- Email: [email protected]
Modeling changes of sensory attributes for individual and mixed fresh-cut leafy vegetables
Postharvest Biology and Technology 38 (2005) 202–212
Modeling changes of sensory attributes for individual and mixed fresh-cut leafy vegetables Andrea M. Piagentini, Julio C. Mendez, Daniel R. Guemes, Mar´ıa E. Pirovani ∗ Instituto de Tecnolog´ıa de Alimentos, Facultad de Ingenier´ıa Qu´ımica, Universidad Nacional del Litoral, C.C. 266-3000 Santa Fe, Argentina Received 10 August 2004; accepted 2 July 2005
Abstract Changes of the main sensory attributes of three fresh-cut leafy vegetables (Iceberg and Romaine lettuce and chicory) were investigated at selected temperatures (2–20 ◦ C). The aim of this work was to develop and apply a kinetic approach to model sensory quality changes in order to establish the appropriate function that describes the time–temperature dependence of each attribute. The changes of the sensory characteristics followed first order reaction kinetics and the temperature dependence of rate constants followed the Arrhenius relationship. The limiting quality factor, which determined the sensory shelf life for the three fresh-cut vegetables assayed, at any temperature, was the general appearance of the products. The activation energies obtained for general appearance were 71.1 kJ mol−1 for fresh-cut Iceberg lettuce, 69.5 kJ mol−1 for fresh-cut Romaine lettuce and 65.7 kJ mol−1 for fresh-cut chicory. Additional experimental tests showed that the predicted and experimental sensory shelf life for individual fresh-cut vegetables at constant temperature were not different (P > 0.05). Under dynamic temperature conditions (sequence of different temperatures), the predicted and experimental values of browning, wilting, and off-odour were also not different (P > 0.05), but the general appearance loss model overestimated the quality loss from 10 to 30%. The models of quality change for individual vegetables were used to predict the sensory shelf life of fresh-cut mixed vegetables. The experimental validation tests proved that these models provide a good approach to evaluate the shelf life of the mixed product. Results showed that the general appearance of fresh-cut Iceberg and Romaine lettuce dominated the sensory perception of mixed product. © 2005 Elsevier B.V. All rights reserved. Keywords: Sensory quality; Shelf life; Iceberg lettuce; Romaine lettuce; Chicory
1. Introduction Leafy vegetables are a well-recognized source of minerals, vitamins, and dietary fibre. The desire of ∗
Corresponding author. E-mail address: [email protected] (M.E. Pirovani).
0925-5214/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.postharvbio.2005.07.001
consumers for fresh-cut vegetables due to their convenience and fresh-like properties (texture, flavour, and appearance) has led to a relatively new area of food preservation named minimally or lightly processed (King and Bolin, 1989; Cantwell, 1992), IV Gamme (Carlin et al., 1990), ready-to-use (Francis et al., 1999), or fresh-cut fruits and vegetables (Gorny,
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
1996). These types of products do not need additional preparation at home and have been slightly modified in their fresh characteristics. In Argentina, their demand has increased markedly in the last years and industry has expectations of further growth. Although it is a relatively new market it represents 20% of the total of vegetable sales in supermarkets. Therefore, this kind of product appears destined to become an important component of the food industry but efforts must be taken to ensure products of high quality. Fresh-cut leafy vegetables have a short shelf life and are exposed to conditions that can destroy their superior quality especially during transport and retailing. Temperature conditions primarily determine the rate of quality degradation and the shelf life of the product causing changes in sensory characteristics, which can influence consumer acceptability. Appearance is the most important attribute evaluated by the consumer in the decision to accept a packaged leafy vegetable (IFT, 1990). Subsequently, when the package is open, aroma or the off-odours, if present, would be also important attributes. Several studies on quality changes of different freshcut vegetables for specific storage conditions have been reported (Bolin et al., 1977; Barriga et al., 1991; Heimdal et al., 1995; Pirovani et al., 1998; Artes et al., 1999) but little is available on sensory quality modeling to predict quality degradation as a function of temperature and time, simultaneously (Vankerschaver et al., 1996; Piagentini et al., 2004). The aim of this work was to develop and apply a kinetic approach to model sensory quality changes in order to establish the appropriate function that describes the time–temperature dependence of each attribute for three fresh-cut leafy vegetables. The kinetic parameters, namely rate constants and activation energies, provide useful information on the quality changes that could occur during commercial handling. By establishing the appropriate quality function, a quantitative tool for shelf life estimation during different conditions of transport and retailing management can be obtained.
2. Materials and methods 2.1. Preparation of fresh-cut vegetables Iceberg lettuce (Latuca sativa L. var. capitata L.), Romaine lettuce (L. sativa L. var. longifolia Lam.),
203
and chicory (Cichorium intybus L.) were obtained from a local farm near Santa Fe (Argentina). Upon arrival, the vegetables were stored at 4 ◦ C and 90% RH, and processed separately 1 day after harvest. Outer, damaged and yellowed leaves, roots, and stems were removed. The remaining leaves were cut in shreds of 5 mm width with a sharp stainless steel knife. The cut vegetable was washed for 4 min in cold water (4–6 ◦ C) containing 100 mg/L available chlorine as sodium hypochlorite and pH 6.8, with a water-toproduce ratio of 18 L/kg. Then, it was rinsed with running tap water (0.2 mg/L total chlorine) for 4 min. The vegetable was centrifuged in a basket-type centrifuge at 540 min−1 for 4 min. Each fresh-cut vegetable sample (70 g for Iceberg and Romaine lettuce, 50 g for chicory) was placed in a semi-rigid polyethylene tray (130 mm × 100 mm × 40 mm) overwrapped with a plasticized PVC film 13 m thick, with a surface area of 0.013 m2 . Gas transmission rates of PVC film for O2 , CO2 , and water vapour were 8.71 × 10−11 and 7.96 × 10−10 mol m−2 s−1 Pa−1 (both at 23 ◦ C and 1.01 × 105 Pa) and 2.96 × 10−4 mol m−2 s−1 (at 37 ◦ C and 90% RH), respectively. Finally, samples were stored at 90% RH and: 1.7, 4.7, 8.9, or 20.3 ◦ C for Iceberg lettuce; 1.4, 4.3, 8.9, or 20.3 ◦ C for Romaine lettuce; 1.6, 4.5, 8.9, or 20.3 ◦ C for chicory. The experiment was repeated three times for each vegetable. 2.2. Sensory analysis The evaluation of the characteristic sensorial properties of the samples was performed by a descriptive test. A trained sensory panel of 6–8 judges, which had previously participated in evaluating fresh vegetable qualities, was required to evaluate sensory quality attributes of fresh-cut vegetable samples. The methodology to select and train panellists was the same described by Pirovani et al. (1998). During the specific training (5 × 30 min sessions), the panellists discussed and agreed on sensory attributes and anchored terms. They evaluated off-odour development, general appearance, wilting, and browning of fresh-cut vegetable each testing day. The judges indicated their perception of each quality attribute intensity on a 150 mm unstructured line, with anchored terms located 15 mm from either end. They scored the perceived intensity of each attribute by placing a vertical line across the unstructured scale
204
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
line. Quantitation was accomplished by measuring the distance from the left end (0.00) to the vertical line, reporting measurements in centimeters. The anchored terms were indicated from left to right as: none and severe for off-odour; very poor and excellent for general appearance; none and very severe for wilting and browning. The panellists were instructed to open the sealed samples and evaluate the off-odour first, and then the other attributes. Each panellist performed the sensory test individually in separate booths with white incandescent lighting (sufficient to provide 700 lx). Samples were coded with three-digit random numbers. The central point of the scale was established as the cut-off score. Therefore, the fresh-cut vegetable was considered as unacceptable when a mean score below 7.5 was reached for general appearance or above 7.5 for the other sensory attributes (Barriga et al., 1991; Piagentini et al., 1997, 2004; Jacxsens et al., 2002).
