Fruit Weight (FW) was determined using a two-decimal precision digital scale (GIbertini, Italy) and expressed in grams. Transversal and longitudinal diameter (TD and LD, respectively) were determined with a TR53307 digital caliper (Turoni, Forlì, Italy) and expressed in millimetres. The percentage of cover colour (CC) was calculated by means of a pre-calibrated Minolta Chroma Meter (Chroma Meter CR-400, Konica Minolta Sensing Inc., Tokyo, Japan). The colour samples were measured against Hunter values: brightness (L* value) from 0 (black) to 100 (white), green (a* value) from -a (green) to +a (red) and yellow (b* value) from -b (blue colour intensity) to +b (yellow colour intensity) [22]. Firmness (F) was determined by means of a TR5325 digital penetrometer (Turoni, Forlì Italia) with an 8 mm diameter plunger and expressed as kg·cm^-2. After the apple juice extraction with a centrifugal squeezer (Ariete, Italy), the Soluble Solids Content (SSC), expressed as °Brix, was measured by means of an ATAGO digital optical refractometer (Atago Co, Ltd., Tokyo, Japan), while the Titratable Acidity (TA), expressed as malic acid (g·l^-1), and the SSC/TA ratio, were measured with a Crison Compact titrator pH-meter (Crison Instruments, SA, Barcelona, Spain).

The apples were cut length-wise and the top was immersed in an iodinated solution (Lugol reagent: 40g KI + 10g L^-1 H₂O) for about 60 seconds. The fruit was then removed from the solution and allowed to air dry for a further 60 seconds. The surface undergoes a colour change, indicating the starch content. This is compared to a reference chart (Eurofru, Ctifl, France) with a scale from 0 (highest starch content) to 6 (no starch content).

The starch state was analysed by Lugol’s iodine, compared with reference charts (Eurofru, CTIFL, France), and expressed on a 10-point scale (0 = maximum starch content, 6 = no starch).

Biometric data were also determined using a specific descriptor list adapted from UPOV (International Union for the Protection of New Varieties of Plants) guidelines [23] to discriminate the different cultivars. In particular, the following descriptors were evaluated: general shape (GS), ribbing (R), stem cavity depth (DS), stem cavity thickness (TS), ocular basin depth (DB), ocular basin thickness (TB), stem length (LS), stem thickness (TS), peel colour (CP), peel cover colour (CP), cover colour type (TC) and flesh colour (CF).

Sensory profiling (UNI 10957) [24] was performed on a subsample of 10 fruits per cultivar by a panel of 10 judges (50% male and 50% female). All panellists were trained in the sensory evaluation of fruits, according to other studies [25]. Twenty different attributes were generated on the basis of the citation frequency (> 60%), as listed below: appearance (A), peel colour (PC), flesh colour (FC), apple odour (AO), herbaceous odour (HO), almond odour (AO), off-odour (OO), sweetness (S), bitterness (B), acidity (AC), astringency (AS), crunchiness (C), mealiness (M), juiciness (J), apple flavour (AF), herbaceous flavour (HF), honey flavour (HOF), almond flavour (ALF), off-flavour (OF) and overall assessment (OA). The judges evaluated the intensity of each attribute by assigning a score between 1 (absence of the descriptor) and 9 (high presence of the descriptor).

Complex carbohydrates have a significantly more elaborate biochemical structure than other carbohydrates (simple sugars with a simpler structure, determined with the Fehling reagent). The carbohydrate content (TSG), either free or present in polysaccharides, was obtained with the anthrone method reported in Loewus studies [26]. Carbohydrates were first hydrolysed into simple sugars using dilute hydrochloric acid. In a hot acidic medium, glucose is dehydrated to hydroxymethylfurfural that with anthrone forms a green-coloured product with an absorption maximum at 630 nm.

Ash content was determined through the procedure described in the Association of Official Analytical Chemists (AOAC) [27-31]. The contents of Ca, Fe, Mg, K, and Na were determined using atomic absorption spectroscopy following wet mineralization. The determination of vitamins (B1, B5, and E) was determined by HPLC after extraction of the lipophilic component from pulp samples through AOAC procedures. The riboflavin (vitamin B2) was extracted in an autoclave with a solution of diluted H₂SO₄ and later, after enzymatic treatment, vitamin B2 (mg%) was quantified was determined through HPLC (for the fluorescent spectra). For the ascorbic acid (vitamin C) determination, the dried methanolic extract (100 mg) was extracted with 10 mL of 1% metaphosphoric acid for 45 min at room temperature and filtered through Whatman No. 4 filter paper. The filtrate (1 mL) was mixed with 9 mL of 2.6–dichlorophenolindophenol, and the absorbance was measured within 30 min at 515 nm against a blank. Ascorbic acid was calculated based on the calibration curve of authentic L-ascorbic acid (0.02 – 0.12 mg mL−1) [32].

