Quality Attributes Prediction of Flame Seedless Grape Clusters Based on Nutritional Status Employing Multiple Linear Regression Technique

Abstract

Flame Seedless grape is considered one of the most popular and favorite grapes for consumers, since it ripens early, and has good cluster quality. Flame seedless grape marketing value depends upon its desirable appearance, berry, cluster size, and shape. Therefore, it is imperative that the cluster yield and quality are enhanced to ensure profitability. In this study, the prediction of physical characteristics of clusters and berries’ color attributes of Flame Seedless grape grown under different culture practices, in particular fertilization treatments, was carried out using nutritional status concentration (leaf mineral elements, total chlorophyll content, total carotenoids content) and multiple linear regression (MLR). The method was based on the development of two indices: the first is called index 1 (%) and was formulated by combing the mineral elements of N, P, K, Ca, and Mg concentrations; and the second is called index 2 (ppm) and was formulated by combing the elements of Fe, Cu, Mn, Zn, and B concentrations in leaf petioles. The results indicated that the established MLR models can obtain variation accuracy, based on values of coefficients of determination (R²) using the test set. The R² values were in the range of 0.9286 to 0.9972 for cluster weight, cluster length, shoulder length, berries’ color attributes (L*, a*, b*, chroma, hue, and color index for red grapes (CIRG)). This study highlighted that during a grown season, leaf mineral elements, total chlorophyll content, and total carotenoids coupled with a MLR model can be used successfully to evaluate the physical characteristics of the cluster and berries’ color attributes of Flame seedless grape. This method is easy, fast and reliable as it retains the physical appearance of the fruits by adjusting the concentration of mineral elements, total chlorophyll content, and total carotenoids in leaves. Moreover, total chlorophyll had the greatest weight of all the predicted quality attributes.

1. Introduction

The grapevine (Vitis vinifera L., Vitaceae) is grown under a wide range of soil and climatic conditions, namely, hot tropical, sub-tropical and temperate regions [1,2,3]. Most of the world’s viticulture regions are located between the latitudes of 40° and 50° N in the northern hemisphere and between the latitudes of 30° and 40° S in the southern hemisphere and are known as temperate climatic belts [1,4]. Grapevine represents the most popular fruit crop in the world [5,6], the first major deciduous fruit crop in Egypt, and is ranked second position amongst the fruit crop trees, following citrus [7]. Grapevine is an important, economically valuable plant the world over, and a likeable fruit, which can be used as a rich source of nutrients in multiple industries [8,9]. It is one of the most economically significant sources for the food industry due to having many nutritive values such as soluble sugars, organic acids, vitamins, aromatic compounds and many antioxidants such as phenolic compounds (phenolic acids, flavonols, anthocyanins, and stilbenes) and β carotene [10,11,12].

Flame Seedless is one of the most valuable grape cultivars in Egypt. In addition, it is one of the earliest ripening red seedless cultivars, which have appeared in the Egyptian market [13]. Moreover, it is a hybrid of different cultivars such as ‘Thompson Seedless’, ‘Cardinal’, and various other grape cultivars. It is a red seedless grape cultivar, early-season, capital-intensive, has proved popular with the consumer thanks to its suitability for the fresh market. It is considered one of the most widely cultivated table grapes among the seedless cultivars worldwide, a popular and favorite grape for the consumer as it ripens early, has good clusters, and has a higher percentage of anthocyanin content [5,14]. Flame seedless grape marketing value depends upon its desirable appearance, berry, cluster size, and shape. As such, it is imperative to enhance its yield and quality to ensure profitability [14,15].

Fruit quality denotes a mixture of properties and features that define the rightness of the fruit to be eaten as fresh or stored for an equitable period without decline and advise a value regarding the consumer’s agreement [16,17,18]. Fruit’s competitiveness in the market can only be improved through enhancement of its quality. In recent years, with the continuous improvement of people’s living standards, they also raised their expectation for the nutritional value and fruit quality [19]. Leaves play primary roles in the photosynthetic process, and essential mineral nutrients during plant development to improvements in productivity [20,21], especially source–sink relationships (mainly leaves vs. reproductive organs) play an important role in fruit growth and quality with respect to fruit size, color, and chemical composition [22].

Nutrient content in a fruit tree’s leaves is a feature of a specific cultivar; it is influenced, within certain limits, by the rootstocks [23,24], by soil and climatic conditions [25], by the phase of vegetation [26,27,28], and particularly by the orchard management practices, i.e., irrigation and fertilization [29,30]. The many irrigation and fertilization methods used in the orchard include the soil application of granular fertilizers or fertigation through drip emitters or different water rates influence of nutrient content in fruit tree leaves [31]. However, the nutritional status of leaves per vine through as a result of its great architectural and physiological plasticity can result in different crop load levels [22] and quality related features of grapes [32]. Therefore, the prediction of quality attributes could enhance the quality characteristics in order to help vineyard management practices to optimize source/sink relationships, favoring fruit development [33]. Several studies found that the leaves’ mineral element content had a significant impact on fruit quality [34]. Mineral nutrients, either in abundance or in deficiency, are among the most important key factors that affect fruit quality [35]. Moreover, minerals influence the quality and yield of various fruits [36].

There is a correlation between fruit quality and mineral nutrients in soil, leaf and fruit [37], and furthermore, there a direct correlation exists between leaf nutrient status and fruit development [31]. Fruit quality prediction can be achieved based on soil mineral element content [38]. Moreover, leaf mineral analysis is a useful tool to diagnosis a fruit apple tree deficiency, and using fruit analysis alone or in combination with leaf analysis often permits more precise prediction of fruit quality [34].

In recent years, for agricultural management, in particular in horticulture production, we have seen the advancement and utilization of mathematical modelling that is more dependable, efficient, and quick, such as the multiple linear regression (MLR), which has been utilized to precisely identify yield and quality attributes of fruits, vegetables, and crops [39,40,41]. However, MLR is a good tool and a fundamental technique for predicting the numerical outcome variable [42] to study the collaboration of the relationships of two or more variables that affect another value. Thus, time and laboratory analysis costs are waived. Moreover, MLR is considered to be a strategic tool used to enhance product planning and predict the quantity and quality of fruit products to fulfill customer needs. However, the quality of the fruit is largely dependent on agricultural practices in the field before and during the growing season, as well as occasionally on commercial experience and conventional methods that are inconsistent. In contrast, the use of predictive models to determine the quality attributes of fruits is sufficient to avoid the need for time and laboratory costs. On the other hand, the MLR algorithm is easier to solve by the availability of Microsoft Excel spreadsheets, and is also used in agricultural production application modeling. For example, peach firmness prediction using MLR was evaluated [42]. Simple and MLR models were successfully established to predict the maturity parameters of ‘Mendez#1’ and ‘Hass’ avocado fruit [43]. In these researches, the MLR model is created by fitting a linear equation to the observed data. Although no causal mechanism is suggested, this statistical method has the main benefit of being able to demonstrate connections between variables [44]. In place of other statistical techniques, multiple regression is simple to set up and has been used to obtain models in a variety of academic domains [45,46,47,48,49,50,51].

Nowadays, many efforts will be established for finding out the best horticultural practices that are responsible for enhancing the yield and fruit quality of Flame Seedless grapes. Therefore, studying the quality prediction of the characteristics of table grapes would be meaningful for the improvement of their logistics management, and could be a valuable reference to overcome the decline in the yield and the inferior grape quality. The impact of various cultural practices, applied in vineyards, such as adjusting crop load by cultivar thinning and shoot topping, as well as increasing crop quality by girdling, topping or tipping, and using growth regulators such as, mainly, gibberellic acid which was appraised by different research papers [52,53,54,55,56] and all these cultural practices cannot be conducted in one vineyard. In Egypt, various cultural practices were employed to enhance the yield and quality of Flame Seedless cultivar [13,57,58,59,60]. In general, the impact of cultural practices on the yield and quality has been studied intensely, with varying results depending on the level, time, cultivar and site [55]. In light of these considerations, our study aimed to emphasis the cluster physical attributes (cluster weight, cluster length, cluster shoulder length) and berries’ color attributes (L*, a*, b*, chroma, hue, and color index for red grapes (CIRG)), of Flame Seedless grape in relation to different agricultural practices, in particular fertilization treatments. Additionally, the study aimed to verify the effectiveness of MLR modeling to build models for the prediction of the mentioned quality attributes using the nutritional status in leaves including leaf mineral elements, total chlorophyll content, and total carotenoids. Furthermore, the study aimed to establish ranges of the nutritional status in leaves for grape cv. Flame Seedless under the climatic soil conditions of QoraKhargin, Gharb El-Nobarya, Beheira Governorate, Egypt.

2. Materials and Methods

2.1. Experimental Site and Plant Materials

This study was carried out during the season of 2021 on seven year old Flame seedless grapevines grown in a private vineyard situated at QoraKhargin, Gharb El-Nobarya, Beheira Governorate, Egypt (GPS co-ordinates: 30°34′55.3″ N, 29°53′25.4″ E). The vines were spaced 4 × 1.75 m apart, grown in a sandy soil, with an average pH of 7.7–7.8, under a drip irrigation system with two lines per vine with four drippers (8 L/h) per vine.

For the present study, nine vineyards were selected that differed from one vineyard to another in applied agricultural practices, in particular, fertilization applications, which included fertilization units, fertilizer types, and fertilization dates [61,62]. However, different vines were selected so that they were as uniform as possible in growth with no visual nutrient deficiency symptoms and almost uniform in their vigor to get samples as follows (i.e., 9 vineyards × 9 replicates = 81 vines).

2.2. Measurements of Studied Parameters

2.2.1. Nutritional Status in the Leaves

Grapevine nutritional status was evaluated by the determination of leaf petiole mineral content and total chlorophyll contents [7]. In total, twenty leaves were picked from the main shoots opposite the basal clusters and were taken from the middle at the beginning of veraison stage per vine according to Summer [63] and Wilde et al. [64]. Blades of the leaves were discarded and petioles were saved to determined different nutrients (N, P, K, Ca, and Mg as %, while Fe, Cu, Mn, Zn and B as ppm). Petioles per vine were washed carefully with tap water and distilled water and subsequently dried at 70 °C until constant weight, then (0.5 g) weight of each sample was digested using H₂SO₄ and H₂O₂ until clear solution was obtained according to Wilde et al. [64]. The digested solutions were quantitively transferred to a 100 mL volumetric flask and completed to 100 mL by distilled water. Nitrogen (N%) was estimated by Micro–Kjeldahl method as described by Chapman and Pratt [65], phosphorus (P) by the method of Wilde et al. [64], potassium (K) by flame photometer according to the method of Peach and Tracey [66]. Calcium (Ca), magnesium (Mg) and iron (Fe), copper (Cu), manganese (Mn) and zinc (Zn) were determined using atomic absorption spectrophotometer “Perkin Elmer 3300” according to Cottenie et al. [67]. Boron (B) was calorimetrically determined using the carmine method according to Hatcher and Wilcox [68]. Samples of five mature and fresh apical fifth and sixth leaves from those leaves opposite to the basal clusters on each shoot were taken to determined chlorophyll A and B, total chlorophyll and carotenoids as (mg/1 g fresh weight) according to Von-Wettstein [69].

