Nondestructive Measurement of Peel Gloss and to Determine Tomato Fruit Ripening and Chilling Injury A. NUSSINOVITCH, G. WARD, and S. LURIE ABSTRACT Mature-green tomatoes showed a gradual decrease in peel gloss during ripening at 20 C. Gloss correlated well with parameters such as a* values and Brix and could possibly serve as criteria for ripeness. When stored at 2 C for 7, 14 or 21 days and then allowed to ripen at 20 C, fluctuations in gloss during ripening were observed. Peel roughness increased as a result of chilling, but did not appear to be related to time elapsed at the chilling temperature and no changes were observed on subsequent ripening. Pink tomatoes showed fluctuations in gloss similar to maturegreen fruit. Chilling had no influence on roughness of pink tomatoes. Key Words: tomatoes, peel gloss, ripeness, chilling injury, roughness INTRODUCTION MANY HORTICULTURAL COMMODITIES of tropical or sub-tropical origin are injured when subjected to below optimal temperatures above freezing. Such damage has been termed chilling injury (Lyons, 1973) and the critical temperature below which injury occurs may range from just above the freezing point of the tissue to 10 to 12 C (Graham and Patterson, 1982). Physiological and biochemical alterations are associated with chilling stress and the extent of such alterations and the ability of the plant to withstand them define the degree of sensitivity or resistance to chilling injury at that temperature (Wang, 1982). Symptoms of chilling injury in most fruits and vegetables appear first on and within the peel (Lyons, 1973). Nondestructive physical techniques could possibly give an early indication of damage. The development of a novel glossmeter suitable for measuring curved surfaces prompted the study of natural gloss properties of fruits and vegetables (Nussinovitch et al., 1996). Gloss is a perceptual attribute of surfaces related to the degree to which a surface simulates a perfect mirror in its capacity to reflect incident light (Hunter, 1952). It was postulated (Corey and Schlimme, 1988a, b) that the increase in rind gloss of Blue Belle watermelons with age and the decrease in rind gloss of Charleston Gray fruits, 29 days after anthesis, may result from changes in quantity and structure of surface waxes. Such changes in gloss were suggested as a possible basis for determination of ripeness of watermelon. This could enable more effective judgment of the appropriate time to harvest watermelons. A similar study on bananas, showed that gloss could be used as an indicator of ripeness (Ward and Nussinovitch, 1996). Our current study was undertaken to determine the effects of chilling on peel gloss of tomatoes and whether changes may provide an early indication of chilling injury extent. Our first objective was to determine gloss properties of tomatoes harvested at mature-green stage, subjected to chilling temperature of 2 C for varying periods of time and then allowed to ripen. Symptoms of chilling injury usually become apparent only after exposure to ripening temperatures. Results could then be compared with typical ripening tomatoes (controls). Chilling sensitivity in tomatoes appears to change during ripening (McColloch Authors Nussinovitch and Ward are with the Institute of Biochemistry, Food Science & Nutrition, The Hebrew University of Jerusalem, Faculty of Agriculture, P.O. Box 12, Rehovot, 76100, Israel. Author Lurie is with the Dept. of Postharvest Storage, Volcani Center, ARO, Bet Dagan, 50250, Israel. et al., 1966; Cook et al., 1958). Therefore we decided to perform a similar experiment on tomatoes harvested at the pink-stage for comparison. Peel roughness may also be influenced by chilling. This property was studied to determine its relationship to chilling stress. may also influence the gloss of a surface and therefore it was of interest to see if these parameters were related to each other. MATERIALS & METHODS DANIELLA 144 TOMATO FRUIT were obtained from plants grown in a greenhouse at 20 C. All plants were grown in artificial soil and supplied with water and fertilizer by drip irrigation. This cultivar is of indeterminate