mination of the reaction order. Then, the effect of temperature on quality attribute changes is modeled. The influence of a constant storage temperature on the reaction rate constant can be described using the Arrhenius equation (Saguy and Karel, 1980; Taoukis et al., 1997): −Ea (2) kq(T ) = k0 exp RT where kq(T) is rate constant for each quality atribute; k0 the pre-exponential factor; Ea the activation energy (J mol−1 ); R the universal gas constant (8.3145 J K−1 mol−1 ); T is the absolute temperature (K). The Ea for each quality attribute is obtained by regression analysis. Another parameter that is often used to describe the relationship between temperature and reaction rate constant is the Q10 value. Q10 is defined as follows: Q10 =
2.3. Weight loss Weight loss was determined on all samples at each testing day and was expressed as a percentage of initial fresh weight. 2.4. Mathematical modeling of quality changes The rate of quality changes of foods can be described by the following general equation: ±
dQ = kq [Q]n dt
(1)
where Q is quality attribute; t the time; n the reaction order; kq is the quality change rate constant for the attribute Q. The sign (+) refers to attributes with increasing values during time (browning, wilting, and off-odour) and the sign (−) to attributes with decreasing values (general appearance). Traditionally, quality change processes of foods stored under controlled environmental conditions are described with zero order and/or first order rate functions (Saguy and Karel, 1980; Taoukis et al., 1997). The regression analysis is used to determine the kinetic order of the quality change rate. An analysis of the coefficient of determination (R2 ), residuals (ei , defined as the difference between observed and fitted data points) and estimated value of the intercept would enable deter-
reaction rate at temperature (T + 10) reaction rate at temperature (T )
(3)
Equivalently, Q10 has been defined as the change of shelf life (ts ), i.e., the time for the failure attribute to reach an unacceptable level when the food is stored at a 10 ◦ C higher temperature. It can be shown that Q10 and activation energy, Ea , are related by the following expression (Singh and Heldman, 1993): Ea 10 ln Q10 = (4) R T (T + 10) The time to reach a sensory characteristic limit of acceptability at a specified storage temperature (average score below 7.5 for general appearance or above 7.5 for browning, wilting, and off-odour) was derived from the prediction models (Eqs. (1) and (2)). For each fresh-cut vegetable, the limiting sensory characteristic (failure attribute) was obtained comparing the time to reach the limit of acceptability of each attribute and temperature. 2.5. Validation of the models Three confirmatory experiments were carried out to validate the fitted models at constant temperature. Samples of each fresh-cut vegetable were prepared in the same way as previously described and stored at different constant temperatures. The sensory shelf life of each sample was determined by evaluating the failure
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
205
attribute. Experimental data were compared to values predicted from the models by a t-test analysis. The combined temperature–quality change models, obtained under constant conditions, were tested to verify their adequacy under a variable temperature sequence. Therefore, these models were used to predict the evolution of sensory attributes of fresh-cut Iceberg lettuce stored at two different temperature profiles. Profile A: 2 h at 21.4 ◦ C; 46 h at 2.3 ◦ C, and finally, 94 h at 5.5 ◦ C. Profile B: 2 h at 22 ◦ C; 46 h at 9.7 ◦ C, and finally 94 h at 5.8 ◦ C. Experiments were repeated twice for each temperature profile. The observed and predicted failure attribute values of fresh-cut Iceberg lettuce were compared. The models to predict changes of general appearance obtained for each vegetable were also tested on mixed vegetable samples. Fresh-cut vegetables were prepared in the way previously described. The mixed vegetable sample was prepared by putting 20 g of each vegetable (Iceberg and Romaine lettuce, and chicory) together in the same tray, wrapping with PVC film and storing at 1.3, 4.3, 8.1, or 20.4 ◦ C. Data from three confirmatory experiments were compared to values predicted by model equations obtained for individual fresh-cut vegetable. 2.6. Statistical analysis All data were analysed using STATGRAPHICS Plus (Manugistics, Inc., Rockville, MD, USA). The average and standard deviation of weight loss values and individual scores of the sensory attribute given by the trained panel were calculated. The data were fitted to the corresponding models and the regression analyses were carried out. The 95% confidence intervals of estimated parameters were calculated. The experimental data obtained during the confirmatory experiments were compared to values predicted by model equation by a t-test analysis.
Fig. 1. Experimental and predicted values of sensory attributes for fresh-cut Iceberg lettuce throughout storage at different temperatures. Bars indicate S.D.
3. Results and discussion 3.1. Sensory characteristics of fresh-cut Iceberg and Romaine lettuces, and chicory The mean scores of sensory quality attributes (general appearance, browning, wilting, and off-odour) for
the three packaged fresh-cut leafy vegetables throughout storage at different temperatures are presented in Figs. 1–3. They showed that browning, wilting, and off-odour scores did not change as rapidly as general appearance scores. Panellists scored general appearance as the overall visual impact of the samples. During
206
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
Fig. 2. Experimental and predicted values of sensory attributes for fresh-cut Romaine lettuce throughout storage at different temperatures. Bars indicate S.D.
Fig. 3. Experimental and predicted values of sensory attributes for fresh-cut chicory throughout storage at different temperatures. Bars indicate S.D.
the process of language development, they agreed that wilting, browning, and all other unexpected visual quality factors would contribute to the general appearance score. At the end of experiments, some panellists indicated that colour changes or decay were included in general appearance evaluation.
As storage time and temperature increased, panellists gave lower scores for general appearance and higher ones for browning, wilting, and off-odour for all the samples. General appearance and browning changed the most during storage for both types of lettuce and for chicory
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
207
chicory. This is in accordance with sensory evaluation of wilting (score around four for both lettuce and six for chicory). 3.2. Modeling sensory characteristics of fresh-cut leafy vegetables
Fig. 4. Weight losses of fresh-cut vegetables during storage at different temperatures.