Samples were broken up into little pieces, and 0.5 g of weight from each sample was taken, homogenized, and extracted with 25% acetone while small amounts of Na₂CO₃ were presented, and then filtered. Acetone was used to wash the residue numerous times until the filtrate was colorless. With acetone at 85%, the extract was finished to a known volume of 10 mL. A sample of this extract was used to determine the colorimetric concentrations of chlorophylls A and B, total chlorophylls, and total carotenoids. Acetone 85% was used as a control. To quantify chlorophylls A, B, and total carotenoids, respectively, the optical density of the filtrate was measured at wave lengths of 662, 644, and 440.5 nm (using a spectrophotometer, model: SP-3000 plus, Optima, Tokyo, Japan) according to Von-Wettstein [69] and Hiscox and Isralastam [70].

The obtained elements of N, P, K, Ca, and Mg (%) were combined in one index called Index 1 (%) as follows:

(1)$\mathrm{Index} 1=\left(\frac{\mathrm{N}\times \mathrm{P}\times \mathrm{K}}{\mathrm{Ca}\times \mathrm{Mg}}\right)$

The obtained elements of Fe, Cu, Mn, Zn and B as ppm were combined in one index called Index 2 (ppm) as follows:

(2)$\mathrm{Index} 2=\left(\frac{\mathrm{Fe}\times \mathrm{Cu}\times \mathrm{Mn}}{\mathrm{Zn}\times \mathrm{B}}\right)$

These indices allowed the number of predictors in the analysis to be limited.

2.2.2. Physical Quality Analysis

A sample of 30 clusters per each vineyard (three clusters/vine) were harvested in the second week of May 2021, when berries reached full color and were transported to a laboratory located at the Food Science and Technology Department, Faculty of Agriculture, Alexandria University, Egypt. Fruit physical characteristics such as the cluster weight, cluster length, and cluster shoulder length (cluster width) were determined using a digital balance and a steel rule, respectively. The berries’ color was assessed using a colorimeter (CR-3000, Konica Minolta, Tokyo, Japan) by recording values of L*, a* and b*, where the L* value corresponds to a dark–bright scale (0, black; 100, white), the a* value to a green–red scale (negative value, greenness; positive, redness) and the b* value to a blue–yellow scale (negative value, blueness; positive, yellowness).

The chroma (C) was calculated using the following formula [71]:

(3)$\mathrm{C}=\sqrt{{\left({\mathrm{a}}^{*}\right)}^2+{\left({\mathrm{b}}^{*}\right)}^2}$

C is the chroma, a measure of the intensity/purity of color (0, achromatic). Chroma indicates the purity or intensity of color, the distance from gray (achromatic) toward a pure chromatic color (0)

The hue angle (h, °) is calculated as follows [72]:

(4)${\mathrm{h}}^{\circ }={\tan }^{-1}\left(\frac{{\mathrm{b}}^{*}}{{\mathrm{a}}^{*}}\right)$

The color index for red grapes (CIRG) was calculated as follows [73]:

(5)$\mathrm{CIRG}=\frac{\left(180-{\mathrm{h}}^{\circ }\right)}{\left(\mathrm{C}+{\mathrm{L}}^{*}\right)}$

2.3. Building of Multiple Linear Regression Models

MLR is one of the statistical techniques used to model the linear relationship between the independent (explanatory) variables and dependent (response) variables. It was employed to determine how the investigated independent variables (leaf mineral elements, which were stated by index 1 and index 2, total chlorophyll content, and total carotenoids) were functionally related to the quality characteristics of Flame Seedless grape clusters and berries. The MLR model was defined as follows:

(6)$\mathrm{Y}={\beta }_0+{\displaystyle \sum }_{\mathrm{i}=1}^{12}{\beta }_{\mathrm{i}}{\mathrm{x}}_{\mathrm{i}}+{\mathsf{\epsilon }}_{\mathrm{i}}$

where Y is the dependent variable (cluster weigh, cluster length, cluster shoulder length, lightness—L*; redness—a*; yellowness—b*, chroma, hue, and color index for red grapes); β₀ is the intercept; β_i is the regression coefficients (i = 1, 2, 3, and 4); x_i is the independent variables (index 1, index 2, total chlorophyll, and total carotenoids in leaves); and ε_i is error term. The experimental data (81 samples) were randomly divided into two groups, with 61 patterns of the data used for training or calibration and the remaining 20 patterns for testing purposes. The enter method, which includes all independent variables in the regression equation, was selected. The model input variables consisted of index 1, index 2, total chlorophyll, and total carotenoids in leaves; the cluster weight, cluster length, cluster shoulder length, lightness—L*, redness—a*, yellowness—b*, chroma, hue, and color index for red grapes were considered as the model’s outputs.

2.4. Statistical and Data Analysis

All data were tested for vineyards effects on analyzed parameters by one-way analysis of variance (ANOVA) technique. Means were separated and compared using least significant difference (LSD) at 0.05 level of probability. The ANOVA was performed using SAS (Statistical Analysis System, version 9.13, SAS Institute Inc., Cary, NC, USA, 2008) [74]. Moreover, the correlation analysis was conducted and used to evaluate the relationship between the measured parameters. The MLR models were also conducted using the SPSS 18.0 software (SPSS Inc., Chicago, IL, USA). It was used to predict the cluster weight, cluster length, cluster shoulder length, lightness—L*, redness—a*, yellowness—b*, chroma, hue, and color index for red grapes. We also evaluated the performance of MLR models by some statistical indices, such as coefficient of determination (R²), mean absolute error (MAE), and root mean square error (RMSE) as follows:

(7)$\mathrm{MAE}=\frac{1}{n}{\displaystyle \sum }_{i=1}^n\left|P_i-O_i\right|$

(8)$\mathrm{RMSE}=\sqrt{\frac{1}{n}{\displaystyle \sum }_{i=1}^n{\left(P_i-O_i\right)}^2}$

where, P_i and O_i are the predicted and measured values, respectively, and n is the total number of points in each dataset. Moreover, scatter graphs are presented for visual correlation of the measured and predicted values.

3. Results and Discussion

3.1. Nutritional Status in the Leaves

Table 1 depicts statistical criteria for the nutritional status in leaf petioles (mineral content) and chlorophyll A, chlorophyll B, total chlorophyll, and total carotenoids in the leaves. Additionally, Table 2 depicts distribution of mean values and standard deviation for N, P, K, Ca, and Mg in leaves of Flame Seedless grape cultivated by different agricultural practices in investigated vineyards. Moreover, Table 3 shows distribution of mean values and standard deviation for Fe, Cu, Mn, and Zn in leaves of Flame Seedless grape cultivated by different agricultural practices in investigated vineyards. Furthermore, Figure 1 depicts distribution of mean values of chlorophyll A, chlorophyll B, and total carotenoids contents in leaves of Flame Seedless grape cultivated by different agricultural practices in investigated vineyards.

By exploring Table 1, it is clear that the maximum values of the N, P, K, Ca, and Mg were 2.58, 0.51, 1.64, 2.89, and 0.69%, respectively; the minimum values were 1.51, 0.11, 1.01, 1.92, 0.21%, respectively, and the average values were 2.00, 0.32, 1.37, 2.44, and 0.46%, respectively. The highest average for N, P, K, Ca, and Mg belonged to vineyard No. 9 (S9), as shown in Table 2 with statistical differences. Furthermore, it is clear that the maximum values of the Fe, Cu, Mn, Zn, and B were 139.00, 39.00, 250.00, 124.00, and 49.00 ppm, respectively (Table 1).

The minimum values of Fe, Cu, Mn, Zn, and B were 41.00, 4.00, 160.00, 28.00, and 16.00 ppm, respectively, and the average values were 88.25, 19.62, 198.26, 76.80, and 34.38 ppm, respectively (Table 1). The highest average for Fe, Cu, Mn, Zn, and B belonged to vineyard No. 9 (S9), as shown in Table 3 with statistical differences. Regarding comparison with other studies, Stino et al. [58] reported range of N as 1.47–2.25%, P as 0.22–0.36%, K as 0.75–1.14%, Ca as 1.48–1.97%, Fe as 121.7–149.4 ppm, Zn as 23.46–37.67 ppm, and Mn as 60.671–90.31 ppm during season 2014 for Flame seedless grapevines in a private vineyard at El-Bostan district, Beheira Governorate, Egypt. In the study of Bassiony and Mahmoud [75] the range of N as 1.21 to 2.25%, P as, 0.20–0.57%, K as 1.16–1.56%, Ca as 1.42–2.68, Mg as 0.20–0.66%, Fe as 50.83–75.21 ppm in Flame seedless leaves affected by magnetic iron and antioxidants applications during 2016 season and cultivated at a commercial vineyard located at Abo Ghalib El-Giza Governorate, Egypt, was reported. Furthermore, in leaves of Flame Seedless as effected by foliarly applied micronutrients, Ali et al. [62] reported for control treatment that Zinc values were 15.3 ± 1.45 and 11.1 ± 1.15 (ppm) for seasons 2018 and 2019, respectively, and Boron values were 21.1 ± 1.15 and 22.3 ± 0.88 ppm and iron 38.6 ± 2.33 and 39.6 ± 1.45 ppm for seasons 2018 and 2019, respectively.

It is clear from Table 1 that the maximum value of chlorophyll A and chlorophyll B was 3.56 and 2.33 mg/g fresh weight, respectively, the minimum value was 2.02 and 1.22 mg/g fresh weight, respectively, and the average value was 2.73 and 1.84 mg/g fresh weight, respectively. The highest average for chlorophyll A and chlorophyll B belonged to vineyard No. 9 (S9), as shown in Figure 1 with statistical differences. In the study of Stino et al. [58], chlorophyll contents of Flame seedless grapevines in 2014 and 2015 seasons ranged from 2.61 to 3.06 mg/g and 2.65 to 3.22 mg/g, respectively, under the influence of bio-stimulants and the lowest chlorophyll contents belonged to vines treated by water. Furthermore, in leaves of Flame Seedless’ as effected by foliarly applied micronutrients, Ali et al. [62] reported for control, i.e., vines were treated by water that Chlorophyll A values were 2.47 ± 0.23 and 2.45 ± 0.14 (mg/g) and Chlorophyll B were 1.32 ± 0.18 and 1.03 ± 0.06 (mg/g) for seasons 2018 and 2019, respectively.

