Quality attributes of Pitanga ( Eugenia uniflora L.) during postharvest storage as affected by ripening stage and storage temperature

Abstract

BACKGROUND:

Pitanga (Eugenia uniflora L.) is a South American berry with a sweet flavor and is rich in bioactive compounds and antioxidants. However, its high water and sugar content, susceptibility to spoilage, and bruising limit its shelf life.

OBJECTIVE:

This study aimed to investigate the physical and chemical changes of pitanga during storage and to identify the optimal harvest time and storage conditions.

METHODS:

Pitangas were harvested at mid or full-ripeness stages, stored at 2°C, 10°C, or 20°C, and analyzed for physical and chemical quality attributes using a kinetic approach.

RESULTS:

The storage temperature of pitanga significantly affected the kinetics of various quality attributes, including weight loss, percentage of damaged and decayed fruit, fruit hardness, total soluble solid content, and color change. However, the ripening stage during harvest only influenced the initial values of a few quality attributes. The bioactive compound content and antioxidant capacity of pitanga were not significantly affected by either storage temperature or ripening stage, except for vitamin C.

CONCLUSIONS:

Pitanga can be preserved at 2°C for 20 days, resulting in a four-fold increase in shelf life compared to current postharvest practices. This finding offers valuable insights into optimizing the postharvest technology of pitanga, thereby increasing its utilization and promoting sustainable production practices.

Keywords

Cold preservation surinam cherry weight loss color bioactive compounds antioxidant capacity

1 Introduction

Berries are known to have a positive effect on human health because of their high concentration of phytochemicals, macro- and micronutrients, as well as their high antioxidant capacity. The consumption of fresh or powdered berries may be considered a tool to prevent obesity, cancer, and cardiovascular and degenerative diseases [1]. The most cultivated berries around the world are strawberry, gooseberry, blueberry, raspberry, and blackberries. North and South American countries are the main producers of those berries. However, South America has many underutilized native berries with a great economic potential as functional foods in fresh and industrial markets [1, 2] and there is a potential to take advantage of the industrial capacity for processing traditional berries. Pitanga (Eugenia uniflora L. Myrtaceae), one of the most promising berries native to South America, is commercially grown in Brazil [3] and utilized as a novelty product by gastronomy industry in Uruguay [4]. In Argentina, it is consumed in natura or processed by forest dwellers and family farmers who live in the natural distribution area of the species [5, 6]. Outside South America, pitanga has been popular in Hawaii, where it is demanded by food processors and chefs [7]. The quality of pitanga is comparable to that of the most cultivated berries worldwide in terms of size, color, flavor, minerals, bioactive compounds, and antioxidant capacity [8 –11].

Previous research has shown that pitanga is highly perishable and susceptible to handling, and has short postharvest life [12, 13]. The promotion of a wider adoption of this species into family farmers or berry farmer’s production systems still has several issues to be addressed, including, but not limited to, availability of selected cultivars, adequate postharvest technology, and correct storage and marketing of good-quality products. Postharvest handling is one of the common constraints on the adoption of indigenous fruit trees by farmers around the world, accounting for up to 60% of post-harvest losses [14]. Previous reports stated that the maximum cold storage period of Pitanga was 5 days at 7°C [15], 8°C [12], 10°C or 14°C [13]. None of those reports evaluated storage temperatures below 7°C, arguing the possible chilling injury in pitanga during storage. Nevertheless, the natural distribution of pitanga reaches 31° S latitude in South America; therefore, the cells of this species could have the genetic information to withstand low temperatures (<10°C). On the other hand, shelf life was related to the ripening stage at harvest and was increased up to 8 days in orange-red pitanga stored at 10°C under modified atmospheres (Santos et al., 2006b). The asynchronous flowering period between plants propagated by seeds and within each plant [16], combined with the short period of conservation, makes postharvest handling of Pitanga a complicated task. This issue also reduces the fresh market period, leaving industrialization as the only option to commercialize large volumes for a long period. Accordingly, this fact may be one of the causes that could discourage pitanga wide cultivation or exploitation for commercialization.

The objective of this work was to study and model the changes in quality of pitangas harvested at two ripening stages and stored at 2, 10, and 20°C, using a kinetic approach. We explored a low-temperature storage conditions not previously explored in pitanga together with the application of kinetics models to analyze the deterioration of the fruit during storage. This approach provides a more comprehensive understanding of the changes in the physicochemical and nutritional properties of pitangas over storage time, for the development of effective preservation strategies. Our results could have significant implications for the commercialization of pitanga, providing greater flexibility in the marketing of this crop and enabling wider adoption by berry producers.

2 Materials and methods

2.1 Plant material and sampling

Pitanga trees were grown from seeds obtained from a natural population in the Chaco forest (28°33‘44.7” S; 59°33‘38.0” W) of Santa Fe province, Argentina, and cultivated since 2013 in the Campo experimental de Cultivos Intensivos y Forestales experimental field of the Universidad Nacional del Litoral (31°24’01.7”S; 60°54’01.7” W). According to the BBCH scale developed for pitanga [16], fruits at phenological stage 89 are classified as “fully ripe”, while fruits at stage 83 are categorized as “mid ripe”. Pitangas at stage 89 have a hue value of 15.4±2.6, pulp hardness of 6±0.3 shores, TSS content of 14.5±0.3 °Brix, and a ratio of 23:1. On the other hand, pitangas at stage 83 have a hue value of 27.7±4, pulp hardness of 8±0.4 shores, TSS content of 10.7±0.2 °Brix, and a ratio of 11:1 [16]. In this study, 630 pitangas were randomly harvested at stage 83 and an additional 630 pitangas at stage 89 from 20 fully productive trees.

Samples were put in plastic containers with a label indicating the ripening stage at harvest and stored in a Coleman 40- Quart Portable PowerChill Thermoelectric Cooler (at 5°C and 95% relative humidity) to prevent quality changes until the beginning of the experiment (2 h after harvest).

2.2 Experimental design and determination of quality attributes

The collected fruits were washed with tap water, sanitized (100 mg.L^–1 of chlorine) and drained on paper towels. Pitangas harvested at both ripening stages were subdivided into 63 samples of 10 fruits each one, which were placed in PET containers with lids. Each sample was randomly assigned to one of the treatments shown in Table 1. Each chamber had a temperature and humidity regulator (Memmert, HPP 260 eco model, Germany). All the samples were stored at 95% RH without light exposure. The experimental design was a completely randomized factorial design, with two factors (ripening stage×temperature), with three replicates per factorial combination (Table 1). The temperatures were chosen to represent an industrial cold chain (2°C), domestic refrigerated storage (10°C), and storage with a slight abuse of temperature at sales points or home (20°C). All containers were put in each refrigerator at the same time (t = 0). At each sampling time, three replicates of each combination (ripening stage and temperature) were removed from the corresponding chamber and the response variables were recorded. Samples stored at 2°C and 10°C were analyzed at 0, 42, 138, 186, 234, 330, and 474 h, and samples stored at 20°C were analyzed at 0, 42, 66, 138, 162, 186, and 210 h.

