Impaginato 27 Adv. Hort. Sci., 2022 36(1): 27­36 DOI: 10.36253/ahsc­10914 Postharvest performance interpretation and storage temperature optimization in some newly introduced melon hybrids M. Alabboud, S. Kalantari (*), F. Soltani Department of Horticultural Sciences, University College of Agriculture and Natural Resources, University of Tehran, Imam‐Zadeh Hasan Blvd., 31587‐77871 Karaj, Iran. Key words: Cucumis melo, principal component analysis, storability, surface response method. Abstract: Temperature is a key factor in melon cold storage. Thus, optimizing storage temperature is an important goal in postharvest research. In this exper­ iment, postharvest attributes of four inbred lines and five derivative hybrids were investigated under three storage temperatures (1, 4, and 12°C). Melon fruit were evaluated for their main characteristics directly after harvest and postharvest changes were monitored through cold storage period. Cluster anal­ ysis results showed that most of the hybrids clustered with their maternal par­ ents illustrating the significant role of cytoplasmic inheritance for the studied traits. Similarly, principal component analysis clustered the studied types into three clusters according to their average postharvest behaviour. The best postharvest performance belonged to inodorus and cantalupensis netted melon with their intercrossing breeds. While the dudaim inbred line and its hybrid scored the highest postharvest changes. Response surface analysis showed that 1.8°C was the optimum storage temperature for inodorus and can­ talupensis clusters, while 5.1°C was the best storage temperature of dudaim cluster. The results of the current study are similar to previous reports for opti­ mum storage temperature in similar melon types. 1. Introduction Melons (Cucumis melo) are among the most important fruit crop worldwide with a yearly production of more than 27 million tons (FAO, 2019). This fruit is of an exclusive importance in Mediterranean and Eastern Asia regions (Garcia­Mas et al., 2012). Melons are usually perish­ able with short shelf life and low storability (Fukuta et al., 2006; Briones et al., 2012). Thus, these fruits do not maintain their marketability for longer than 10 days in room temperature conditions. Therefore, introduc­ ing new melon types with enhanced storability, longer shelf life, and bet­ ter shipping potentials is among the main aims of melon breeders. Additionally, investigating postharvest behaviour and optimizing storage (*) Corresponding author: kalantaris@ut.ac.ir Citation: ALABBOUD M., KALANTARI S., SOLTANI F., 2022 ­ Postharvest performance interpretation and sto‐ rage temperature optimization in some newly introduced melon hybrids. ­ Adv. Hort. Sci., 36(1): 27­36. Copyright: © 2022 Alabboud M., Kalantari S., Soltani F. This is an open access, peer reviewed article publi­ shed by Firenze University Press (http://www.fupress.net/index.php/ahs/) and distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Competing Interests: The authors declare no competing interests. Received for publication 23 May 2021 Accepted for publication 14 January 2022 AHS Advances in Horticultural Science https://doi.org/10.36253/ahsc-10914 http://www.fupress.net/index.php/ahs/ http://creativecommons.org/licenses/by/4.0/ http://creativecommons.org/licenses/by/4.0/ http://creativecommons.org/licenses/by/4.0/ Adv. Hort. Sci., 2022 36(1): 27­36 28 variables of this fruit is of high importance. Persian melons (Cucumis melo var. inodorus) are reported to be a valuable material in melon breeding for postharvest purposes. These types of melons have firmer mesocarps compared to cantalupensis types and thus, endure shipping and handling better (Welbaum, 2015). For instance, including the line ‘Cm­ UTKH’ in breeding programs by crossing with other cantalupensis melons reportedly introduced new types with promising postharvest attributes (Alabboud et al., 2020; Shajari et al., 2021. Therefore, investigat­ ing these newly introduced intergroup hybrids for their postharvest potentials under different storage temperature might be of high importance. Weight losses, firmness losses, soluble solids fluc­ tuations and colour changes are the main posthar­ vest attributes in stored melons. Furthermore, tem­ perature is the main factor affecting these attributes throughout cold storage (De Arruda et al., 2003; Yang et al., 2003; Žnidarĉiĉ et al., 2013; Hatami et al., 2019). Therefore, minimizing postharvest changes through optimized storage