Tommy Atkins mango ( Mangifera indica L . ) postharvest quality with cassava starch , chitosan and pectin based coatings

Bearing in mind the use of natural and biodegradable products, this study aimed to evaluate physicochemical quality of ‘Tommy Atkins’ mangoes coated with a cassava starch, chitosan and pectin based edible coatings. The experiment was conducted using simplex-centroid mixture design. Compounds concentration corresponded to the independent variables studied, encoded as factors x1, x2 and x3: 1– cassava starch (CS) 2%; 2– chitosan (CH) 2%; and 3– pectin (PE) with 2% of water polymers. Seven coating compositions (100% CS; 100% CH; 100% PE; 50% CS + 50% CH; 50% CS + 50% PE; 50% CH + 50% PE and 33.33% CS + 33.33% CH 33.33% PE (center point)), and uncoated control treatment with five replications per treatment, with four replications having five fruits treated in the center point were used. Fruits were stored for 28 days at 13°C and 90% relative humidity (RH). Physiological loss, appearance, pulp color: lightness, chroma and hue angle, fruit firmness, soluble solids content, pH, titratable acidity, and ratio were evaluated as dependent variables. Significant difference occurred with Scott-Knott tests and response surface for all the analyzed variables. Increase in physiological loss, soluble solids content were observed from fruits of all treatments. Decrease in appearance, fruit firmness, lightness in pulp color were observed. Fruits coated with mixtures of chitosan at 50 and 33.33% showed lower physiological loss. Coated fruits from all treatments presented lower fruit firmness than the control, except for SC + PE (1:1) treatment which provided similar result with the control group. Coating based in cassava starch and chitosan at 50% delayed the buildup of soluble solids and ensured reduction and/or maintenance of the variables of quality analyzed.


INTRODUCTION
Mango (Mangifera indica L.) is a tropical fruit, of climacteric behavior, with big economic importance in Brazilian national market and worldwide, due to its attractive color, nutritional value, taste and specific aroma (Singh et al., 2013).As visual quality, regular size, and firmness are the characteristics with most commercial importance, "Tommy Atkins" mango, in this context, is the most cultivated and exported cultivar in Brazil, because of its yield, good capacity in adapting to different environment and postharvest conservation (Carvalho et al., 2004;Cohen et al., 2001).In 2012, Brazil was in the seventh position among major mango producing countries, with an area of 73.3 hectares and 1,575,735 tons produced (FAO, 2016).Out of that, 127,002,229 tons were exported, generating 138 millions of dollars in Brazilian exportations (MDIC, 2016).
Having in mind that mango trade to distant markets requires studies to extend its storage life, it is necessary to focus on: maturation stage at harvest, correct harvest procedures, and postharvest handling, as well as proper storage conditions (Osorio and Fernie, 2013;Razzaq et al., 2013).As a tropical product, mango shows sensibility to low temperatures, causing injuries when kept under temperatures below 12°C (Nair and Singh, 2003;Narayana et al., 2012).
The association of refrigeration and modified atmosphere packaging (the use of plastic films, wax or edible coatings) or controlled atmosphere packaging has been used to reduce deterioration, extend shelf life and sustain mango"s quality (Singh and Singh, 2012).Edible coating works as a barrier, reducing gas exchange between fruit and atmosphere, results in a modified intern atmosphere (high CO 2 concentrations, and low O 2 ), and so decrease in water loss (Terry et al., 2011;Oliveira, 2014;Aquino et al., 2015;Petriccione et al., 2015).
The use of edible packages is important due to modern consumer being more responsible, preferring more natural, renewable and biodegradable products (Jiménez et al., 2012).Edible coatings used in postharvest are biodegradable, created from renewable sources, perfectly adjusting itself in the ecosystem and avoiding environmental pollution (Campos et al., 2011;Pascall and Lin, 2013).
In these terms, cassava starch, chitosan and pectin are polysaccharides being studied as feedstock in edible coatings production (Wills and Goulding, 2015), building a resistant and transparent layer, giving a nice and bright look to the fruit, and making them commercially attractive (Jiménez et al., 2012).The use of chitosan coatings reduced physiological loss and delayed fruit firmness loss in mangoes during storage (Cissé et al., 2015), such as the use of polysaccharides (methylcellulose, hydroxypropyl cellulose, carboxymethyl cellulose and chitosan) on sustaining citrus fruits quality (Arnon et al., 2014).Another positive effect of polysaccharide coatings (pectin and chitosan) was pointed by the decrease in respiration rate and ethylene production and, as a consequence, extended mango"s shelf life (Medeiros et al., 2012).
Mangoes stored only in refrigerated environment reached the stage of consumption 20 days after refrigeration and four days at room temperature.For "Tommy Atkins" mangoes, 12°C combined with a 5% of CO 2 + 5% of O 2 kept fruits in good marketable conditions for 31 days.These results supported the possibility of long sea shipping (21 to 23 days), and reducing costs.
Mango"s postharvest life under refrigeration at 10 and 13°C and relative humidity (RH) of 85% varies from 2 to 3 weeks depending on maturation stage (Neves, 2009), while when it was wax coated or plastic packed, it increased its life between 28 and 35 days.However, Santos et al. (2012) verified postharvest life in "Tommy Atkins" mangoes of 28 days after coating with a 2% cassava starch coating.This study"s purpose was to evaluate effects of cassava starch (CS), chitosan (CH), and pectin (PE) isolated and mixtures of edible coatings on sustaining "Tommy Atkins" mangoes postharvest quality.