Table 1
Experimental design and number of replicates of Pitanga per ripening stage, storage temperature, and sampling time point

Sampling time points

Ripening Storage t ₀ t ₁ t ₂ t ₃ t ₄ t ₅ t ₆ TRT^b TRRS^c

stage temperature (°C) Number of replicates^a

83 2 3 3 3 3 3 3 3 21 63

10 3 3 3 3 3 3 3 21

20 3 3 3 3 3 3 3 21

89 2 3 3 3 3 3 3 3 21 63

10 3 3 3 3 3 3 3 21

20 3 3 3 3 3 3 3 21

TRSP^d 18 18 18 18 18 18 18 126

		Sampling time points
83	2	3	3	3	3	3	3	3	21	63
	10	3	3	3	3	3	3	3	21
	20	3	3	3	3	3	3	3	21
89	2	3	3	3	3	3	3	3	21	63
	10	3	3	3	3	3	3	3	21
	20	3	3	3	3	3	3	3	21
TRSP^d		18	18	18	18	18	18	18		126

Notes: ^aA replicate was composed of 10 Pitangas inside a PET container. ^bTotal number of replicates per storage temperature. ^cTotal number of replicates per ripening stage at harvest. ^dTotal number of replicates per sampling time point. t₀ was the beginning of the experiment and t₆ was the end.

The following response variables were analyzed;

Weight loss (%): it was calculated as follows: WL (%) = (W₀ - W_n)/W₀ * 100, where; W_n is the weight of all pitangas in the container at t = n (h) and W₀ is the weight of all pitangas in the container at t = 0 (h).

Percentage of damaged fruits (%): a pitanga was considered “damaged” when at least one of the following symptoms was observed: dehydration wrinkles, fruit shriveling, loss of gloss, pitting, and brown shrunken areas. We evaluated the incidence 1 , but not the severity 2 of dehydration on each sample.

Percentage of decayed fruits (%): a pitanga was considered “decayed” when it presented soft rot accompanied by a collapse and water soaking, and/or the appearance of gray masses of conidia. We evaluated the incidence 1 , but not the severity 2 of rot on each sample.

Fruit hardness was evaluated using a SHORE-C durometer (Ludwig, Germany). The determinations were done on opposite sides of the equatorial axis of 10 pitangas. The mean values are reported and expressed in shores.

Total soluble solid (TSS) content was evaluated from the juice obtained of each pitanga sample using a refractometer and expressed in Brix degrees (%).

Color determinations were done on opposite sides of the equatorial axis of 10 pitangas. The mean values of those measurements on the CIE L*C*h color space are reported. CIE L*C*h values were measured using a Minolta Chroma-meter (Model CR-400, Minolta, Tokyo, Japan), calibrated using the standard white tile. D65/10° was used as the illuminant/viewing geometry and the specular component was excluded (SCE). L* (Lightness), C* (Chroma value), and h (hue angle) are reported. When h > 275°, the hue was converted using h-360, which resulted on negatives values of h, allowing a proper fit of the kinetic models. In addition, ΔE* was calculated according to [18].

Total anthocyanin content was determined by the pH differential method, according to [19]. Results are expressed as mg kg^–1 of pelargonidin-3-glucoside on a fresh weight basis.

Total phenolic content was determined using the Folin–Ciocalteu reagent, according to [18]. Results are expressed as mg kg^–1 of gallic acid on a fresh weight basis.

Vitamin C (VitC) was determined according to [20]. Results are expressed as mg kg^–1 of ascorbic acid on a fresh weight basis.

Antioxidant capacity was determined by analyzing the ability of samples to scavenge the free-radical DPPH, according to [21]. The results are expressed as mmol kg^–1 of Trolox on a fresh weight basis. In addition, the FRAP assay was conducted by monitoring the absorbance change at 593 nm caused by the reduction of the Fe^3 +-TPTZ complex to the ferrous form at pH 3.6 [21]. FRAP values were obtained by comparing the absorbance change in the samples with those obtained from increasing concentrations of Fe^2 + (0.1–0.9 mmol FeSO₄/L). Results are expressed as mmol kg^–1 of Fe^2 + on a fresh weight basis.

Since deterioration rates of pitangas varied with storage temperature, the sampling times differed at each temperature to obtain at least seven experimental data during the experiment.

Additionally, oxygen (O₂) and carbon dioxide (CO₂) gas composition within PET containers were monitored by injecting 1 mL of gas from the headspace (Hamilton syringe) into an SRI gas chromatograph fitted with a thermal conductivity detector. A CTRI column (cat. N° 8700, Alltech Associates Inc.) was used. The operation conditions were set according to [22]: oven temperature was 55°C, the injector and detector were held at 110°C and the hydrogen carrier flow was 65 mL/min. The evolution of oxygen and carbon dioxide concentration at each temperature condition is shown in Fig. 3.

Fig. 3

Container headspace CO₂ and O₂ concentrations of pitangas harvested at 83 and 89 ripening stages and stored at 2°C (red solid line), 10°C (green dotted line) and 20°C (blue dash dotted line). A) Evolution of CO₂ as a function of storage time. B) Evolution of O₂ as a function of storage time.

2.3 Data analysis

To determine the kinetics of change of pitanga quality attributes, the following general equation was used: $\pm \frac{dQ}{dt} = k_{q} \times [Q]^{n}$ (1)

Where Q is the quality attribute; t the time; n the reaction order, and kq is the quality change rate constant for the attribute Q. The signs (+) and (–) refer to attributes with increasing and decreasing values over time, respectively. Traditionally, reactions involved in food deterioration were characterized by zero-order or first-order kinetics models [23]. By integrating equation 1, the corresponding kinetic equations are obtained. When n = 0, the reaction is zero order and the corresponding expression is: $Q = Q_{0} \pm k_{q} \times t$ (2)

Where Q₀ is the initial value of the attribute, Q is the value of the attribute at time t, kq is the apparent rate constant (h^–1), and t is the time (h).

When n = 1, the reaction is first order and is represented by the following equation: $Q = Q_{0} \pm e^{\pm k_{q} \times t}$ (3)

Where Q₀ is the initial value of the attribute, Q is the value of the attribute at time t, kq the apparent rate constant (h^–1), and t is the time (h). To evaluate the effect of storage temperature on the quality attributes tested, we used the Q₁₀ factor that is defined by [23]: $Q_{10} = {(\frac{k_{2}}{k_{1}})}^{\frac{10}{(T_{2} - T_{1})}}$ (4)

Where k₂ is the apparent rate constant (h^–1) at T₂, k₁ is the apparent rate constant (h^–1) at T₁ and T is the storage temperature (°C).

