Comparative study of polyphenolic content and antioxidant capacity in fruits of Arbutus unedo, A. andrachne and their natural hybrid A.× andrachnoides

Fruits of the three Arbutus species native in the Eastern Mediterranean region, i.e. Arbutus andrachne, A. unedo and their natural hybrid A. × andrachnoides, collected from two different regions in Greece (Kalamos and Varympompi), were analysed comparatively in order to determine their phenolic profile and measure the antioxidant activity using two different single electron transfer-based assays, ferric reducing ability (FRAP) and 2,2-diphenylpicrylhydrazil radical scavenging capacity (DPPH). The pomological characteristics of the hybrid were similar to A. andrachne, while A. unedo had about 2.0-2.5 times larger fruits, with 20% more moisture being 175% heavier. All three species and their biotypes from the two areas had high phenolic content and antioxidant properties and were clearly pulled in different groups. Total flavanols scored a very strong correlation to antioxidant capacity. Highest total flavanols concentrations were found in A. andrachne fruits, while the hybrid had similar concentrations with A. unedo. In samples from Kalamos, concerning condensed tannins, total flavones, flavonols, phenolics and antioxidant capacity FRAP, as well as total anthocyanin and delphinidin 3-O-galactoside, the hybrid presented average values compared to the two parental species, while A. andrachne had the highest values. A strong correlation was found between ortho-diphenols and total soluble solids, as well as of total phenolics and flavonoids with total anthocyanins. The polyphenolic content and the antioxidant capacity varied in different biotypes. A. andrachne biotype from Kalamos showed the highest antioxidant capacity and anthocyanin potential. All A. unedo biotypes and A. × andrachnoides from Varympompi showed higher concentrations of ascorbic acid compared to the other biotypes.


Introduction
Fruits were always a significant part of human diet. The increasing food industries competition in nowadays led to the introduction of new fruiting species for cultivation, such as the temperate forest minor fruits, some of which have a relative wide scale of cultivation since the 15th century. Furthermore, temperate being poor in lipids and Na (Mosele et al., 2011) and of higher values than other red small fruits (Guerreiro et al., 2013). The antioxidant content of A. unedo is rated highly between 28 common fruits studied (Garcia-Alonso et al., 2004) and is characterized by a 'sui-generis' sweet tart flavour due to high sugar and acid content making this suitable for different food applications (Alarcao-E-Silva et al., 2001). Most plant parts of A. unedo have been used in folk medicine (Bonvehi et al., 1994;González et al., 2010;Sansanelli et al., 2017;Morgado et al., 2018) with promising results for industrial use (Salem et al. 2018). In the NW Turkey five strawberry tree genotypes were selected for extensive cultivation (Celikel et al., 2008). A. andrachne has also been employed in folk medicine and there are reports on its constituents (Sakar et al., 1991).
Thus, although both A. unedo and A. andrachne have been studied for their industrial value (Issa et al., 2008;Barros et al., 2010;Afifi-Yazar et al., 2011;Mendes et al., 2011;Miguel et al., 2014;Abidi et al., 2015), there are no data about the nutritional and phytochemical characteristics for use in the food and medicinal industry of their hybrid A. × andrachnoides. In a previous work, it was found greater phenotypic resemblance and genetic similarity of A. × andrachnoides with A. andrachne than with A. unedo. We found intriguing to see if this pattern will apply also to nutritional and phytochemical characteristics of the three species. In addition to the scientific interest, research on this issue coincides with the encouragement of the Food and Agriculture Organization for the use of underutilized species by enhancing the exploitation of their nutritional and pharmaceutical properties (FAO, 2010). The existence of in vitro propagation protocols and phylogenetic studies  that could reveal the discrimination of suitable genotypes of these three Arbutus species will further facilitate their introduction to the pharmaceutical and food industry.
The present research aims to study comparatively the fruits of the three natives in the Eastern Mediterranean region Arbutus spp., i.e. A. andrachne, A. unedo and their natural hybrid A. × andrachnoides, collected from two different areas, in terms of phenolic content and antioxidant activity, in order to estimate them as sources of nutritional and pharmaceutical components. 3