at all temperatures assayed. Jacxsens et al. (2002) found that general appearance (freshness) and the colour were the properties of packaged mixed lettuce that presented the largest changes. In the present work, wilting, and off-odour were important defects only in samples stored at higher temperatures (8.9 and 20.3 ◦ C). Weight loss of fresh-cut Iceberg and Romaine lettuce and chicory was less than 1.7% after 6 days of storage for temperatures lower than 4.7 ◦ C (Fig. 4). When storage temperature was 8.9 ◦ C, the weight loss after 6 days of storage was approximately 2–2.5% for both types of lettuce but significantly higher (3.7%) for
The first step in selecting an appropriate model to represent the quality changes was to perform regression analysis to determine the kinetic order. The performance of the fitted models was analysed. Based on the best coefficient of determination (R2 ), first order was the apparent order of the quality change reactions in the majority of the cases (Table 1). Further analysis determined the estimated value of the intercept of the first order model. Finally, the plots of residuals versus predicted values (not shown) for each model (zero and first order) indicated that the distribution around zero was more random for the first order model. Therefore, all these results were taken into account to select the first order reaction model to describe the quality changes of these leafy vegetables. Other researchers (Vankerschaver et al., 1996; Piagentini et al., 2004) also found that first order was the most adequate kinetic model for fitting sensory characteristic changes in minimally processed vegetables. Curves in Figs. 1–3 show the evolution of each predicted attribute using the fitted models. Tables 2–4 show the rate constants for each attribute and temperature and the activation energies for packaged fresh-cut Iceberg and Romaine lettuce, and chicory, respectively. Results indicated that for all the attributes the dependence of rate constants on temperature followed the Arrhenius relationship (R2 values ranged between 0.700 and 0.998). The Q10 values for each attribute and fresh-cut leafy vegetable are presented in Table 5. The activation energies of general appearance (71.1 and 69.6 kJ mol−1 ) and wilting (66.9 and 65.1 kJ mol−1 ) for fresh-cut Iceberg and Romaine lettuce, respectively, showed the highest values compared to other sensory characteristics indicating that these attributes are the most influenced by temperature. On the other side, lower activation energies of browning signify smaller temperature sensitivity for this attribute (Ea = 47.5 and 39.6 kJ mol−1 for Iceberg and Romaine lettuce, respectively). Therefore, a reduction in temperature would benefit general appearance and wilting
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
208
Table 1 Coefficient of determination (R2 ) for zero and first order models for sensory attribute of fresh-cut Romaine and Iceberg lettuces and chicory Attribute
Temperature (◦ C)a
Iceberg lettuce
Romaine lettuce
Chicory
Zero
First
Zero
First
Zero
First
General appearance
A B C D
0.8500 0.9849 0.9341 0.9165
0.9239 0.9937 0.9731 0.9974
0.9595 0.7799 0.8101 0.8597
0.9370 0.7513 0.9445 0.9892
0.4316 0.8178 0.9466 0.9164
0.4198 0.7879 0.7812 0.9133
Wilting
A B C D
0.7814 0.9687 0.9271 0.9843
0.8868 0.977 0.9153 0.9706
0.7821 0.5614 0.7081 0.9674
0.7874 0.6277 0.7626 0.8888
0.0070 0.3351 0.8923 0.8228
0.0105 0.3477 0.9119 0.893
Browning
A B C D
0.7902 0.9853 0.9408 0.9532
0.9349 0.9570 0.8550 0.8227
0.9313 0.7017 0.8154 0.9304
0.9714 0.7286 0.9140 0.7671
0.2810 0.6551 0.9169 0.9016
0.3212 0.6917 0.9555 0.8948
Off-odour
A B C D
0.6264 0.8539 0.8113 0.9649
0.8284 0.8615 0.8331 0.9832
0.7715 0.8069 0.652 0.8899
0.7920 0.8795 0.7522 0.8109
0.5152 0.5477 0.8930 0.9078
0.5768 0.6691 0.9216 0.9066
Fresh-cut Iceberg lettuce: A = 1.7 ◦ C, B = 4.7 ◦ C, C = 8.9 ◦ C, D = 20.3 ◦ C; fresh-cut Romaine lettuce: A = 1.4 ◦ C, B = 4.3 ◦ C, C = 8.9 ◦ C, D = 20.3 ◦ C; fresh-cut chicory: A = 1.6 ◦ C, B = 4.5 ◦ C, C = 8.9 ◦ C, D = 20.3 ◦ C. a
Table 2 Rate constant (kq ) from sensory attributes of fresh-cut Iceberg lettuce for the first order model and the respective activation energies (Ea ) from the Arrhenius equation Attribute
Temperature (◦ C)
Rate constanta , kq (day−1 )
Coefficient of determination (R2 )
General appearance
1.7 4.7 8.9 20.3
0.065 ± 0.019 0.161 ± 0.013 0.235 ± 0.054 0.576 ± 0.090
0.9239 0.9937 0.9731 0.9974
Ea (kJ mol−1 ) Wilting
71.1
Browning
66.9
47.5
Ea (kJ mol−1 ) a
kq ± confidence interval at 95%.
65.3
0.9349 0.9570 0.8550 0.8227 0.9968
0.094 ± 0.038 0.104 ± 0.042 0.172 ± 0.107 0.535 ± 0.213
1.7 4.7 8.9 20.3
0.8868 0.9770 0.9153 0.9706 0.9979
0.178 ± 0.042 0.227 ± 0.048 0.287 ± 0.164 0.671 ± 0.484
1.7 4.7 8.9 20.3
Ea (kJ mol−1 ) Off-odour
0.081 ± 0.026 0.118 ± 0.018 0.185 ± 0.078 0.528 ± 0.280
1.7 4.7 8.9 20.3
Ea (kJ mol−1 )
0.9168
0.8284 0.8615 0.8331 0.9832 0.9888
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
209
Table 3 Rate constant (kq ) from sensory attributes of fresh-cut Romaine lettuce for the first order model and the respective activation energies (Ea ) from the Arrhenius equation Attribute
Temperature (◦ C)
Rate constanta , kq (day−1 )
Coefficient of determination (R2 )
General appearance
1.4 4.3 8.9 20.3
0.075 ± 0.015 0.123 ± 0.058 0.132 ± 0.047 0.579 ± 0.183
0.9370 0.7513 0.9445 0.9892
0.053 ± 0.021 0.089 ± 0.056 0.105 ± 0.052 0.367 ± 0.155
0.7874 0.6277 0.7626 0.8888
0.125 ± 0.016 0.140 ± 0.070 0.198 ± 0.072 0.373 ± 0.222
0.9714 0.7286 0.9140 0.7671
0.078 ± 0.030 0.131 ± 0.040 0.147 ± 0.075 0.373 ± 0.204
0.7920 0.8795 0.7522 0.8109
Ea (kJ mol−1 ) Wilting
69.6 1.4 4.3 8.9 20.3
Ea (kJ mol−1 ) Browning
65.1 1.4 4.3 8.9 20.3
Ea (kJ mol−1 ) Off-odour
0.9744
39.6 1.4 4.3 8.9 20.3
Ea (kJ mol−1 ) a
0.9585
51.4
0.9962
0.9632
kq ± confidence interval at 95%.
more than it would reduce browning development on both types of packaged fresh-cut lettuce. With respect to the losses of quality due to the development of off-odour, the Iceberg lettuce is more sensitive to temperature than Romaine lettuce. In the case of chicory, wilting is the most temperature sensitive attribute (Ea = 92.9 kJ mol−1 ) and Q10 = 4.0), followed by off-odour, browning, and general appearance. As it was previously mentioned, the shelf life of a fresh-cut vegetable was defined as the time of storage at which any one of the sensory attributes scored 7.5. Once the estimates of the quality rate constants and their temperature dependency were obtained, the time to reach the cut-off score could be predicted for any temperature within the experimental range (Fig. 5). When several reactions with different Ea are important to food quality, it is possible that each of them will predominantly define quality for a different temperature range (Taoukis et al., 1997). For the three fresh-cut vegetables assayed, the limiting quality fac-
tor at any temperature was the visual impact evaluated by the panellists as general appearance. However, it should be noted that any of the attributes could be considered the failure attribute between 15 and 20 ◦ C for chicory. Finally, curves from the general appearance models could be used for shelf-life prediction of these freshcut vegetables as a function of temperature (Fig. 5). It can be seen that fresh-cut chicory has the longest shelf life at low temperature. 3.3. Validation of the fitted models Three challenge tests were carried out at constant temperature for each fresh-cut vegetable to validate the shelf-life prediction models from the kinetics of general appearance loss. Adequate agreement between predicted values and experimental results was found, validating the models (Table 6). The possible utilization of these quality loss models when the fresh-cut vegetables are stored under dynamic
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
210
Table 4 Rate constant (kq ) from sensory attributes of fresh-cut chicory for the first order model and the respective activation energies (Ea ) from the Arrhenius equation Attribute
Temperature (◦ C)
Rate constanta , kq (day−1 )
Coefficient of determination (R2 )
General appearance
1.6 4.5 8.9 20.3
0.017 ± 0.009 0.037 ± 0.019 0.156 ± 0.060 0.219 ± 0.103
0.4198 0.7879 0.9573 0.9133
0.004 ± 0.019 0.037 ± 0.025 0.180 ± 0.078 0.405 ± 0.213
0.0150 0.3477 0.9119 0.893
0.036 ± 0.026 0.068 ± 0.045 0.261 ± 0.078 0.422 ± 0.220
0.3212 0.6917 0.9555 0.8948
0.080 ± 0.068 0.078 ± 0.055 0.313 ± 0.127 0.640 ± 0.313
0.5768 0.6691 0.9216 0.9066
Ea (kJ mol−1 )
65.7
Wilting
1.6 4.5 8.9 20.3
Ea (kJ mol−1 )
92.9
Browning
1.6 4.5 8.9 20.3
Ea (kJ mol−1 )
0.8298
69.0
Off-odour
1.6 4.5 8.9 20.3
Ea (kJ mol−1 ) a
0.6990
0.7650
81.2
0.8337
kq ± confidence interval at 95%.