MATERIALS AND METHODS
The present study was carried out at Laboratório de Tecnologia de Alimentos (Laboratory of Food Technology), Universidade Federal Rural do Semi-Árido -UFERSA.Materials used for films production are cassava starch, chitosan, citrus pectin, and distilled white glycerin.
Mangoes from "Tommy Atkins" cultivar were collected for this experiment, harvested at physiological maturity, at Fazenda Finobrasa Agroindustrial S/A, located at Ipanguaçu-RN.After transportation to the Laboratory of Food Technology at Universidade Federal Rural do Semi-Árido, fruits were selected, washed with chlorinated water (100 ppm) and dried at room temperature.
Fruits were identified and sorted according to the experimental design used (response surface in simplex-centroid mixture design), with five replications in each trial and seven coating types, and the uncoated control treatment.
Response surface methodology describes dependent variable (Y) behavior when there are changes in the independent variables, within the studied interval (Rodrigues and Iemma, 2005).Results for mixtures design are represented by a response surface ternary triangle, where basal vertices correspond to measured responses to pure components.Mean points in the edges represent responses for binary mixtures, and points within the triangle correspond to ternary mixtures (Bruns et al., 2001).
Cassava starch (CS), chitosan (CH) and pectin (PE) coatings were made with 20 g of polymer, 2 g of plasticizer (glycerol), in 978 g of distilled water (CS and PE), and 978 g of acetic acid (1) in a pH of 3. Cassava starch solution was stirred and heated to a temperature of 70°C, during 15 min, using a stirrer with a heating plate while chitosan and pectin solutions were homogenized with the stirrer for 45 min.After preparation of each solution, separately, mixtures were set and refrigerated at 25°C.Fruits were individually immersed for one minute in its respective treatment (coating) solution and dried at room temperature for one hour, just followed by storage at a 13±1°C and 90±5% RH for 28 days.The following physical and physicochemical analysis were evaluated at the end of storage period.

External appearance (EA)
Six trained people evaluated fruits using a visual and subjective *Corresponding author.E-mail: thiago@ufersa.edu.br.

Physiological loss (PL)
This is measured by the difference between mass at the start and each measurement period, expressed in percentage (%).

Fruit firmness (FF)
Measurement was made based on penetration, using a McCormick FT 327 penetrometer (8 mm-diameter tip) in equatorial regions of the fruit, two per fruit.Some of the epicarp was removed.Results are expressed in Newton.

Total soluble solids content (SS)
This was determined after passing the whole fruit through a blender and measured with a digital Pallete PR-100 refractometer (AttagoCo.Ltd., Japan), with automatic temperature correction and reading range from 0 to 32 °Brix.Results are expressed in percentage (%) (AOAC, 1992).

