INTRODUCTION
Currently, there is a notable increase in the use of predictive models in the design of processes in the field of agronomy. These mathematical tools allow not only accurately describing the behavior of variables of interest, but also designing processes and evaluating transformations with different quality variables such as color, appearance, level of contamination, weight loss and fruit yield. With proper manipulation of the variables, these models can be compared and evaluated in practice easily and with a small margin of error. In this sense, their wide range of utility has allowed both the use of various pre-harvest systems for the control of pests and diseases and conservation treatments for the correct postharvest handling of fruits and vegetables (Díaz et al., 2019; Bhattacharjee et al., 2020; Martin et al., 2021).
The consequences of the decrease in the quality of the fruit are a risk, not only for the producers due to the obvious economic losses but also for the consumer due to the action of pathogenic microorganisms and toxin producers. Strawberry (Fragaria vesca) is one of the most important crops in Ecuador that stands out for its high nutritional value and sensory attributes, highly appreciated by consumers. However, the characteristics of the fruit cause its rapid postharvest deterioration (Idzwana et al., 2020; Pombo et al., 2011). This deterioration is due to factors of different nature (physical, chemical, and biological); however, the appearance of microorganisms is the factor that most reduces its postharvest useful life.
UV-C radiation is one of the minimum postharvest conservation treatments capable of substantially reducing the microbial load without affecting the organoleptic properties of the fruit (Idzwana et al., 2020; Urban et al., 2016). UV-C has been used successfully in the postharvest of numerous fruits and vegetables such as strawberries, pineapple, peaches, tomatoes, cucumbers, and broccoli (Formica et al., 2017; Liu et al., 2018; Yang et al., 2014). This conservation treatment produces resistance to deterioration not only due to its ability to inhibit microbial growth on the fruit surface but also due to the stimulation of metabolites that delay the senescence of the fruit at low doses of UV-C (hormonal effect) (Andrade et al., 2013; Liao et al., 2016).
UV-C has a significant influence (p <0.05) in the main causes of deterioration of both fruits and vegetables, causing a delay in senescence. However, uneven surfaces of fruits and vegetables can vary the effectiveness of UV-C. Other factors of variation are the variety of the fruit, the season of the year, and the storage temperature (Cote et al., 2013; Hakguder y Unluturk, 2018; Petrielli et al., 2019). Combined UV-C treatments with gases, chemicals, modified atmospheres and light pulses can also increase the effectiveness of the UV-C response to fruit senescence (Allende et al., 2007; Xu et al., 2019; Kim et al., 2010; Lu et al., 2018; Marquenie et al., 2002; Moreno et al., 2016).
Several studies of UV-C treatment in strawberries indicate a high sensitivity of the fruit to these treatments (Forges et al., 2018; Ortiz et al., 2018; Severo et al., 2015). However, few investigations include a mathematical model that can accurately predict the postharvest shelf life change of strawberries based on the treatment applied. The objective of the present work was to obtain a mathematical model that allows effectively evaluating the combined effect of different doses of UV-C radiation, storage time and temperature, on the postharvest shelf life of strawberries (Fragaria vesca L.) var. Big Bear.
MATERIALS AND METHODS
The fruits (Fragaria vesca) were selected without defects due to the uniformity of size, color, and degree of maturation from a systematic sampling (N = 20000; n = 380 and k = 52) according to the quality protocol of the packing plant "FRESTAR". The sample size was determined for a finite population (Montgomery, 2010). Once selected, the fruits were placed in a 20 L radiation chamber containing a UV-C mercury lamp (15 W) and treated with a variable dose of 2-7 kJ/m2 at different exposure times (from 10 to 15 min). Subsequently, they were stored in plastic containers of approximately 0.35 L capacity at different storage temperatures in the 12-20 ° C range.
Every day all the fruits of each container (20) were analyzed according to four key indicators that characterize the deterioration process in the strawberry: firmness, weight, contamination, and color. The quality of the fruit (C) was determined as the percentage change rate (1) that establishes the average values of each of the four indicators defined as the average of the sum of the percentage ratio of each indicator.