growth type and continues producing flowers and fruits on the upper branches as it grows. When lower fruits are harvested the vine is pruned so that it will continue to grow upwards and produce more fruit. In the first part of the study, tomatoes were harvested at mature-green stage and transferred directly to a chilled storage chamber, where they remained at 2 C for 7, 14 or 21 days. When chilling was terminated, the tomatoes were allowed to ripen at 20 C, the typical temperature for study of tomato shelf life (Hobson and Davies, 1971). All measurements were made at 2-day intervals after removal from the storage chamber. Control fruit consisted of fruit harvested at mature-green stage and allowed to ripen at 20 C. In the second part of the study, fruit were harvested at pink stage and transferred directly to a chilled storage chamber, where they remained at 2 C for 7, 14 or 21 days. On termination of chilling, all measurements were made immediately. Control consisted of fruit harvested at pink stage and stored at 12 C for 7, 14 or 21 days. Tomatoes are commercially stored at this temperature. Gloss measurement Tomatoes were positioned in a curved surface glossmeter (Fig. 1) and illuminated by a light beam from a helium-neon laser at incident angles of 45 and. A semi-conductive plate collected all reflected light from the surface of the tomatoes and a video recorder (Sony, Japan) positioned directly facing the plate recorded the images. The recorded images were relayed to a computer where they were analyzed by a special computer program which translated them into the goniophotometric curve form of light intensity (arbitrary units) vs distance or light scattering (pixels). The widths of the curves at 50% intensity were measured as an indicator of gloss (Nussinovitch et al., 1996). For comparison of gloss, strips of peel from the same tomatoes were removed using a peeler, flattened and examined using a flat surface glossmeter, capable of measuring gloss at coincident angles of 20, and 85 (Triple Angle Novo-Gloss, Rhopoint Instrumentation Ltd, Germany). All results were recorded in gloss units, which are relative to a highly polished plane surface of black glass, which served as the standard and was arbitrarily assigned a gloss value near to 100, differing according to angle used. Flat surface gloss measurements were performed in a cabinet of 100% humidity, to prevent drying of peel. A total of 50 measurements were made by both methods at randomly chosen locations on the peels of 5 tomatoes. Strips of peel were removed from the tomatoes and laid on a flat surface. was measured using a Portable Surface Tester (Mitutoyo Corporation, Japan). All measurements were made perpendicular to the vertical axis of the tomato. Readings were given as the arithmetic mean deviation of the profile, Ra in µm for an evaluation Volume 61, No. 2, 1996JOURNAL OF FOOD SCIENCE383
PEEL GLOSS AND ROUGHNESS OF TOMATOES... Fig. 1(A) and (B)Curved surface glossmeter. 1. Video Recorder; 2. Laser; 3. Adjustable support; 4. Sample; 5. Support plate; 6. Semi-conductive plate; 7. Computer; 8. Typical goniophotometric curve. length of 12.5 mm at a speed of 0.5 mm/sec. This was performed at 100% humidity to prevent peel drying. A total of 50 measurements were made at random locations on the peels of 5 tomatoes. and chemical analysis was analyzed and monitored at different visible stages of ripening using a Minolta Chroma Meter CR-100, Japan. All colors within the range of human perception were converted and compared using parameters L*, a* and b*. In this color notation system, the L* indicates lightness/darkness, a* indicates red-green (axis) and b* yellow-blue (axis). The instrument was calibrated using a white standard plate. Total soluble solids content was determined using a digital refractometer (PR- 100, Atago, Japan). The ph was measured using a digital ph meter (El- Hama, Israel). Chilling injury Parallel studies were carried out to determine the extent of chilling injury. Fruit (3 replicates of 10 fruit) were examined at the time of removal from the chilling temperature and during the subsequent 10 days, while