Both zero- (Equation 2) and first-order (Equation 3) models were adjusted by multiple regression procedures using R language and environment for statistical computing. When adjusting regression models for each response variable, each storage temperature was analyzed separately. Zero-order models were adjusted using the function lm()from package stats. When heteroscedasticity was detected, the model was adjusted again, but using the function gls(),and the variance structure was modeled with the function varIdent from package nlme. Normality was evaluated using shapiro.test() function and homoscedasticity was tested using both leveneTest() and barlett.test() functions of package stats. Additionally, graphical assessments examined normality and homoscedasticity using Q-Q and studentized residuals vs. predictors plots. Model fit assessment, measures of influence, and residual diagnostics were done using specific functions from package olsrr [24] according to [25].

First-order models were adjusted using nls() from package gravity. Assumptions of the models were tested according to [25]. When zero- and first-order models were significant for a given response variable, the best model was chosen using anova()function, Akaike Information Criterion, and Bayesian Information Criterion. Tables were performed using stargazer() function [26] and graphics were performed using ggplot2() function [27].

Additionally, the correlations between damaged fruits versus weight loss and fruit hardness versus weight loss were adjusted by linear regression analysis using the function lm()from package stats. In adjusting these models, the effect of storage temperature and ripening stage were evaluated as fixed factors. Normality, homoscedasticity, model fit assessment, measures of influence, and residual diagnostics were conducted as previously described. On the other hand, the correlation between total anthocyanin content and hue angle was evaluated using cor.test()function from package stats.

3 Results

3.1 Fruit quality

The rate of weight loss was affected by storage temperature (k_2°C < k_10°C < k_20°C), but with no significant differences between ripening stages (k₈₃ = k₈₉) (Fig. 1A). The increase in storage temperature significantly increased the rate of weight loss in pitanga (Fig. 1A). Thus, weight loss rose by 15.7 times at 10°C with respect to 2°C according to the Q₁₀ parameter (equation 4), but by 2.4 times at 20°C with respect to 10°C. These facts indicate that weight loss of pitanga harvested at both 83 and 89 ripening stages was very sensitive to storage temperature.

Fig. 1

Weight loss and percentage of damaged and decayed pitanga fruits harvested at 83 (empty circles) and 89 (solid circles) ripening stages and stored at 2°C (red solid line), 10°C (green dotted line) and 20°C (blue dot dashed line). A) Experimental and predicted weight loss (WL-%) as a function of storage time (ST –h). The plotted lines correspond to zero-order models. WL_2°C = 0.0021*ST (R² = 0.30. Std. error = 1.57. N = 32). WL_10°C = 0.019*ST (R² = 0.87. Std. error = 1.34. N = 30). WL_20°C = 0.046*ST (R² = 0.93. Std. error = 1.51. N = 31). The dotted line parallel to the x axis corresponds to a weight loss of 1.7%. B) Experimental and predicted percentage of damaged fruits (DF-%) relative to storage time (ST –h). The plotted lines correspond to zero-order models. DF_2°C = 0.022*ST (R² = 0.74. Std. error = 1.14. N = 24). DF_10°C = 0.082*ST (R² = 0.60. Std. error = 17.4. N = 28). DF_20°C = 0.382*ST (R² = 0.54. Std. error = 37.2. N = 27). The dotted line parallel to the x axis corresponds to a proportion of damaged fruits of 10%. C) Experimental and predicted percentage of decayed fruits (DeF-%) relative to storage time (ST –h). Plotted lines correspond to zero-order models. DeF_2°C = 0 for all ST (R² = 0.13. Std. error = 1.14. N = 24). DeF_10°C = 0 for all STs (R² = 0.10. Std. error = 1.4. N = 28). DeF_20°C = 0.135*t (R² = 0.49. Std. error = 17.2. N = 24). The dotted line parallel to the x axis corresponds to a 10% proportion of decayed fruits. Plotted lines for 2°C and 10°C storage temperatures are overlapped.

The rate of damaged fruits during storage was also affected by the storage temperature (k_2°C < k_10°C < k_20°C), and there were no significant differences between ripening stages (k₈₃ = k₈₉) (Fig. 1B).

On the other hand, the presence of decayed fruits was only observed at 20°C (Fig. 1C). No symptoms or signs of decay during storage time were detected in pitanga harvested at both ripening stages and stored at 2°C and 10°C showed (Q₀= 0% and k = 0% h^–1, Fig. 1C).

Figure 2A shows the relationship between weight loss and percentage of damaged fruits for pitangas, irrespective of ripening stages and storage temperatures.

Fig. 2

A) Damaged fruits (DF-%) as a function of weight loss (WL–%) for pitangas harvested at 83 (empty circles) and 89 (solid circles) ripening stages and stored at different temperatures. The plotted line corresponds to the following regression model: DF = 5.74 * WL. (R² = 0.52. Std. error = 10, N = 75). B) Fruit hardness (Fh-shores) as a function of weight loss (WL –%) and ripening stage at harvest for pitangas collected at 83 (empty circles) and 89 (solid circles) ripening stages. Plotted lines correspond to the following regression models: Fh₈₃ = 10.04 - 0.70 * WL for pitangas collected at 83 ripening stage and Fh₈₉ = 7.47 - 0.70 * WL for pitangas collected at 89 ripening stage. For both models, R² = 0.50. Std. error = 1.32. N = 39. Note that since the interaction between weight loss and ripening stage at harvest was not significant, both models have the same slope. The vertical dotted line indicates the threshold of 1.7% in both figures.

According to the regression model (Fig. 2A), for each 1% increase in pitanga weight loss during storage at any temperature, the percentage of damaged fruits increased by approximately 5.7%. Therefore, considering 10% of damaged fruits as a reasonable threshold (Fig. 2A), pitanga could withstand up to 1.7% weight loss. Under these conditions, pitanga could be stored for up to 2 days at 20°C and 4 days at 10°C, regardless of the ripening stage at harvest (Fig. 1B). However, at 2°C pitanga could be stored for at least 20 days without exceeding 10% of damaged fruits (Fig. 1B).

Figure 2B shows the relationship between fruit hardness and percentage of weight loss for pitanga harvested at each ripening stage. An inverse correlation between fruit hardness and weight loss was observed, depending on the ripening stage (Fig. 2B). For each 1% increase in weight loss, fruit hardness dropped 0.7 shores in fruits harvested at both ripening stages (Fig. 2B). These results show that for a given hardness level, weight loss was greater for fruits harvested at stage 83.