Titratable acidity (TA)
This was determined by titration of an aliquot of 10 g of juice, in duplicate, which were added 40 mL of distilled water and titrated with NaOH (0.02 N) solution until pH equals 8.1 (using a digital pHmeter).Results were expressed in percentage of citric acid, according to IAL (1985) methodology.

pH (potential hydrogen)
This was measured in the fruit"s juice in duplicate, using a digital pHmeter (AOAC, 1992).

SS/TA ratio
Result of division between average values of soluble solids content and titratable acidity average.Response surface statistical analysis of individual effects on experimental factors and their interactions on mangos" attributes of quality through 28 storage days, coated with cassava starch, chitosan and pectin isolated or in mixtures, were ran with Programa Statistica 7.0 software (STATSOFT, 2004).Adjustments of regression models in response surface were made.The model choice followed the significance criteria and its parameters, errors regression significance and coefficient determination (R²).Data was subjectted to variance analysis using SISVAR software (Ferreira, 2003).Level of treatment factors were compared by Scott-Knott at 5% of probability.

RESULTS AND DISCUSSION
Table 1 shows the experimental arrangement used; it was verified that trials from 7 to 10 originated from four replications made at central point to determine experimental error.It was considered that the response shows homocedasticity within the experimental domain studied.Table 2 shows model"s coefficients, standard errors estimate, Student test (t test) and p-value to main effects, and interactions of quadratic and cubic models used to analyze variables" description in mangos coated with different biopolymers compositions.Figure 1 shows a Pareto graph of standardized effects for comparison of factors significance and their interactions with variables.
Main effects are statistically significant at 95% of confidence: x1*x2 double interaction in physiological loss, appearance, and pulp lightness, x1*x3 interaction in chroma and hue angle of pulp color, and x1*x2*x3 cubic interaction in appearance, may be considered significant at 95% of confidence.Other interactions between factors are not too significant.
Best subset selection method was used to evaluate factors and interactions that must be disregarded from the models.Used as criteria of factors and interactions selection to set the mathematical model, was adjusted Rsquared value maximization.Adjusted R-squared is calculated from R-squared adjustment of each subset from the model considering the amount of variables in the model and sample size.
Variance analysis (Table 3) shows that the regressions are significant at 5% of probability for analyzed variables.The models may explain 96.11; 99.80; 98.29; 82.10 and 91.77% of R-squared.During the period of storage, a significant difference of physiological loss occurred within treatments. Figure 2 presents response surface contour curves (a) obtained with application of mathematical model and Scott-Knott test (b) at 5% of probability to physiological loss.
Significant differences were observed in physiological loss of fruits within treatments, as shown in Figure 2. Physiological loss occurs due to the transpiration process, which may be increased with increase in temperature and decrease in relative humidity (Chitarra and Chitarra, 2005).Physiological loss of coated fruits is related to an attribute of water vapor barrier.Polysaccharides such as cassava starch, chitosan and pectin are excellent barriers to oxygen and carbon Table 1.Experimental arrangements and results for physiological loss (%) -Y1, external appearance (Y2), pulp lightness color (L) (Y3), fruit firmness (Newton) (Y4), soluble solids content (%) (Y5), pH (Y6), titratable acidity (% of citric acid) (Y7), and SS/TA ratio (Y8) in "Tommy Atkins" mangoes coated with chitosan, cassava starch and pectin coatings.dioxide; however, in isolated forms, they are hydrophilic and have high permeability to water vapor, which differ from each polymer matrix (Medeiros et al., 2012;Cissé et al., 2015).The coating may be optimized by combination with hydrocolloids turns the coatings formation more homogenous (Gao et al., 2013;Castañeda, 2013;Silva, 2015) or with lipidic materials (Xu et al., 2005;Oliveira, 2014).
When response surface quadratic model is analyzed for weigh loss on mango fruits during 28 storage days, it is seen that the model is adequate to describe the mixture of all three constituents, because the graph for ternary mixtures with quadratic model already have a higher synergic interaction in binary mixtures with cassava starch and chitosan.