Firmness was determined, using the FHT200 brand digital fruit penetrometer (± 0.5% + 2 digits of precision), as the average of the maximum puncture force when drilling the sample on the radial and axial axes. To evaluate the weight, the digital scale EM-KBS2 was used with a precision of 0.01 g. The contamination level (CL) was determined using a subjective visual scale of five points that indicate by approximation the proportion of the contaminated area of the fruit (0: 100% of the surface without alteration, 1: CL < 10 % slight alteration, 2: 10 % ≤ CL <25 % moderate alteration, 3: 25 % ≤ CL ≤ 50 % severe alteration and 4: CL > 50 % very serious alteration). The color variation ΔE was determined using a digital photo colorimeter, using the CIE-L*a*b* system and the equation proposed by Chen & Ramaswamy (2002):
where:
ΔL, Δa, and Δb represent the deviations of the individual color values (L- luminosity, a-red/green coordinate, and b-yellow/blue coordinate) of a fully ripe strawberry sample.
As for most foods, the relations of dependence of the variables in time obey to equations of the first-order and zero-order (Fu & Labuza, 1993; Gacula Jr, 1975). To define the postharvest life of the strawberry, the change in quality (%) over time was considered as follows:
where;
k: |
is the kinetic constant |
Q: |
is the quality of the fruit at a time t and |
n: |
is the order of the reaction |
By integrating equation (3) two equations of different order are obtained (4, 5). Qo represents the quality of the fruit at the beginning when t = 0, and Q is the quality of the fruit at time t, with kc being the specific velocity constant for the zero-order reaction (n = 0) and k the specific velocity constant for the first-order reaction (n = 1). Both equations represent a straight line but first order equation (5) was considered as representing the best fit of our data (higher R2 coefficient (R2=0,87)):
To calculate the value of the specific rate constant k, expression 5 was used, obtaining the value k (slope of the graph) from the experimental data of C. The limit value of C established for calculating postharvest life of the controls and treatments was 10% of quality, which is the time t in which ten percent of the postharvest quality has been lost (determined by the variation in the indicators: firmness, weight, contamination, and color), was defined as an indicator of quality loss. The value for t (10%) was selected as it represented the time when the quality losses of the strawberry, were visually appreciable (time in which 10% of the quality of the strawberry is lost by visual appreciation). This time was obtained through expression 6.
The postharvest life (Pl) of the treated strawberries was expressed as the difference of t (10%) between controls and treatments, where tc represents the postharvest time of the strawberry when 10% of the quality of the controls has been lost, and tt represents the postharvest time of the strawberry when 10% of the quality in the treatments has been lost (5). These values were obtained experimentally from triplicate tests.
Experimental Design
The Response Surface method was used in the statistical program Design Expert v.11 that uses the least-squares technique to fit the data obtained to first (8) and second-order (9) polynomial approximation equations. For this case, the full factorial Central Compound Design (CCD) was selected, whose matrix is composed of ten factorial points, five axial points, and four replicas in central points.
The design matrix and the combination of factors, including the axial and central points with their replicas, resulted in 19 runs. The design was rotatory to guarantee consistent variance and orthogonal response (α = 1.681) (Table 1). For this purpose, the 19 experimental runs were randomly executed and, in order to respect the randomness, the execution order for the sequencing of experiments was taken exactly as indicated by the Design Expert program.
The analysis of variance (ANOVA) was used to evaluate the statistical significance of the models and the Fisher’s statistical test (F-test) determined which of the factors significantly affected postharvest life of Fragaria vesca. To this purpose the significance and the magnitude of the effects of each variable (UV-C dose, exposure time and temperature) and their possible interactions on the postharvest life were estimated. The variable effects with a p-value higher than 0.05 or with less than 95% of significance were discarded and a new analysis of variance was performed for the reduced model (Montgomery, 2010).