Moisture Content (MC), TSG content, and ASH content were expressed as g%, while minerals and vitamins were expressed as mg%. All measurements were done in three replicates.

The physico-chemical and sensory data were tested for differences between the cultivars using the one-way analysis of variance (ANOVA: general linear model) using XLSTAT software version 9.0 (Addinsoft, Paris, France). The differences between cultivars were tested, followed by Tukey’s multiple range test for p ≤ 0.05.

Our data showed significant differences between the different apples in a genotype-dependent manner. In particular, Table 1 shows the differences in physico-chemical characteristics.

As expected, the new clones showed a higher weight than the old varieties. Probably the detected weight was influenced by the thinning of the fruits, which was done in a single step. For this reason, the fruits might be slightly undersized [33]. According to the commercial classification in size classes (EC Reg. n.85/2004) [34], considering both transverse diameter and weight, all observed genotypes would be included in the extra category (Table 2). Similar data were observed by Lo Bianco and Farina [8].

Firmness showed very high values in all observed fruits, showing a good aptitude for cold storage. Old varieties ranged from 3.12 (Faccia Rossa) to 4.68 (San Giuseppe). The new clones varied from 2.83 (Sinfonia) to 4.46 (Luresweet Redlove^®). The old varieties have values above 3.5 in three cases because they are probably derived from old spontaneous hybridisation with Cydonia oblonga [8, 35].

Concerning colorimetric analyses (Fig. 1 and Table 3), for the old genotypes, the highest percentage of cover colour (CC%) was recorded in Cannamelo, Turco, Majesti and Luresweet Redlove^® which was almost uniformly red, while the lowest was recorded in Shinano Gold, being yellow apple. For brightness (L*), the highest value was recorded in Faccia Rossa, San Giuseppe, Shinano Gold and Sinfonia. For the parameter a* (red = +100, green= -100), the highest value was recorded in the new clones (Majesti, Luresweet Redlove^® and Sinfonia). For b* (yellow = +100, blue= -100). Faccia Rossa, Turco, Shinano Gold, and Sinfonia showed the highest value. The results of colorimetric analyses showing that fruits from the new clones are potentially more appreciable for the fresh consumption market could be due to good visual characteristics, which correspond to the maximum brightness index, in agreement with Ivascu et al. [36] and Iglesias et al. [37], according to whom consumers are strongly attracted to the more colourful fruits, as well as the San Giuseppe among the old varieties (Fig. 1).

As for the juiciness analysed in a sample of 100 g of fruit (Fig. 2), all the new clones showed high values, whereas only Faccia Rossa and San Giuseppe reached the lowest values.

With regard to SSC, all the observed genotypes show very high values. New clones varied from 12.38 to 16.66 Brix°, whereas old varieties varied from 13.31 to 17.16 Brix°. In particular, Shinano Gold and Luresweet Redlove^® have the highest content matched only by Cannamelo among the older varieties. As for TA, the differences are also large within each genotype group. The values range from 4.1 (Turco) to 7.44 g of malic acid per litre (Cannamelo) in old varieties, while in the new clones, they vary between 4.46 (Shinano Gold) and 8.93 (Luresweet Redlove^®). The SSC/TA ratio is strongly influenced by the high variability of the two chemical parameters and, in particular, of TA. The genotypes can be grouped into three subgroups i) value slightly lower than 2 (Virchiata, Majesti, Luresweeet Redlove^®), ii) values higher than 2 (Cannamelo, San Giuseppe and Sinfonia), and iii) values higher than 3 (Faccia Rossa, Turco and Shinano Gold).

Fig. (1). Colour table obtained from the results of the L*a*b* colour recorded in the CIEL*a*b* colour space and converted to the red/green/blue (RGB) scale through the www.e-paint.co website.

Fig. (2). Soluble solids content (SSC), titratable acidity (TA), and their ratio (SSC/TA) of the observed old varieties and new clones. Values indicate the means, and the bars indicate the standard deviations of the replicates (n = 30).

Considering the data measured with Lugol’s test for starch content, the values vary on a scale between 0.6 and 1.0 (data not shown), indicating a fairly similar ripening stage. This small variability is closely related to the genotype and, although there are no specific tables for old varieties and new clones, we have tried to conform to what we found in the literature [38].