Furthermore, it is clear from Table 1 that the maximum value of total carotenoids was 5.89 mg/g fresh weight, the minimum value was 3.25 mg/g fresh weight and the average value was 4.58 mg/g fresh weight. The highest average for total carotenoids was belonged to vineyard No. 9 (S9), as shown in Figure 1 with statistical differences. However, among values, the coefficient of variation in Cu content was the largest (54.33%), and that of Ca content was the smallest (11.82%), as shown in Table 1 indicating that the difference of Cu content in different vineyards was significant.

3.2. Cluster Quality Attributes of Flame Seedless Grape

Table 4 illustrates statistical criteria of experimental results of quality attributes (cluster weight, cluster length, cluster shoulder length, berries’ color attributes (L*, a*, b*, chroma, hue, and color index for red grapes (CIRG) of Flame Seedless grape cultivated by different agricultural practices. Furthermore, Table 5 indicates distribution of mean values of such attributes in investigated vineyards. By exploring Table 4, it is clear that the maximum value of the cluster weight was 439 g, the minimum was 325 g and the average value was 395.67 g and the highest average (436.67 g) belonged to vineyard No. 9 (S9), as shown in Table 5 with statistical differences. Similar physical properties variability was reported for clusters of Flame Seedless grapes by other authors, however, a comparison of the presented data was not available in the literature data. In the study of Bassiony and Mahmoud [75], a range of cluster weight as 210.45–455.68 g was reported for Flame seedless tress that was affected by magnetic iron and antioxidant applications during the 2016 season and cultivated at a commercial vineyard located at Abo Ghalib El-Giza Governorate, Egypt. Stino et al. [58] reported the range of cluster weight as 632.1–747.9 g during the 2014 season for Flame seedless grapevines in a private vineyard at El-Bostan district, Beheira Governorate, Egypt.

The maximum value of the cluster length was 17.50 cm, the minimum was 13.50 cm and the average value was 15.82 cm, as shown in Table 4. The highest average of the cluster length (17.46 cm) belonged to vineyard No. 9 (S9), as shown in Table 5 with statistical differences. In the study of Bassiony and Mahmoud [75], the range of cluster length was 15.33–26.31 cm for Flame seedless tress and was affected by magnetic iron and antioxidants applications during the 2016 season and cultivated at a commercial vineyard located at Abo Ghalib El-Giza Governorate, Egypt. Stino et al. [58] reported range of cluster length as 16.67–24.0 cm, cluster width as 7.08–10.08 cm during season 2014 for Flame seedless grapevines in a private vineyard at El-Bostan district, Beheira Governorate, Egypt.

The maximum value of the cluster shoulder length was 13.60 cm, the minimum value was 9.20 cm and the average value was 11.47 cm, as shown in Table 4. The highest average (13.52 cm) belonged to vineyard No. 9 (S9), as shown in Table 5 with statistical differences. The coefficient of variation in cluster shoulder length was 12.92%, and that of cluster weight and cluster length was 8.61% and 8.12%, respectively, and statistically, the investigated vineyards had significant effect on all parameters as shown in Table 5. In conclusion, cluster weight, cluster length, cluster shoulder length of Flame Seedless grape can be significantly improved when the content of available N, P, K, Ca, Mg, Fe, Cu, Mn, and Zn, chlorophyll A, chlorophyll B, and total carotenoids was increased in the leaves. However, by exploring the variations in parameters, we could say that the optimum level of culture practices recommended by the national ministry of agriculture in Egypt was adopted in vineyard No. 9.

It is clear that the maximum value of the L* was 46.45, the minimum was 25.65 and the average value was 35.13 (Table 4). The highest average of the L* (45.90) belonged to vineyard No. 1 (S1), as shown in Table 5 with statistical differences. Additionally, the maximum value of the a* was 10.99, the minimum value was 5.24 (Table 4), and the average value was 8.58. The highest average of the a* (10.88) belonged to vineyard No. 9 (S9), as shown in Table 5 with statistical differences. Furthermore, the maximum value of the b* was 12.96, the minimum value was 3.12 and the average value was 7.83 (Table 4). The highest average of the b* (12.83) belonged to vineyard No. 1 (S1), as shown in Table 5 with statistical differences. The coefficient of variation in L* was 19.33%, and that of a* and b* was 20.10% and 40.61%, respectively, and statistically, the vineyards had significant effect on L*, a*, and b* as shown in Table 5.

It is clear that the maximum value of chroma was 14.01, the minimum value was 11.09 and the average value was 12.12, as shown in Table 4. The highest average chroma (13.89) belonged to vineyard No. 1 (S1), as shown in Table 5. Additionally, the maximum value of hue was 67.85°, the minimum value was 1.86 and the average value was 41.15°, as shown in Table 4. The highest average of hue (67.47°) belonged to vineyard No. 1 (S1), as shown in Table 5. Moreover, the maximum value of CIRG was 4.43, the minimum value was 1.86 and the average value was 3.07, as shown in Table 4. The highest average of CIRG (4.43) belonged to vineyard No. 9 (S9), as shown in Table 5. The cluster color was classified into five categories according to the CIRG values, i.e., green–yellow (<1.5), pink (1.5 to 2.5), red (2.5 to 3.5), violet (3.5 to 4.5), and dark violet (>5.0) [76]. We found that the coefficients of variation for all fruit quality parameters were in the range of 7.79 to 40.61 (Table 4), meaning that fruit quality from different vineyards is relatively non-identical.

3.3. Correlation Analysis

Fruit quality was one of the market’s most important core competitiveness factors, which not only affected fruit price, but also fruit sales volume [77]. It was caused by a combination of multiple factors, especially the individual and combined effects of mineral nutrients [78]. However, the leaf is a vital component in the plant’s exchange of matter and energy. Thus, enhancement of leaf nutrients content such as N, P, K, Ca, Mg and Fe led to improve quality of grape berries’ physical properties [75]. Therefore, Pearson’s correlation was used to assess the relationships among fruit quality variables [79,80]. As in Table 6, there are some high positive correlations and some negative correlations. The positive high correlations were observed between N, P, K, Ca, Mg, Fe, Cu, Mn, Zn, B, index 1, index 2, chlorophyll A, chlorophyll B, total chlorophyll, and total carotenoids and cluster weight, cluster length, and cluster shoulder length, berry a*, and berry CIRG by r values in the range of 0.874 to 0.988, 0.898 to 0.994, and 0.937 to 0.993, 0.894 to 0.995, and 0.959 to 0.998, respectively. This suggests that increasing N, P, K, Ca, Mg, Fe, Cu, Mn, Zn, B, total chlorophyll, and total carotenoids amounts increase the cluster weight, cluster length, and cluster shoulder length, berry a*, and berry CIRG and give evidence that these elements are the major factor responsible for enhancing cluster and berries’ characteristics. However, several mineral nutrients such as nitrogen (N), potassium (K), phosphorous (P), calcium (Ca), and boron (B) can influence fruit quality and disorders of an apple [34]. Additionally, fruit diameter increased with increasing leaf N concentration [81] in another study, inducing leaf nutrient imbalance, eventually impairing fruit quality parameters [82].

The negative high correlations between N, P, K, Ca, Mg, Fe, Cu, Mn, Zn, B, index 1, index 2, chlorophyll A, chlorophyll B, total chlorophyll, and total carotenoids and berries’ color attributes (L*, b*, chroma, and hue) were noticed by r values in the range of −0.925 to −0.996, −0.942 to −0.997, −0.882 to −0.971, −0.944 to −0.998, respectively. These findings showing that all the elements do contribute significantly to the berries’ color attributes (L*, b*, chroma, and hue).

The positive high correlations between cluster length, cluster shoulder length, a*, and CIRG and cluster weight were observed by r = 0.992, r = 0.971, r = 0.992, and r = 0.954 as shown in Table 7, respectively. In other words, by increasing the cluster length, cluster shoulder length, berry a*, and berry CIRG, the cluster weight is increased and vice versa. On the other hand, negative high correlations between berries’ color attributes (L*, b*, chroma, and hue) and cluster weight were noticed by r = −0.981, r = −0.975, r = −0.978, and r = −0.973 as shown in Table 7, respectively. Additionally, positive high correlations between berry a*, and berry CIRG and cluster shoulder length were detected by r = 0.978 and r = 0.993 as shown in Table 7, respectively. Furthermore, negative high correlations between berries’ color attributes (L*, b*, chroma, and hue) and cluster shoulder length were observed by r = −0.994, r = −0.997, and r = −0.995, as shown in Table 7, respectively. In other words, by increasing the berries’ color attributes (L*, b*, chroma, and hue), the cluster shoulder length is decreased and vice versa. These correlations and leaf mineral analyses are useful tools that help diagnosis fruit tree deficiency and, using fruit analysis alone or in combination with leaf analysis, often permit a more precise prediction of fruit quality [34]. However, the increase in soil K fertilization raised fruit mass, diameter, titratable acidity, red color, and fruit K contents [83]. Furthermore, sufficient levels of leaf K increase the surface color of apples [84]. However, Boron is an essential trace element for the normal growth of fruits such as apples [85,86].

3.4. Prediction of Cluster Quality Attributesof Flame Seedless Grape Cultivated by Different Agricultural Practices

In order to test the hypothesis that cluster quality attributes of Flame Seedless grapes cultivated by different agricultural practices can be estimated, regardless of index 1, index 2, total chlorophyll, and total carotenoids, all the assessed variables were included. The original data set was divided into a training set (with 75% of the data) and test set (with 25% of the data) in a full random mode. The training set was used to establish the relationship between the analyzed variables (described above) and to train the regression model. The final multiple regression model is presented in Equation (9).

