy rough glossy rough glossy rough 2: pink blush on ± ± ± ± ± ± 1.33 yellow ground 3: pink ± ± ± ± ± ± : red ± ± ± ± ± ± : reddish purple 6.90 ± ± ± ± ± ± : dark-purple 6.10 ± ± ± ± ± ± 0.54

10 422 JARIMOPAS, PUSHPARIKSHA, AND SINGH x coordinates glossy rough stage of ripening 0.40 y coordinates glossy rough stage of ripening 0.40 z coordinates glossy rough stage of ripening Figure 5 The x, y, z coordinates of glossy- and rough-surface mangosteen fruits.

11 DAMAGE OF MANGOSTEEN AND QUALITY GRADING z coordinates glossy rough x coordinates 0.35 (a) stage 2 z coordinates glossy rough x coordinates (b) stage 6 Figure 6 The clusters of the chromaticity coordinates (x, z) of glossy- and rough-surface mangosteens. Defective fruits show one sector of a normal color and the other the color of the previous stage. As mentioned previously, the tristimulus values X, Y, and Z of mangosteens decreased considerably from stage 2 to stage 3. The differences between tristimulus values for the stage 2 previous stage sector (X 1, Y 1, and Z 1 ) and the stage 3 normal sector (X 2, Y 2, and Z 2 ) were highest. Figure 7 shows the changes in tristimulus values over time for fruits with non-uniform color development. However, no significant difference remained at the end of the maturity development (at storage days = 6). The tristimulus values of the two sectors became closer. Hence, the changes in light reflectance on the first day of testing were of most interest. Based on the hypothesis that mangosteens with color ratio of X 1 /X 2, Y 1 /Y 2, and Z 1 /Z 2 >1 could be considered defective, it was found that the probability of correctly

12 424 JARIMOPAS, PUSHPARIKSHA, AND SINGH Tristimulus values X1 Y1 Z1 X2 Y2 Z storage day(s) Figure 7 The changes in two sectors of a non-uniform color development mangosteen. (X 1, Y 1, Z 1 refer to previous stage sector; X 2, Y 2, Z 2, refer to normal sector). predicting spoilage using color ratio of X 1 /X 2, Y 1 /Y 2, and Z 1 /Z 2 >1 was 61.1%, 60%, and 56.8 % respectively. However, this might be inadequate because some non-uniform color fruits were found to be healthy. Accordingly, a color ratio of X 1 /X 2, Y 1 /Y 2 and Z 1 / Z 2 >1.05, 1.10, 1.15, 1.20, 1.25, 1.30, 1.35, 1.40, 1.45, and 1.50 was considered. As can be seen in Figure 8, the detection of defective fruits using X 1 /X 2 >1.25 and Y 1 /Y 2 >1.45 was better (67.4 and 66.4 % respectively). Using X 1 /X 2 >1.25, misidentification of defective fruits as sound was at 8.4% and that of sound fruit as defective was at 24.2%, while errors for both categories were recorded at 16.8% using Y 1 /Y 2 >1.45. Therefore, % correct prediction X1/X2 Y1/Y2 Z1/Z ratio Figure 8 Percentages of correct symptom predictions using the color ratio.