temperature conditions was also evaluated. Successive temperature changes are usual during commercial transporting and retailing even though a cold distribution chain is intended. Two additional experimental assays with fresh-cut lettuce were done. Panellists evaluated quality losses at days 3 and 6. Simultaneously, the models were used to predict the attributes of the samples based on the time–temperature storage profiles experimentally measured (Table 7). The predicted values for browning, wilting, and off-odour showed adequate agreement with the experimental results. In the case of general appearance, the predicted values
were lower than the experimental ones; therefore, it could be assumed that the model slightly overestimates the general appearance losses when dynamic temperature conditions are used. Table 6 Predicted and experimental shelf life of fresh-cut vegetables Type of vegetable
Temperature (◦ C)
Sensory shelf life (days) Predicted
Experimental (range)
Iceberg lettuce
1.7 4.3 8.9 20.3
6.4 4.8 2.9 0.9
8 4–6 2–3 0.5–1
Romaine lettuce
1.4 4.3 8.9 20.3
7.2 5.2 3.2 1.0
8–9 2–5 2–5 1–2
4.5 8.9 20.3
9.2 5.9 2.0
9–10 4–6 2–3
Table 5 Q10 values for each attribute and fresh-cut leafy vegetables over the temperature range 1–20 ◦ C Attribute
Iceberg lettuce
Romaine lettuce
Chicory
Chicory
General appearance Wilting Browning Off-odour
2.9 2.7 2.0 2.6
2.9 2.7 1.8 2.2
2.7 4.0 2.8 3.4
Initial general appearance quality (Q0 ) for Iceberg lettuce: 13.3, for Romaine lettuce: 12.9 and for chicory: 12.5.
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
211
Table 7 Experimental and predicted attributes values for fresh-cut Iceberg lettuce under dynamic temperature conditions Temperature profile
Attribute
Day 3
Day 6
Experimentala
Predicted
Experimentala
Predicted
A
General appearance Browning Wilting Off-odour
11.6 2.5 2.0 2.8
± ± ± ±
0.9 0.8 0.9 0.2
9.1 3.0 2.5 3.9
8.5 ± 1.5 5.5 ± 1.1 3.7 ± 0.7 4.2 ± 0.5
6.1 6.0 3.6 5.6
B
General appearance Browning Wilting Off-odour
8.1 5.7 2.7 3.5
± ± ± ±
1.8 1.7 1.1 1.9
7.2 3.9 3.0 4.6
6.0 ± 1.0 8.4 ± 1.9 4.1 ± 1.3 5.6 ± 1.9
4.8 7.8 4.4 6.9
A: 2 h at 21.4 ◦ C + 46 h at 2.3 ◦ C + 94 h at 5.5 ◦ C; B: 2 h at 22.0 ◦ C + 46 h at 9.7 ◦ C + 94 h at 5.8 ◦ C. a Mean value ± S.D.
Fig. 6. Experimental shelf life of mixed vegetables and predicted shelf life of individual fresh-cut vegetables at different temperatures. Bars indicate S.D.
Another possible application of these models is for the prediction of shelf life of fresh-cut mixed vegetables from the individual vegetable quality change models. The superposition of individual shelf-life curves (Fig. 6) for the three products showed that Iceberg and Romaine lettuce would be the first to fail in general appearance during storage. The experimental results with mixed vegetables validated this conclusion showing that these two models provide an adequate technique for evaluating the shelf life of the mixed products.
4. Conclusions
Fig. 5. Influence of temperature on the predicted time to reach the limit of acceptability (7.5) for all sensory attributes.
The present work allows prediction of sensory changes of selected fresh-cut leafy vegetables stored at different temperatures by means of their kinetic
212
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
constants and their temperature relationship. Taking into account that visual attributes and off-odour are important components of fresh-cut leafy vegetable quality and that sensory testing is expensive and time consuming, these mathematical models will be a useful tool to predict quality loss or shelf life under a broad range of storage temperatures (2–20 ◦ C, approximately) often observed in the distribution chain of these product. However, shelf life prediction under dynamic temperature conditions should be taken as a first estimate of the likely behaviour of the product.
Acknowledgements This study was partly supported by CAI + D of Universidad Nacional del Litoral (Santa Fe, Argentina). The authors thank Mr. Miguel Ranieri and Mrs. Silvina Lassa (Monte Vera, Santa Fe, Argentina) for providing the raw vegetables used in the experiments.
References Artes, F., Martinez, J.A., Marin, J.G., 1999. Quality changes in minimally processed Romaine lettuce as affected by several treatments. In: Hagg, M., Ahvenainen, R., Evers, A.M., Tiilikkala, K. (Eds.), Agri-Food Quality II. Quality Management of Fruits and Vegetables. Finland, pp. 115–118, Special Publication No. 229. Barriga, M.I., Tracy, G., Willemot, C., Simard, R.E., 1991. Microbial changes in shredded iceberg lettuce stored under controlled atmosphere. J. Food Sci. 56, 1586–1588. Bolin, H.R., Stafford, A.E., King Jr., A.D., Huxsoll, C.C., 1977. Factors affecting the storage stability of shredded lettuce. J. Food Sci. 42, 1319–1321. Cantwell, M., 1992. Postharvest handling systems: minimally processed fruits and vegetables. In: Kader, A.A. (Ed.), Postharvest Technology of Horticultural Crops. Division of Agriculture and Natural Resources, Cooperative Extension University of California, Bekerly, CA, pp. 277–281, Special Publication 3311.
Carlin, F., Nguyen-The, C., Varoquaux, P., 1990. La conservation des produits de la 4eme gamme. Ind. Alimentaires Agricoles 107, 931–943. Francis, G.A., Thomas, C., O’Beirne, D., 1999. The microbiological safety of minimally processed vegetables. Int. J. Food Sci. Technol. 34, 1–22. Gorny, J.R., 1996. Fresh-cut product preparation. Fresh-cut products: maintaining quality and safety. In: Postharvest Horticulture Series No. 10. Postharvest Outreach Program. Department of Pomology, University of California Davis (Ed), CA, pp. 7.2–7.7. Heimdal, H., Falk K¨uhn, B., Poll, L., Larsen, L.M., 1995. Biochemical changes and sensory quality of shredded and MA-packaged iceberg lettuce. J. Food Sci. 60, 1265–1268/1276. IFT, 1990. Quality of fruits and vegetables. A scientific status summary of IFT expert panel on food safety and nutrition. Food Technol. 44, 99–106. Jacxsens, L., Devlieghere, F., Debevere, J., 2002. Temperature dependence of shelf-life as affected by microbial proliferation an sensory quality of equilibrium modified atmosphere packaged fresh produce. Postharvest Biol. Technol. 26, 59–73. King, A.D., Bolin, H.R., 1989. Physiological and microbiological storage stability of minimally processed fruits and vegetables. Food Technol. 43, 132–135/139. Piagentini, A.M., Pirovani, M.E., G¨uemes, D.R., 2004. Cin´etica de deterioro de la calidad de repollo fresco cortado. Cien. Tecnol. Alimentaria 4, 169–176. Piagentini, A.M., Pirovani, M.E., G¨uemes, D.R., Di Pentima, J.H., Tessi, M.A., 1997. Survival and growth of Salmonella hadar on minimally processed cabbage as influenced by storage abuse conditions. J. Food Sci. 62, 616–619/631. Pirovani, M.E., Piagentini, A.M., G¨uemes, D.R., Di Pentima, J.H., 1998. Quality of minimally processed lettuce as influenced by packaging and chemical treatment. J. Food Qual. 21, 475– 484. Saguy, I., Karel, M., 1980. Modeling of quality deterioration during food processing and storage. Food Technol. 34, 78–85. Singh, R., Heldman, D. (Eds.), 1993. Introduction to Food Engineering, second ed. Academic Press, San Diego, USA. Taoukis, P.S., Labuza, T.P., Saguy, I., 1997. Kinetics of food deterioration and shelf life prediction. In: Valentas, K.J., Rotstein, E., Singh, R.P. (Eds.), The Handbook of Food Engineering Practice. CRC Press, NY, USA, pp. 363–405 (Chapter 9). Vankerschaver, K., Willocx, F., Smout, C., Hendrickx, M., Tobback, P., 1996. Modeling and prediction of visual shelf life of minimally processed endive. J. Food Sci. 61, 1094–1098.
