Influence of the type of packaging on textural properties of minimally processed yellow Peruvian roots

Submitted on March 27 , 2014 and approved on May10 ,2016. 1 Part of the Master ’s dissertation of the first author . Work funded by CNPq. 2 Universidade Federal de Viçosa, Departamento de Engenharia Agrícola, Viçosa, Minas Gerais, Brazil. lara.santanaf@gmail.com; copace@ufv .br Universidade Federal de São João Del Rei, Departamento de Engenharia de Alimentos, Sete Lagoas, Minas Gerais, Brazil. mateusjunq@yahoo.com.br Universidade Federal de Viçosa, Departamento de Fitotecnia, Viçosa, Minas Gerais, Brazil. f finger@ufv.br Universidade Federal de Viçosa, Departament de Estatística, Viçosa, Minas Gerais, Brazil. cecon@ufv .br *Corresponding author: lara.santanaf@gmail.com Influence of the type of packaging on textural properties of minimally processed yellow Peruvian roots 1


INTRODUCTION
For fruits and vegetables consumed in natura, texture and color are the most important quality attributes for market value definition (Liu et al., 2009).
Factors that affect texture can substantially change after harvest, due to intercellular adhesion change, conversion of starch into sugars, loss of water and cellular wall force (Toivonen & Brummel, 2008).In minimally processed products, several handling techniques and conservation treatments are applied to reduce texture changes, like refrigerated conservation and use of plastic wrap.
Food texture can be measured by physical tests: punching, penetration, compression, shearing and relaxation, allowing the gathering of data related to consistency and resistance of vegetable tissues by the application of any force and the specific deformation or total test time (Bourne, 2003).
Relaxation is an important test for food texture evaluation, usually used for the study of the viscoelastic behavior of biological materials (Cespi et al., 2007), and the modeling of a mathematical instrument is basic for analysis of this behavior (Del Nobile et al., 2007;Fustier et al., 2009).
According to Resende & Corrêa (2007), specific models can satisfactorily describe the variation of rheological parameters during food ripening, providing not only practical texture rates, but also underlying characteristics to processes that occur during storage.
Other methods have been developed to determine food texture properties, being worth highlighting the Instrumental Texture Profile, which has been efficiently applied for a large food scale.The Instrumental Texture Profile Analysis (TPA) simulates the conditions to which food is submitted along the mastication process and is offered as a way of helping food researchers to obtain descriptive and quantitative sensorial data regarding texture characteristics.
Thus, the aim of this study was to evaluate the texture alterations that occur during the storage period of minimally processed yellow Peruvian roots by texture profile analysis (TPA) and relaxation, using four types of plastic packaging, combined to refrigeration during storage.

MATERIALS AND METHODS
The present study was developed in the Laboratory of Physical Properties and Quality Evaluation of Agricultural products of the National Grain Storage Training Center (CENTREINAR), and in the Post-harvest Laboratory, of the Phytotechnology Department and Minimum Processing Unit, both at the Federal University of Viçosa, UFV (MG, Brazil).
We used yellow Peruvian roots from cultivar Amarela de Carandaí, obtained at the Central Food Supply (Ceasa, MG, Brazil).The tubercles with damages, visible deformations or diseases were removed to obtain a homogeneous product.
The roots were selected, washed in running water with neutral detergent to remove dirt.Subsequently, they were taken to the minimum processing room, refrigerated at 18 ± 1 °C, and sanitized in solution with 200 mg L -1 active chlorine, for 15 min.Next, they were peeled and manually cut into ± 1 cm thick slices and immediately placed in frozen water.For final sanitization, the slices were immersed in a solution containing 200 mg L -1 active chlorine, at 5 ± 2 °C, for 15 min, with subsequent rinsing in solution containing 3 mg L -1 active chlorine, at 5 ± 2 °C for 15 min and immersed in a 3% ascorbic acid solution and 3% citrus acid for 5 min.Water excess was removed by centrifugation at 2000 rpm for 15 s and the slices were conditioned in the following packaging types: expanded polystyrene trays covered with 16 µm-thick polyvinylchloride (PVC) film; 10 µm-thick high density polyethylene (HDPE) packaging bags; 20 µmthick polypropylene (PP) packaging bags and 70 µm-thick multilayer polyolefin vacuum bags.The samples were maintained at 5 ± 2 ºC and 90 ± 5% of relative humidity for a period of 12 days in a biochemical oxygen demand (B.O.D.) incubator.
The texture profile analysis was carried out according to Beléia & Pereira (2004), using a texture analyzer, model TA.HD (Stable Micro System), also known as texturometer.The samples were pressed at 20% of the initial height using compression with constant speed of 0.02 m min -1 with an aluminum cylinder of 100 mm diameter.The interest parameters were hardness and adhesiveness, which were automatically calculated by the software program Texture Expert for Windows ® from force curves (N) x time (s) produced during the test.
The experiment was delineated in four groups, containing the following packaging types: 1) expanded polystyrene trays covered with polyvinylchloride film (PVC); 2) high-density polyethylene (HDPE) packaging bags; 3) polypropylene (PP) packaging bags; and 4) multilayer polyolefin (PML) packaging bags for vacuum.The groups were subdivided based on chronological evaluation (from 0 to 12 days) in a completely randomized delineation with four replications.The data were analyzed by analysis of variance and regression.To compare means of qualitative factor, Turkey's test was used, at 5% significance level.For the quantitative factor, models were chosen based on significance of regression coefficients using t test, coefficient of determination (R 2 ) and biological behavior.
The relaxation test was also carried out using the texturometer.A cylindrical flat probe was used, with 100 mm diameter, at a test speed of 0.02 m min -1 and constant force of 4 N.For the relaxation curve modeling, the generalized Maxwell model was employed, equation 1: (1) where: σ (t): force normalized at time t, dimensionless; σe: force normalized at balance (t = ), dimensionless; A 1 , A 2 , A 3 : model constants, dimensionless; t: time, s; and τ 1 , τ 2 , τ 3 : relaxation time, s.
In order to verify the adjustment degree of models to describe the relaxation of yellow Peruvian root slices, the magnitudes of determination coefficient (R 2 ), relative mean error (P) (Equation 2) and standard deviation estimates were used (SE) (Equation 3).