temperature is the ulti­ mate goal in cold storage research. However, moni­ toring these attributes in large populations can be a laborious quest especially when investigating opti­ mum storage temperature. Thus, a more appropriate experimental method should be used in order to minimize experiment size and maintain an accept­ able accuracy. For this purpose, principal component analysis (PCA) is considered a useful tool to explain diversity within a studied group of samples with mini­ mum loss of information. Many researches imple­ mented PCA for better understanding of melon fruit characteristics (Obando et al., 2008; Maietti et al., 2012; Farcuh et al., 2020).On the other hand, response surface method (RSM) is a direct applica­ tion of regression theory to optimize an experimental input (Barton, 2013) with fewer number of experi­ ments (Wani et al., 2012) which fits the goal of postharvest studies in melon. Therefore, the main goals of this study were to investigate postharvest attributes of some newly introduced melon hybrids under different storage temperatures, and to find the optimum storage tem­ perature for these hybrids through the application of response surface regression optimization. 2. Materials and Methods Plant material The experiment was carried out for the seasons of 2019 and 2020 at the Research Station of the Department of Horticultural Sciences (University of Tehran) in Karaj, Iran. Four melon inbred lines (Table 1) with five of the intercrossing hybrids were used in this experiment (Fig. 1). The seed of inbred lines and hybrids were planted in seedling trays in greenhouse and transformed to the field three weeks after germi­ nation. The seedlings were planted in rows with 1.5 m distance between rows and 0.8 m between plants on each row. Ripening was determined by the devel­ opment of ½ to ¾ of abscission layer in types that tend to develop an abscission layer and after 42 days post­anthesis in types that lacks the ability to devel­ op an abscission layer. The fruit were harvested at ripening and transferred to the cold storage facility at the Department of Horticultural Sciences. The har­ vested fruit were stored under three different tem­ perature (1, 5, and 12°C) and a relative humidity of 90­95% for a period of one month. Fruit attributes monitoring The main fruit characteristics were monitored Table 1 ­ The morphological characterizations of the inbred lines used for the crosses Line name Group and origin Climacteric behaviour Fruit shape Rind Mesocarp Volatiles Origin Cm­UTKH Inodorus, Non­climacteric Oval elongated Yellow­Shallow netted Greenish white Odourless Khorasan (South­eastern part of Iran) Cm­UTA Cantaloupensis x Inodorus derived inbred line Non­climacteric Oval­Round Dark yellow ­ Without net White Odorous (Medium) Abadan (Southern part of Iran) Cm­UTJ Cantaloupensis Non­climacteric Ova ­Round Green ­ Netted Green Odourless Alborz (North­western part of Iran) Cm­UTZ Dudaim Climacteric Round flatted Dark yellow with orange areas ­ Without net White Odorous (High) Kerman (South­eastern part of Iran) Alabboud et al. ‐ Storage temperature optimization in newly introduced melon types 29 directly after harvest and throughout storage period in 10 days intervals (i.e. days 0, 10, 20, and 30). Fruit weight was measured using an electronic scale. TSS% in the juice of extracted from a longitudinal section of the fruit was measured using a handheld refrac­ tometer. Mesocarp firmness was measured using a handheld penetrometer equipped with an 8­mm tip. The firmness was expressed as an average of three reads along the longitudinal surface of the fruit after the removal of the rind. Rind colour was expressed in CIE­lab scale as lightness (L*), chroma (C*) and hue angle (h°) using a chromameter (Konica Minolta, Japan). The changes in fruit attributes throughout storage were presented as percentages and were calculated using the following formulas: Weight loss (%) = (FW0 ­ FWA)/FW0 (1) Firmness loss (%) = (MF0 ­ MFA)/ MF0 (2) Where FW0 and FWA are fruit weight (g) at har­ vest and at analysis time, respectively; while MF0 and MFA are mesocarp firmness (Kg cm­2) at harvest and at analysis time, respectively. Colour changes were calculated using colour dif­ ference formulas according to Sharma et al. (2005) eq. 3 and using the same system described in Mohi­ Alden et al. (2021). ΔC = [(L* ­ L0*)2 + (a* ­ a0*)2 + (b* ­ b0*)2]1⁄2 (3) where ΔC is colour changes during cold storage L0*, a0* and b0* are the values of colour parameters at harvest, and L*, a* and b* are the values of colour parameters after cold storage. Data analysis The two­factor experiment (evaluation date and storage temperature) was carried out in a split plot design. Fruit attributes mean at harvest were com­ pared using Fisher’s LSD test (P<0.05) with Minitab 19 software. Furthermore, fruit attributes were clus­ tered using Ward’s method on NCSS 12. The averages postharvest measurement for the studied lines and breeds were analysed using principal component analysis (PCA). The best Response surface method (RSM) was used to calculate the optimized storage temperature for each PCA cluster. 