Modeling changes of sensory attributes for individual and mixed fresh-cut leafy vegetables Andrea M. Piagentini, Julio C. Mendez, Daniel R. Guemes, Mar´ıa E. Pirovani ∗ Instituto de Tecnolog´ıa de Alimentos, Facultad de Ingenier´ıa Qu´ımica, Universidad Nacional del Litoral, C.C. 266-3000 Santa Fe, Argentina Received 10 August 2004; accepted 2 July 2005
Abstract Changes of the main sensory attributes of three fresh-cut leafy vegetables (Iceberg and Romaine lettuce and chicory) were investigated at selected temperatures (2–20 ◦ C). The aim of this work was to develop and apply a kinetic approach to model sensory quality changes in order to establish the appropriate function that describes the time–temperature dependence of each attribute. The changes of the sensory characteristics followed first order reaction kinetics and the temperature dependence of rate constants followed the Arrhenius relationship. The limiting quality factor, which determined the sensory shelf life for the three fresh-cut vegetables assayed, at any temperature, was the general appearance of the products. The activation energies obtained for general appearance were 71.1 kJ mol−1 for fresh-cut Iceberg lettuce, 69.5 kJ mol−1 for fresh-cut Romaine lettuce and 65.7 kJ mol−1 for fresh-cut chicory. Additional experimental tests showed that the predicted and experimental sensory shelf life for individual fresh-cut vegetables at constant temperature were not different (P > 0.05). Under dynamic temperature conditions (sequence of different temperatures), the predicted and experimental values of browning, wilting, and off-odour were also not different (P > 0.05), but the general appearance loss model overestimated the quality loss from 10 to 30%. The models of quality change for individual vegetables were used to predict the sensory shelf life of fresh-cut mixed vegetables. The experimental validation tests proved that these models provide a good approach to evaluate the shelf life of the mixed product. Results showed that the general appearance of fresh-cut Iceberg and Romaine lettuce dominated the sensory perception of mixed product. © 2005 Elsevier B.V. All rights reserved. Keywords: Sensory quality; Shelf life; Iceberg lettuce; Romaine lettuce; Chicory
1. Introduction Leafy vegetables are a well-recognized source of minerals, vitamins, and dietary fibre. The desire of ∗
Corresponding author. E-mail address: [email protected] (M.E. Pirovani).
0925-5214/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.postharvbio.2005.07.001
consumers for fresh-cut vegetables due to their convenience and fresh-like properties (texture, flavour, and appearance) has led to a relatively new area of food preservation named minimally or lightly processed (King and Bolin, 1989; Cantwell, 1992), IV Gamme (Carlin et al., 1990), ready-to-use (Francis et al., 1999), or fresh-cut fruits and vegetables (Gorny,
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
1996). These types of products do not need additional preparation at home and have been slightly modified in their fresh characteristics. In Argentina, their demand has increased markedly in the last years and industry has expectations of further growth. Although it is a relatively new market it represents 20% of the total of vegetable sales in supermarkets. Therefore, this kind of product appears destined to become an important component of the food industry but efforts must be taken to ensure products of high quality. Fresh-cut leafy vegetables have a short shelf life and are exposed to conditions that can destroy their superior quality especially during transport and retailing. Temperature conditions primarily determine the rate of quality degradation and the shelf life of the product causing changes in sensory characteristics, which can influence consumer acceptability. Appearance is the most important attribute evaluated by the consumer in the decision to accept a packaged leafy vegetable (IFT, 1990). Subsequently, when the package is open, aroma or the off-odours, if present, would be also important attributes. Several studies on quality changes of different freshcut vegetables for specific storage conditions have been reported (Bolin et al., 1977; Barriga et al., 1991; Heimdal et al., 1995; Pirovani et al., 1998; Artes et al., 1999) but little is available on sensory quality modeling to predict quality degradation as a function of temperature and time, simultaneously (Vankerschaver et al., 1996; Piagentini et al., 2004). The aim of this work was to develop and apply a kinetic approach to model sensory quality changes in order to establish the appropriate function that describes the time–temperature dependence of each attribute for three fresh-cut leafy vegetables. The kinetic parameters, namely rate constants and activation energies, provide useful information on the quality changes that could occur during commercial handling. By establishing the appropriate quality function, a quantitative tool for shelf life estimation during different conditions of transport and retailing management can be obtained.
2. Materials and methods 2.1. Preparation of fresh-cut vegetables Iceberg lettuce (Latuca sativa L. var. capitata L.), Romaine lettuce (L. sativa L. var. longifolia Lam.),
203
and chicory (Cichorium intybus L.) were obtained from a local farm near Santa Fe (Argentina). Upon arrival, the vegetables were stored at 4 ◦ C and 90% RH, and processed separately 1 day after harvest. Outer, damaged and yellowed leaves, roots, and stems were removed. The remaining leaves were cut in shreds of 5 mm width with a sharp stainless steel knife. The cut vegetable was washed for 4 min in cold water (4–6 ◦ C) containing 100 mg/L available chlorine as sodium hypochlorite and pH 6.8, with a water-toproduce ratio of 18 L/kg. Then, it was rinsed with running tap water (0.2 mg/L total chlorine) for 4 min. The vegetable was centrifuged in a basket-type centrifuge at 540 min−1 for 4 min. Each fresh-cut vegetable sample (70 g for Iceberg and Romaine lettuce, 50 g for chicory) was placed in a semi-rigid polyethylene tray (130 mm × 100 mm × 40 mm) overwrapped with a plasticized PVC film 13 m thick, with a surface area of 0.013 m2 . Gas transmission rates of PVC film for O2 , CO2 , and water vapour were 8.71 × 10−11 and 7.96 × 10−10 mol m−2 s−1 Pa−1 (both at 23 ◦ C and 1.01 × 105 Pa) and 2.96 × 10−4 mol m−2 s−1 (at 37 ◦ C and 90% RH), respectively. Finally, samples were stored at 90% RH and: 1.7, 4.7, 8.9, or 20.3 ◦ C for Iceberg lettuce; 1.4, 4.3, 8.9, or 20.3 ◦ C for Romaine lettuce; 1.6, 4.5, 8.9, or 20.3 ◦ C for chicory. The experiment was repeated three times for each vegetable. 2.2. Sensory analysis The evaluation of the characteristic sensorial properties of the samples was performed by a descriptive test. A trained sensory panel of 6–8 judges, which had previously participated in evaluating fresh vegetable qualities, was required to evaluate sensory quality attributes of fresh-cut vegetable samples. The methodology to select and train panellists was the same described by Pirovani et al. (1998). During the specific training (5 × 30 min sessions), the panellists discussed and agreed on sensory attributes and anchored terms. They evaluated off-odour development, general appearance, wilting, and browning of fresh-cut vegetable each testing day. The judges indicated their perception of each quality attribute intensity on a 150 mm unstructured line, with anchored terms located 15 mm from either end. They scored the perceived intensity of each attribute by placing a vertical line across the unstructured scale
204
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
line. Quantitation was accomplished by measuring the distance from the left end (0.00) to the vertical line, reporting measurements in centimeters. The anchored terms were indicated from left to right as: none and severe for off-odour; very poor and excellent for general appearance; none and very severe for wilting and browning. The panellists were instructed to open the sealed samples and evaluate the off-odour first, and then the other attributes. Each panellist performed the sensory test individually in separate booths with white incandescent lighting (sufficient to provide 700 lx). Samples were coded with three-digit random numbers. The central point of the scale was established as the cut-off score. Therefore, the fresh-cut vegetable was considered as unacceptable when a mean score below 7.5 was reached for general appearance or above 7.5 for the other sensory attributes (Barriga et al., 1991; Piagentini et al., 1997, 2004; Jacxsens et al., 2002).