Fruits coated with film-forming solution composed of 5 to 33.33 % of chitosan, and having cassava starch or pectin as the others components, provided lower physiological loss within treatments, under the experimental conditions.Notwithstanding, mangoes coated with mixtures of equal parts (1:1) of chitosan and cassava starch showed decrease of physiological loss (27.27%).According to Castañeda (2013), in refrigerated storage of apple postharvest conservation, mixture of cassava starch and chitosan at equal proportions made better coatings, due to better rearrangement of polymeric chain, proved by the scanning electron microscopy, which ensures a more efficient water vapor barrier.
Coatings with composition of CS, and CS+PE provided similar effect of physiological loss to control treatment fruits.On the other side, the other compositions resulted in positive effect in reducing fruit physiological loss as compared to the control treatment.Mixture coatings of CS+CH, CH+PE, and CS+CH+PE had less physiological loss than mixtures with PE and CH.
The differences of physiological loss between coated and uncoated mangoes are primarily related to water vapor barrier provided by the coatings, where each polymer has different properties and structures which gives different structural arrangements and different formation of polymeric chain (Wills and Goulding, 2015).
Tommy Atkins and Supresa cultivars of mangos, coated with cassava starch at 3 and 2%, respectively, presented lower physiological loss as compared to lower concentrations and to control treatment (Santos et al., 2011;Scanavaca Júnior et al., 2007).Medeiros et al. (2012) coated "Tommy Atkins" mangos with chitosan and pectin coating, kept for 45 days at 4°C and 90% of RH, and observed less physiological loss at 28 days to coated fruits.Cissé et al. (2015) also had lower physiological loss as compared to control treatment fruits in mango fruits from Kent cultivar stored for 8 days coated with 1 and 1.5% chitosan.
Reduction in external appearance scores was observed in fruits from all treatments at day 28 (Figure 3a and b).Fruits coated with CH, PE and CS+CH sustained higher external appearance scores when compared with the control treatment, with statistically equal scores, differing from fruits from other treatments, which had lower external appearance scores.The positive effect of coatings with CH, PE and CS+CH is due to a brighter mango skin resulting in a good-looking fruit and also having important effect on retarding senescence processes which is caused by the barrier layer that blocks gases.Ethylene, the main hormone associated with ripening process, has its effect reduced when CO 2 concentration within the cell is higher than 5% (Cissé et al., 2015).For each polymer of isolated form or in mixture, there are three elements in their formation or in the molecular structure of aqueous gel: (a) junction zones, where polymeric molecules are together; (b) polymer interjunction segments that are relatively mobile; and (c) water bonded to polymeric net.A junction zone may involve covalent bonds, electrostatic, of hydrogen or hydrophobic interactions (Thakur et al., 1997).
Supporting this study results, Medeiros et al. ( 2012) indicated that mangos coated with pectin and chitosan had better appearance and lower physiological loss than fruits from control group.One of the reasons for appearance change after harvest is the physiological loss due to transpiration, which may cause fruits to wilt, making them unattractive in market.
Another factor that may affect fruits appearance is chilling injuries, just as mechanic injuries, microorganisms attack, physiological disturbs, and also wilt caused by some coatings (Lima et al., 2012).One way to avoid this event is reducing ethylene production (Wills and Goulding, 2015).Chien et al. (2007) verified reduction on external appearance of nine mango fruits when stored for seven days at 6°C and 80% RH.Fruits coated with chitosan at concentrations of 0.5, 1 and 1.5% obtained values of 5.79, 6.42 and 6.02, while uncoated fruits had value of 3.85.Similar results for appearance with the use of coatings were detected in eggplant coated with cassava starch by Souza et al. (2010); Formosa papaya coated with carnaúba palm wax by Fernandes et al. (2012), showed better scores for external appearance at the end of storage period as compared to fruits from the control treatment.Lima et al. ( 2012) reported divergent results using cassava starch in "Tommy Atkins" mangoes, obtaining scores lower than the control treatment.Fruit appearance evaluation is extremely useful to estimate the time of commercialization, because fruit must get to large consumer markets with acceptable visual quality for consumption and marketing.Figure 4 presents response surface contour curves (a) obtained from mathematical model and Scott-Knott test (b) at 5% of probability for pulp color: lightness (L).Mango fruits pulp color suffered a decrease in lightness and at day 28 of storage, demonstrating that ripening process occurred and was represented by the changes in pulp variables during storage period.