3. Results Fruit characteristics and heterosis analysis The main fruit characteristics of parents and F1 progeny were investigated (Table 2). The result showed that the parent ‘Cm­UTKH’ recorded the highest fruit length and fruit shape values and, the parent ‘Cm­UTA’ scored the highest results in terms of fruit diameter, mesocarp thickness, and cavity diameter. The highest firmness results were observed in ‘Cm­UTJ’ parent with 11.67 Kg cm­2 fol­ lowed by its hybrid ‘Cm­UTJKH’ (10.98 Kg cm­2), with no significant differences between the parent and the hybrid. The highest fruit weight, and TSS% were observed in the hybrids ‘Cm­UTKHA’ and ‘Cm­UTJKH’ with 2033.91 g and 15.3%, respectively. As for colour attributes, the highest L* and hue values were recorded in the parent ‘Cm­UTA’, while the highest chroma values were observed in the hybrid ‘Cm­ UTJKH’ and the reciprocal ‘Cm­UTKHJ’. Moderately low heterosis results were recorded for fruit diameter, cavity diameter, firmness, TSS, and colour attributes. However, F1 progeny expedited higher heterosis values levels for fruit length (33.17%), fruit shape (22.71%), mesocarp thickness (17.48%), and weight (29.35%). Cluster analysis of fruit attributes Ward linkage cluster analysis of fruit attributes at day 0 can be seen in figure 2. The studied genotypes were clustered in three clusters the first cluster con­ tained the hybrid ‘Cm­UTZKH’ with its maternal par­ ent. The second cluster contained ‘Cm­UTJKH’ with Fig. 1 ­ The studied melon inbred lines (P) and their hybrids (F1). The first part in each hybrid name represents the mater­ nal parent while the second represent the male parent. Breed Maternal parent Paternal parent Cm­UTKHA Cm­UTKH Cm­UTA Cm­UTAKH Cm­UTA Cm­UTKH Cm­UTKHJ Cm­UTKH Cm­UTJ Cm­UTJKH Cm­UTKH Cm­UTJ Cm­UTZKH Cm­UTZ Cm­UTKH Adv. Hort. Sci., 2022 36(1): 27­36 30 its maternal parent. While the final cluster constitut­ ed of the ‘Cm­UTKHA’, ‘Cm­UTKHJ’, and their mater­ nal parent in addition to the hybrid ‘Cm­UTAKH’. On the other hand, ‘Cm­UTA’ was not clustered with any of the studied types. As for the studied fruit traits, two major clusters were noticed. The first contained TSS, chroma, firm­ ness, and lightness (L). The second cluster contained the rest of the studied attributes (fruit length, fruit shape, fruit diameter, mesocarp thickness, weight, and cavity diameter), while hue was not clustered with any other attribute. Principal component analysis results The average postharvest performance data under three different storage temperature was subjected to principal component analysis (PCA) (Fig. 3). Scree dis­ tribution of factors showed that the first two factors were accounted for 68.7% of the total variance with 49.3% for the first factor (PC1) and 19.4% for the sec­ ond (PC2). Additionally, the accumulated effect of the first three factors was accounted for 79.1% of the total variance among the studied types (Table 3). The most significant positive contributors to the PC1 were chroma, TSS, firmness, and colour changing index, while hue and firmness loss were the most sig­ nificant negative contributors to this factor. On the other hand, PC2 was substantially positively influ­ enced by fruit weight. As for genotype clustering according to biplot of postharvest performance, three major clusters were noticed. The green cluster located on the positive Types with similar letters in each column have no significant differences for the monitored trait in that column (P<0.05). h% represents heterosis in each trait compared to med­parents. Table 2 ­ Fruit characteristics for the studied melon types directly after harvest (Day 0 observation). FL (Fruit length), FD (Fruit diameter, FS (Fruit shape), MT (Mesocarp thickness), CD (Cavity diameter), W (weight), F (Firmness), TSS (Total soluble solids), L* (Lightness), H (Hue), C (Chroma) Type FL (cm) FD (cm) FS MT (cm) CD (cm) W (g) F (Kg cm­2) TSS (%) L* H C Cm­UTKH 27.77 a 11.33 cd 2.50 a 2.65 cd 6.09 b 1717.42 b 4.77 e 11.70 b 71.15 d 69.72 d 62.35 c Cm­UTA 13.81 e 14.82 a 0.93 f 3.77 a 7.14 a 1520.51 c 9.84 b 6.33 e 78.24 a 78.77 a 60.37 d Cm­UTJ 12.88 f 11.94 c 1.09 e 3.01 b 5.88 bc 1026.36 d 11.67 a 11.65 b 72.81 cd 22.06 e 65.21 b Cm­UTD 6.21 g 7.57 e 0.80 f 1.05 e 5.44 c 201.84 f 5.63 de 6.67 e 68.72 e 70.99 cd 44.45 f Cm­UTAKH 21.65 c 13.44 b 1.61 c 3.66 a 6.30 b 1738.41 b 7.46 c 8.70 d 74.23 c 75.76 b 60.13 d Cm­UTKHA 22.29 c 13.15 b 1.69 c 3.54 a 5.98 bc 2033.91 a 6.51 cd 10.60 c 70.92 d 71.69 c 60.12 d Cm­UTJKH 16.71 d 10.64 d 1.59 c 2.92 bc 4.82 d 917.40 de 10.98 ab 15.30 a 76.19 b 17.17 f 75.38 a Cm­UTKHJ 25.98 b 13.79 b 1.95 b 2.78 bcd 6.97 a 1771.08 b 5.20 e 10.33 c 73.24 c 12.71 g 74.49 a Cm­UTZKH 14.36 e 11.12 cd 1.32 d 2.49 d 6.16 b 760.31 e 7.38 c 6.80 e 72.74 cd 74.36 b 55.36 e h% 33.17 8.87 22.71 17.48 ­1.49 29.35 ­5.91 13.85 1.01 ­16.64 12.05 Fig. 2 ­ Ward’s linkage cluster analysis for the studied melon types and their fruit characteristics. P1, P2, P3, and P4 represent the studied lines Cm­UTKH, Cm­UTA, Cm­UTJ, and Cm­UTZ, respectively while two letters genotypes represent F1 direct and reciprocal breeds. FL (Fruit length), FD (Fruit diameter, FS (Fruit shape), MT (Mesocarp thickness), CD (Cavity diameter), W (weight), F (Firmness), TSS (Total soluble solids), L* (Lightness), H (Hue), C (Chroma). The colors of the heat map represent the normalized values of raw data to Z score. Thus, red colors or yellow colors represent values higher or lower than the average value for each studied characteristic, respectively. Alabboud et al. ‐ Storage temperature optimization in newly introduced melon types 31 quadrates of PC1 and contained ‘Cm­UTJ’ parent with its hybrid ‘Cm­UTJKH’ and reciprocal ‘Cm­UTKHJ’. The blue cluster located on the positive PC2 and negative PC2 quadrate and was constituted of the parents ‘Cm­UTKH’ and ‘Cm­UTA’ with their hybrids ‘Cm­ UTKHA’ and ‘Cm­UTAKH’. The last cluster was located on the quadrate determined by negative PC1 and PC2 and contained the parent ‘Cm­UTZ’ and its hybrid ‘Cm­UTZKH’ (Fig. 3). Response surface analysis throughout storage period Contour plotting was used to demonstrate fruit postharvest performance during cold storage period can be seen in figure 4 and figure 5. The most pro­ nounced weight loss was observed in the parent ‘Cm­ UTZ’ under 12°C and after 20 days in in cold storage with more than 30% weight loss. On the other hand, the highest firmness losses were recorded under 12°C and after 30 days in cold storage with the line ‘Cm­UTA’ and its hybrids ‘Cm­UTAKH’ and ‘Cm­ UTKHA’ with more than 80% firmness loss in the par­ ent and more than 75% firmness loss in both hybrids. Fig. 3 ­ Principal component analysis for the studied melon. Loading factors represent the average performance of the studied types calculated as the mean for each loa­ ding variable of each type during storage period (the ave­ rage of 0, 10, 20, and 30 days in cold storage). W (Weight), WL (Weight loss), F (Firmness), FL (Firmness loss). TSS (Total soluble solids), L (Lightness), H (Hue), C (Chroma), CC (Color changes). Table 3 ­ Principal component analysis unrotated factor load­ ings and communalities Variable Factor1 Factor2 Factor3 Weight 0.17 0.901 0.141 Weight loss ­0.594 ­0.433 ­0.418 Firmness 0.754 ­0.328 ­0.464 Firmness loss ­0.705 0.354 ­0.139 TSS 0.868 ­0.034 0.131 Lightmess 0.51 0.484 ­0.661 Hue ­0.852 0.347 ­0.108 Chroma 0.911 0.334 ­0.02 Colour changes 0.651 ­0.215 0.2 % Var 49.3 19.4 10.4 Accumulated 49.3 68.7 79.1 Fig. 4 ­ Contour plots of postharvest changes during cold storage period for the studied inbred lines. Each column of contour plots represents an inbred line (From left to right ‘Cm­UTKH’, ‘Cm­UTA’, ‘Cm­UTJ’, and ‘Cm­UTZ’, respectively). Each row of contour plots represents a postharvest attribute (From top to bottom Weight loss, TSS, Firmness loss, and Color changes, respectively). A significant difference between contour levels is present when the difference between contour levels are larger than 1.32, 1.67, 5.97, and 1.02 for Weight loss, TSS, Firmness loss, and Color changes, contours respectively (P<0.05). 