RESULTS AND DISCUSSION
CCD Experimental Design Analysis
The results obtained showed average values of postharvest life of the strawberry of 2.7 ± 1.12 days, indicating that the differences between treatments and controls are notable. This result also shows that UV-C treatments (in variable doses of 4-15 kJ/m2) and exposure times (15-30 min) take approximately 3.6 days to lose their quality attributes (depending on the weight, contamination, and indicators of color firmness) in relation to the controls (Table 2).
Run | A: UV-C dose (kJ/m2) | B: time (min) | C: Temperature (ºC) | Postharvest life (days) |
---|---|---|---|---|
1 | 9.5 | 35.1 | 18.0 | 3.01 |
2 | 15 | 30.0 | 24.0 | 3.54 |
3 | 15 | 30.0 | 12.0 | 4.01 |
4 | 4 | 15.0 | 24.0 | 1.27 |
5 | 9.5 | 22.5 | 18.0 | 3.54 |
6 | 9.5 | 22.5 | 18.0 | 3.44 |
7 | 18.7 | 22.5 | 18.0 | 4.54 |
8 | 15 | 15.0 | 12.0 | 3.87 |
9 | 4 | 30.0 | 12.0 | 1.82 |
10 | 4 | 30.0 | 24.0 | 1.37 |
11 | 9.5 | 22.5 | 18.0 | 3.55 |
12 | 9.5 | 22.5 | 07.9 | 3.02 |
13 | 15 | 15.0 | 24.0 | 3.34 |
14 | 0.25 | 22.5 | 18.0 | 0.46 |
15 | 4 | 15.0 | 12.0 | 1.58 |
16 | 9.5 | 22.5 | 18.0 | 3.58 |
17 | 9.5 | 22.5 | 18.0 | 3.25 |
18 | 9.5 | 22.5 | 28.0 | 1.51 |
19 | 9.5 | 09.8 | 18.0 | 2.48 |
Similar studies determined that the UV-C dose of 9 kJ/m2 and UVB at 72 h/15 °C improved the stability of broccoli flowers (Formica et al., 2017; Lu et al., 2018). The same effect of UV-C has been observed in pears, pineapples, blueberries, amaranth, spinach, leeks, onion, tomato, persimmon, and cucumber in doses that vary between 1.7 kJ/m2 to 39.6 kJ/m2, and storage temperatures of 4-24 ºC (Liu et al., 2018; Petrielli et al., 2019; Gogo et al., 2018; Imaizumi et al., 2018; Sari et al., 2016).
In strawberries, exposure to UV-C (250 nm) in doses of 0.4-15 kJ/m2 has shown substantial changes in the visible characteristics that limit their postharvest life (Jin et al., 2017) and delay contamination fungal. This effect at the same time produces an increase in the expression of genes related to the defense of the host against attacks by microorganisms (Forges et al., 2018). If the UV-C doses are increased, this effect is cyclically enhanced at different times of the postharvest (Hakguder y Unluturk, 2018; Ortiz et al., 2019).
Source | Sum of Squares | df | Mean Square | F-value | p-value | |
---|---|---|---|---|---|---|
22,57 | 6 | 3,76 | 111,73 | < 0.0001 | significant | |
A-Dose (UV-C) | 17,78 | 1 | 17,78 | 528,00 | < 0.0001 | |
B-Time exposure | 0,1808 | 1 | 0,1808 | 5,37 | 0,0390 | |
C-Temperature | 1,35 | 1 | 1,35 | 40,20 | < 0.0001 | |
A² | 1,32 | 1 | 1,32 | 39,09 | < 0.0001 | |
B² | 0,6843 | 1 | 0,6843 | 20,32 | 0,0007 | |
C² | 2,11 | 1 | 2,11 | 62,81 | < 0.0001 | |
0,4040 | 12 | 0,0337 | ||||
Lack of Fit | 0,3314 | 8 | 0,0414 | 2,28 | 0,2221 | not significant |
Pure Error | 0,0727 | 4 | 0,0182 | |||
22,98 | 18 | |||||
0,1835 | 0,9824 | |||||
2,80 | 0,9736 | |||||
6,56 | 0,9296 | |||||
34,4567 |
In the ANOVA test, the effect of each variable on the postharvest life of Fragraria vesca was obtained (Table 3). Once the terms of the quadratic model that were not significant (p> 0.05), (AB, AC, and BC) were eliminated, the terms (A, C, B, A², B², C²) significant (p <0.05) were obtained. The F value (111.73) obtained implies that the quadratic model is significant (p <0.05), its F value for lack of fit (2, 28) was not significant (p> 0.05) in relation to the pure error and there is a 22.21% probability that an out-of-fit F-value occurs due to noise (this implies a good fit of the model to the data).