RESULTS AND DISCUSSION
The texture attributes of yellow Peruvian root slices of the four packaging bags studied are presented in Table 1, according to TPA data.The packaging bags presented statistical difference for hardness.
Figure 1 shows the evolution for hardness (a) and adhesiveness (b) values of slices along storage.
Regarding hardness, a decrease was observed in all packaging bags, and slices conditioned in PP and vacuum packaging bags presented higher hardness value, which allows inferring that at the sensory level, greater intensity force is required to compress food in mouth between molars compared to slices conditioned in other packaging bags.Nunes et al. (2011), in a study with minimally processed yellow Peruvian roots, treated with different antioxidants and stored at 5 ºC for fifteen days, also verified reduction of hardness of slices during storage.According to Chitarra & Chitarra (2005), hardness is associated with the force required for the product to reach a given deformation, giving an idea of cell structure transformations, cell cohesion and biochemical alterations, occurred during the product useful life as a consequence of loss of turgor and/or action of cell wall hydrolytic enzymes.
The adhesiveness profile increased (in module) as the estimate for slices conditioned in vacuum packaging bags was lower, and for slices conditioned in PVC, it was higher at the respective storage times, requiring more work to separate the compression probe from the sample.According to Junqueira et al. (2010), the adhesiveness increase is possibly associated with microbial growth in the sample surface.They noticed that, from day 6, minimally processed cassava sticks were sticky during handling, showing high correlation with the UV fluorescence test, detecting presence of pseudomonas.

Storage days
Rev. Ceres, Viçosa, v. 63, n.3, p. 291-296, mai/jun, 2016  Table 2 shows parameters of the generalized Maxwell model to describe the relaxation process of yellow Peruvian roots during storage.
The isolated use of the coefficient of determination must not be used as evaluation criterion of non-linear models (Kashaninejad et al. 2007;Cunningham et al., 2007).Hence, P and SE values were calculated (Equations 2 and 3).The estimated data variability lower than 10% indicates better model adjustment for practical purposes, which is recommended for non-linear models (Mohapatra & Rao, 2005;Cunningham et al., 2007).P and SE values were satisfactory, indicating that the Maxwell model represents the behavior of relaxation curves of yellow Peruvian root slices during storage.

Time Parameters
Other authors, using the Maxwell model with different numbers of elements, also reported better representation of experimental relaxation data for different products by this equation (Del nobile et al., 2007;Rodríguez-Sandoval et al., 2009;Sadowska et al., 2009;Bellido & Hatcher, 2009;Bhattacharya, 2010;Campus et al., 2010).This fact demonstrates the wide use of this model for the study of viscoelastic alterations.
It is noticed, in the results of Table 2, that slices conditioned in PP and vacuum packaging bags are more elastic than slices conditioned in other packaging bags, as the parameter óe represents the elastic component of higher magnitude in the model.Therefore, higher óe values show more elastic and better quality products (Bellido & Hatcher, 2009).This can be seen in Figure 2, which shows the relaxation curves of values estimated by the generalized Maxwell model of yellow Peruvian root slices conditioned in all packaging bags studied, at different storage times.It is noticed that, by fixing a time during analysis, slices conditioned in PP (c) and vacuum (d) packaging bags present a higher normalized force than slices conditioned in PVC (a) and HDPE (b) packaging bags.
The relative normalized force at day 0 in a determined test time is higher than the normalized forces referring to day 2 of storage, and so on.This is due to the fact that slices from the first storage days are firmer than slices from the subsequent days, since they present, at each day, a less rigid structure, with less normalized force.

CONCLUSIONS
Slices conditioned in polypropylene and vacuum packaging bags presented higher hardness value.
The generalized Maxwell model represented well the relaxation process in all studied packaging bags, appropriately describing the alterations of yellow Peruvian root slices during storage.
Yellow Peruvian roots slice conditioned in polypropylene and vacuum packaging bags were more elastic than slices conditioned in other packaging bags.
by model; n: Number of data observed; and GLR: Degrees of Freedom of models (number of parameters of model minus one).

Figure 1 :
Figure 1: Observed values for the attributes hardness (a) and adhesiveness (b) in yellow Peruvian root slices conditioned in the packaging bags studied (PVC, HDPE, PP and VACUUM).

Figure 2 :
Figure 2: Relaxation curves of yellow Peruvian root slices conditioned in PVC (a), HDPE (b), PP (c) and vacuum (d) packaging bags, at different storage times.

Table 1 :
Result of the Texture Profile Analysis (TPA) of yellow Peruvian root slices conditioned in packaging bags studied during storage For the same variable, means followed by same small letter in line do not differ statistically by Tukey's test at 5% significance level.

Table 2 :
Values of estimated standard deviation (SE), relative mean error (P), coefficient of determination (R 2 ) and parameters (σe, A1, A2, A3, τ1, τ2 and τ3) of the generalized Maxwell model obtained by the adjustment of experimental data from relaxation process of yellow Peruvian root slices conditioned in all packaging bags **Significant at 1% significance level by t-test.