32 Adv. Hort. Sci., 2022 36(1): 27­36 There were no significant colour changes during the first 10 days of storage under any of storage tem­ perature in any of the studied lines except for ‘Cm­ UTZ’. However, all the studied hybrids showed signifi­ cant colour changes after 10 days in cold storage and under all temperature except for ‘Cm­UTKHA’ which showed significant colour changes only after 20 days in storage and under higher storage temperature (>5°C). colour changes were steep and more pro­ nounced in the hybrids ‘Cm­UTKHJ’ and ‘Cm­UTJKH’. There was no defined trend in TSS changes throughout storage period in the studied genotypes. A general decreasing trend was noticed in the lines ‘Cm­UTJ’ and ‘Cm­UTZ’ and similar decreasing overall trends were observed in all the studied hybrids. On the other hand, minimal TSS fluctuations were observed in the line ‘Cm­UTA’. Furthermore, the line ‘Cm­UTKH’ experienced significant decrease in TSS during the first 10 days of storage under 12°C and then remained constant for the rest of storage dura­ tion. However, the line ‘Cm­UTKH’ kept an almost constant TSS under 4°C throughout storage period. Response surface model was used in order to cal­ culate the optimized storage temperature of the clus­ ters obtained via PCA analysis. The monitoring date and storage temperature were set to be the continu­ ous factors of the model, while weight loss (WL), firmness loss (FL), colour changes (CC), and TSS were set as responses. Stepwise method was used to omit unnecessary sources for better fit of each model (Table 4). Then, the optimized storage temperature for each cluster was calculated by setting optimiza­ tion goals of WL, FL, and CC to a minimum and TSS to a maximum. The results showed that the optimized temperatures were 1.8 for the green and blue clus­ ters, and 5.1 for the red cluster. 4. Discussion and Conclusions Fruit characteristics Cluster analysis showed that fruit shape was more correlated with fruit length than diameter even though these three attributes were located in the same cluster (Fig. 2). This relation was also observ­ able in (Table 2) since the line ‘Cm­UTKH’ recorded the highest FL and FS values, while the line ‘Cm­UTZ’ recoded the lowest. In fact, both these traits had also a high heterosis values with 33.17% and 22.71% for FL and FS, respectively, which is similar to Monforte et al. (2005) reports of significant correlation between fruit shape and length. The same research suggested that the heterosis in fruit shape is a func­ tion of fruit elongation. On the other hand, the segregation fruit diameter, mesocarp thickness, fruit weight, and cavity diameter might indicate that the more rounded the melon is, the thicker the mesocarp and thus, the heavier fruit will be. The relatively high heterosis in mesocarp thickness and fruit weight (17.48% and 29.35%, Fig. 5 ­ Contour plots of postharvest changes during cold storage period for the studied hybrids. Each column of contour plots repre­ sents a hybrid (From left to right ‘Cm­UTKHA’, ‘Cm­UTAKH’, ‘Cm­UTKHJ’, ‘Cm­UTJKH’, and ‘Cm­UTZKH’, respectively). Each row of contour plots represents a postharvest attribute (From top to bottom Weight loss, TSS, Firmness loss, and Color changes, respec­ tively). A significant difference between contour levels is present when the difference between contour levels are larger than 1.32, 1.67, 5.97, and 1.02 for Weight loss, TSS, Firmness loss, and Color changes, contours respectively (P<0.05). Alabboud et al. ‐ Storage temperature optimization in newly introduced melon types 33 respectively) and the low heterosis in cavity diameter might in favour of the previous debate. The current results showed that the hybrids ‘Cm­UTKHA’, ‘Cm­ UTAKH’, and ‘Cm­UTKHJ’ scored positive weight het­ erosis values compared to best parents, while ‘Cm­ UTJKH’ and ‘Cm­UTZKH’ that scored negative weight heterosis compared to mid parent results. This result is similar to Monforte et al. (2005) and Mohammadi et al. (2014) reports of positive to negative heterosis for melon fruit weight, while the general positive het­ erosis for fruit weight is similar to this reported by Feyzian et al. (2009). Furthermore, the results refer to the importance of the line ‘Cm­UTKH’ in breeding programs to increase average fruit weight. It was noticed that each hybrid clustered with its maternal parent, except for the hybrid ‘Cm­UTAKH’ which was however the closest to its maternal parent ‘Cm­UTA’ although they were not clustered together (Fig. 2). This observation refers to the high maternal impact of the used lines and the cytoplasmic inheri­ tance for the studied traits. The cytoplasmic inheri­ tance and the high similarities between hybrids and maternal parents was previously reported for fruit length, fruit diameter, fruit shape, average fruit