The concentration of oxygen and carbon dioxide in the headspace of the containers (Fig. 3) shows that at 20°C there was a moderate increase in the concentration of carbon dioxide (approximately 5 KPa) and an equivalent reduction in oxygen concentration (from 20.8 to 15 KPa). The modified atmosphere generated in this case was probably not sufficient to affect the preservation of Pitangas. At 2 and 10°C, the changes observed in the gas composition were minimal. No passive modified atmosphere was generated under these conditions.

The initial values (Q₀) of fruit hardness, TSS content (Table 2), and color parameters (Fig. 4) were affected by the ripening stages at harvest (Q_0 (83) ≠ Q_0 (89)), as expected. However, the rates of change in those response variables were not significantly different between ripening stages (k₈₃ = k₈₉), except for rate of the chroma changes at 2°C and 10°C storage temperatures (Fig. 4B and E).

Table 2

Predicted initial values (Q₀) and rate of changes (k) in fruit hardness, total soluble solid content, hue angle and total anthocyanin content of Pitangas collected at 83 and 89 ripening stages (according to BBCH scale) and stored at 2, 10 and 20°C

Attribute	Temperature (°C)	Q _0 (83)	Q _0 (89)	k _(83 - 89) ^a	R²	Residual Std. Error	F statistic
Fruit hardness (shores)	2	7.9	6.5	0	0.13	1.8	6.1^**
	10	8.2	6.3	0	0.27	2.7	4.4^**
	20	10.9	8.2	–0.045	0.84	1.2	58.8^***
Total soluble solid content (%)	2	11.1	14.4	0	0.58	1.5	53.3^***
	10	11.4	15.6	0	0.44	1.1	27.8^***
	20	10.9	15.0	–0.016	0.57	2.1	17.5^***
Total anthocyanin content (mg Kg^-1) of pelargonidin-3-glucoside	2	509	2060	0	0.89	150	66.2^***
	10	894	1610	0	0.56	185	12.6^***
	20	1312	1683	0	0.03	275	0.2^ns

Notes: The optimal regression model for all attributes was found to be a zero-order equation of the form Q = Q₀ + k * ST, where ST represents the storage time in h. In this equation, the value of k denotes the rate of change for hardness (in shores h^–1), total soluble content (in % h^–1), and total anthocyanin content (in mg kg^–1 h^–1). ^aThe unique value for k indicates no effect of the ripening stage on the rate of changes of the attributes tested. k_(83 - 89) = 0 indicates no significant change in the parameters (Q₍83) or Q₍89)) until the end of the experiment. ^***Significant at the 1 % level. ^**Significant at the 5 % level. ^nsNot significant.

Fig. 4

Experimental and predicted values for CIE LCh color parameters relative to storage time (ST - h) for Pitangas harvested at 83 (empty circles, red solid line) and 89 (solid circles, blue solid line) ripening stages, and stored at 2°C (A, B, C), 10°C (D, E, F) and 20°C (G, H, I). The plotted lines correspond to zero-order models. A) H₈₃ = 36.5 - 0.069 * ST and H₈₉ = 18.8 - 0.069 * ST (R² = 0.57. Std. error = 9.8. N = 24). B) C₈₃ = 47.1 - 0.056 * ST and C₈₉ = 5.9 + 0.0069 * ST (R² = 0.96. Std. error = 3.2. N = 23). C) L₈₃ = 21.7 + 0.057 * ST and L₈₉ = 9.66 + 0.057 * ST (R² = 0.54. Std. error = 6.4. N = 25). D) H₈₃ = 43.3 - 0.13 * ST and H₈₉ = 28.3 - 0.13 * ST (R² = 0.47. Std. error = 8.9. N = 22). E) C₈₃ = 48.1 - 0.13 * ST and C₈₉ = 5.8 + 0.005 * ST (R² = 0.97. Std. error = 2.8. N = 19). F) L₈₃ = 15.9 + 0.067 * ST and L₈₉ = 6.0 + 0.065 * ST (R² = 0.51. Std. error = 7.3. N = 21). G) H₈₃ = 23.4 + 0* ST and H₈₉ = 10.2 + 0* ST (R² = 0.62. Std. error = 5.3. N = 14). H) C₈₃ = 24.8 - 0.038 * ST and C₈₉ = 6.5 + 0.038 * ST (0.82. Std. error = 4.8. N = 14). I) L₈₃ = 22.5 - 0.065 * ST and L₈₉ = 13.0 - 0.065 * ST (R² = 0.61. Std. error = 4.9. N = 14).

The rate of change in fruit hardness was affected by storage temperature (Table 2). Only pitanga stored at 20°C showed a loss of fruit hardness during storage (k_20°C = –0.045 shores h^–1) (Table 2). No significant change in fruit hardness during storage time was observed at 2°C or 10°C (Table 2).

TSS showed the same pattern as fruit hardness (Table 2). Pitanga harvested at 89 stage had higher TSS content than those harvested at 83 stage (Q_0 (83) < Q_0 (89)). The rates of TSS change were affected by storage temperature regardless of the ripening stage at harvest. Pitanga stored at 20°C showed a loss of TSS content of –0.016% h^–1 during storage, whereas such loss was not observed at the lower storage temperatures (k_2°C = k_10°C = 0% h^–1).

3.2 Fruit color

With respect to hue, pitanga harvested at the 89 ripening stage had a redder hue than those collected at the 83 ripening stage (h_0 (83) > h_0 (89)), as expected (Fig. 4. See also Figs. 1 2 of supplementary material). The rates of change in hue of pitanga stored at 2°C and 10°C were –0.069 h^–1 and –0.080 h^–1, respectively. In both cases, fruits became redder during storage, regardless of the ripening stage at harvest. However, hue did not shown a change over time for pitanga stored at 20°C (k_20°C = 0.0 h^–1) (Fig. 4A, D and G).

The initial chroma values depended on the ripening stage at harvest (C_0 (83) > C_0 (89)) and the rates of chroma change were different during storage at 2° and 10°C depending on the ripening stage (k₈₃ ≠ k₈₉). However, the rates of chroma change were the same (k₈₃ = k₈₉ = 0.038 h^–1) at both ripening stages at 20°C (see Fig. 4B, E and H). As an overall tendency, the chroma of pitanga harvested at 83 ripening stage decreased during storage. On the contrary, the chroma of pitanga harvested at stage 89 did not change significantly during storage at 2°C and 10°C, but varied significantly at 20°C.