(9)$\mathrm{Y}={\beta }_0+{\beta }_1\times \mathrm{Index}\hspace{0.33em}1+{\beta }_2\times \mathrm{Index}\hspace{0.33em}2+{\beta }_3\times \mathrm{total}\hspace{0.33em}\mathrm{chlorophyll}\hspace{0.33em}+{\beta }_4\times \mathrm{total}\hspace{0.33em}\mathrm{carotenoids}\hspace{0.33em}$

The established MLR models were highly significant and specifically described the influence of the inputs on the quality of the product. Table 8 indicates the values of regression coefficients and coefficients of determination (R²) of regression equation for modeling cluster weight, cluster length, cluster shoulder length, berries L*, a*, b*, chroma, hue, and CIRG of Flame Seedless grape. The regression analysis between the measured and predicted quality attributes yielded R² in the range of 0.941 to 0.998 on the training set. Our results are an important step towards easy quality attributes estimation in the vineyard.

A relationship between the components and the quality feature can be established with the aid of the straight forward and well-known MLR technique. The ability to simply express linear relationships between variables and to ignore other types of correlations is one apparent drawback of MLR models, making them a constrained approach to building mathematical models. Despite being viewed as ineffectual, MLR models are frequently employed in the modelling process and successfully validated. A study employing the MLR model to predict the flavor of orange using 70 samples is an example of this, and the results revealed that the prediction performance of the MLR model had a strong significant relationship between orange taste and vitamin C, red color, and blue color as reported by Mokarram et al. [87]. Furthermore, eight aroma properties of Dianhong tea were predicted using MLR produced R² > 0.95 [46]. Gajewski and Arasimowicz [88] predicted overall quality score for eggplant fruits with MLR. Kavand et al. [89] used MLR to estimate aril browning in pomegranate genotypes. However, all studies proved that MLR has usefulness, simplicity, and rapidity in preparation.

As shown in Table 8, increasing total chlorophyll resulted in a decrease in berries’ color attributes (L*, b*, chroma, and hue). Camargo et al. [90] claimed that the overall color changes have a close relationship with chlorophyll degradation, which was indicated by the green color of the skin. Moreover, increasing total carotenoids led to a decrease in a* as Schweiggert et al. [91] discussed the correlation of a* value with the carotenoid development in the fruits that were responsible for the reddish color of the skin.

Measured and predicted cluster weight, cluster length, and cluster shoulder length in MLR models and the coefficient of determination (R²) are presented in Figure 2 of Flame Seedless grape using a test set. All scatter plots showed a low dispersion in the distribution pattern of data and high accuracy in predicting the cluster weight, cluster length, and cluster shoulder length. These results suggest that MLR models effectively predict the cluster weight, cluster length, and cluster shoulder length, however, similar findings were reported by Stangiersk et al. [44], who used the MLR model to simulate changes in overall quality of processed Gouda cheese. They discovered that MLR models were highly significant with high R², making them an effective tool to predict quality deterioration. Although regression analysis is straight forward to implement, quick to model, and particularly helpful when the variable region is not overly complex, other more sophisticated data analytics techniques are probably necessary to benefit from highly multiplex and multidimensional hyperspectral datasets [92]. In this study, an MLR model is obtained, which is the top performing model when applied to the testing dataset described in this paper. As is obvious from the figures, the R² values were higher than 0.90 using test sets that prove the best performance of the MLR method to estimate quality attributes.

Color is an important feature that dictates the quality and consumer preferences of many fresh fruits and vegetables. In color measurement of fruits, the CIE L*, a*, and b* color space is widely used since it is a uniform color scale [93] and could be predicted by another parameters. However, measured and predicted berries L*, a*, and b*, chroma, hue, and CIRG in MLR models and the R² are presented in Figure 3 of Flame Seedless grape berries using a test set. All scatter plots showed a low dispersion in the distribution pattern of data and high accuracy in predicting the L*, a*, b*, chroma, hue, and CIRG. In the study of Aredo et al. [94], employed of MLR for the prediction of external quality attributes of orange presented a high performance for L* (R² = 0.898), a* (R² = 0.952), and b* (R² = 0.922).

A strength of the study was the emphasis the various leaf mineral content as well as chlorophyll and total carotenoids with different culture practices, as shown in the previous sections. In our study, the values of the various parameters were distributed fairly evenly over all the intervals, with a wide range of variation. At the same time, the use of a small number of samples has represented a non-weakness in the accuracy of the estimates of the cluster and berries’ quality parameters, although the small number of samples could be one of the reasons why some regression models did not perform highly [95]. Furthermore, the developed process can contribute to the valorization and design of equipment for this grape cultivar, making it available to the consumers through its consumption.

3.5. The Sensitivity Analysis of the Input Parameters on the Flame Seedless Grape Quality Attributes

We obtained reliable prediction models by constructing the MLR models, which can accurately predict the fruit quality attributes based on the content of mineral elements in the leaves as well as total chlorophyll and total carotenoids. To elucidate the influence of certain parameter on fruit quality indicators, we conducted a sensitivity analysis of the prediction MLR models to explore which parameter had a greater impact on fruit quality indicators. In the sensitivity analysis, RMSE value represented the relative contribution of the input value to the output value in the MLR model that was evaluated. The higher the RMSE value was, the higher the importance of the eliminated input value, that is, the greater the influence of the eliminated parameters on the fruit quality attributes. As shown in Table 9, in the prediction models of all quality attributes based on large RMSE, this indicates that total chlorophyll had the greatest weight. However, Ta et al. [92] found that one of the most critical variables impacting photosynthetic ability and nitrogen status, both of which affect crop harvest, is leaf chlorophyll content. Furthermore, according to Gitelson and Merzlyak [96], the amount of chlorophyll in leaves is a measure of the photosynthetic rate, nitrogen content, and general health of crops as well as the degree of senescence. It is important for nutrient content monitoring, crop growth monitoring, quality assessment, and yield estimation [97]. However, according to Hendry and Price [98], chlorophylls tend to be photo-oxidized under high photo-irradiation. Photo-oxidation is an irreversible process and directly involves the light receptor pigments, which, by absorbing a great deal of light, remain excited longer and interact with O₂, producing free radicals, such as superoxide (O₂), which may destroy the pigments [99].

4. Conclusions

The study findings showed that there were variations among nutritional status in leaves of Flame seedless grapes according to vineyards. The average values of N, P, K, Ca, and Mg were 2.00, 0.32, 1.37, 2.44, and 0.46%, respectively. The average values of Fe, Cu, Mn, Zn, and B were 88.25, 19.62, 198.26, 76.80, and 34.38 ppm, respectively. The average values Chlorophyll A and Chlorophyll B were 2.73 and 1.84 mg/g fresh weight. Additionally, the average value of total carotenoids was 4.58 mg/g fresh weight. Furthermore, in this research, an attempt was made to predict the quality attributes of Flame Seedless grape clusters and berries produced under different agricultural practices based on leaves nutritional status and MLR technique. The results indicated that the established MLR models can predict cluster weight, cluster length, cluster shoulder length, berries’ color attributes (L*, a*, b*, chroma, hue, and color index for red grapes (CIRG) with higher correlation between measured and predicted attributes. Total chlorophyll had the greatest contribution on prediction of all quality attributes. The advantage of developed MLR models was the definition of ranges of leaves’ nutritional status that helped establish a management tool to get the best quality of Flame Seedless grape clusters. We recommended the study of the effect of different agricultural practices applied in the vineyards by modeling using machine learning techniques.

Author Contributions

Conceptualization, M.A.-S. and L.S.-P.; Methodology, M.A.-S. and D.H.E.; Software, M.A.-S., A.M.A.-S. and A.M.A.; Formal analysis, M.A.-S., A.M.A.-S., A.M.A., L.S.-P. and D.H.E.; Investigation, M.A.-S. and D.H.E.; Resources, M.A.-S., A.M.A.-S. and D.H.E.; Data curation, M.A.-S., A.M.A. and D.H.E.; Writing—original draft preparation, M.A.-S., A.M.A.-S., A.M.A., L.S.-P. and D.H.E.; Writing—review and editing, all authors. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

All data are presented in this article in the form of figures and tables.

Conflicts of Interest

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures and Tables

Figure 1. Distribution of mean values of chlorophyll A, chlorophyll B, and total carotenoids contents in leaves of Flame Seedless grape cultivated by different agricultural practices in investigated vineyards [different letters indicate that means are significantly different from each other (p < 0.05)].

Figure 2. Regression analysis between measured and predicted cluster weight, cluster length, and cluster shoulder length for the test set.

Figure 3. Regression analysis between measured and predicted berries L*, a*, and b*, chroma, hue, and CIRG for the test set.

Table 1

Statistical criteria of experimental results for the nutritional status in leaf petioles and leaves of Flame Seedless grape cultivated by different agricultural practices.

Variables	Statistical Criteria
	Minimum	Maximum	Mean	Standard Deviation	Coefficient of Variation (%)	Skewness	Kurtz
N (%)	1.51	2.58	2.00	0.34	16.80	0.28	−1.22
P (%)	0.11	0.51	0.32	0.12	37.91	−0.27	−1.19
K (%)	1.01	1.64	1.37	0.20	14.48	−0.35	−1.20
Ca (%)	1.92	2.89	2.44	0.29	11.82	−0.22	−1.05
Mg (%)	0.21	0.69	0.46	0.15	32.83	−0.25	−1.14
Fe (ppm)	41.00	139.00	88.25	32.70	37.05	−0.01	−1.42
Cu (ppm)	4.00	39.00	19.62	10.66	54.33	0.35	−1.12
Mn (ppm)	160.00	250.00	198.26	26.00	13.12	0.39	−0.92
Zn (ppm)	28.00	124.00	76.80	32.46	42.26	−0.02	−1.29
B (ppm)	16.00	49.00	34.38	9.82	28.56	−0.31	−0.99
Chlorophyll A (mg/g fresh weight)	2.02	3.56	2.73	0.51	18.72	0.24	−1.36
Chlorophyll B (mg/g fresh weight)	1.22	2.33	1.84	0.36	19.67	−0.30	−1.26
Total chlorophyll (mg/g fresh weight)	3.25	5.89	4.58	0.87	18.97	0.03	−1.36
Total carotenoids (mg/g fresh weight)	2.23	3.39	2.78	0.39	14.09	0.23	−1.42
Index 1 (%)	0.40	1.11	0.78	0.20	25.89	−0.16	−1.17
Index 2 (ppm)	59.88	221.87	125.36	47.14	37.60	0.54	−0.91
No. of points	81	81	81	81	81	81	81

Table 2

Distribution of mean * values of N, P, K, Ca, and Mg in leaves petioles of Flame Seedless grape cultivated by different agricultural practices in investigated vineyards.