weight, mesocarp thickness, cavity diameter, TSS, and firmness (Y. Hassan Al­Hamdany, 2013; Shajari et al., 2021) while other study reported a non­signifi­ cant reciprocal effect for fruit weight (Feyzian et al., 2009). Postharvest performance In the current work, PCA clustered the studied genotypes according to their average postharvest performance into three clusters with each hybrid clustered with its maternal parent. The high similari­ ties in postharvest behaviour between hybrids and their maternal parent can be also attributed to the significant reciprocal effects for the studied traits, which were previously reported in similar population (Alabboud et al., 2020). By comparing loading variables to samples distri­ bution, it can be concluded that the ‘Cm­UTJ’ line with its hybrid ‘Cm­UTJKH’ and reciprocal ‘Cm­UTKHJ’ had lower weight loss and firmness loss during cold storage compared to ‘Cm­UTZ’ and its hybrid ‘Cm­ UTZKH’. Thus, the former samples had a better postharvest performance compared later (Fig. 3) which can be also noticed in (Fig. 4) and (Fig. 5). colour changes in the line ‘Cm­UTJ’ and its hybrids are attributed to chroma increase which indicates more saturated colour during cold storage. According to PCA biplot (Fig. 3), the samples with higher TSS content were characterized by lower weight and firmness losses during cold storage; therefore, the types with higher TSS content per­ formed better throughout cold storage. Previously a negative correlation was reported between sucrose content in blueberry and Polygalacturonase activity (Wang et al., 2020). Additionally, postharvest shelf Model Response surface regression Sources Optimization goal Optimized storage temperatureLinear Square interaction Green cluster WL D ­ ­ Minimize 1.8 FL D, T T x T D x T Minimize 1.8 CC T T x T ­ Minimize 1.8 TSS T ­ ­ Maximize 1.8 Blue cluster WL D, T D x D, T x T D x T Minimize 1.8 FL D, T T x T ­ Minimize 1.8 CC D, T T x T D x T Minimize 1.8 TSS D, T T x T ­ Maximize 1.8 Red cluster WL D, T D x D ­ Minimize 5.1 FL D, T T x T ­ Minimize 5.1 CC D, T ­ D x T Minimize 5.1 TSS T T x T ­ Maximize 5.1 WL= Weight loss, FL= Firmness loss), CC= Colour changes, TSS =Total soluble solids, D= Test date, T= Storage temperature. Cluster colours (green, blue, and red) refer to the principal component analysis (PCA) clustering shown in figure 3. Table 4 ­ Storage temperature optimization results for the clusters obtained by principal component analysis using response surface method Adv. Hort. Sci., 2022 36(1): 27­36 34 life of roquette leaves was extended in relation to higher sucrose content (Clarkson et al., 2005). Furthermore, it was reported that sucrose can act as a protective signal in fresh cut melon against wound­ ing signal (Wu et al., 2020). Considering the fact that sucrose is the major component in melons’ TSS con­ tent (Burger et al., 2000), it can be assumed that the current observed correlation between TSS content and enhanced postharvest performance is related to a higher sucrose level. However, this relation should be investigated thoroughly. Storage temperature optimization The decrease in fruit weight and firmness in addi­ tion to colour changes are among the usual observa­ tions throughout cold storage of fruit and vegetables. Weight loss is mainly attributed to water evaporation during storage period (lal Basediya et al., 2013) which is supposed to increase by increasing storage temper­ ature as can be seen in all the studied types (Fig. 4 and 5). The differences in weight loss between differ­ ent types during storage phase should correlated with fruit characteristics. For instance, the line ‘Cm­UTJs’ and its related hybrids are characterized with thicker rind and suberized periderm tissues which can effi­ ciently block water evaporation (Nishizawa et al., 2017) especially when considering the non­climacteric behaviour of the aforementioned types compared to other climacteric types such as ‘Cm­UTZ’. Firmness loss is the result of cell wall deterioration due to vari­ ous enzymes activity throughout storage (Qi et al., 2011; Wu et al., 2020). The current results showed lower and slower firmness loss under lower storage temperature, which is usually observed in stored melon (Hatami et al., 2019) due to the slower overall metabolism and the lower enzymatic activity. The currently observed fluctuation in TSS during cold storage phase was observed in full melon fruit (Hatami et al., 2019) and in fresh cut slices (Chong et al., 2015). In fact, most of the