mination of the reaction order. Then, the effect of temperature on quality attribute changes is modeled. The influence of a constant storage temperature on the reaction rate constant can be described using the Arrhenius equation (Saguy and Karel, 1980; Taoukis et al., 1997): −Ea (2) kq(T ) = k0 exp RT where kq(T) is rate constant for each quality atribute; k0 the pre-exponential factor; Ea the activation energy (J mol−1 ); R the universal gas constant (8.3145 J K−1 mol−1 ); T is the absolute temperature (K). The Ea for each quality attribute is obtained by regression analysis. Another parameter that is often used to describe the relationship between temperature and reaction rate constant is the Q10 value. Q10 is defined as follows: Q10 =
2.3. Weight loss Weight loss was determined on all samples at each testing day and was expressed as a percentage of initial fresh weight. 2.4. Mathematical modeling of quality changes The rate of quality changes of foods can be described by the following general equation: ±
dQ = kq [Q]n dt
(1)
where Q is quality attribute; t the time; n the reaction order; kq is the quality change rate constant for the attribute Q. The sign (+) refers to attributes with increasing values during time (browning, wilting, and off-odour) and the sign (−) to attributes with decreasing values (general appearance). Traditionally, quality change processes of foods stored under controlled environmental conditions are described with zero order and/or first order rate functions (Saguy and Karel, 1980; Taoukis et al., 1997). The regression analysis is used to determine the kinetic order of the quality change rate. An analysis of the coefficient of determination (R2 ), residuals (ei , defined as the difference between observed and fitted data points) and estimated value of the intercept would enable deter-
reaction rate at temperature (T + 10) reaction rate at temperature (T )
(3)
Equivalently, Q10 has been defined as the change of shelf life (ts ), i.e., the time for the failure attribute to reach an unacceptable level when the food is stored at a 10 ◦ C higher temperature. It can be shown that Q10 and activation energy, Ea , are related by the following expression (Singh and Heldman, 1993): Ea 10 ln Q10 = (4) R T (T + 10) The time to reach a sensory characteristic limit of acceptability at a specified storage temperature (average score below 7.5 for general appearance or above 7.5 for browning, wilting, and off-odour) was derived from the prediction models (Eqs. (1) and (2)). For each fresh-cut vegetable, the limiting sensory characteristic (failure attribute) was obtained comparing the time to reach the limit of acceptability of each attribute and temperature. 2.5. Validation of the models Three confirmatory experiments were carried out to validate the fitted models at constant temperature. Samples of each fresh-cut vegetable were prepared in the same way as previously described and stored at different constant temperatures. The sensory shelf life of each sample was determined by evaluating the failure
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
205
attribute. Experimental data were compared to values predicted from the models by a t-test analysis. The combined temperature–quality change models, obtained under constant conditions, were tested to verify their adequacy under a variable temperature sequence. Therefore, these models were used to predict the evolution of sensory attributes of fresh-cut Iceberg lettuce stored at two different temperature profiles. Profile A: 2 h at 21.4 ◦ C; 46 h at 2.3 ◦ C, and finally, 94 h at 5.5 ◦ C. Profile B: 2 h at 22 ◦ C; 46 h at 9.7 ◦ C, and finally 94 h at 5.8 ◦ C. Experiments were repeated twice for each temperature profile. The observed and predicted failure attribute values of fresh-cut Iceberg lettuce were compared. The models to predict changes of general appearance obtained for each vegetable were also tested on mixed vegetable samples. Fresh-cut vegetables were prepared in the way previously described. The mixed vegetable sample was prepared by putting 20 g of each vegetable (Iceberg and Romaine lettuce, and chicory) together in the same tray, wrapping with PVC film and storing at 1.3, 4.3, 8.1, or 20.4 ◦ C. Data from three confirmatory experiments were compared to values predicted by model equations obtained for individual fresh-cut vegetable. 2.6. Statistical analysis All data were analysed using STATGRAPHICS Plus (Manugistics, Inc., Rockville, MD, USA). The average and standard deviation of weight loss values and individual scores of the sensory attribute given by the trained panel were calculated. The data were fitted to the corresponding models and the regression analyses were carried out. The 95% confidence intervals of estimated parameters were calculated. The experimental data obtained during the confirmatory experiments were compared to values predicted by model equation by a t-test analysis.
Fig. 1. Experimental and predicted values of sensory attributes for fresh-cut Iceberg lettuce throughout storage at different temperatures. Bars indicate S.D.
3. Results and discussion 3.1. Sensory characteristics of fresh-cut Iceberg and Romaine lettuces, and chicory The mean scores of sensory quality attributes (general appearance, browning, wilting, and off-odour) for
the three packaged fresh-cut leafy vegetables throughout storage at different temperatures are presented in Figs. 1–3. They showed that browning, wilting, and off-odour scores did not change as rapidly as general appearance scores. Panellists scored general appearance as the overall visual impact of the samples. During
206
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
Fig. 2. Experimental and predicted values of sensory attributes for fresh-cut Romaine lettuce throughout storage at different temperatures. Bars indicate S.D.
Fig. 3. Experimental and predicted values of sensory attributes for fresh-cut chicory throughout storage at different temperatures. Bars indicate S.D.
the process of language development, they agreed that wilting, browning, and all other unexpected visual quality factors would contribute to the general appearance score. At the end of experiments, some panellists indicated that colour changes or decay were included in general appearance evaluation.
As storage time and temperature increased, panellists gave lower scores for general appearance and higher ones for browning, wilting, and off-odour for all the samples. General appearance and browning changed the most during storage for both types of lettuce and for chicory
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
207
chicory. This is in accordance with sensory evaluation of wilting (score around four for both lettuce and six for chicory). 3.2. Modeling sensory characteristics of fresh-cut leafy vegetables
Fig. 4. Weight losses of fresh-cut vegetables during storage at different temperatures.
at all temperatures assayed. Jacxsens et al. (2002) found that general appearance (freshness) and the colour were the properties of packaged mixed lettuce that presented the largest changes. In the present work, wilting, and off-odour were important defects only in samples stored at higher temperatures (8.9 and 20.3 ◦ C). Weight loss of fresh-cut Iceberg and Romaine lettuce and chicory was less than 1.7% after 6 days of storage for temperatures lower than 4.7 ◦ C (Fig. 4). When storage temperature was 8.9 ◦ C, the weight loss after 6 days of storage was approximately 2–2.5% for both types of lettuce but significantly higher (3.7%) for