Vineyards	N (%)		P (%)		K (%)		Ca (%)		Mg (%)
	Mean	Standard Deviation	Mean	Standard Deviation	Mean	Standard Deviation	Mean	Standard Deviation	Mean	Standard Deviation
S1	1.53 i	0.021	0.12 i	0.007	1.03 i	0.016	1.94 i	0.019	0.21 i	0.005
S2	1.65 h	0.022	0.18 h	0.010	1.12 h	0.011	2.11 h	0.032	0.27 h	0.024
S3	1.76 g	0.020	0.23 g	0.005	1.22 g	0.026	2.24 g	0.019	0.35 g	0.016
S4	1.84 f	0.019	0.28 f	0.007	1.32 f	0.022	2.35 f	0.017	0.43 f	0.010
S5	1.91 e	0.040	0.35 e	0.005	1.40 e	0.020	2.46 e	0.021	0.48 e	0.008
S6	2.10 d	0.035	0.39 d	0.005	1.48 d	0.011	2.56 d	0.018	0.53 d	0.009
S7	2.25 c	0.032	0.42 c	0.007	1.53 c	0.011	2.65 c	0.028	0.58 c	0.008
S8	2.44 b	0.027	0.46 b	0.004	1.58 b	0.010	2.75 b	0.014	0.63 b	0.007
S9	2.55 a	0.018	0.49 a	0.012	1.63 a	0.009	2.86 a	0.023	0.68 a	0.008
LSD (5%)	0.025		0.0065		0.0148		0.0205		0.0107

* Different letters indicate that means are significantly different from each other (p < 0.05).

Table 3

Distribution of mean * values of Fe, Cu, Mn, and Zn in leaves petioles of Flame Seedless grape cultivated by different agricultural practices in investigated vineyards.

Vineyards	Fe (ppm)		Cu (ppm)		Mn (ppm)		Zn (ppm)		B(ppm)
	Mean	Standard Deviation	Mean	Standard Deviation	Mean	Standard Deviation	Mean	Standard Deviation	Mean	Standard Deviation
S1	42.67 i	1.32	5.33 i	0.71	163.00 i	1.66	29.33 i	1.12	17.00 i	0.87
S2	47.56 h	1.01	8.00 h	1.12	170.56 h	1.67	34.67 h	1.00	23.33 h	0.50
S3	58.44 g	0.73	11.56 g	0.53	177.56 g	1.13	52.78 g	0.83	27.89 g	0.60
S4	76.00 f	1.66	14.56 f	0.88	186.00 f	1.22	67.67 f	2.92	31.78 f	0.67
S5	86.78 e	1.39	17.00 e	0.87	194.22 e	1.20	75.33 e	0.71	35.56 e	0.53
S6	107.33 d	1.73	23.33 d	0.50	205.89 d	2.62	87.11 d	1.62	38.89 d	0.60
S7	115.44 c	1.74	25.56 c	0.53	215.44 c	1.94	102.78 c	0.83	41.89 c	0.60
S8	122.11 b	1.05	33.67 b	1.00	225.89 b	2.32	118.56 b	0.53	44.78 b	0.67
S9	137.89 a	1.05	37.56 a	0.73	245.78 a	2.86	123.00 a	0.87	48.33 a	0.50
LSD (5%)	1.229		0.751		1.871		1.262		0.560

* Different letters indicate that means are significantly different from each other (p < 0.05).

Table 4

Statistical criteria of experimental results of quality attributes of Flame Seedless grape cultivated by different agricultural practices.

Quality Attributes	Statistical Criteria
	Minimum	Maximum	Mean	Standard Deviation	Coefficient of Variation (%)	Skewness	Kurtz
Cluster weight (g)	325.00	439.00	395.67	34.07	8.61	−0.60	−0.85
Cluster length (cm)	13.50	17.50	15.82	1.29	8.12	−0.40	−1.18
Cluster shoulder length (cm)	9.20	13.60	11.47	1.48	12.92	−0.07	−1.43
Berry L*	25.65	46.45	35.13	6.79	19.33	0.18	−1.35
Berry a*	5.24	10.99	8.58	1.72	20.10	−0.48	−0.80
Berry b*	3.12	12.96	7.83	3.18	40.61	0.07	−1.28
Berry chroma	11.09	14.01	12.12	0.94	7.79	0.68	−0.89
Berry hue (°)	15.85	67.85	41.15	16.63	40.41	0.04	−1.24
Color index for red grapes (CIRG)	1.86	4.43	3.07	0.85	27.52	0.15	−1.33
No. of points	81	81	81	81	81	81	81

Table 5

Distribution of mean * values of cluster weight, cluster length, cluster shoulder length, color attributes of berries of Flame Seedless grape cultivated by different agricultural practices in investigated vineyards.

Vineyards	Cluster Weight	Cluster Length	Cluster Shoulder Length	L*	a*	b*	Chroma	Hue	CIRG
	(g)	(cm)	(cm)	(−)	(−)	(−)	(−)	(°)	(−)
S1	332.11 i	13.610	9.29 i	45.90 a	5.32 i	12.83 a	13.89 a	67.47 a	1.88 i
S2	354.44 h	14.24 h	9.62 h	43.16 b	6.74 h	11.51 b	13.34 b	59.67 b	2.13 h
S3	374.89 g	14.84 g	10.32 g	40.76 c	7.48 g	10.18 c	12.64 c	53.69 c	2.37 g
S4	390.56 f	15.51 f	10.78 f	37.88 d	8.24 f	9.07 d	12.25 d	47.72 d	2.64 f
S5	404.89 e	16.21 e	11.51 e	33.83 e	8.86 e	7.92 e	11.88 e	41.78 e	3.02 e
S6	415.44 d	16.51 d	12.24 d	31.88 f	9.39 d	6.39 f	11.36 f	34.24 f	3.37 d
S7	421.78 c	16.83 c	12.72 c	29.37 g	9.91 c	5.22 g	11.20 g	27.81 g	3.75 c
S8	430.22 b	17.20 b	13.24 b	27.60 h	10.38 b	4.12 h	11.17 g	21.64 h	4.08 b
S9	436.67 a	17.46 a	13.52 a	25.77 i	10.88 a	3.19 i	11.34 f	16.35 i	4.41 a
LSD (5%)	3.2061	0.0596	0.0666	0.2707	0.0841	0.0518	0.0712	0.3119	0.0165

* Different letters indicate that means are significantly different from each other (p < 0.05).

Table 6

Pearson’s correlation coefficients (r) describing the correlations of nutritional status, index 1, index 2 and Cluster attributes and berries’ color attributes of Flame Seedless grape cultivated by different agricultural practices.

Variables	Cluster Weight	Cluster Length	Cluster Shoulder Length	L*	a*	b*	Chroma	Hue	CIRG
N	0.937	0.952	0.981	−0.971	0.957	−0.986	−0.895	−0.988	0.992
P	0.987	0.995	0.991	−0.996	0.992	−0.994	−0.965	−0.993	0.984
K	0.988	0.994	0.988	−0.992	0.991	−0.990	−0.971	−0.988	0.977
Ca	0.982	0.989	0.988	−0.992	0.991	−0.994	−0.955	−0.994	0.985
Mg	0.984	0.991	0.988	−0.991	0.989	−0.993	−0.956	−0.992	0.982
Fe	0.963	0.979	0.993	−0.991	0.971	−0.993	−0.937	−0.992	0.992
Cu	0.932	0.950	0.977	−0.968	0.951	−0.982	−0.885	−0.984	0.990
Mn	0.933	0.952	0.975	−0.970	0.953	−0.982	−0.882	−0.984	0.990
Zn	0.967	0.981	0.993	−0.989	0.977	−0.995	−0.938	−0.994	0.992
B	0.987	0.992	0.986	−0.991	0.995	−0.992	−0.961	−0.992	0.980
Index 1	0.959	0.969	0.976	−0.976	0.970	−0.978	−0.937	−0.979	0.973
Index 2	0.874	0.898	0.937	−0.925	0.894	−0.942	−0.821	−0.944	0.959
Chlorophyll A	0.939	0.961	0.988	−0.982	0.958	−0.990	−0.901	−0.991	0.998
Chlorophyll B	0.985	0.994	0.989	−0.995	0.991	−0.992	−0.967	−0.991	0.981
Total chlorophyll	0.965	0.982	0.995	−0.994	0.978	−0.997	−0.935	−0.998	0.998
Total carotenoids	0.937	0.956	0.986	−0.978	0.956	−0.988	−0.906	−0.989	0.995

Table 7

Pearson’s correlation coefficients (r) describing the correlations among quality attributes of Flame Seedless grape cultivated by different agricultural practices.

Quality Attributes	Cluster Weight	Cluster Length	Shoulder Length	Berry L*	Berry a*	Berry b*	Berry Chroma	Berry Hue	Berry CIRG
Cluster weight	1.000	0.992	0.971	−0.981	0.992	−0.975	−0.978	−0.973	0.954
Cluster length		1.000	0.985	−0.994	0.992	−0.987	−0.974	−0.985	0.973
Cluster shoulder length			1.000	−0.994	0.978	−0.997	−0.950	−0.995	0.993
Berry L*				1.000	−0.987	0.995	0.957	0.994	−0.990
Berry a*					1.000	−0.986	−0.966	−0.986	0.970
Berry b*						1.000	0.949	1.000	−0.995
Berry chroma							1.000	0.942	−0.917
Berry hue								1.000	−0.996
Berry CIRG									1.000

Table 8

Coefficients of determination (R²) and regression coefficients values of regression equation (Equation (9)) for modeling cluster weight, cluster length, cluster shoulder length, berries L*, a*, b*, C, h, and CIRG of Flame Seedless grape.

Independent Variables	Regression Coefficients Symbol	Cluster Weight (g)	Cluster Length (cm)	Shoulder Length (cm)	L*	a*	b*	Chroma	Hue °	CIRG
Intercept	${\beta }_0$	286.607	11.366	4.123	64.991	1.731	23.697	15.289	125.497	-1.494
Index 1	${\beta }_1$	60.133	1.279	1.134	−4.877	2.16	−2.01	−2.21	−9.874	0.127
Index 2	${\beta }_2$	−0.185	−0.005	0	0.012	−0.007	0.001	0.01	−0.003	0.002
Total chlorophyll	${\beta }_3$	78.089	2.868	1.58	−11.024	3.29	−3.472	−2.242	−17.009	0.601
Total carotenoids	${\beta }_4$	−98.116	−3.273	−0.257	8.27	−3.23	0.547	2.717	0.611	0.517
R 2		0.976	0.991	0.991	0.995	0.981	0.996	0.941	0.996	0.998
No. of data points		61	61	61	61	61	61

Table 9

The sensitivity analysis of the input parameters on cluster weight, cluster length, cluster shoulder length, berries L*, a*, b*, chroma, hue, and CIRG.