Finally, the fruits of each variety were subjected to analyses regulated by the UPOV system. These analyses allowed the attribution of an adjective to each fruit, which was visually detected (Table 4).

According to the results of the statistical analysis, the most appreciated cultivars were Faccia Rossa and Cannamelo, which presented appreciable characteristics of flesh crispness, skin and flesh colour, apple flavour and especially sweet taste, as shown in Fig. (3a) (old cultivars),. On the contrary, Virchiata and San Giuseppe showed less appreciable characteristics, showing higher values of honey taste, almond smell, and acidity, which are rather negative characteristics for the sensory analysis of apples. In addition, low sweetness values were recorded.

Fig. (3). Sensory evaluation of the observed old varieties and new clones; Descriptors legend: As: Aspect, PC: Peel Colour, FC: Flesh Colour, F: Firmness, AO: Apple Odour, HEO: Herbal Odour, HO: Honey Odour, ALO: Almond Odour, OO: Off Odour, S: Sweetness, A: Acid, B: Bitterness, AST: Astringency, C: Crispness, FL: Flouriness, J: Juiciness, AF: Apple Flavour, HEF: Herbal Flavour, HF: Honey Flavour, ALF: Almond Flavour, OF: Off Flavour, OA: Overall Acceptability; The * symbol indicates significant differences found by statistical analysis (p ≥ 0.05).

With regard to the new clones (Fig. 3b), the highest percentage of positive evaluations occurred in Shinano Gold, which presented good characteristics in terms of crispness, apple taste, and smell, firmness and sweetness. In addition, positive values were found concerning the acid content of the fruit, which was only perceived in small quantities, and concerning the bitter taste, which was not perceived. An interesting result was found for the variety Luresweet Redlove^® due to its particular flesh colour.

The moisture content (MC - Table 5) is highest in Turco, Majesti, Faccia Rossa, Shinano Gold and Sinfonia, which were also the juiciest. It is obvious that the higher the MC of fruit, the fresher it appears [39]. Moisture content is an important factor as it can influence the texture and overall acceptability, but on the other hand, it can directly influence the shelf-life of the fruits. In fact, moisture loss during the long-term storage of apples causes a direct economic loss due to a decrease in saleable weight. The highest values of carbohydrate content (TSG) were observed in San Giuseppe and Virchiata, while the ASH content is very variable.

The results concerning the analysis of proximate compounds (Table 6) showed significant genotype-dependent differences, according to other studies [40, 41]. In all observed genotypes, Mg and Na values are in line with those found on the USDA nutrient database [40]; Ca and Fe have higher values and, K and Na have lower values. In particular, with regard to the Ca content, Majesti (new clones) has the highest value while, among the old varieties, Virchiata has the highest Fe value. The Mg content is similar between new and old varieties, but the concentration is higher in the old varieties. In terms of K content, the old varieties show the highest values and, among them, Turco has relevant content. On the contrary, in terms of Na content, the new clones and, in particular, Sinfonia and Majesti, have the highest values.

Concerning the vitamin content (Table 7), the results show that all the analysed groups have a great variability between all genotypes. In particular, the old varieties generally have higher values than the new clones. Among the new clones, Sinfonia has similar values to the old varieties and, in particular, Virchiata, Cannamelo, and Turco. The activity of vitamins, known as “phytonutrients,” is commonly included in the anti-oxidant properties of many fruits or vegetables. Thiamine, (vitamin B1) is one of the water-soluble vitamins and should be taken regularly through the diet. Vitamin B1 contributes to the important process of converting glucose into energy. Vitamin B1 deficiency can also lead to heart disease, neurological disorders, dilated pupils, and hypersensitivity of the spine. Vitamin B5 exists in food in bound and free forms and is involved in many biological functions. Its deficiency is associated with metabolic and energy disorders in humans [42]. Epidemiological studies show that diets rich in fruit and vegetable with vitamin C are associated with a lower risk of cardiovascular disease, stroke, cancer, and increased longevity [43-45]. Furthermore, plant-based foods contributes vitamin E to the human diet, which, among other essential nutrients, may act against free radicals [46, 47]. Indeed, scientific evidence shows that diets based on vitamin E-rich plant products, such as the traditional Mediterranean diet, could prevent coronary heart disease [48] and cancer (World Cancer Research Fund) [49]. Riboflavin (vitamin B2) is important for the prevention of cardiovascular disease, cancer, incipient diabetic nephropathy, and retinopathy and, this value is better in new cultivars, whereas old cultivars have the lowest value.