The effect of temperature can be evaluated through the Q₁₀ value (Equation 4). This value was 2.8 when it was calculated on the rates of chroma change at 2 and 10°C in pitanga harvested at 83 ripening stage. This result means that the kinetics of decrease in chroma increases 2.8 times when the temperature increases 8°C. The Q₁₀ value was greater for chroma in pitanga harvested at 89 ripening stage (Q₁₀= 7.6).

With respect to lightness, pitangas collected at the 89 ripening stage were darker than those harvested at the 83 ripening stage (Fig. 4C, F, and I). In addition, pitangas stored at 2°C or 10°C became slightly lighter during storage, but those kept at 20°C became darker, regardless of the ripening stage (Fig. 4C, F and I). According to ΔE*, pitangas showed a perceptible change in color as the storage time increased at all temperatures (ΔE * >10).

3.3 Bioactive compounds and antioxidant capacity

The rate of changes in total phenolic content (TPC) and antioxidant capacity (DPPH and FRAP) of pitanga were not affected by the ripening stage, storage time or temperature (k_2°C = k_10°C = k_20°C = 0) (Table 3). The initial total phenolic content (TPC) and antioxidant capacity were similar for both ripening stages at harvest (Q_0 (83) = Q_0 (89)) and remained constant until the end of the experiment at all temperatures tested (Table 3).

Table 3
Predicted initial values (Q₀) and rate of changes (k) in some bioactive compounds and the antioxidant capacity of Pitangas collected at 83 and 89 ripening stages according to BBCH scale and stored at 2, 10 and 20°C

Attribute Temperature (°C) Q _0(83-89) ^a k _(83-89) ^b R² Residual Std. Error F statistic

Total phenolic content (mg kg^–1 gallic acid equivalent) 2 933 0 0.13 28.8 1.54^ns

10 878 0 0.087 27.7 0.86^ns

20 950 0 0.041 25.0 3.48^ns

Vitamin C (mg kg^–1 ascorbic acid) 2 49 0 0.07 2.3 0.67^ns

10 25 0 0.20 2.0 1.53^ns

20 63 –0.17 0.87 0.389 26.30^***

DPPH (mmol kg^–1 Trolox) 2 9.3 0 0.03 0.33 0.31^ns

10 9.5 0 0.10 0.26 1.20^ns

20 9.1 0 0.24 0.30 1.57^ns

FRAP (mmol kg^–1 of Fe^2 +) 2 17.0 0 0.03 0.51 0.31^ns

10 15.9 0 0.003 0.60 0.03^ns

20 23.6 0 0.063 0.30 5.32^ns

Attribute	Temperature (°C)	Q _0(83-89) ^a	k _(83-89) ^b	R²	Residual Std. Error	F statistic
Total phenolic content (mg kg^–1 gallic acid equivalent)	2	933	0	0.13	28.8	1.54^ns
	10	878	0	0.087	27.7	0.86^ns
	20	950	0	0.041	25.0	3.48^ns
Vitamin C (mg kg^–1 ascorbic acid)	2	49	0	0.07	2.3	0.67^ns
	10	25	0	0.20	2.0	1.53^ns
	20	63	–0.17	0.87	0.389	26.30^***
DPPH (mmol kg^–1 Trolox)	2	9.3	0	0.03	0.33	0.31^ns
	10	9.5	0	0.10	0.26	1.20^ns
	20	9.1	0	0.24	0.30	1.57^ns
FRAP (mmol kg^–1 of Fe^2 +)	2	17.0	0	0.03	0.51	0.31^ns
	10	15.9	0	0.003	0.60	0.03^ns
	20	23.6	0	0.063	0.30	5.32^ns

Notes: The optimal regression model for all attributes was found to be a zero-order equation of the form Q = Q_0(83-89) + k_83-89 * ST, where ST represents the storage time in h. In this equation, the value of k denotes the rate of change for total phenolic content, vitamin C (in mg kg^–1 h^–1), FRAP and DPPH (in mmol kg^–1 h^–1). ^aThe unique value for Q indicates no effect of ripening stage on initial values of the attributes tested. ^bThe unique value for k indicates no effect of the ripening stage on the rate of changes of the attributes tested. k_(83 - 89) = 0 indicates no significant change in the parameters (Q₍₈₃₎ or Q₍₈₉₎) until the end of the experiment. ^***Significant at the 1 percent level. ^nsNot significant.

On the contrary, the rate of changes in vitamin C content was affected by storage time only at 20°C (k_20°C = –0.17 mg kg^–1 h^–1), regardless of the ripening stage at harvest (Table 3). This fact demonstrates that refrigerated storage is essential for preserving vitamin C content, due to its sensitivity to high temperatures.

Unlike in TPC and vitamin C, the initial values of total anthocyanin content (TAC) were statistically different between ripening stages (Table 2). Thus, TAC was significantly higher in pitangas harvested at the 89 than in those harvested at the 83 ripening stage (TAC_o (83) < TAC_o (89)) and remained constant during storage (k₍₈₃₎ = k₍₈₉₎ = 0) (Table 2).

To sum up, we found that TAC, TPC and antioxidant capacity remained constant during the storage period, irrespective of the temperature and the ripening stage at harvest. Furthermore, except for TAC, the initial values of TPC, vitamin C and antioxidant capacity were not dependent on the ripening stage at harvest.

4 Discussion

Pitanga has exhibited a long flowering period that leads to the simultaneous existence of fruits at different stages of maturity and flowers in some geographic locations [15, 16]. Under such conditions, it is difficult to pick large quantities of ripe fruits per harvest day. This fact, combined with the short shelf life previously reported, represents a constraint on pitanga commercialization. A larger supply of fully ripe pitanga could be achieved with homogeneous plantations of selected cultivars, as well as an extension of the shelf life of fully ripe pitanga. Until domestication and breeding of cultivars is achieved, a shelf-life extension could have an immediate impact on the adoption of pitanga cultivation or the sustainable exploitation of this wild resource. Our results demonstrate that pitanga can be stored at 2°C for at least 20 d without loss of visual quality, fruit hardness, soluble solid content, bioactive compound content, or antioxidant capacity, regardless of the ripening stage at harvest. Thus, by reducing the storage temperature from 7–10°C to 2°C, a four-fold increase in the shelf life of pitanga compared to previous reports is achieved [12 , 15]. In pitanga stored at 10°C, weight loss was a limiting factor for shelf life, although a container with low water vapor permeability was used. Weight loss had a direct impact on fruit damage (Fig. 1A) and fruit hardness (Fig. 2B).