MLR Model		Cluster weight	Cluster length	Cluster shoulder length	Berry L*	Berry a*	Berry b*	Berry Chroma	Berry hue	Berry CIRG
MLR without index 1	RMSE	5.909	0.134	0.148	0.506	0.253	0.219	0.244	1.1	0.038
	R 2	0.969	0.989	0.99	0.994	0.978	0.995	0.929	0.996	0.998
MLR without index 2	RMSE	5.891	0.138	0.139	0.487	0.258	0.199	0.267	1.006	0.049
	R 2	0.969	0.988	0.991	0.995	0.977	0.996	0.915	0.996	0.997
MLR without total chlorophyll	RMSE	8.45	0.271	0.193	1.04	0.364	0.355	0.293	1.758	0.063
	R 2	0.937	0.954	0.983	0.976	0.954	0.987	0.898	0.989	0.994
MLR without total carotenoids	RMSE	6.838	0.189	0.14	0.587	0.275	0.2	0.254	1.005	0.044
	R 2	0.959	0.978	0.991	0.992	0.974	0.996	0.923	0.996	0.997

References

1. Iland, P.; Gago, P.; Callard, A.; Dry, P. A Taste of the World of Wine; Patrick Iland Wine Promotions: Adelaide, Australia, 2010.

2. Demir, K.O.K. A review on grape growing in tropical regions. Turk. J. Agric. Nat. Sci.; 2014; 1, pp. 1236-1241.

3. Taskesenlioglu, M.Y.; Ercisli, S.; Kupe, M.; Ercisli, N. History of Grape in Anatolia and Historical Sustainable Grape Production in Erzincan Agroecological Conditions in Turkey. Sustainability; 2022; 14, 1496. [DOI: https://dx.doi.org/10.3390/su14031496]

4. Kupe, M.; Ercisli, S.; Baron, M.; Sochor, J. Sustainable viticulture on traditional ‘Baran’ training system in Eastern Tukey. Sustainability; 2021; 13, 10236. [DOI: https://dx.doi.org/10.3390/su131810236]

5. Mansour, A.H.A.; Mahmoud, G.E.; Mohamed, A.A. Fungal biodiversity, total loss and berry quality of flame seedless grape during storage as affected with natural oils pre- harvest spraying. Fayoum J. Agric. Res. Dev.; 2019; 33, pp. 340-355.

6. Alkan, A.; Abdullah, M.Ü.; Abdullah, H.O.; Assaf, M.; Zhou, H. A smart agricultural application: Automated detection of diseases in vine leaves using hybrid deep learning. Turk. J. Agric. For.; 2021; 45, pp. 717-729. [DOI: https://dx.doi.org/10.3906/tar-2007-105]

7. Abdel-Sattar, M.; Haikal, A.M.; Hammad, S.E. Meloidogyne incognita population control and nutritional status and productivity of Thompson seedless grapevines managed with different treatments. PLoS ONE; 2020; 15, e0239993. [DOI: https://dx.doi.org/10.1371/journal.pone.0239993]

8. Xu, C.; Yagiz, Y.; Zhao, L.; Simonne, A.; Lu, J.; Marshall, M.R. Fruit quality, nutraceutical and antimicrobial properties of 58 Muscadine grape varieties (Vitis rotundifolia Michx.) grown in the United States. Food Chem.; 2017; 215, pp. 149-156. [DOI: https://dx.doi.org/10.1016/j.foodchem.2016.07.163]

9. Chitarrini, G.; Zulini, L.; Masuero, D.; Vrhovsek, U. Lipid, phenol and carotenoid changes in ‘Bianca’ grapevine leaves after mechanical wounding: A case study. Protoplasma; 2017; 254, pp. 2095-2106. [DOI: https://dx.doi.org/10.1007/s00709-017-1100-5]

10. Fuleki, T.; Ricardo-da-silva, M.J. Effects of cultivar and processing method on the contents of catechins and procyanidins in grape juice. J. Agric. Food Chem.; 2003; 51, pp. 640-646. [DOI: https://dx.doi.org/10.1021/jf020689m]

11. Liang, Z.; Owens, C.L.; Zhong, G.Y.; Cheng, L. Polyphenolic profiles detected in the ripe berries of Vitis vinifera germplasm. Food Chem.; 2011; 129, pp. 940-950. [DOI: https://dx.doi.org/10.1016/j.foodchem.2011.05.050]

12. Gohari, G.; Zareei, E.; Kulak, M.; Labib, P.; Mahmoudi, R.; Panahirad, S.; Jafari, H.; Mahdavinia, G.; Juárez-Maldonado, A.; Lorenzo, J.M. Improving the berry quality and antioxidant potential of Flame Seedless grapes by foliar application of chitosan–phenylalanine nanocomposites (CS–Phe NCs). Nanomaterials; 2021; 11, 2287. [DOI: https://dx.doi.org/10.3390/nano11092287] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34578605]

13. Lo’ay, A.A.; El-khateeb, A.Y. Evaluation the effect of rootstocks on postharvest berries quality of ‘Flame Seedless’ grapes. Sci. Hortic.; 2017; 220, pp. 299-302. [DOI: https://dx.doi.org/10.1016/j.scienta.2017.04.006]

14. Lo’ay, A.A.; EL-Boray, M.S. Improving fruit cluster quality attributes of ‘Flame Seedless’ grapes using preharvest application of ascorbic and salicylic acid. Sci. Hortic.; 2018; 233, pp. 339-348. [DOI: https://dx.doi.org/10.1016/j.scienta.2018.02.010]

15. Fidelibus, M.; El-kereamy, A.; Zhuang, G.; Haviland, D.; Hembree, K.; Stewart, D. Sample Costs to Establish and Produce Table Grapes. San Joaquin Valley South. Flame Seedless, Early Maturing; UC Agricultural Issues Center: Davis, CA, USA, 2018.

16. ElMasry, G.; Wang, N.; ElSayed, A.; Ngadi, M. Hyperspectral imaging for nondestructive determination of some quality attributes for strawberry. J. Food Eng.; 2007; 81, pp. 98-107. [DOI: https://dx.doi.org/10.1016/j.jfoodeng.2006.10.016]

17. Cen, H.; Lu, R.; Mendoza, F.A.; Ariana, D.P. Peach maturity/quality assessment using hyperspectral imaging-based spatially resolved technique. Sensing for Agriculture and Food Quality and Safety III, Proceedings of the SPIE Defense, Security, and Sensing, Orlando, FL, USA, 25–29 April 2011; SPIE: Bellingham, WA, USA, 2011; Volume 8027, pp. 165-179.

18. Milosevic, T.; Milosevic, N.; Glisic, I. Early tree performances, precocity and fruit quality attributes of newly introduced apricot cultivars grown under western Serbian conditions. Turk. J. Agric. For.; 2021; 45, pp. 819-833. [DOI: https://dx.doi.org/10.3906/tar-2010-39]

19. Zhao, H.; Yao, Y.; Yang, L.; Tian, Y.; Tian, M.; Liu, X. Pitaya research status and its development prospect. Shandong Chem. Ind.; 2018; 47, pp. 52-54. [DOI: https://dx.doi.org/10.19319/j.cnki.issn.1008-021x.2018.14.019]

20. Rustioni, L.; Grossi, D.; Brancadoro, L.; Failla, O. Iron, magnesium, nitrogen and potassium deficiency symptom discrimination by reflectance spectroscopy in grapevine leaves. Sci. Hortic.; 2018; 241, pp. 152-159. [DOI: https://dx.doi.org/10.1016/j.scienta.2018.06.097]

21. Amaro, A.C.E.; Souza, E.R.; Santos, L.S.; Baron, D.; Ono, E.O.; Rodrigues, J.D. Foliar application of plant growth regulators to ‘Crimson Seedless’ grapevine influences leaf age. Aust. J. Crop Sci.; 2020; 14, pp. 422-430. [DOI: https://dx.doi.org/10.21475/ajcs.20.14.03.p1874]

22. Wu, B.; Niu, N.; Li, J.; Li, S. Leaf: Fruit ratio affects the proteomic profile of grape berry skins. J. Am. Soc. Hortic. Sci.; 2013; 138, pp. 416-427.

23. Ponchia, G.; Fila, G.; Gardiman, M. Effects of rootstock and inter stem on growth, productivity and mineral nutrition of ‘Golden Delicious’ apple trees. ActaHorticulturae (ISHS); 1997; 448, pp. 107-112.

24. Ikinci, A.; Bolat, I.; Ercisli, S.; Kodad, O. Influence of rootstocks on growth, yield, fruit quality and leaf mineral element contents of pear cv. ‘Santa Maria’ in semi-arid conditions. Biol. Res.; 2014; 47, pp. 1-8. [DOI: https://dx.doi.org/10.1186/0717-6287-47-71] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25723734]

25. Blažek, J.; Hlušičková, I. Orchard performance and fruit quality of 50 apple cultivars grown or tested in commercial orchards of the Czech Republic. Hortic. Sci.; 2007; 34, pp. 96-106. [DOI: https://dx.doi.org/10.17221/1888-HORTSCI]

26. Sadowski, A.; Kepka, M.; Lenz, F.; Engel, G. Effect of fruit load on leaf nutrient content of apple trees. ActaHorticulturae (ISHS); 1995; 383, pp. 67-72.

27. Weinbaum, S.A.; Brown, P.H.; Rosecrance, R.C.; Picchioni, G.A.; Niederholzer, F.J.A.; Youseffi, F.; Muraoka, T.T. Necessity for whole tree excavations in determining patterns and magnitude of macronutrient uptake by mature deciduous fruit trees. ActaHorticulturae (ISHS); 2001; 564, pp. 41-49. [DOI: https://dx.doi.org/10.17660/ActaHortic.2001.564.3]

28. Holb, I.J.; Gonda, I.; Vágó, I.; Nagy, P.T. Seasonal dynamics of nitrogen, phosphorus, and potassium contents of leaf and soil in environmental friendly apple orchards. Soil Sci. Plant Anal.; 2009; 40, pp. 694-705. [DOI: https://dx.doi.org/10.1080/00103620802693359]

29. Hipps, N.A. Effects of nitrogen, phosphorus, water and pre-planting soil sterilisation on growth and yield of Queen Cox/M9 apple trees. ActaHorticulturae (ISHS); 1997; 448, pp. 125-132. [DOI: https://dx.doi.org/10.17660/ActaHortic.1997.448.19]

30. Nagy, P.T.; Holb, I.J. Study on the macronutrient content of apple leaves in an organic apple orchard. J. Cent. Eur. Agric.; 2006; 7, pp. 329-336.