being held at 20 C. Firmness was measured using a Durometer, which nondestructively monitored the force to compress the flesh 4 mm. The percentage of rots and chilling injury which developed on fruits during ripening following storage was determined visually. Chilling injury was evidenced by surface pitting. At the end of shelf life, a section from each of the fruits was removed and 3 replicates of 10 sections were ground in a blender. Brix was determined using a digital refractometer and acidity by titrating 2 ML of juice to ph 8.2 with 0.01N NaOH. RESULTS & DISCUSSION COLOR, ROUGHNESS AND GLOSS characteristics of the peel of tomatoes harvested at the mature-green stage and left to ripen at 20 C were compared (Table 1). A gradual decrease in L* value was observed from 50.8 at the turning stage of ripening to 42.7 at the red stage of ripening, corresponding to darkening of the peel. The a* value gradually increased from 9.2 at mature-green stage to 19.2 at the red stage corresponding to change in peel color from green to red. The b* value increased as the peel yellowed and then decreased as ripening proceeded and yellow color was masked. These results were related to changes in tomato pigments during ripening (Hobson and Davies, 1971). The green color of immature tomato fruit is attributable to a mixture of chlorophylls. With incipient ripening, yellow pigments ( -carotene and xanthophylls) are produced and become more apparent as chlorophyll decreases. A subsequent rapid accumulation of the red pigment lycopene influences the fruit color, although the yellows are reinforced by lutein and lycoxanthin. Tomato pigments have been the focus of many reports (Hobson and Davies, 1971; Grierson and Kader, 1986). A typical roughness profile of the surface of a tomato at the red stage of ripening was developed (Fig. 2). Peel roughness of tomatoes did not change during ripening and remained stable at about 2 µm Ra, denoting a smooth surface. This lack of change contrasted with banana peel, where an increase in roughness from 3.4 to 9.5 µm Ra was observed during ripening (Ward and Nussinovitch, 1996). Goniophotometric reflection curves were obtained from the curved surface glossmeter at a coincident angle of for control tomatoes (Fig. 3). In such curves, correlation of dispersion with gloss is inverse. Where dispersion is high, gloss is low and conversely. Thus the narrower the width of the curve, the greater the gloss value (Nimeroff, 1952). The width of the curve at 50% intensity was defined as gloss value in the curved surface glossmeter (Nussinovitch et al., 1996). Thus the smaller the value the greater the gloss. The opposite would apply for the flat surface glossmeter. The flat surface glossmeter was used for reference since the only published information on gloss involves measurements made using flat surface glossmeters. We recognized that flattening of the peel may affect the surface gloss of the tomato due to changing of the natural structure. A gradual decrease in peel gloss was observed as tomatoes ripened from mature-green to red at all coincident angles, with both methods. It is well known that high coincident angles are more suitable for matt surfaces (Hunter, 1952) Thus we assumed that for a slightly glossy surface such as tomato peel an inbetween angle () would be more suitable. Correlation coefficients between gloss at different angles and other parameters were determined (Table 2). With a*, good correlation coefficients were obtained for all coincident angles. Since color is the generally accepted criterion for determining tomato ripeness (Hobson and Davies, 1971), our results suggest that peel gloss may have a potential value as a criterion for ripening. The high correlation coefficients obtained with L*, ph and Brix supported this possibility. The b* value did not correlate well with gloss. To validate results, correlation was determined between gloss as measured by flat surface glossmeter and gloss as measured by the curved surface meter. However, this was only possible for results obtained at the angle. A significant correlation coefficient ( 0.954, significant at P 