In turn, weight loss was strongly conditioned by storage temperature, regardless of the ripening stage at harvest (Fig. 2A). Our results agree with previous reports for pitanga from tropical provenances [12 , 15], and for other tropical and temperate berries [28]. Santos et al. [13] found an increase in the rate of weight loss with the rise in storage temperature, and more than 10% of pitangas stored at 10°C showed dehydration damage after the second day of storage. In their experiment, the authors considered that between 16% and 30% of pitangas were unmarketable after the fourth day, coinciding with a weight loss greater than 5%. The “Zill Dark” pitanga cultivar showed a noticeable weight loss (≈20%) during 10 d of storage at 7–10°C [15]. An increment in dehydration damage that turned pitanga “Zill Dark” unmarketable was observed after 5 days of storage, which corresponded to a weight loss between 6% and 8%. Mélo et al. [12] found that fully ripe pitanga showed a weight loss of 3% on the fifth day and 12% on the tenth day of storage at 8°C. After 10 d of storage, the fully ripe pitanga showed 18% of damaged fruits; thus, they concluded that the optimal storage period was 5 days. We found a slower rate of weight loss than Santos et al. [13], Griffits Jr. et al. [15] and Mélo et al. [12], but almost the same shelf life in pitanga stored at 10°C (about 4 days). It is possible that our approach to determining the proportion of damaged fruits was conservative. We classified as “damaged” a pitanga with at least one damage sign, regardless of the severity of that damage. However, as shown in Fig. 2C and 2F of supplementary material, the appearance of pitangas stored at 10°C for almost 10 days was quite acceptable, which coincided with a 4.5–5% of weight loss (Fig. 1A). However, a weight loss of 4–5% led to a 27.8% –37.5% loss of fruit hardness, depending on the ripening stage (Fig. 2B). Hence, weight loss should be carefully monitored. Furthermore, a 5% weight loss produced a loss of visual quality in blueberry and raspberry [29].

We found no differences in the rates of weight loss or percentage of damaged fruits at either ripening stage at harvest (Fig. 1A and B). Our results seem to agree with those obtained by Santos et al. [13], who did not observe clear differences in the rate of weight loss between orange (physiological maturity) and red pitangas (fully ripe) in their experiments. Furthermore, the rate of weight loss was not affected by the ripening stage at harvest in sweet cherry cv. “Sweetheart” [30]. By contrast, Mélo et al. [12] found a higher weight loss in fully ripe red pitangas than in orange pitangas (12% and 2%, respectively). Other tropical berries like acerola showed a higher weight loss when harvested fully ripe than when harvested at physiological maturity and stored at 12°C [28]. The difference in weight loss during refrigerated storage under high humidity conditions was related to the respiration rate for pomegranate fruit cv “Wonderfull” [31]. Our previous results showed that there were no differences in the respiration rate of pitanga harvested at the 83 (mid ripe) or 89 (fully ripe) ripening stage and stored at room temperature [32]. Based on the pattern observed in the evolution of CO₂ atmospheres within containers from both ripening stages at all temperatures (Fig. 3), we hypothesize that the lack of differences in weight loss between ripening stages was due to the possible absence of differences in respiration rates, as stated by Lufu et al. [31]. On the other hand, we found a moderate variability in weight loss at all storage temperatures, which can be attributed to the origin of the samples for the experiment; indeed, all the samples were taken from different plants produced by seed. High intra-population genetic variability was observed in wild populations of pitanga [33], which means that there is very high genetic and phenotypic variability between plants produced by seeds. Furthermore, a high variability in weight loss was observed between different clones (cultivars) in other tropical berries like acerola [28] and in temperate berries like blueberry [34] and haskap berry [35].

On the other hand, pitanga was characterized as a non-climacteric fruit [36]; therefore, no important changes are expected to occur in fruit hardness or TSS content, regardless of the ripening stage at harvest and the storage time within each storage temperature. However, the loss of fruit hardness was strongly influenced by weight loss (Fig. 2B), which in turn was affected by storage temperature (Fig. 1A). The correlation between weight loss and fruit softening was also demonstrated in rabbiteye blueberry (Vaccinum virgatum cv. “Centurion”) [37] and later confirmed in “Bluecrop” and “Sierra” highbush blueberry cultivars (V. corymbosum) [34]. However, since we did not evaluate the evolution of the cell wall components, we cannot rule out the effect of enzymatic degradation of the cell wall on pitanga softening. Consequently, the decrease in fruit hardness of pitanga stored at 20°C –compared to lower temperatures–could be due to the combined effect of a very high rate of weight loss (Fig. 1A) and of the enzymatic cell wall degradation related to the fruit itself and decay (Fig. 1C). Raspberry stored at between 0°C and 15°C showed a noticeable loss of pulp firmness (–30%) after 12 days, regardless of the temperature, which was attributed to enzymatic degradation of the cell walls [38]. However, in raspberries, it was reported a high weight loss (>5%) and fungal decay incidence (>30%) during storage, regardless of the storage temperature [38]. Hence, weight loss rate and fungal decay incidence could be acting as confounding effects on firmness loss in raspberry, as we hypothesized for pitangas stored at 20°C. In addition, the decrease in TSS content observed in pitangas stored at 20°C can be related to deterioration by decay (Fig. 1C) and fermentation (Fig. 3 of the supplemental material). Our results agree with previous works reporting a retention of TSS in non-climacteric berries, such as raspberry [38], sweet cherry [30], blackberry [39], and pitanga [40], regardless of storage temperatures. Our results also agree with reports on other berries that could be classified as climacteric, like acerola [28] and haskap berry [35].

Visible changes in color parameters of berries relative to storage time at a given storage temperature were reported in acerola [28] and raspberries [38], two fruits with opposite ripening physiology. In addition, our results on the effect of the interaction between storage temperature and storage time (Fig. 4) agree with findings for many red raspberry cultivars [41]. Our results also agree with previous reports that showed a direct correlation between anthocyanin content and color enhancement in several berries [42].

Bioactive compounds and antioxidant capacity are both important commercial traits of berries for consumption as fresh or processed products [1, 2]. For some species, the nutritional value of berries depends on the effects of the cultivar, growing location, crop technology, ripening stage at harvest, and postharvest handling [43]. Our research demonstrates that bioactive compound content and antioxidant capacity were not affected by storage temperature or storage time, except for vitamin C (Tables 2 3). Our results on retention of total phenolic content, total anthocyanin content, and antioxidant capacity during storage agree with values reported for strawberry, raspberry, sweet cherry, sour cherry, red currant, blackberries, and other fruits stored at refrigerated or room temperatures [39 , 45]. Pitanga vitamin C content was not affected by storage time under refrigerated conditions (T < 10°C) or ripening stage at harvest (Table 3), as previously reported [12, 40]. Our results also agree with those observed in raspberry [38] and other fruits [44]. However, the absence of HPLC analysis may lead to a limited comprehension of the stability of specific polyphenols during storage, which could hinder the development of effective preservation strategies aimed at maintaining the nutritional and health benefits of pitanga.