Study of mango conservation using cassava starch performed by Serpa et al. (2014) verified that there was no effect of coatings on L variables, causing reduction of pulp lightness in "Palmer" mangoes from 80.58 to 70.07 during a 10-days storage at room temperature.
When the cubic model for L, from response surface of  pulp color of mangos during 28 days of storage was analyzed, it was shown that the models are adequate to describe the mixture of the three components, because in the presented graphs for ternary mixtures with cubic model, there is a higher interaction of mixtures with the three polymers while in the graph with quadratic model, it is noted, a higher synergic interaction in binary mixtures that involves cassava starch and pectin.Lightness of mango"s pulp was influenced by coatings at day 28.There was significant difference with Scott-Knott test within treatments.Fruits coated with CS and CS+PE presented values of pulp lightness higher than the other treatments.Lightness is related to brightness, which varies from 0 (black) to 100% (white) (Ferreira, 2003).Different results were found by Azerêdo (2013) using different coatings of cassava starch and chitosan in "Tommy Atkins" mangoes where coated fruits presented higher values of pulp lightness at the end of storage period.Other studies carried with Tommy Atkins (Amariz et al., 2010) andPalmer (Braz et al., 2007) cultivars showed decrease in pulp brightness during storage period as well, indicating quality conservation (Chien et al., 2007).
Pectin and cassava starch created a semipermeable coat around the fruit that reduces gases exchange with the environment, modifying the atmosphere inside the fruit (Wills and Goulding, 2015), interrupting carotenoids synthesis process.Similar results were reported in "Tommy Atkins" mangoes treated with 1.0% CMC + 0. 2% dextrin, 0.8% CMC + 0.5% dextrin and dextrin (Amariz et al., 2010).
Main effects are statistically significant at 95% confidence.For fruit firmness, soluble solids content, titratable acidity and SS/TA ratio variables, the x 1 *x 3 double interaction and x 1 *x 2 *x 3 cubic interaction may be considered significant at 95% of confidence, as well as x 1 *x 2 double interaction for pH and soluble solids.Other interactions between factors are not so significant.
Figure 5 shows response surface contour curves (a) obtained from mathematical model application and Scott-Knott test (b) at 5% probability for fruit firmness (Newton).Fruits coating did not avoid decrease in fruit firmness of mango fruits during storage period (Figure 5b).A decrease of 55.16% was observed in fruit firmness, from zero (121.88N) to day 28 of storage (55.16N) to control fruits; even though, fruits were still marketable.Among coated fruits, it was verified that the proportion of cassava starch and pectin showed higher values than other coatings (Figure 5a), not differing from control fruits.
x 1 *x 2 *x 3 cubic interaction, x 2 *x 3 double interaction and isolated cassava starch and pectin coatings presented the lowest values of fruit firmness at day 28 of storage, which has no effect on holding reactions that reduce fruit firmness of fruits.
Firmness is considered one of the most important attributes of fruits quality, since it affects transport resistance, postharvest conservation techniques, and microorganism"s attacks (Wills and Goulding, 2015).It is also one of the characteristics of texture and corresponds to the level of plant tissue compressive strength.It is related to the composition and pectin solubilization from cellular wall as well as middle lamella (Chitarra and Chitarra, 2005).
Firmness is also related to the amount of water in the cells, or cell turgor (Chitarra and Chitarra, 2005), which decreases with storage.Firmness decrease is also associated with the insoluble pectin fractions conversion to soluble forms during maturation.Through maturation, proto pectin enzymes and pectin methylesterase are responsible for hydrolysis and solubilization of pectic substances, contributing to firmness reduction (Wills and Goulding, 2015).Lima et al. (2012) verified in postharvest conservation of "Tommy Atkins" mangoes, after 15 days of storage at 10°C and 88% of RH, higher firmness of control fruits as compared to fruits coated with 3% cassava starch + Anise (Pimpinella anisum) extract.Cissé et al. (2015) also verified that coating systems with chitosan and lactoperoxidase in all coating treatments with concentrations of 1 to 1.5%, isolated or in mixtures, ensured higher firmness in control fruits of "Kent" mango.
Results reported by Azerêdo et al. (2016) differ, where fruits coated with cassava starch and chitosan in similar concentrations to this study showed, at 29 days of storage, firmness values higher than uncoated fruits.An increase in soluble solids content of fruits was observed in all treatments at day 28 (Figure 6a and b).