31. Jivan, C.; Sala, F. Relationship between tree nutritional status and apple quality. Hortic. Sci.; 2014; 41, pp. 1-9.

32. Usha, K.; Singh, B. Effect of macro and micro-nutrient spray on fruit yield and quality of grape (Vitis vinifera L.) cv. Perlette. ActaHorticulturae; 2002; 594, pp. 197-202. [DOI: https://dx.doi.org/10.17660/ActaHortic.2002.594.21]

33. Lebon, E.; Pellegrino, A.; Louarn, G.; Lecoeur, J. Branch development controls leaf area dynamics in grapevine (Vitisvinifera) growing in drying soil. Ann. Bot.; 2006; 98, pp. 175-185. [DOI: https://dx.doi.org/10.1093/aob/mcl085]

34. Fallahi, E.; Fallahi, B.; Neilsen, G.H.; Neilsen, D.; Peryea, F.J. Effects of mineral nutrition on fruit quality and nutritional disorders in apples. ActaHorticulturae (ISHS); 2008; 868, pp. 49-60. [DOI: https://dx.doi.org/10.17660/ActaHortic.2010.868.3]

35. Huang, X.; Wang, H.; Qu, S.; Luo, W.; Gao, Z. Using artificial neural network in predicting the key fruit quality of loquat. Food Sci. Nutr.; 2021; 9, pp. 1780-1791. [DOI: https://dx.doi.org/10.1002/fsn3.2166] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33747488]

36. Guo-yi, W.; Xin-zhong, Z.; Yi, W.; Xue-feng, X.; Zhen-hai, H. Key minerals influencing apple quality in Chinese orchard identified by nutritional diagnosis of leaf and soil analysis. J. Integr. Agric.; 2015; 14, pp. 864-874. [DOI: https://dx.doi.org/10.1016/S2095-3119(14)60877-7]

37. Yu, X.; Wen, T.; Ma, S.; Liu, D.; Hu, W.; Liu, Y.; Yang, L. Correlation between fruit quality and mineral nutrients in soil, leaf and fruit of ‘Jinsha’ pomelo. Acta Agric. Univ. Jiangxiensis; 2021; 43, pp. 70-81. [DOI: https://dx.doi.org/10.13836/j.jjau.2021009]

38. Sun, H.; Huang, X.; Chen, T.; Zhou, P.; Huang, X.; Jin, W.; Liu, D.; Zhang, H.; Zhou, J.; Wang, Z. et al. Fruit quality prediction based on soil mineral element content in peach orchard. Food Sci. Nutr.; 2022; 10, pp. 1756-1767. [DOI: https://dx.doi.org/10.1002/fsn3.2794]

39. Niedbała, G. Application of multiple linear regression for multi-criteria yield prediction of winter wheat. J. Res. Appl. Agric. Eng.; 2018; 63, pp. 125-131.

40. Torkashvand, A.M.; Ahmadi, A.; Nikravesh, N.L. Prediction of kiwifruit firmness using fruit mineral nutrient concentration by artificial neural network (ANN) and multiple linear regressions (MLR). J. Integr. Agric.; 2017; 16, pp. 1634-1644. [DOI: https://dx.doi.org/10.1016/S2095-3119(16)61546-0]

41. Niazian, M.; Sadat-Noori, S.A.; Abdipour, M. Artificial neural network and multiple regression analysis models to predict essential oil content of ajowan (Carum copticum L.). J. Appl. Res. Med. Aromat. Plants; 2018; 9, pp. 124-131. [DOI: https://dx.doi.org/10.1016/j.jarmap.2018.04.001]

42. Ivanovski, T.; Zhang, G.; Jemrić, T.; Gulić, M.; Matetić, M. Fruit firmness prediction using multiple linear regression. Proceedings of the 2020 43rd International Convention on Information, Communication and Electronic Technology (MIPRO); Opatija, Croatia, 28 September–2 October 2020; pp. 1306-1311. [DOI: https://dx.doi.org/10.23919/MIPRO48935.2020.9245213]

43. Shezi, S.; Magwaza, L.S.; Tesfay, S.Z.; Mditshwa, A. Simple and multiple linear regression models for predicting maturity of ‘Mendez#1’ and ‘Hass’ avocado fruit harvested from inside and outside tree canopy positions. Int. J. Fruit Sci.; 2020; 20, pp. 1969-1983. [DOI: https://dx.doi.org/10.1080/15538362.2020.1839626]

44. Stangierski, J.; Weiss, D.; Kaczmarek, A. Multiple regression models and Artificial Neural Network (ANN) as prediction tools of changes in overall quality during the storage of spreadable processed Gouda cheese. Eur. Food Res. Technol.; 2019; 245, pp. 2539-2547. [DOI: https://dx.doi.org/10.1007/s00217-019-03369-y]

45. Tortoe, C.; Orchard, J.; Beezer, A. Multilinear regression approach in predicting osmo-dehydration processes of apple, banana and potato. J. Food Process. Technol.; 2011; 2, pp. 1-6. [DOI: https://dx.doi.org/10.4172/2157-7110.1000122]

46. Pang, X.; Qin, Z.; Zhao, L.; Cheng, H.; Hu, X.; Song, H.; Wu, J. Development of regression model to differentiate quality of black tea (Dianhong): Correlate aroma properties with instrumental data using multiple linear regression analysis. Int. J. Food Sci. Technol.; 2012; 47, pp. 2372-2379. [DOI: https://dx.doi.org/10.1111/j.1365-2621.2012.03112.x]

47. Sundari, R.; Hadibarata, T.; Abdul Malik, R.F.; Aziz, M. Multiple linear regression (MLR) modeling of wastewater in Urban region of Southern Malaysia. J. Sustain. Sci. Manag.; 2012; 8, pp. 93-102.

48. Lieverloo, J.H.M.; Roode, M.; Fox, M.B.; Zwieterung, M.H.; Wells-Bennik, M.H.J. Multiple regression model for thermal inactivation of Listeria monocytogenes in liquid food products. Food Control; 2013; 29, pp. 394-400. [DOI: https://dx.doi.org/10.1016/j.foodcont.2012.05.078]

49. Eregno, A.F.E. Multiple linear regression models for estimating microbial load in a drinking water source case from the Glomma River, Norway. Master’s Thesis; Department of Mathematical Sciences and Technology, Norwegian University of Life Sciences: Ås, Norway, 2014.

50. Kerna, C.; Stefan, T.; Jörg, H. Multiple linear regression modeling: Prediction of cheese curd dry matter during curd treatment. Food Res. Int.; 2019; 121, pp. 471-478. [DOI: https://dx.doi.org/10.1016/j.foodres.2018.11.061]

51. Amoriello, T.; Ciccoritti, R.; Ferrante, P. Prediction of Strawberries’ Quality Parameters Using Artificial Neural Networks. Agronomy; 2022; 12, 963. [DOI: https://dx.doi.org/10.3390/agronomy12040963]

52. Bravdo, B.A. Effect of cultural practices and environmental factors on wine production and quality. ActaHorticulturae (ISHS); 2004; 652, pp. 119-124. [DOI: https://dx.doi.org/10.17660/ActaHortic.2004.652.13]

53. Kamiloǧlu, Ö. Influence of some cultural practices on yield, fruit quality and individual anthocyanins of table grape cv. ‘HorozKarasi’. J. Anim. Plant Sci.; 2011; 21, pp. 240-245.

54. Conte, A.; Mastromatteo, M.; Antonacci, D.; Del Nobile, M.A. Influence of cultural practices and packaging materials on table grape quality. J. Food Process Eng.; 2012; 35, pp. 701-707. [DOI: https://dx.doi.org/10.1111/j.1745-4530.2010.00616.x]

55. Isci, B.; Gokbayrak, Z.; Keskin, N. Effects of cultural practices on total phenolics and vitamin C content of organic table grapes. S. Afr. J. Enol. Vitic.; 2015; 36, pp. 191-197. [DOI: https://dx.doi.org/10.21548/36-2-952]

56. Ghantous, G.; El Sebaaly, Z.; Kfoury, L.; Rizkallah, J.; Sassine, Y.N. Effect of different cultural practices on the performance of the red wine grape cultivar ‘Malbec’ recently grown in Bekaa, Lebanon. ActaHorticulturae; 2020; 1276, pp. 7-14. [DOI: https://dx.doi.org/10.17660/ActaHortic.2020.1276.2]

57. Hegazi, A.; Sallam, A.E. Cluster and berry characteristics of flame seedless grapes under different environmental conditions in Egypt. ActaHorticulturae; 2003; 603, pp. 201-204. [DOI: https://dx.doi.org/10.17660/ActaHortic.2003.603.27]

58. Stino, R.G.; Ali, M.A.; Abdel-Mohsen, M.A.; Maksoud, M.A.; Thabet, A.Y.I. Quality attributes of Flame seedless grapes as affected by some bio-stimulants. Int. J. ChemTech Res.; 2017; 10, pp. 273-288.

59. Abd EL-Rahman, M.M.A.; Khodair, O.A.; Hamed, M.H. Impact of organic, bio fertilization and humic acid on growth and fruiting of Flame Seedless grapevines under sandy soil conditions. J. Plant Prod. Mansoura Univ.; 2021; 12, pp. 171-177. [DOI: https://dx.doi.org/10.21608/jpp.2021.58050.1015]

60. Mosa, W.F.A.; Salem, M.Z.M.; Al-Huqail, A.A.; Ali, H.M. Application of glycine, folic acid, and moringa extract as bio-stimulants for enhancing the production of ‘Flame Seedless’ grape cultivar. BioResources; 2021; 16, pp. 3391-3410. [DOI: https://dx.doi.org/10.15376/biores.16.2.3391-3410]

61. Abd El-Razek, E.; Treutter, D.; Saleh, M.M.S.; El-Shammaa, M.; Fouad, A.A.; Abdel-Hamid, N. Effect of nitrogen and potassium fertilization on productivity and fruit quality of ‘crimson seedless’ grape. Agric. Biol. J. N. Am.; 2011; 2, pp. 330-340. [DOI: https://dx.doi.org/10.5251/abjna.2011.2.2.330.340]

62. Ali, I.; Wang, X.; Abbas, W.M.; Hassan, M.U.; Shafique, M.; Tareen, M.J.; Fiaz, S.; Ahmed, W.; Qayyum, A. Quality responses of table grapes ‘Flame Seedless’ as effected by foliarly applied micronutrients. Horticulturae; 2021; 7, 462. [DOI: https://dx.doi.org/10.3390/horticulturae7110462]

63. Summer, M.E. Diagnosis and recommendation integrated system (DRIB) as a guide to orchard fertilization. Hort. Abst.; 1985; 55, 7502.