The results of our work may have a positive effect on pitanga commercialization by allowing growers to combine fully ripe pitangas harvested on different days and market them for a longer period without loss of visual quality or nutritional value.

5 Conclusions

Pitangas could be stored under refrigeration (2°C) for at least 20 d without substantial loss of quality, fruit hardness, antioxidant capacity, and soluble solid, total phenolic, total anthocyanin, and vitamin C contents. This period represents a four-fold increase in the storage time compared to previous reports that evaluated the refrigerated storage of pitangas at 8–10°C. Thus, our results demonstrate that refrigerated storage below 10°C has more benefits in terms of fruit quality and allows a significant improvement in marketing strategies.

CRediT authorship contribution statement

First author: Investigation, Methodology, Data curation, Writing –original draft, R language. Second author: Writing –review & editing, Project administration, Funding acquisition. Third author: Writing –review & editing, Project administration, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Footnotes

Acknowledgments

The sampling were supported by the Universidad Nacional del Litoral through the Programa de Documentación, Conservación y Valoración de la Flora Nativa. The laboratory determinations were supported by the Agencia Nacional de Promoción Científica y Tecnológica under Grants PICT 2016-0295, PICT 2017-406 and PICT 2017-2265. We thank Prof. J.D. Cortez Latorre for vitamin C determinations.

Data availability

Data will be made available on request.

The supplementary material is available in the electronic version of this article: .

Incidence was defined as the number of pitangas that were visibly damaged/diseased, relative to the total number of pitangas within each container, according to [].

Severity was defined as the area or volume of each pitanga that was visibly damaged/diseased, relative to the total pitanga surface or volume, according to [].

References

Schreckinger

, Lotton

, Lila

, de Mejia

. Berries from South America: A comprehensive review on chemistry, health potential, and commercialization. Journal of Medicinal Food. 2010;13(2):233–46.

Schulz

, Seraglio

SKT

, Brugnerotto

, Gonzaga

, Costa

ACO

, Fett

. Composition and potential health effects of dark-colored underutilized Brazilian fruits –A review. Food Research International. 2020;137:109744.

Vizzotto

, Cabral

, Santos

. 13 - Pitanga (Eugenia uniflora L.) A2 - Yahia, Elhadi M. In: Postharvest Biology and Technology of Tropical and Subtropical Fruits [Internet]. Sawston, Cambridge, UK: Woodhead Publishing; 2011 [cited 2016 Aug 12]. pp. 272-88e. (Woodhead Publishing Series in Food Science, Technology and Nutrition). Available from: https://www-sciencedirect-com.web.bisu.edu.cn/science/article/pii/B9780857090904500135s

Rosano

. Recetario de frutos nativos del Uruguay. 1st ed. Montevideo, Uruguay: Ministerio de Educación y Cultura; 2012, p. 98.

Furlan

, Pochettino

, Hilgert . Frontiers | Management of Fruit Species in Urban Home Gardens of Argentina Atlantic Forest as an Influence for Landscape Domestication | Plant Science. Frontiers in Plant Science [Internet]. 2017 [cited 2020 Feb 14];8. Available from: https://www.frontiersin.org/articles/10.3389/fpls.2017.01690/full

Schmeda-Hirschmann

, Feresin

, Tapia

, Hilgert

, Theoduloz

. Proximate composition and free radical scavenging activity of edible fruits from the Argentinian Yungas. J Sci Food Agric. 2005;85(8):1357–64.

Griffis

JL Jr

, Sams

, Manners

, McDonald

, Radovich

. A progress report on commercialization in the USA of purple-fruited pitanga (Eugenia uniflora L.), an underutilized fruit crop. Acta Horticulturae. 2013;(979):807–14.

Bagetti

, Facco

EMP

, Piccolo

, Hirsch

, Rodriguez-Amaya

, Kobori

, et al. Caracterização físico-quimica e capacidade antioxidante de pitanga (Eugenia uniflora L.). Ciênc Tecnol Aliment Campinas. 2011;31(1):147–54.

Castro

, Rozycki

, Latorre

JDC

, Alesso

, Exner

EDL

, Piagentini

, et al. Variation of Eugenia uniflora L. fruit quality attributes in the south of the Argentine Gran Chaco. Forests, Trees and Livelihoods. 2020;29(2):130–42.

10.

Castro

, Cortez-Latorre

, Cerino

, Gabriel

, Piagentini

, Pirovani

. Compuestos bioactivos y capacidad antioxidante en distintas poblaciones santafesinas de Eugenia uniflora. In: VII Congreso Internacional de Ciencia y Tecnología de Alimentos (CICYTAC 2018) - Libro de Resúmenes. Córdoba, Argentina, 2018, pp. 184-5.

11.

Griffis

Jr , Sams

, Manners

, McDonald

, Radovich

. Pupple-fruited pitanga - antioxidant levels and flavors of mature fruits vary considerably among closely related cultivars. In: Proc 3rd IS on Guava and Other Myrtaceae [Internet]. Petrolina, Brazil: Acta Hort.; 2012 [cited 2018 Apr 4]. pp. 209-15. Available from: https://www.actahort.org/books/959/959_26.htm

12.

Mélo de

E A

, Lima de

VLAG

, Nascimento

PP do

. Temperature and the storage of surinam cherries. ScientiaAgricola. 2000;57(4):629–34.

13.

Santos dos

, Silva de

S M

, Nunes Mendoça

, Cunha Filgueiras

. Storage of suriname cherry under modified atmosphere and refrigeration: II –Quality and posharvest conservation. Revista Brasileira de Fruticultura. 2006;28(1):42–5.

14.

Akinnifesi

, Leakey

RRB

, Ajayi

, Sileshi , Tchoundjeu

, Matakala

, et al., editors. Indigenous fruit trees in the tropics: Domestication, utilization and commercialization. CAB International; 2008, p. 458.

15.

Griffis

JL Jr

, McDonald

, Manners

, Bingham

. Some effects of refrigerated storage on postharvest quality of ripe fruits of the tropical, purple-fruited pitanga (Eugenia uniflora L.). Acta Horticulturae. 2015;(1088):113–8.

16.

Castro

, Abelvey

, Richard

, Gabriel

, Álvarez

, Buyatti

, et al. Estadios fenológicos dedesarrollo en “Ñangapirí” (Eugenia uniflora L.) de acuerdo a la escala BBCH. Revista FAVE - CienciasAgrarias. 2019;17(2):7–21.

17.

Madden

, Hughes

. Sampling for plant disease incidence. Phytopathology^®. 1999;89(11):1088–103.

18.

Piagentini

, Pirovani

. Total phenolics content, antioxidant capacity, physicochemical attributes, and browning susceptibility of different apple cultivars for minimal processing. International Journal of Fruit Science. 2017;17(1):102–16.