The predicted R² model of 0.92 is in reasonable agreement with the adjusted R² of 0.97; (the difference is less than 0.2) indicating that in the model space the factors UV-C dose, temperature, and exposure time can explain 97% of the postharvest life variations. The precision measure (AP = 34, 45) was higher than 4, which indicates an adequate signal. The coefficient of variation (CV = 6.56) reveals the reliability of the experiments (Table 3). The predicted response values versus the actual response values (experimental values) show that at each point, the model correctly predicted the corresponding values (Figure 1a). The normal probability indicates that the residuals follow a straight line with little scatter data; therefore, they follow a normal distribution. (Figure 1b). In general, the relationship between the predictions and the experimental values and the diagnosis of the residuals were adequate (R² = 0.92), which implies that the model obtained can be used to navigate through the design space. (Fig 1 a, and b).
Mathematical model, the second-order (in actual and coded terms)
a) Postharvest life = 3,465 + 1,140A + 0,115B -0,314C -0,3105A2 -0,223B2 -0,393C2.
b) Postharvest life = -4,389 + 0,402 UVC-Dose-C + 0,194time + 0,341Temperature -0,010 UVC-Dose-C^2 -0,003 (time)2 -0,010 (Temperature)2
The mathematical expression in terms of coded factors can be used to make predictions about postharvest life for given levels of each factor (a). The coded expression is useful to identify the relative impact of the factors (UV-C dose, temperature, and time) by comparing the coefficients of the factors. It is observed that temperature (C) has a linear negative effect on postharvest life (coefficient preceded by a negative sign -0.314), while UVC dose and exposure time (A; B) have a linear positive effect (coefficients preceded by positive signs 0.14 and 0.11, respectively) on postharvest life, negative quadratic terms (A; B) indicate a curvature where the values of the factors determined the response to the maximum value. From this point on, the response values could begin to gradually increase or begin to decrease (design limit)
The expression obtained in terms of actual factors (b) can be used to make predictions about postharvest life for given levels of each factor, such as UV-C dose, temperature, and exposure time, provided the levels are specified in the original factor units for each. However, this expression should not be used to determine the relative impact of each factor because the coefficients are scaled to accommodate the units of each factor and the intersection is not in the center of the design space (Table 5).
In the statistical program, the numerical optimization maximized the postharvest life response to a limit lower than 10 (minimum acceptable value) and a limit higher than 20 (Montgomery, 2010) Under these conditions, maximum desirability of 73% is obtained in the following factorial coordinates (A, B, C) = (15; 24, 4; 15, 5). At this theoretical point, three experimental replicas were carried out. The results obtained are compared with the results of expression b (Table. 4) and show a good agreement between the calculated (4.30) and experimental (3.78) values, which confirms that this model can be used to navigate for the design space (Cote et al., 2013).
Factors | Values |
---|---|
A- UV- C Dose | 15 |
B-Time exposure | 30 |
C-Temperature | 12 |
Optimized response* | 4,37 |
Experimental response** | 3,78 |
*Equation b results **Mean value of three experimental measurements.
The overall mean response for all executions was adequate. The coefficients are properly adjusted around an average value (when the factors are orthogonal, the variance inflation factor (VIF) is 1; if VIF is greater than 1, it indicates multicollinearity (the higher the VIF, the more severe the correlation of factors). This indicates that the expected change in postharvest life per unit changes in value by one factor when all the remaining factors remain constant (Table 5).