64. Wilde, S.A.; Corey, R.B.; Lyer, I.G.; Voigt, G.K. Soil and Plant Analysis for Tree Culture; 3rd ed. Oxford Publishing Co.: New Delhi, India, 1985; pp. 1-218.

65. Chapman, H.D.; Pratt, P.P. Method of Analysis for Soils Plants and Water; Division of Agricultural Sciences, University of California: Berkeley, CA, USA, 1965; pp. 168-173.

66. Peach, K.; Tracey, I.M.V. Modern Methods of Plants Analysis; Springer: Berlin/Heidelberg, Germany, 1955; Volume 4.

67. Cottenie, A.; Verloo, M.; Kiekens, L. Chemical Analysis of Plant and Soil; Laboratory of Analytical and Agrochemistry, State University of Belgium: Leuven, Belgium, 1982.

68. Hatcher, J.T.; Wilcox, L.V. Colorimetric determination of boron using carmine. Anal. Chem.; 1950; 22, pp. 567-569. [DOI: https://dx.doi.org/10.1021/ac60040a017]

69. Von Wettstein, D. Chlorophyll-letale and der sub-mikroskopishe formweschsel derplastiden. Exp. Cell Res.; 1957; 12, pp. 427-433. [DOI: https://dx.doi.org/10.1016/0014-4827(57)90165-9]

70. Hiscox, J.D.; Isralstam, G.F. A method for the extraction of chlorophyll from leaf tissue without maceration. Can. J. Bot.; 1979; 57, pp. 1332-1334. [DOI: https://dx.doi.org/10.1139/b79-163]

71. McGuire, R.G. Reporting of objective color measurements. HortScience; 1992; 27, pp. 1254-1255. [DOI: https://dx.doi.org/10.21273/HORTSCI.27.12.1254]

72. Sancho, L.E.; Yahia, E.M.; Tellez, M.A.; Aguilar, G.A.G. Effect of maturity stage of papaya Maradol on physiological and biochemical parameters. Am. J. Agric. Biol. Sci.; 2010; 2, pp. 194-203. [DOI: https://dx.doi.org/10.3844/ajabssp.2010.194.203]

73. Carreño, J.; Martinez, A.; Almela, L.; Fernandez-Lopez, J.A. Proposal of an index for the objective evaluation of the colour of red table grapes. Food Res. Int.; 1995; 28, pp. 373-377. [DOI: https://dx.doi.org/10.1016/0963-9969(95)00008-A]

74. The SAS System for Windows; Version 9.13 SAS Institute Inc.: Cary, NC, USA, 2004.

75. Bassiony, S.S.; Mahmoud, M.A. Impact of magnetic iron applications and foliar spray of some antioxidants on growth, and productivity of “Flame seedless” grapevines under salt affected soil. Ann. Agric. Sci. Moshtohor (ASSJM); 2019; 57, pp. 113-128. [DOI: https://dx.doi.org/10.21608/assjm.2019.42222]

76. Carreño, J.; Faray, S.; Martinez, A. Effects of girdling and covering mesh on ripening, color and fruit characteristics of ‘Italia’ grapes. J. Hortic. Sci. Biotechnol.; 1998; 73, pp. 103-106. [DOI: https://dx.doi.org/10.1080/14620316.1998.11510951]

77. Cun, D.; Pu, J.; Gao, J.; Li, J.; Yue, J.; Du, Y.; Li, D.; Li, X.; Chen, L.; Zhou, X. et al. The relationship between mineral element content in the peel and pulp of Yunnan Bingtang orange and fruit quality. Jiangsu Agric. Sci.; 2020; 48, pp. 207-214. [DOI: https://dx.doi.org/10.15889/j.issn.1002-1302.2020.24.040]

78. Aular, J.; Natale, W. Nutrição mineral e qualidade do fruto de algumasfrutíferastropicais: Goiabeira, mangueira, bananeira e mamoeiro. Rev. Bras. Frutic.; 2013; 35, pp. 1214-1231. [DOI: https://dx.doi.org/10.1590/S0100-29452013000400033]

79. Borges, R.M.E.; Coêlho de Lima, M.A.; Franklin de Melo, N. Correlations between fruit yield and sensory and nutritional quality of pumpkin (Cucurbita moschata Duch.). Aust. J. Crop Sci.; 2019; 13, pp. 1676-1682. [DOI: https://dx.doi.org/10.21475/ajcs.19.13.10.p1914]

80. Farinelli, D.; Portarena, S.; da Silva, D.F.; Traini, C.; da Silva, G.M.; da Silva, E.C.; da Veiga, J.F.; Pollegioni, P.; Villa, F. Variability of fruit quality among 103 Acerola (Malpighiaemarginata D. C.) phenotypes from the subtropical region of Brazil. Agriculture; 2021; 11, 1078. [DOI: https://dx.doi.org/10.3390/agriculture11111078]

81. Dris, R.; Niskanen, R.; Fallahi, E. Relationships between leaf and fruit minerals and fruit quality attributes of apples grown under northern conditions. J. Plant Nutr.; 1999; 22, pp. 1839-1851. [DOI: https://dx.doi.org/10.1080/01904169909365760]

82. Li, Y.; Han, M.-Q.; Lin, F.; Ten, Y.; Lin, J.; Zhu, D.-H.; Guo, P.; Weng, Y.-B.; Chen, L.-S. Soil chemical properties, Guanximiyou pummelo leaf mineral nutrient status and fruit quality in the southern region of Fujian province, China. J. Soil Sci. Plant Nutr.; 2015; 15, pp. 615-628. [DOI: https://dx.doi.org/10.4067/S0718-95162015005000029]

83. Hunsche, M.; Brackmann, A.; Ernani, P.R. Effect of potassium fertilization on the postharvest quality of ‘Fuji’ apples. Pesqui. Agropecu. Bras.; 2003; 38, pp. 489-496. [DOI: https://dx.doi.org/10.1590/S0100-204X2003000400007]

84. Zhang, L.X.; Zhang, L.S.; Li, B.Z.; Han, M.Y. Mineral nutrition elements and their roles in growth and development of apple trees in arid areas. J. Northwest For. Univ.; 1997; 22, pp. 111-115. (In Chinese)

85. Shi, Y.H.; Liu, P. A review of advances in physiological function of boron in plants. Subtrop. Plant Sci.; 2002; 31, pp. 64-69.

86. Wojcik, P.; Wojcik, M.; Klamkowski, K. Response of apple trees to boron fertilization under conditions of low soil boron availability. Sci. Hortic.; 2008; 116, pp. 58-64. [DOI: https://dx.doi.org/10.1016/j.scienta.2007.10.032]

87. Mokarram, M.; Amin, H.; Khosravi, M.R. Using adaptive Neuro-fuzzy inference system and multiple linear regression to estimate orange taste. Food Sci. Nutr.; 2019; 7, pp. 3176-3184. [DOI: https://dx.doi.org/10.1002/fsn3.1149]

88. Gajewski, M.; Arasimowicz, D. Sensory quality of eggplant fruits (Solanum melongena L.) as affected by cultivar and maturity stage. Pol. J. Food Nutr. Sci.; 2004; 54, pp. 249-254.

89. Kavand, M.; Arzani, K.; Barzegar, M.; Mirlatifi, S.M. Pomegranate (Punica granatum L.) fruit quality attributes in relation to aril browning disorder. J. Agric. Sci. Technol.; 2020; 22, pp. 1053-1065.

90. Camargo, R.J.; Tadini, C.C.; Sabato, S.F. Physical-chemical analyses of irradiated papayas (Carica papaya L.). Radiat. Phys. Chem.; 2007; 76, pp. 1866-1868. [DOI: https://dx.doi.org/10.1016/j.radphyschem.2007.02.103]

91. Schweiggert, R.M.; Steingass, C.B.; Mora, E.; Esquivel, P.; Carle, R. Carotenogenesis and physicochemical characteristics during maturation of red fleshed papaya fruit (Carica papaya L.). Food Res. Int.; 2011; 44, pp. 1373-1380. [DOI: https://dx.doi.org/10.1016/j.foodres.2011.01.029]

92. Ta, N.; Chang, Q.; Zhang, Y. Estimation of apple tree leaf chlorophyll content based on machine learning methods. Remote Sens.; 2021; 13, 3902. [DOI: https://dx.doi.org/10.3390/rs13193902]

93. Demir, B.; Gürbüz, F.; Eski, I.; Kuş, Z.A. Data mining approach for prediction of fruit color properties. J. Agric. Fac. Atatürk Univ.; 2018; 49, pp. 37-43.

94. Aredo, V.; Velásquez, L.; Carranza-Cabrera, J.; Siche, R. Predicting of the quality attributes of orange fruit using hyperspectral images. J. Food Qual. Hazards Control; 2019; 6, pp. 82-92.

95. Alwosheel, A.; van Cranenburgh, S.; Chorus, C.G. Is your dataset big enough? Sample size requirements when using artificial neural networks for discrete choice analysis. J. Choice Model.; 2018; 28, pp. 167-182. [DOI: https://dx.doi.org/10.1016/j.jocm.2018.07.002]

96. Gitelson, A.A.; Merzlyak, M.N. Remote estimation of chlorophyll content in higher plant leaves. Int. J. Remote Sens.; 1997; 18, pp. 2691-2697. [DOI: https://dx.doi.org/10.1080/014311697217558]

97. Peng, Y.; Nguy-Robertson, A.; Arkebauer, T.; Gitelson, A.A. Assessment of Canopy Chlorophyll Content Retrieval in Maize and Soybean: Implications of Hysteresis on the Development of Generic Algorithms. Remote Sens.; 2017; 9, 226. [DOI: https://dx.doi.org/10.3390/rs9030226]

98. Hendry, G.A.F.; Price, A.H. Stress Indicators: Chlorophylls and Carotenoids. Methods in Comparative Plant Ecology; Hendry, G.A.F.; Grime, J.P. Chapman Hall: London, UK, 1993; pp. 148-152.

99. Taiz, L.; Zeiger, E. Fisiologia Vegetal; 5th ed. Artmed: Porto Alegre, Brazil, 2013; 954p.