x 1 *x 2 *x 3 cubic interaction with mixtures of the three polymers in equal proportions showed the highest values of soluble solids content at day 28 of storage, not holding the reactions that breaks down carbohydrates and stored reserve of the fruit, differently from x 1 *x 2 and x 2 *x 3 double interaction which had lower values of soluble solids content than the other treatments analyzed at day 28 of storage.
Significant difference was reported for soluble solids content between analyzed treatments.Mango fruits coated with 1:1 mixture of chitosan and cassava starch and 1:1 of pectin and cassava starch had lower values of soluble solids content within treatments, under the experimental conditions.Mangoes coated with mixtures of equal parts (1:1) of chitosan and cassava starch had a 27% reduction of soluble solids content related to control group.This occur due to an specific characteristic of combination coating formulation which avoids physiological loss and gas exchanges with the increase of CO 2 levels and O 2 reduction, what prevents polysaccharides and stored reserves of mango fruits from breaking down.
According to Medeiros et al. (2012), soluble solids content is used as maturity indicators and determines fruit quality, having a great role in its taste.Results showed, in general, that fruits coated with a homogeneous coating of ½ chitosan and cassava starch sustained their soluble solids content, as compared to control fruits group.Fruits from control treatment showed soluble solids content variation from 8.81 to 13.64%, while coated fruits held values between 8.81 and 9.96%, confirming that coatings reduced metabolic activity of the fruit, slowing down its ripening.
Normally, soluble solids content increases during maturation due to fruit"s polysaccharides degradation.There is also an increase in soluble solids content when water loss is increased, causing sugar concentration to increase in fruits tissues.Reduced concentration of O 2 influences respiration processes causing decline in ethylene production and, consequently, expanding maturation period (Kader, 2002).
This sugar buildup is due to the starch degradation (Cissé et al., 2015).The increase in soluble solids content indicates maturation evolution (Silva, 2015).This way, retarding the increase in soluble solids content implies extending the conservation period (Medeiros et al., 2012).Through the mango"s maturation, starch degradation occurs due, mainly, to the α-amylase and β-amylase enzymes actions.Sugars found in mango are glucose, fructose and sucrose, with the last one being found in larger quantity than the others, and having a major contribution to soluble solids content (Cissé et al., 2015).When the buildup of soluble solids is retarded, there is a delay in ripening process.This is probably due to metabolic activity reduction (Silva et al., 2001).This attribute delay may be achieved with biodegradable coatings such as chitosan (Cissé et al., 2015), cassava starch (Serpa el al., 2014), and mixture of cassava starch and chitosan (Azerêdo et al., 2016).Medeiros et al. (2012) reported in "Tommy Atkins" mangoes coated with chitosan and pectin, a significant effect during the experimental period until day 28, an increase from 11 to 17%, with buildup of soluble solids in control fruits, while coated mangoes maintained themselves relatively stable during this period, from 11 to 13%. Figure 7 shows response surface contour curves (a) obtained from mathematical model application and Scott-Knott test (b) at 5% of probability for pH.
A decrease in pH from 3.55 at timing of zero to 3.40 at day 28 of storage from control fruits was observed (Figure 5b).Within the coated fruits, it was verified that 1:1 proportion of cassava starch and pectin and isolated cassava starch showed lower values then other coatings (Figure 5a), not differing from control fruits.All these treatments reported high physiological loss at the end of storage period and higher soluble solids content values due to breaking down of stored reserves, that causes increase in acidity with the solutes concentration and breaking down of stored reserves (amid and pectic substances) which reduced pH values.
x 1 *x 2 double interaction and chitosan and pectin isolated coatings reported the highest values of pH at day 28, higher than the control fruits.This result shows that a change occurred in the atmosphere around the fruit, caused by the use of coatings that promoted a semipermeable coating on fruits surface, modifying endogenous concentration of CO 2 and O 2 and slowing down the maturation process.