The first step in selecting an appropriate model to represent the quality changes was to perform regression analysis to determine the kinetic order. The performance of the fitted models was analysed. Based on the best coefficient of determination (R2 ), first order was the apparent order of the quality change reactions in the majority of the cases (Table 1). Further analysis determined the estimated value of the intercept of the first order model. Finally, the plots of residuals versus predicted values (not shown) for each model (zero and first order) indicated that the distribution around zero was more random for the first order model. Therefore, all these results were taken into account to select the first order reaction model to describe the quality changes of these leafy vegetables. Other researchers (Vankerschaver et al., 1996; Piagentini et al., 2004) also found that first order was the most adequate kinetic model for fitting sensory characteristic changes in minimally processed vegetables. Curves in Figs. 1–3 show the evolution of each predicted attribute using the fitted models. Tables 2–4 show the rate constants for each attribute and temperature and the activation energies for packaged fresh-cut Iceberg and Romaine lettuce, and chicory, respectively. Results indicated that for all the attributes the dependence of rate constants on temperature followed the Arrhenius relationship (R2 values ranged between 0.700 and 0.998). The Q10 values for each attribute and fresh-cut leafy vegetable are presented in Table 5. The activation energies of general appearance (71.1 and 69.6 kJ mol−1 ) and wilting (66.9 and 65.1 kJ mol−1 ) for fresh-cut Iceberg and Romaine lettuce, respectively, showed the highest values compared to other sensory characteristics indicating that these attributes are the most influenced by temperature. On the other side, lower activation energies of browning signify smaller temperature sensitivity for this attribute (Ea = 47.5 and 39.6 kJ mol−1 for Iceberg and Romaine lettuce, respectively). Therefore, a reduction in temperature would benefit general appearance and wilting
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
208
Table 1 Coefficient of determination (R2 ) for zero and first order models for sensory attribute of fresh-cut Romaine and Iceberg lettuces and chicory Attribute
Temperature (◦ C)a
Iceberg lettuce
Romaine lettuce
Chicory
Zero
First
Zero
First
Zero
First
General appearance
A B C D
0.8500 0.9849 0.9341 0.9165
0.9239 0.9937 0.9731 0.9974
0.9595 0.7799 0.8101 0.8597
0.9370 0.7513 0.9445 0.9892
0.4316 0.8178 0.9466 0.9164
0.4198 0.7879 0.7812 0.9133
Wilting
A B C D
0.7814 0.9687 0.9271 0.9843
0.8868 0.977 0.9153 0.9706
0.7821 0.5614 0.7081 0.9674
0.7874 0.6277 0.7626 0.8888
0.0070 0.3351 0.8923 0.8228
0.0105 0.3477 0.9119 0.893
Browning
A B C D
0.7902 0.9853 0.9408 0.9532
0.9349 0.9570 0.8550 0.8227
0.9313 0.7017 0.8154 0.9304
0.9714 0.7286 0.9140 0.7671
0.2810 0.6551 0.9169 0.9016
0.3212 0.6917 0.9555 0.8948
Off-odour
A B C D
0.6264 0.8539 0.8113 0.9649
0.8284 0.8615 0.8331 0.9832
0.7715 0.8069 0.652 0.8899
0.7920 0.8795 0.7522 0.8109
0.5152 0.5477 0.8930 0.9078
0.5768 0.6691 0.9216 0.9066
Fresh-cut Iceberg lettuce: A = 1.7 ◦ C, B = 4.7 ◦ C, C = 8.9 ◦ C, D = 20.3 ◦ C; fresh-cut Romaine lettuce: A = 1.4 ◦ C, B = 4.3 ◦ C, C = 8.9 ◦ C, D = 20.3 ◦ C; fresh-cut chicory: A = 1.6 ◦ C, B = 4.5 ◦ C, C = 8.9 ◦ C, D = 20.3 ◦ C. a
Table 2 Rate constant (kq ) from sensory attributes of fresh-cut Iceberg lettuce for the first order model and the respective activation energies (Ea ) from the Arrhenius equation Attribute
Temperature (◦ C)
Rate constanta , kq (day−1 )
Coefficient of determination (R2 )
General appearance
1.7 4.7 8.9 20.3
0.065 ± 0.019 0.161 ± 0.013 0.235 ± 0.054 0.576 ± 0.090
0.9239 0.9937 0.9731 0.9974
Ea (kJ mol−1 ) Wilting
71.1
Browning
66.9
47.5
Ea (kJ mol−1 ) a
kq ± confidence interval at 95%.
65.3
0.9349 0.9570 0.8550 0.8227 0.9968
0.094 ± 0.038 0.104 ± 0.042 0.172 ± 0.107 0.535 ± 0.213
1.7 4.7 8.9 20.3
0.8868 0.9770 0.9153 0.9706 0.9979
0.178 ± 0.042 0.227 ± 0.048 0.287 ± 0.164 0.671 ± 0.484
1.7 4.7 8.9 20.3
Ea (kJ mol−1 ) Off-odour
0.081 ± 0.026 0.118 ± 0.018 0.185 ± 0.078 0.528 ± 0.280
1.7 4.7 8.9 20.3
Ea (kJ mol−1 )
0.9168
0.8284 0.8615 0.8331 0.9832 0.9888
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
209
Table 3 Rate constant (kq ) from sensory attributes of fresh-cut Romaine lettuce for the first order model and the respective activation energies (Ea ) from the Arrhenius equation Attribute
Temperature (◦ C)
Rate constanta , kq (day−1 )
Coefficient of determination (R2 )
General appearance
1.4 4.3 8.9 20.3
0.075 ± 0.015 0.123 ± 0.058 0.132 ± 0.047 0.579 ± 0.183
0.9370 0.7513 0.9445 0.9892
0.053 ± 0.021 0.089 ± 0.056 0.105 ± 0.052 0.367 ± 0.155
0.7874 0.6277 0.7626 0.8888
0.125 ± 0.016 0.140 ± 0.070 0.198 ± 0.072 0.373 ± 0.222
0.9714 0.7286 0.9140 0.7671
0.078 ± 0.030 0.131 ± 0.040 0.147 ± 0.075 0.373 ± 0.204
0.7920 0.8795 0.7522 0.8109
Ea (kJ mol−1 ) Wilting
69.6 1.4 4.3 8.9 20.3
Ea (kJ mol−1 ) Browning
65.1 1.4 4.3 8.9 20.3
Ea (kJ mol−1 ) Off-odour
0.9744
39.6 1.4 4.3 8.9 20.3
Ea (kJ mol−1 ) a
0.9585
51.4
0.9962
0.9632
kq ± confidence interval at 95%.
more than it would reduce browning development on both types of packaged fresh-cut lettuce. With respect to the losses of quality due to the development of off-odour, the Iceberg lettuce is more sensitive to temperature than Romaine lettuce. In the case of chicory, wilting is the most temperature sensitive attribute (Ea = 92.9 kJ mol−1 ) and Q10 = 4.0), followed by off-odour, browning, and general appearance. As it was previously mentioned, the shelf life of a fresh-cut vegetable was defined as the time of storage at which any one of the sensory attributes scored 7.5. Once the estimates of the quality rate constants and their temperature dependency were obtained, the time to reach the cut-off score could be predicted for any temperature within the experimental range (Fig. 5). When several reactions with different Ea are important to food quality, it is possible that each of them will predominantly define quality for a different temperature range (Taoukis et al., 1997). For the three fresh-cut vegetables assayed, the limiting quality fac-
tor at any temperature was the visual impact evaluated by the panellists as general appearance. However, it should be noted that any of the attributes could be considered the failure attribute between 15 and 20 ◦ C for chicory. Finally, curves from the general appearance models could be used for shelf-life prediction of these freshcut vegetables as a function of temperature (Fig. 5). It can be seen that fresh-cut chicory has the longest shelf life at low temperature. 3.3. Validation of the fitted models Three challenge tests were carried out at constant temperature for each fresh-cut vegetable to validate the shelf-life prediction models from the kinetics of general appearance loss. Adequate agreement between predicted values and experimental results was found, validating the models (Table 6). The possible utilization of these quality loss models when the fresh-cut vegetables are stored under dynamic
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
210
Table 4 Rate constant (kq ) from sensory attributes of fresh-cut chicory for the first order model and the respective activation energies (Ea ) from the Arrhenius equation Attribute
Temperature (◦ C)
Rate constanta , kq (day−1 )
Coefficient of determination (R2 )
General appearance
1.6 4.5 8.9 20.3
0.017 ± 0.009 0.037 ± 0.019 0.156 ± 0.060 0.219 ± 0.103
0.4198 0.7879 0.9573 0.9133
0.004 ± 0.019 0.037 ± 0.025 0.180 ± 0.078 0.405 ± 0.213
0.0150 0.3477 0.9119 0.893
0.036 ± 0.026 0.068 ± 0.045 0.261 ± 0.078 0.422 ± 0.220
0.3212 0.6917 0.9555 0.8948
0.080 ± 0.068 0.078 ± 0.055 0.313 ± 0.127 0.640 ± 0.313
0.5768 0.6691 0.9216 0.9066
Ea (kJ mol−1 )
65.7
Wilting
1.6 4.5 8.9 20.3
Ea (kJ mol−1 )
92.9
Browning
1.6 4.5 8.9 20.3
Ea (kJ mol−1 )
0.8298
69.0
Off-odour
1.6 4.5 8.9 20.3
Ea (kJ mol−1 ) a
0.6990
0.7650
81.2
0.8337
kq ± confidence interval at 95%.