19.

Heo

, Lee

. Strawberry and its anthocyanins reduce oxidative stress-induced apoptosis in PC12 cells. J Agr Food Chem. 2005;53:1984–9.

20.

Van de Velde

, Pirovani

, Cámara

, Güemes

, Bernardi

. Optimization and validation of aUV–HPLC method for vitamin C determination in strawberries (Fragaria ananassa Duch.), using experimentaldesigns. Food Anal Method. 2012;5:1097–114.

21.

Villamil-Galindo

, Van de Velde

, Piagentini

. Extracts from strawberry by-products rich in phenolic compounds reduce the activity of apple polyphenol oxidase. LWT. 2020;133:110097.

22.

Piagentini

, Güemes

, Pirovani

. Mesophilic aerobic population of fresh-cut spinach as affected by chemical treatment and type of packaging film. Journal of Food Science. 2003;68(2):602–6.

23.

Piagentini

, Mendez

, Guemes

, Pirovani

. Modeling changes of sensory attributes for individual and mixed fresh-cut leafy vegetables. Postharvest Biology and Technology. 2005;38(3):202–12.

24.

Hebbali

. olsrr: Tools for Building OLS Regression Models. [Internet]. 2018. Available from: https://CRAN.R-project.org/package=olsrr

25.

Amat Rodrigo

. Introducción a la Regresión Lineal Múltiple [Internet]. RPubs by RStudio. 2016 [cited 2022 May 24]. 407 Available from: https://rpubs.com/Joaquin_AR/226291

26.

Hlavac

. Stargazer: Well-Formatted Regression and Summary Statistics Tables. [Internet]. 2018. Available from: https://CRAN.R-project.org/package=stargazer

27.

Wickham

. ggplot2: Elegant Graphics for Data Analysis [Internet]. Springer-Verlag New York,; 2016. Available from: https://ggplot2.tidyverse.org

28.

Ribeiro

, de Freitas

. Maturity stage at harvest and storage temperature to maintain postharvest quality of acerola fruit. Scientia Horticulturae. 2020;260:108901.

29.

Nunes

, Emond

. Relationship between weight loss and visual quality of fruits and vegetables. Proc Fla State Hort Soc. 2007;120:235–45.

30.

Horvitz

, Yommi

, Camelo

, Godoy

. Efecto del estado de madurez y atmósferas modificadas sobre la calidad de cerezas cv. Sweetheart. Revista de la Facultad de Ciencias Agrarias. 2004;36(2):39–48.

31.

Lufu

, Ambaw

, Opara

. The contribution of transpiration and respiration processes in the mass loss of pomegranate fruit (cv. Wonderful). Postharvest Biology and Technology. 2019;157:110982.

32.

Castro

, Cortez Latorre

, Piagentini

, Pirovani

. Patrón respiratorio de frutos de“Ñangapirí”: Resultados preliminares. In: Libro de Resúmenes del II Congreso Argentino deBiología y Tecnología Poscosecha. Santiago del Estero (Argentina).: UNSE; 2019, p. 78.

33.

Salgueiro

, Felix

, Caldas

, Margis-Pinheiro

, Margis

. Even population differentiation for maternal and biparental gene markers in Eugenia uniflora, a widely distributed species from the Brazilian coastal Atlantic rain forest. Diversity Distrib. 2004;10(3):201–10.

34.

Liu

, Wang

, Shu

, Liang

, Fan

, Sun

. Changes in fruit firmness, quality traits and cell wall constituents of two highbush blueberries (Vaccinium corymbosum L.) during postharvest cold storage. Scientia Horticulturae. 2019;246:557–62.

35.

Leisso

, Jarrett

, Richter

, Miller

. Fresh haskap berry postharvest quality characteristics and storage life. Can J Plant Sci. 2021;101(6):1051–63.

36.

Akamine

, Goo

. Respiration and ethylene production in fruits of species and cultivars of Psidium and species of Eugenia. J Amer Soc Hort Sci. 1979;104(5):632–5.

37.

Paniagua

, East

, Hindmarsh

, Heyes

. Moisture loss is the major cause of firmness change during postharvest storage of blueberry. Postharvest Biology and Technology. 2013;79:13–9.

38.

Tezotto-Uliana

, Berno

, Gómez

ACS

. Evaluación del comportamiento poscosecha de Frambuesas en diferentes condiciones de almacenamiento refrigerado. Revista Iberoamericana de Tecnología Postcosecha. 2018;19(1):12.

39.

, Frei

, Kennedy

, Zhao

. Effects of refrigerated storage and processing technologies on the bioactive compounds and antioxidant capacities of ‘Marion’ and ‘Evergreen’ blackberries. LWT - Food Science and Technology. 2010;43(8):1253–64.

40.

Santos dos

, Silva de

S M

, Alves

. Storage of suriname cherry under modified atmosphere and refrigeration: I-postharvest chemical changes. Revista Brasileira de Fruticultura. 2006;28(1):36–41.

41.

Robbins

, Moore

. Color change in fresh red raspberry fruit stored at 0, 4. 5, or 20C. HortScience. 1990;25(12):1623–4.

42.

Olivares

, Contreras

, Muñoz

, Rivera

, González-Agüero

, Retamales

, et al. Relationship amongcolor development, anthocyanin and pigment-related gene expression in ‘Crimson Seedless’ grapes treated withabscisic acid and sucrose. Plant Physiology and Biochemistry. 2017;115:286–97.

43.

Talcott

. Chemical components of berry fruits. In: Zhao Y, editor. Berry fruit Value-added products for health promotion. 1st ed. Boca Raton, FL, USA: CRS Press. Taylor & Francis Group; 2007, pp. 73-104. (Food science and technology).

44.

Kevers

, Falkowski

, Tabart

, Defraigne

, Dommes

, Pincemail

. Evolution of antioxidant capacity during storage of selected fruits and vegetables. J Agric Food Chem. 2007;55(21):8596–603.

45.

Piljac-Zegarac

, Samec

. Antioxidant stability of small fruits in postharvest storage at room and refrigerator temperatures. Food Research International. 2011;44(1):345–50.

		Sampling time points
Ripening	Storage	t ₀	t ₁	t ₂	t ₃	t ₄	t ₅	t ₆	TRT^b	TRRS^c
stage	temperature (°C)	Number of replicates^a
83	2	3	3	3	3	3	3	3	21	63
	10	3	3	3	3	3	3	3	21
	20	3	3	3	3	3	3	3	21
89	2	3	3	3	3	3	3	3	21	63
	10	3	3	3	3	3	3	3	21
	20	3	3	3	3	3	3	3	21
TRSP^d		18	18	18	18	18	18	18		126