The change in pH is due to sugar formation, acids, and concentration of solutes due to physiological loss throughout maturation.pH increases with the period of storage of fruits (Moraes et al., 2012).Results clearly shows the relationship between pH and acidity, with increase in pH (Figure 7) as the acidity decreases (Figure 8), making the fruit more palatable (Serpa et al., 2014) and more mature.
The use of some coatings decelerates pH changes, affecting the acidity (Moraes et al., 2012).Possibly, these events occur due to reduction in respiration process (Trigo et al., 2012).Azerêdo (2013) verified pH reduction with ripening of fruits up to the twentieth day of storage, followed by a significant increase at day 29, existing differences between coated fruits and control treatment.Increase in titratable acidity was observed from 0.73 at timing zero to 0.85% of citric acids at day 28 of storage for control fruits (Figure 5b).Some authors reported that during the storage period, mangoes from "Tommy Atkins" cultivar had increase in values of acidity followed by a significant reduction with the use of coatings or modified atmosphere packaging (Yamashita et al., 2001;Jeronimo et al., 2007;Ribeiro et al., 2009;Amariz et al., 2010).The high physiological loss and soluble solids content of control fruits at the end of storage period cause, firstly, solute concentration and rapid increase in acidity due to breakage of stored reserves (amid and structural substances) from control fruits.
Titratable acidity of fruits was verified to be altered with coating type.Fruits that had high values of soluble solids content and physiological loss showed higher acidity, caused by the loss of solutes concentration and increase in organic acids from the breakdown of substances such as amid and stored reserves of cell walls.
x 1 *x 2 *x 3 cubic interaction and x 1 *x 3 double interaction showed lower values of acidity caused by a lower loss of weight and soluble solids content, and higher values of acidity due to greater weight and soluble solids content values with a statistical difference for each coating in isolated forms at day 28 of storage.Cissé et al. (2015) considered that the ripening reduced acidity in coated and uncoated mango fruits, which was not seen in this study.According to Chitarra and Chitarra (2005), normally, content of organic acids decreases with fruits maturation due to its use as respiratory substrate or in sugar conversion, however phenolic compounds have acidy nature as well, maybe contributing to acidity in some way.According to Alves et al. (2000), an increase in acidity is caused by release of galacturonic acids that increases with fruit ripening by the action of pectin methylesterase and polygalacturonase enzymes.
Fruits that reported decrease in titratable acidity at the end of storage period are probably related to the use of acids as carbon skeleton structure in respiratory process as explained by Kays (1991).According to Chitarra and Chitarra (2005), the content of acids in plants may decrease with maturation, because of transformation of substrates to synthesize phenolic compounds, lipids, and natural aromas.
Similar behavior for titratable acidity reported in treatments with higher values were also detected by Amariz et al. (2010) with coatings based in carboxymethyl cellulose and dextrin in "Tommy Atkins" mangoes under refrigeration due to increase in values of acidity to coated and control fruits at the twentieth day, with reduction of values throughout the storage period.Figure 9 shows response surface contour curves (a) obtained from mathematical model application and Scott-Knott test (b) at 5% of probability for SS/TA ratio.
Significant difference was observed for SS/TA ratio within analyzed treatments, graph of contour curves and Scott-Knott test (b) at 5% reported in Figure 9a and b.Increase was reported in SS/TA ratio from 12.17 at zero time to 15.95 at day 28 of storage in control fruits (Figure 5b).It is verified that fruits SS/TA ratio changed with the type of coating.Fruits that had reduced values showed low soluble solids content and/or titratable acidity at the end of storage.There was significant effect for x 1 *x 2 *x 3 cubic interaction.x 1 *x 2 double interaction presented higher and lower values of SS/TA ratio, not having statistical difference for each coating in isolated form at day 28 of storage.
The response surface contour curve shows that the use of coatings in mixtures with cassava starch and chitosan and pectin at 1:1 proportion had beneficial effects in SS/TA ratio.The beneficial effects of these coatings on reducing maturation velocity of mangoes may be verified when SS/TA ratio values are analyzed.
Azerêdo (2016) verified, in "Tommy Atkins" mango fruits storage for 32 days, fruits coated with cassava starch had lower value as compared to the control fruits and other treatments.SS/TA ratio is one of the most used indexes to determine maturation, being taste indicator (Chitarra and Chitarra, 2005).