0.05) provided validation of the results. Since gloss as measured at by both methods gave very good correlations with other parameters, all subsequent gloss measurements were taken at that angle only. Gloss, roughness and color characteristics of the peel of tomatoes harvested at mature green stage, stored at 2 C for 7 (Table 3), 14 (Table 4) or 21 days (Table 5) and then allowed to ripen at 20 C were compared. values indicated the failure to ripen normally, which is a symptom of chilling injury in tomatoes (McColloch et al., 1966). With all tomatoes examined, including controls, a gradual increase in a* value was observed, corresponding to the change in color from green to red. However, for tomatoes subjected to chilling temperatures this in- 384JOURNAL OF FOOD SCIENCEVolume 61, No. 2, 1996
Time/Stage b Table 1Changes in some physical and chemical parameters of mature-green tomatoes, a Gloss (flat) 20 85 Gloss (curved) 45 ph Brix 0 days/mg 49.8 3.0 9.2 0.7 21.1 2.7 2.2 0.4 12.9 2.2 11.1 2.0 60.1 2.6 48.6 4.2 1.8 0.2 3.51 5 2 days/t 50.8 1.6 3.7 0.9 28.3 2.4 2.6 0.1 8.2 1.0 5.1 1.0 70.1 6.8 60.9 7.6 2.3 0.3 3.53 5 4 days/p 46.6 1.2 9.1 3.5 25.0 2.2 1.3 0.1 6.4 0.7 3.4 0.3 82.8 3.7 59.7 4.7 1.7 0.4 3.66 5.2 6 days/r 42.7 1.5 19.2 2.4 24.5 2.1 1.4 0.1 5.3 0.3 1.8 0.3 85.2 4.2 66.8 5.6 2.2 0.4 3.84 5.5 a Results for color, gloss, and roughness are means of 50 measurements SD, taken at different locations on 5 fruit. With the curved surface glossmeter, results are in pixels where an increase in value indicates a decrease in gloss. For the flat surface glossmeter, results are in gloss units where an increase in value indicates an increase in gloss. b MG mature green, T turning, P pink, R red. Fig. 2Typical roughness profile for a control tomato at red stage of ripening. crease seemed to be delayed. After 6 days ripening at 20 C control tomatoes had an a* value of 19.2 and those subjected beforehand to 2 C for either 7, 14 or 21 days had a* values of 5.5, 4.2 and 8.1 respectively. With controls, a decrease in L* value from 50.8 at 2 days to 42.7 at 6 days of ripening was observed, corresponding to darkening of the peel on ripening. This trend was not so apparent in tomatoes after storage at 2 C. Slight fluctuations occurred in the L* values of these tomatoes and similar fluctuations in b* values were observed. Fluctuations were also observed in gloss properties of such tomatoes, although the extent of these did not seem related to the period of time stored at 2 C. Changes in gloss may be a result of changes in the quantity and structure of epicuticular waxes present on the peel (Corey and Schlimme, 1988a, b). Fluctuations observed may be a consequence of irregularities in normal changes in epicuticular wax on the tomato peel. Sensitivity to chilling injury may vary with location on the tomato peel due to differences in thickness of the cutin layer and of the peel. Thus it can be seen that unequal changes in quantity and structure of wax could result. Consequently, irregularities in gloss values would be apparent. All gloss values are presented as means SD of 50 measurements. Any such nonuniform changes caused by chilling would be indicated by higher coefficients of variance (Table 6). In general, those tomatoes subjected to chilling temperatures prior to ripening showed larger values than controls. Differences in gloss properties of those mature-green tomatoes allowed to ripen at 20 C and those allowed to ripen only after chilling, provide a potential for use of gloss measurement as an indicator of chilling injury. Peel roughness values of mature green tomatoes subjected to chilling temperatures and then allowed to ripen were compared (Tables 3, 4, and 5). Immediately after removal, peel roughness of tomatoes subjected to 2 C for 7, 14 or 21 days was 4.7, 3.6 and 4.5 µm respectively. No significant changes were observed on subsequent ripening. Since peel roughness of typical ripening tomatoes had remained stable at around 2 µm Ra, these differences suggest that an increase in peel roughness is a result from chilling of mature-green tomatoes, that is apparent immediately after removal from the