13 DAMAGE OF MANGOSTEEN AND QUALITY GRADING 425 the color ratio X 1 /X 2 >1.25 seemed more appropriate for a grading system because it was at this level that the proportion of defective fruits misidentified as sound was lowest. Manual grading of non-uniform color mangosteens as defective fruit was typically achieved at the error of 39%. Application of the determined mechanical and optical properties would be (a) the mechanical property like coefficient of friction could be used to design conveying and grading system for glossy- and rough-surface mangosteens; (b) the dimension ratio could be used to classify the undersize mangosteens out of the harvested fruits; and (c) the color ratio could be used to grade the translucent flesh mangosteens. CONCLUSION Data on the physical properties of mangosteens as they related to quality grading were obtained for large, medium, small, undersize, sound, and defective fruits. The defects included rough surface, translucent flesh and gummosis. Of special significance to present study was that the optical properties of mangosteen could also be affected by external and internal fruit defects. By measuring light reflected from fruits with defects and comparing it with reflections from sound fruits, defective fruits could be sorted from non-damaged fruits. Postharvest damage of mangosteens at wholesale level was also quantified. The dimension ratio was proposed as a better sizing indicator for undersized fruits than the weight threshold. The mechanical and color properties of rough-surface fruits, which made up to about 50% of production yield, were evaluated. It was possible to use the coefficient of static friction with plexiglass or the y value of the chromaticity coordinate for classifying rough- and glossy-surface fruits. The rind color not only indicated the degree of fruit maturity but also identified internal disorders such as translucent flesh and gummosis. It was found that non-uniform color on the skin of the fruits provided probable indication of internal disorders. Accordingly, color measurement allowed nondestructive prediction of the fruit s internal condition of mangosteens, and thus a color ratio was suggested for the purpose of detection of defects. Further research will involve validation of the suggested color threshold in order to design an experimental setup for detecting defects. ACKNOWLEDGMENTS The authors would like to acknowledge the Postgraduate Education and Research Development Project in Postharvest Technology, Chiang Mai University, and Kasetsart University Research and Development Institute, Thailand, for sponsoring this research. Deep appreciation is given to Dr. P. Chen, Professor Emeritus, Department of Biological and Agricultural Engineering, University of California, Davis, USA for his valuable guidance. REFERENCES 1. Department of Agricultural Extension. Fruits production of the eastern region of Thailand in 2001; DOAE: Bangkok, (in Thai) 2. Noichinda, S. Effect of modified atmosphere condition on quality and storage life of mangosteen fruit. M.S. Thesis; Kasetsart University: Bangkok, (in Thai). 3. Thailand Institute of Scientific and Technological Research. Color indices of mangosteen fruit;tistr: Bangkok, (in Thai)

14 426 JARIMOPAS, PUSHPARIKSHA, AND SINGH 4. Dangcham, S. Mechanism of flesh translucent disorder development of mangosteen fruit. M.S. Thesis; Kasetsart University: Bangkok, (in Thai) 5. Laiwisetkul, S. Factors influencing the development of translucent disorder in mangosteens. M.S. Thesis; Kasetsart University: Bangkok, (in Thai) 6. Lakhanatinnawong, W. The study on chemical composition, cell viability and influence of water on flesh translucent disorder in mangosteen. M.S. Thesis; Kasetsart University: Bangkok, (in Thai) 7. Jarimopas, B.; Toomsaengtong, S.; Inprasit, C. Design and testing of mangosteen rotating disk sizing machine. Journal of Food Engineering 2007, 79, Jarimopas, B. Post-harvest Machinery and Packaging of Fruit; Department of Agricultural Engineering, Kamphaengsaen Engineering Faculty: Nakorn Pathom, Thailand, 2001; 134 pp. 9. Jarimopas, B.; Kongwatananon, K.; Rangdang, C.; Yamashita, R. Mangosteen sizing machine. Kasetsart J. 1988, 22, Wang, J; Teng, B. Firmness evaluation by drop impact characteristics for peach. International Journal of Food Properties 2006, 9 (3), Bodria, L.; Fiala, M.; Guidetti, R.; Oberti, R. Optical techniques to estimate the ripeness of red-pigmented fruits. Transactions of the ASAE 2004, 47 (3), Raju, P.S.; Jagannath, J.H.; Ashok, N.; Das Gupta, D.K.; Bawa, A.S. Nondestructive monitoring of ripeness in mango cultivars by acoustic resonance spectroscopy. International Journal of Food Properties 2006, 9 (3), Jha, S.N.; Matsuoka, T.; Miyauchi, K. Surface gloss and weight of eggplant during storage. Biosystems Engineering 2002, 81 (4), Quevedo, R.; Aguilera, J.M. Characterization of food surface roughness using the glistening points method. Journal of Food Engineering 2004, 65, Peleg, K. Produce handling, packaging and distribution; AVI Publishing: Westport, CT, 1985; 642 pp. 16. Chen, P. Quality evaluation technology for agricultural products. Invited paper for presentation of the International Conference on Agricultural Machinery Engineering, Seoul, Korea. November 12 15, Mohsenin, N.N., Physical properties of plant and animal materials; Gordon and Breach Science Publishers: New York, 1986; 742 pp. 18. Sitkei, G. Mechanics of agricultural materials; Akademiai Kiado: Budapest, Jarimopas, B.; Terdwongworakul, A.; Timkham, P.; Roongratanaubol, K. Semi-automatic device determining coefficient of friction and rolling angle of vegetable and fruit. Thai Society of Agricultural Engineering Journal 2004, 11 (1), Gunasekaran, S. Nondestructive food evaluation techniques to analyze properties and quality; Marcel Dekker: New York, Chen, P.; Squire, E.F. An evaluation of the coefficient of friction and abrasion damage of oranges on various surfaces. Transactions of the ASAE 1971, 14 (6),