observations where TSS fluctuated were under lower storage tempera­ ture (1°C) (Fig. 4 and 5). Therefore, the most con­ vincing explanation is that fruit metabolism under lower temperature was lower than that of higher storage temperature; therefore, water evaporation from fruit resulted in higher concentrations of solu­ ble solids. Response surface method (RSM) is a combination of experimental design, analysis of regression and stochastic response optimization (Barton, 2013). The main advantage of RSM analysis compared to other optimization methods is that fewer number of exper­ iments are needed to monitor the interaction of the independent variables on the response (Wani et al., 2012). PCA clustering results were used to produce response surface models. The postharvest related data of each cluster members were the inputs while the target was to find the storage temperature which minimizes weight losses, firmness losses and color changes while maintaining the highest TSS possible. RSM analysis showed that the optimized storage temperature for both green (netted cantalupensis melon) and blue clusters (inodorus) members was 1.8°C. This result is similar to previous recommended storage temperature for various inodorus group culti­ vars (Yang et al., 2003) where a storage temperature of 3°C was recommended, and netted cantalupensis group cultivars (De Arruda et al., 2003; Žnidarĉiĉ et al., 2013) where a storage temperature of 2­3°C resulted in the best storage outcomes. On the other hand, the optimum storage temperature for the red cluster (dudaim) was 5.1°C which is similar to the previous observations in dudaim group including the line ‘Cm­UTZ’ where 5°C resulted in better colour preservation and less weight, firmness, and TSS loss­ es during cold storage period (Hatami et al., 2019). Therefore, RSM can be considered a feasible method of storage temperature optimization especially in newly established populations and hybrids. The current study showed that the hybrid melon types of inodorus x cantalupensis and dudaim x inodorus crosses had more similarities with their maternal parent. These similarities were obvious not only on fruit characteristics directly after harvest, but also in the later postharvest performance in cold storage. This result illustrated the importance of cytoplasmic inheritance in melon. The best posthar­ vest performance was that of crosses between inodorus x netted cantaloupe, while the worst per­ formance was related to ‘Cm­UTZ’ line of the group dudaim and its hybrid with the line ‘Cm­UTKH’. The current reported work flow of using PCA followed by RSM showed promising results in optimizing storage temperature which was similar to previous literature. Therefore, the use of this workflow is highly recom­ mended in optimizing postharvest inputs especially when testing for newly introduced hybrids where the experimental work volume might be overwhelming. References ALABBOUD M., KALANTARI S., SOLTANI F., 2020 ­ Analysis of general and specific combining ability of postharvest Alabboud et al. ‐ Storage temperature optimization in newly introduced melon types 35 attributes in Melon. ­ J. Agric. Sci. Technol., 22(60): 1523­1535. BARTON R.R., 2013 ­ Response surface methodology. ­ In: GASS S.I., and M.C. FU (eds.) Encyclopedia of opera‐ tions research and management science. Springer, Boston, MA, USA. BRIONES A., UBEDA­IRANZO J., HERNÁNDEZ­GÓMEZ L., 2012 ­ Spirits and liqueurs from melon fruits (Cucumis melo L.), pp. 183­196. ­ In: ZERESHKI S. (ed.) Distillation advances from modeling to applications. IntechOpen limited, UK. BURGER Y., SHEN S., PETREIKOV M., SCHAFFER A. A., 2000 ­ The contribution of sucrose to total sugar content in melons. ­ Acta Horticulturae, 510: 479­486. CHONG J.X., LAI S., YANG H., 2015 ­ Chitosan combined with calcium chloride impacts fresh‐cut honeydew melon by stabilising nanostructures of sodium‐carbon‐ ate‐soluble pectin. ­ Food Control, 53: 195­205. CLARKSON G.J.J., ROTHWELL S.D., TAYLOR G., 2005 ­ End of day harvest extends shelf life. ­ HortScience, 40(5): 1431­1435. DE ARRUDA M.C., JACOMINO A.P., KLUGE R.A., AZZOLINI M., 2003 ­ Temperatura de armazenamento e tipo de corte para melão minimamente processado. ­ Rev. Bras. Frutic., 25(1): 74­76. FAO, 2019 ­ FAOSTAT. Statistical database. ‐ FAO, Rome, Italy. FARCUH M., COPES B., LE­NAVENEC G., MARROQUIN J., JAUNET T., CHI­HAM C., CANTU D., BRADFORD K.J., VAN DEYNZE A., 2020 ­ Texture diversity in melon (Cucumis melo L.): Sensory and physical assessments. ­ Postharvest Biol. Techn., 159(2020): 1­9. FEYZIAN E., DEHGHANI H., REZAI A.M., JAVARAN M.J., 2009 ­ Diallel cross analysis for maturity and yield‐relat‐ ed traits in melon (Cucumis melo L.). ­ Euphytica, 168(2): 215­223. FUKUTA S., MIZUKAMI Y., ISHIDA A., KANBE M., 2006 ­ Development of loop‐mediated isothermal amplifica‐ tion (LAMP)‐based SNP markers for shelf‐life in melon (Cucumis melo L.). ­ J. Appl. Genet., 47(4): 303­308. GARCIA­MAS J., BENJAK A., SANSEVERINO W., BOURGEOIS M., MIR G., GONŹALEZ V. M., HEŃAFF E., CAM̂ARA F., COZZUTO L., LOWY E., ALIOTO T., CAPELLA­GUTIEŔREZ S., BLANCAE J., CAÑIZARES J., ZIARSOLO P., GONZALEZ­ IBEAS D., RODRIǴUEZ­MORENO L., DROEGE M., DU L., PUIGDOMEǸECH P., 2012 ­ The genome of melon (Cucumis melo L.). ­ PNAS USA, 109(29): 11872­11877. HATAMI M., KALANTARI S., SOLTANI F., BEAULIEU J., 2019 ­ Storability, quality changes, and general postharvest behavior of dudaim melon harvested at two maturity stages. ­ HortTechnology, 29(3): 241­250. LAL BASEDIYA A., SAMUEL D. V. K., BEERA V., 2013 ­ Evaporative cooling system for storage of fruits and vegetables‐a review. ­ J. Food Sci. Technol., 50(3): 429­ 442. MAIETTI A., TEDESCHI P., STAGNO C., BORDIGA M., TRAVAGLIA F., LOCATELLI M., ARLORIO M., BRANDOLI­ NI V., 2012 ­ Analytical traceability of melon (Cucumis Melo var Reticulatus): Proximate composition, bioac‐ tive compounds, and antioxidant capacity in relation to cultivar, plant physiology state, and seasonal variabili‐ ty. ­ J. Food Sci., 77(6): C646­C652. MOHAMMADI R., DEHGHANI H., KARIMZADEH G., 2014 ­ Genetic analysis of yield components, early maturity and total soluble solids in Cantaloupe (Cucumis melo L. subsp. melo var cantalupensis Naudin). ­ Yüzüncü Yıl Üniversitesi Tarım Bilimleri Dergisi, 24(1): 79­86. MOHI­ALDEN K., ALABBOUD M., SOLTANI F., KALANTARI S., 2021 ­ Developing a postharvest color changes iden‐ tification system of melon rind using image processing. – DYSONA. Appl. Sci., 2(2): 13­20. MONFORTE A.J., EDUARDO I., ABAD S., ARUS P., 2005 ­ Inheritance mode of fruit traits in melon: Heterosis for fruit shape and its correlation with genetic distance. ‐ Euphytica, 144(1­2): 31­38. NISHIZAWA T., PUTHMEE T., KAWAMATA R., AIKAWA T., MOTOMURA Y., 2017 ­ Does the thickness of net tissues affect the waterproofing ability of musk melon (Cucumis melo L.) fruit?. ­ Acta Horticulturae, 1179: 51­ OBANDO J., FERNÁNDEZ­TRUJILIO J.P., MARTÍNEZ J.A., ALARCÓN A.L., EDUARDO I., ARÚS P., MONFORTE A.J., 2008 ­ Identification of melon fruit quality quantitative trait loci using near‐isogenic lines. ­ J. Amer. Soc. Hort. Sci., 133(1): 139­151. QI H., HU W., JIANG A., TIAN M., LI Y., 2011 ­ Extending shelf‐life of Fresh‐cut “Fuji” apples with chitosan‐coat‐ ings. ­ Innov. Food Sci. Emerg. Technol., 12(1): 62­66. SHAJARI M., SOLTANI F., BIHAMTA M.R., ALABBOUD M., 2021 ­ Genetic analysis and inheritance of floral and fruit traits in melon (Cucumis melo L.) in full diallel cross. ­ Plant Breed., 2021: 1­11 SHARMA G., WU W., DALAL E.N., 2005 ­ The CIEDE2000 color‐difference formula: Implementation notes, sup‐ plementary test data, and mathematical observations. ­ Color Res. Appl., 30(1): 21­30. WANG S., ZHOU Q., ZHOU X., ZHANG F., JI S., 2020 ­ Ethylene plays an important role in the softening and sucrose metabolism of blueberries postharvest. ­ Food Chem., 310: 125965. WANI T.A., AHMAD A., ZARGAR S., KHALIL N.Y., DARWISH I.A., 2012 ­ Use of response surface methodology for development of new microwell‐based spectrophoto‐ metric method for determination of atrovastatin calci‐ um in tablets. ­ Chem. Cent. J., 6(1): 1­9. WELBAUM G., 2015 ­ Vegetable production and practices. ­ CABI, Wallingord, Oxfordshire, UK, pp. 486. WU Z., TU M., YANG X., XU J., YU Z., 2020 ­ Effect of cutting and storage temperature on sucrose and organic acids metabolism in postharvest melon fruit. ­ Postharvest Biol. Technol., 161: 111081. Y. HASSAN AL­HAMDANY S., 2013 ­ Combining ability for yield and its components in melon (Cucumis melo L.) Adv. Hort. Sci., 2022 36(1): 27­36 36 depending on full diallel cross. ­ Mesopotamia Journal of Agriculture, 41(1): 91­105. YANG B., SHIPING T., HONGXIA L., JIE Z., JIANKANG C., YONGCAI L., WEIYI Z., 2003 ­ Effect of temperature on chilling injury, decay and quality of Hami melon during storage. ‐ Postharvest Biol. Technol., 29(2): 229­232. ŽNIDARĈIĈ D., ŠIRCELJ H., KACJAN MARŠIĆ N., 2013 ­ The influence of temperature and storage time on can‐ taloupe melons physicochemical quality. ­ Ital. J. Food Sci., 25(4): 459.