Conclusion
Mangoes ripening, fruit firmness, external appearance, lightness of pulp color and pH of all treatments reduced.Soluble solids content and SS/TA ratio increased during the 28 days of storage.The coating based in cassava starch and chitosan at 50% proportion was more efficient in containing physiological loss, slowing down the soluble solids buildup and sustaining the pH and SS/TA ratio of "Tommy Atkins" mango fruits.Coated fruits had sustained external appearance and quality to commercialize product at the end of storage period.
Cassava starch and chitosan coating provided better conservation of "Tommy Atkins" mango fruits for a 28 days period of storage at 13°C and 90% of RH.

Figure 2 .
Figure 2. Response surface contour curves (a) and Scott-Knott test (b) for physiological loss at the twenty-eighth day of storage (13°C and 90% of RH).

Figure 4 .
Figure 4. Response surface contour curve (a) and Scott-Knott test (b) to lightness of pulp at 28 day of storage (13°C and 90% of RH).

Figure 5 .
Figure 5. Response surface contour curves (a) and Scott-Knott test (b) for fruit firmness (Newton) at twenty-eighth day of storage (13°C and 90% of RH).

Figure 6 .
Figure 6.Response surface contour curves (a) and Scott-Knott test (b) for soluble solids content (%) at the twenty-eighth day of storage (13°C and 90% of RH).

Figure 7 .
Figure 7. Response surface contour curves (a) and Scott-Knott test (b) for pH at twenty-eighth day of storage (13°C and 90% of RH).

Figure 8
Figure 8 presents response surface contour curves (a) obtained from mathematical model application and Scott-Knott test (b) at 5% of probability for titratable acidity.Increase in titratable acidity was observed from 0.73 at timing zero to 0.85% of citric acids at day 28 of storage for control fruits (Figure5b).Some authors reported that during the storage period, mangoes from "Tommy Atkins" cultivar had increase in values of acidity followed by a significant reduction with the use of coatings or modified atmosphere packaging(Yamashita et al., 2001;Jeronimo et al., 2007;Ribeiro et al., 2009;Amariz et al., 2010).The high physiological loss and soluble solids content of control fruits at the end of storage period cause, firstly, solute concentration and rapid increase in acidity due to breakage of stored reserves (amid and structural substances) from control fruits.Titratable acidity of fruits was verified to be altered with coating type.Fruits that had high values of soluble solids content and physiological loss showed higher acidity, caused by the loss of solutes concentration and increase in organic acids from the breakdown of substances such as amid and stored reserves of cell walls.x 1 *x 2 *x 3 cubic interaction and x 1 *x 3 double interaction showed lower values of acidity caused by a lower loss of weight and soluble solids content, and higher values of acidity due to greater weight and soluble solids content values with a statistical difference for each coating in isolated forms at day 28 of storage.Cissé et al. (2015) considered that the ripening reduced acidity in coated and uncoated mango fruits, which was not seen in this study.According toChitarra and Chitarra (2005), normally, content of organic acids decreases with fruits maturation due to its use as respiratory substrate or

Figure 8 .
Figure 8. Response surface contour curves (a) and Scott-Knott test (b) for titratable acidity (% of citric acid) at 28 day of storage (13°C and 90% of RH).

Figure 9 .
Figure 9. Response surface contour curves (a) and Scott-Knott test (b) for SS/TA ratio at 28 day of storage (13°C and 90% of RH).

Table 3 .
Variance analysis for adjustment of models to physiological loss (%), appearance, lightness of pulp color (L), chroma (C), and hue angle (H) in "Tommy Atkins" mangoes coated with chitosan, cassava starch and pectin coatings.