chilling chamber. Fig. 3Typical goniophotometric curves for control tomatoes. Parallel studies relating the extent of chilling injury and increased sensitivity to fungal rots showed that without storage at 2 C, neither of these symptoms appeared. Following 7 days at 2 C, the amount of visible chilling injury increased until 10 6% of all tomatoes showed signs (evidenced by surface pitting) after 10 days subsequent storage at 20 C. Following 14 days at 2 C, the same results were observed, yet the amount of chilling injury increased to 45 10% after being stored at 20 C for 10 days. A further increase to 60 11% chilling injury was noted after 21 days at 2 C followed by 10 days at 20 C. Rots caused by both Botrytis cinerea and Alternaria alternata were only evident following 14 and 21 days at 2 C. The percentage of affected tomatoes increased with subsequent storage time at 20 C. After 10 days, 45 9% of tomatoes subjected to 2 C for 14 days and 75 9% of those subjected to 2 C for 21 days showed fungal rots. These results supported previous reports that low temperature storage of tomatoes increased susceptibility to fungal rot (McColloch and Worthington, 1952). Volume 61, No. 2, 1996JOURNAL OF FOOD SCIENCE385
PEEL GLOSS AND ROUGHNESS OF TOMATOES... Table 2Correlation coefficients between gloss, measured at different coincident angles and other parameters of ripening tomatoes Method Angle ph of pulp Brix of pulp Flat surface 20 0.867 0.831 a 0.258 0.802 0.797 Glossmeter 85 0.717 0.698 0.896 a 0.883 a 0.574 0.603 0.803 0.791 0.762 0.750 Curved surface 45 0.830 0.957 b 0.355 0.875 0.842 Glossmeter 0.669 0.839 b 0.641 0.784 0.750 a Significant at P 0.05 b Significant at P 0.01 Table 3Changes in color, gloss and roughness of mature-green tomatoes stored at 2 C for 1 wk and then a 0 51.1 3.4 9.6 0.9 21.3 2.0 8.7 1.0 48.6 2.9 4.7 1.4 2 47.4 3.2 6.2 0.4 16.1 1.9 14.5 3.1 39.9 8.1 3.3 0.8 4 48.8 1.7 1.4 3.8 26.4 1.4 4.8 0.3 51.9 12.9 3.7 0.8 6 46.4 1.4 5.5 4.9 25.0 3.6 4.7 0.5 50.5 13.9 5.2 1.8 8 45.1 1.9 10.9 5.0 28.5 1.7 8.7 1.6 45.9 5.3 4.6 0.9 indicates a decrease in gloss. For the flat surface glossmeter, results are in gloss units where an increase in value indicates an increase in gloss. Table 4Changes in color, gloss and roughness of mature-green tomatoes stored at 2 C for 2 wk and then a 0 52.6 4.0 9.8 1.4 23.6 3.4 5.9 0.4 46.4 5.3 3.6 0.9 2 49.9 2.7 5.0 1.6 17.4 2.3 9.1 3.0 50.1 5.8 4.1 0.8 4 53.7 2.9 3.2 3.0 31.5 2.2 10.6 2.9 48.4 6.5 3.8 1.1 6 48.8 1.3 4.2 4.5 28.9 4.1 5.7 0.8 44.6 7.2 4.3 0.8 8 43.8 1.8 14.3 3.3 25.3 1.4 12.9 2.2 53.7 11.6 4.2 1.1 indicates a decrease in gloss. For the flat surface glossmeter, results are in gloss units where an increase in value indicates an increase in gloss. Table 5Changes in color, gloss and roughness of mature-green tomatoes stored at 2 C for 3 weeks and then a Gloss (Flat) 0 49.6 3.6 8.8 1.6 20.9 2.2 7.7 1.1 46.8 6.2 4.5 1.7 2 50.0 2.3 3.4 4.9 22.2 2.5 5.2 0.9 56.2 14.4 6.1 1.3 4 47.2 2.5 2.5 8.1 23.8 4.8 9.8 1.6 71.8 9.1 4.6 1.2 6 45.8 3.7 8.1 5.6 24.1 2.2 4.1 1.1 72.1 14.3 5.1 0.9 indicates a decrease in gloss. For the flat surface glossmeter, results are in gloss units where an increase in value indicates an increase in gloss. Only 4 stages were examined since the fruit became rotten before they had fully ripened. Table 6Coefficients of variance (%) of gloss as measured at of mature-green tomatoes harvested and the and of mature-green tomatoes after chilling stress Gloss measured: Flat surface glossmeter Curved surface glossmeter Days of ripening 0 2 4 6 8 0 2 4 6 8 Mature-green tomatoes 17.1 12.2 10.9 5.7 8.6 12.5 7.9 8.4 7 days at 2 C and then 11.5 21.4 6.3 10.6 18.4 6.0 20.3 24.9 27.5 11.5 14 days at 2 C and then 6.8 33.0 27.4 14.0 17.1 11.4 11.6 13.4 16.1 21.6 21 days at 2 C and then 14.3 17.3 16.3 26.8 13.2 25.6 12.7 19.8 Unstored tomatoes softened more than those subjected to chilling stress. Firmness (N) of unstored tomatoes decreased from 53.5 4.4 after 3 days at 20 C to 35.7 4.9 after 10 days. Following 7 days at 2 C, firmness decreased from 51.3 4.1 after 3 days at 20 C to 42.9 5.5 after 10 days. Similar results were obtained for