temperature conditions was also evaluated. Successive temperature changes are usual during commercial transporting and retailing even though a cold distribution chain is intended. Two additional experimental assays with fresh-cut lettuce were done. Panellists evaluated quality losses at days 3 and 6. Simultaneously, the models were used to predict the attributes of the samples based on the time–temperature storage profiles experimentally measured (Table 7). The predicted values for browning, wilting, and off-odour showed adequate agreement with the experimental results. In the case of general appearance, the predicted values
were lower than the experimental ones; therefore, it could be assumed that the model slightly overestimates the general appearance losses when dynamic temperature conditions are used. Table 6 Predicted and experimental shelf life of fresh-cut vegetables Type of vegetable
Temperature (◦ C)
Sensory shelf life (days) Predicted
Experimental (range)
Iceberg lettuce
1.7 4.3 8.9 20.3
6.4 4.8 2.9 0.9
8 4–6 2–3 0.5–1
Romaine lettuce
1.4 4.3 8.9 20.3
7.2 5.2 3.2 1.0
8–9 2–5 2–5 1–2
4.5 8.9 20.3
9.2 5.9 2.0
9–10 4–6 2–3
Table 5 Q10 values for each attribute and fresh-cut leafy vegetables over the temperature range 1–20 ◦ C Attribute
Iceberg lettuce
Romaine lettuce
Chicory
Chicory
General appearance Wilting Browning Off-odour
2.9 2.7 2.0 2.6
2.9 2.7 1.8 2.2
2.7 4.0 2.8 3.4
Initial general appearance quality (Q0 ) for Iceberg lettuce: 13.3, for Romaine lettuce: 12.9 and for chicory: 12.5.
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
211
Table 7 Experimental and predicted attributes values for fresh-cut Iceberg lettuce under dynamic temperature conditions Temperature profile
Attribute
Day 3
Day 6
Experimentala
Predicted
Experimentala
Predicted
A
General appearance Browning Wilting Off-odour
11.6 2.5 2.0 2.8
± ± ± ±
0.9 0.8 0.9 0.2
9.1 3.0 2.5 3.9
8.5 ± 1.5 5.5 ± 1.1 3.7 ± 0.7 4.2 ± 0.5
6.1 6.0 3.6 5.6
B
General appearance Browning Wilting Off-odour
8.1 5.7 2.7 3.5
± ± ± ±
1.8 1.7 1.1 1.9
7.2 3.9 3.0 4.6
6.0 ± 1.0 8.4 ± 1.9 4.1 ± 1.3 5.6 ± 1.9
4.8 7.8 4.4 6.9
A: 2 h at 21.4 ◦ C + 46 h at 2.3 ◦ C + 94 h at 5.5 ◦ C; B: 2 h at 22.0 ◦ C + 46 h at 9.7 ◦ C + 94 h at 5.8 ◦ C. a Mean value ± S.D.
Fig. 6. Experimental shelf life of mixed vegetables and predicted shelf life of individual fresh-cut vegetables at different temperatures. Bars indicate S.D.
Another possible application of these models is for the prediction of shelf life of fresh-cut mixed vegetables from the individual vegetable quality change models. The superposition of individual shelf-life curves (Fig. 6) for the three products showed that Iceberg and Romaine lettuce would be the first to fail in general appearance during storage. The experimental results with mixed vegetables validated this conclusion showing that these two models provide an adequate technique for evaluating the shelf life of the mixed products.
4. Conclusions
Fig. 5. Influence of temperature on the predicted time to reach the limit of acceptability (7.5) for all sensory attributes.
The present work allows prediction of sensory changes of selected fresh-cut leafy vegetables stored at different temperatures by means of their kinetic
212
A.M. Piagentini et al. / Postharvest Biology and Technology 38 (2005) 202–212
constants and their temperature relationship. Taking into account that visual attributes and off-odour are important components of fresh-cut leafy vegetable quality and that sensory testing is expensive and time consuming, these mathematical models will be a useful tool to predict quality loss or shelf life under a broad range of storage temperatures (2–20 ◦ C, approximately) often observed in the distribution chain of these product. However, shelf life prediction under dynamic temperature conditions should be taken as a first estimate of the likely behaviour of the product.
Acknowledgements This study was partly supported by CAI + D of Universidad Nacional del Litoral (Santa Fe, Argentina). The authors thank Mr. Miguel Ranieri and Mrs. Silvina Lassa (Monte Vera, Santa Fe, Argentina) for providing the raw vegetables used in the experiments.
References Artes, F., Martinez, J.A., Marin, J.G., 1999. Quality changes in minimally processed Romaine lettuce as affected by several treatments. In: Hagg, M., Ahvenainen, R., Evers, A.M., Tiilikkala, K. (Eds.), Agri-Food Quality II. Quality Management of Fruits and Vegetables. Finland, pp. 115–118, Special Publication No. 229. Barriga, M.I., Tracy, G., Willemot, C., Simard, R.E., 1991. Microbial changes in shredded iceberg lettuce stored under controlled atmosphere. J. Food Sci. 56, 1586–1588. Bolin, H.R., Stafford, A.E., King Jr., A.D., Huxsoll, C.C., 1977. Factors affecting the storage stability of shredded lettuce. J. Food Sci. 42, 1319–1321. Cantwell, M., 1992. Postharvest handling systems: minimally processed fruits and vegetables. In: Kader, A.A. (Ed.), Postharvest Technology of Horticultural Crops. Division of Agriculture and Natural Resources, Cooperative Extension University of California, Bekerly, CA, pp. 277–281, Special Publication 3311.
Carlin, F., Nguyen-The, C., Varoquaux, P., 1990. La conservation des produits de la 4eme gamme. Ind. Alimentaires Agricoles 107, 931–943. Francis, G.A., Thomas, C., O’Beirne, D., 1999. The microbiological safety of minimally processed vegetables. Int. J. Food Sci. Technol. 34, 1–22. Gorny, J.R., 1996. Fresh-cut product preparation. Fresh-cut products: maintaining quality and safety. In: Postharvest Horticulture Series No. 10. Postharvest Outreach Program. Department of Pomology, University of California Davis (Ed), CA, pp. 7.2–7.7. Heimdal, H., Falk K¨uhn, B., Poll, L., Larsen, L.M., 1995. Biochemical changes and sensory quality of shredded and MA-packaged iceberg lettuce. J. Food Sci. 60, 1265–1268/1276. IFT, 1990. Quality of fruits and vegetables. A scientific status summary of IFT expert panel on food safety and nutrition. Food Technol. 44, 99–106. Jacxsens, L., Devlieghere, F., Debevere, J., 2002. Temperature dependence of shelf-life as affected by microbial proliferation an sensory quality of equilibrium modified atmosphere packaged fresh produce. Postharvest Biol. Technol. 26, 59–73. King, A.D., Bolin, H.R., 1989. Physiological and microbiological storage stability of minimally processed fruits and vegetables. Food Technol. 43, 132–135/139. Piagentini, A.M., Pirovani, M.E., G¨uemes, D.R., 2004. Cin´etica de deterioro de la calidad de repollo fresco cortado. Cien. Tecnol. Alimentaria 4, 169–176. Piagentini, A.M., Pirovani, M.E., G¨uemes, D.R., Di Pentima, J.H., Tessi, M.A., 1997. Survival and growth of Salmonella hadar on minimally processed cabbage as influenced by storage abuse conditions. J. Food Sci. 62, 616–619/631. Pirovani, M.E., Piagentini, A.M., G¨uemes, D.R., Di Pentima, J.H., 1998. Quality of minimally processed lettuce as influenced by packaging and chemical treatment. J. Food Qual. 21, 475– 484. Saguy, I., Karel, M., 1980. Modeling of quality deterioration during food processing and storage. Food Technol. 34, 78–85. Singh, R., Heldman, D. (Eds.), 1993. Introduction to Food Engineering, second ed. Academic Press, San Diego, USA. Taoukis, P.S., Labuza, T.P., Saguy, I., 1997. Kinetics of food deterioration and shelf life prediction. In: Valentas, K.J., Rotstein, E., Singh, R.P. (Eds.), The Handbook of Food Engineering Practice. CRC Press, NY, USA, pp. 363–405 (Chapter 9). Vankerschaver, K., Willocx, F., Smout, C., Hendrickx, M., Tobback, P., 1996. Modeling and prediction of visual shelf life of minimally processed endive. J. Food Sci. 61, 1094–1098.