Factor | Coefficient Estimate | Df | Standard Error | 95% CI Low | 95% CI High | VIF |
---|---|---|---|---|---|---|
Intercept | 3,47 | 1 | 0,0820 | 3,29 | 3,64 | |
A-Dose (UV-C) | 1,14 | 1 | 0,0497 | 1,03 | 1,25 | 1,0000 |
B-exposure time | 0,1151 | 1 | 0,0497 | 0,0069 | 0,2232 | 1,0000 |
C-Temperature | -0,3148 | 1 | 0,0497 | -0,4230 | -0,2066 | 1,0000 |
A² | -0,3105 | 1 | 0,0497 | -0,4187 | -0,2023 | 1,04 |
B² | -0,2239 | 1 | 0,0497 | -0,3321 | -0,1157 | 1,04 |
C² | -0,3936 | 1 | 0,0497 | -0,5018 | -0,2854 | 1,04 |
The response surface, the contour lines and the cube graphs show that the postharvest life of the strawberry increases as the dose increases and the exposure to UV-C as the temperature decreases (Fig 2). A maximum zone is verified where the best postharvest life conditions are obtained (4.37 days). At this point, the designed system verifies the best conditions of the study variables to extend the postharvest life of the strawberry (Figure 2).
The observed changes are representative of the ranges of the study system and indicate that, under optimal conditions, the treated fruits take around four days to lose 10% of their quality compared to the controls. The upward curvature of the model allows inferring that the maximum value obtained is not absolute and other areas with better response behavior could be found by increasing the UV-C dose. This implies that the postharvest life of the strawberry could be extended at higher doses of UV-C and lower temperatures.
Several causes are attributed to the delay of senescence induced by UV-C, among them there are compounds called phenylpropanoids, which prevent the direct action of microorganisms on the cell parenchyma (Liu et al., 2018). Other bioactive compounds involved are phytoalexins and polyamines that cause the inactivation of enzymes that soften the tissues of the fruit, generating mechanisms that can slow down its senescence (Severo et al., 2015). Causally, there are several mechanisms that describe the action of UV-C on biological systems that in turn can delay the damage due to the cold weather and soft rot of fruit caused by microorganisms (Formica et al., 2017; Petrielli et al., 2019).
UV-C has been shown to generate mutations in the DNA molecule of microorganisms by directly inhibiting microbial growth, but at the same time, it also stimulates certain genomic regions of the fruit. (Pombo et al., 2011). These regions encode enzymes that slow fruit senescence (Urban et al., 2016). On the other hand, there are components of the membrane (phospholipids, glycolipids, proteins and lignin) that are sensitive to the short wave ultraviolet range, where they absorb energy promoting permanent metabolic changes. In these conditions, the cell walls of the fruit are reinforced, offering greater firmness (Ortiz et al., 2019).
The effect of cold also affects the delay of the metabolism of the fruit (Andrade et al., 2013). As the temperature drops, microbial growth is inhibited thus preventing contamination of the fruit. Low temperatures also regulate respiratory processes that directly influence ripening (Liao et al; 2016), delaying the deterioration of the fruit and the senescence process (Liao et al., 2016).
In this study, the combined effect of low temperatures and UV radiation-induced favorable changes in strawberry postharvest life. The adequate combination of the study factors probably induced a cascade mechanism on the plant tissue, which led to the prolongation of the postharvest life of strawberry (Formica et al., 2017).
CONCLUSION
It was demonstrated that the Design of the Central Compound is a good tool to evaluate the effects and interactions of UV-C dose, temperature and time, in the postharvest life of Fragaria vesca.
The empirical model obtained was statistically significant and showed good agreement between the experimental and predicted values, obtaining a maximum response of 4 days above the controls.
The model indicates that a better response could be found by increasing the UV-C dose.
The applications of this work emphasize the importance of the combined effect of UV-C and low temperatures in the postharvest life of the strawberry.