tomatoes stored for 14 or 21 days, also indicating an apparent abnormal softening. Storage had no effect on the Brix or titratable acidity. Gloss, color, roughness, ph, and Brix of tomatoes harvested at pink stage and then stored at either 2 C or 12 C for 7, 14 or 21 days were also compared (Table 7). As noted with maturegreen tomatoes harvested and stored at 20 C, pink tomatoes harvested and stored at 12 C showed a decrease in gloss as they ripened. This provides additional evidence that gloss may have potential as an indicator of ripening. The increase in a* value indicated that ripening was taking place. Peel roughness of tomatoes in this experiment remained stable regardless of time or whether tomatoes were stored at 2 C or 12 C. Values were similar to those for typical ripening mature-green tomatoes. This seemed to indicate that chilling had no influence on roughness of pink tomatoes in contrast to mature green tomatoes. Pink tomatoes did not exhibit typical 386JOURNAL OF FOOD SCIENCEVolume 61, No. 2, 1996
Storage Time 0 L* value 43.6 1.2 Table 7Changes in some physical and chemical parameters immediately after storage a* value 9.0 2.8 b* value 21.3 1.6 6.5 1.4 61.7 7.1 1.8 0.3 1wk2 C 50.0 1.6 9.5 1.9 31.3 2.4 5.1 1.0 71.5 9.8 1.4 0.1 4.9 3.79 2wk2 C 47.9 2.2 7.4 4.1 27.2 2.0 9.5 1.4 67.3 7.9 1.8 0.3 4.4 3.7 3wk2 C 47.7 3.9 5.7 4.4 26.7 4.3 4.6 1.8 77.7 12.8 2.2 0.7 4.8 3.77 1wk12 C 41.3 2.4 16.9 1.3 20.8 3.2 5.6 0.9 66.2 8.5 1.7 0.2 5.2 3.83 2wk12 C 40.1 1.4 20.9 2.5 20.9 1.8 4.4 1.4 69.0 13.4 2.1 0.5 5.4 3.92 indicates a degrease in gloss. For the flat surface glossmeter, results are in gloss units where an increase in value indicates an increase in gloss. Brix 4.9 ph 3.82 chilling injury symptoms either. Susceptibility to chilling injury changes during ripening (McColloch et al., 1966; Cook et al., 1958). CONCLUSIONS PEEL GLOSS MEASUREMENTS may serve as criteria for ripeness of tomatoes as well as indicators for chilling injury. Peel roughness in mature-green tomatoes is influenced by chilling and differences were apparent immediately after removal from the chilling chamber. REFERENCES Cook, C.C., Parson, C.S., and McColloch, L.P. 1958. Methods to extend storage of fresh vegetables aboard ships of the U.S. Navy. Food Tech. 12: 548. Corey, K.A. and Schlimme, D.V. 1988a. Relationship of rind gloss and groundspot color to flesh quality of watermelon fruits during maturation. Scientia Hort. 34: 211-218. Corey, K.A. and Schlimme, D.V. 1988b. Changes in the epicuticular wax on watermelon fruits during ripening. HortScience 23(4): 730-731. Graham, D. and Patterson, B.D. 1982. Responses of plants to low non-freezing temperatures: Proteins, metabolism and acclimitation. Annu. Rev. Plant Physiol. 33: 347-372. Grierson, D. and Kader, A.A. 1986. Fruit ripening and quality. Ch. 6, in The Tomato Crop, J.G. Atherton and J. Rudich (Ed.), p. 241-280. Chapman and Hall, London. Hobson, G.E. and Davies, J.N. 1971. The tomato. Ch. 13, in The Biochemistry of Fruits and their Products, A.C. Hulme (Ed.), Vol. 2, p. 437-482. Academic Press, London. Hunter, R.S. 1952. Gloss Evaluation Of Materials. ASTM Bulletin No. 186, p. 48-55. Lyons, J.M. 1973. Chilling injury in plants. Annu. Rev. Plant Physiol. 24: 445-466. McColloch, L.P. and Worthington, J.T. 1952. Low temperature as a factor in the susceptibility of mature-green tomatoes to alternaria rot. Phytopathology 42: 425-427. McColloch, L.P., Yeatman, J.N., and Loyd, P. 1966. changes and chilling injury of pink tomatoes held at various temperatures. USDA Mktg. Res. Rpt. 735. Nimeroff, I. 1952. Analysis of goniophotometric reflection curves. J. Research, National Bureau of Standards 48(6): 441-448. Nussinovitch, A., Ward, G., and Mey-Tal, E. 1996. Gloss of Fruits and Vegetables. J. Food Sci. Technol. (In press). Wang, C.Y. 1982. Physiological and biochemical responses of plants to chilling stress. HortScience 17(2): 173-181. Ward, G. and Nussinovitch, A. 1996. Peel gloss as a potential indicator of banana ripeness. J. Food Sci. Technol. (Accepted). Ms received 9/6/95; revised 10/2/95; accepted 10/18/95. Volume 61, No. 2, 1996JOURNAL OF FOOD SCIENCE387