Materials and Methods
Sampling area and sample plant selection Arbutus unedo, A. andrachne and A. × andrachnoides fruits were collected from two typical macchia forests located in the prefecture of Attica, Greece: Kalamos (450 alt., lat. 38°16'22.1"N, long. 23°52'13.3"E) and Varympompi (400 m alt., lat. 38°08'13.6"N, long. 23°47'41.0"E). At each forest, fruits from three shrubs of each Arbutus species were collected (i.e. total of six different populations) (Table 1). Morphological characters were used for plant identification. Shrubs of similar size (2-3 m in height and diameter) at least 100 m apart from each other were selected. Each sample contained 500-600 g of randomly collected fruits from each individual plant. Each sample was considered as one replication. The fruits were collected fully ripened, based on the fruit colour (Herrera, 1998). A. unedo and A. andrachne fruits were collected in December 2017 and A. × andrachnoides fruits were collected in January 2018. The fruits were collected in the morning, packed in plastic bags, carried to the laboratory in a portable refrigerator and stored at -80 °C. Measurements included the determination of fruit pomological characteristics, soluble solids, pH, total titratable acidity, total phenols, total anthocyanins and total carotenoids, as well as the determination of individual profiles in anthocyanins and organic acids. As soon as fruits were taken to the laboratory the weight of each fruit (Mettler Toledo AB104-S/FACT Analytical Balance), as well as its length and width (largest diameter vertical to length) was recorded (with a digital pachymeter). Fruit moisture was calculated as the difference between fruit fresh and dry weight (fruits were dried for five days at 70 o C).

Sample preparation
For the fruit phenolic extraction, approximately 100 frozen fruits were manually homogenized. Afterwards, 1 g fresh homogenized tissue was mixed for two times with 4 mL extraction medium water/methanol/acetone/HCL (19/40/40/1), homogenized for 1 min at 8,000 rpm in an Ultra Turrax homogenizer and the extracts were shaken in controlled temperature vacuum at 150 rpm for 30 min at room temperature (25 °C) and one more time with 4 mL extraction medium methanol/water/HCL (80/19/1) to prepare the extracts. The supernatants were collected after centrifugation of the samples at 4,500 rpm for 10 min. The extraction procedure was repeated two more times; the fractions were combined and the supernatants were stored at -80 °C until analysis.

4
Soluble solids, pH and total titratable acidity For the determination of fruit juice soluble solids ATAGO N1-a refractometer with a 0-32 °Brix measurement range with 0.28 °Brix increments were used (no temperature compensation). Total titratable acidity was measured by titration with 0.1 N NaOH solution and was expressed as citric acid. For the measurement of soluble solids and total acidity, 3.5 g homogenized tissue of fresh fruits were mixed with 10.5 ml deionized water and then the samples were centrifuged at 4,500 rpm for 10 min (Mitcham et al., 1996).
Τotal phenols and anthocyanins Somers and Evans, (1977) method with some modifications was used for the measurement of total phenols and anthocyanins. An extract-aliquot was dissolved in methanol 1:10. 0.5 mL of the dissolved extract was added to 10 mL HCL 1M in order to measure anthocyanins and total phenols. Tubes kept in the dark for 3 h and afterwards the concentration both of total anthocyanins and phenols was determined spectrophotometrically at 520 and 280 nm. Known solutions of catechin and malvidin 3-O glucoside (12.5-200 μg mL -1 ) were used to construct a calibration curve to estimate the concentration of total phenols and total anthocyanins respectively and the data are expressed as mean (milligrams of catechin equivalent per gram of fresh matter).

Total flavonoid content
The colorimetric method (Dewanto et al., 2002) was used for the total flavonoid content determination as following. At first an aliquot of the extract was dissolved in methanol 1:10 and afterwards 0.25 mL of the dissolved extract or catechin standard solution was mixed with 1.25 mL of distilled water in a test tube followed by addition of 75 μL of 5% NaNO2 solution. 6 min later, 150 μL of a 10% AlCl3. 6H2O solution was added to stand for another 5 min followed by addition of 0.5 mL of 1 ml NaOH. The mixture was filled with distilled water to 2.5 mL, mixed well and absorbance was measured immediately against the blank at 510 nm. Known solutions of catechin (12.5-200 μg mL -1 ) were used to construct a calibration curve to estimate the concentration of total flavonoids and the data are expressed as mean (milligrams of catechin equivalent per gram of fresh matter).
Total flavanols, flavones and flavonols The dimethylaminocinnamaldehyde (DMACA) method (Vivas et al., 1994) was used for the estimation of the total flavanol content. Extract (0.2 mL), diluted 1:40 with methanol, was added into a 15-mL tube and then 1 mL of DMACA solution (0.1% in 1 N HCl in MeOH) was added. The mixture was vortexed, allowed to react at room temperature for 10 min and the absorbance at 640 nm was read against the blank. Known solutions of catechin (12.5-200 μg mL -1 ) were used to construct a calibration curve to estimate the concentration of total flavanols and the data are expressed as mean (milligrams of catechin equivalent per gram of fresh matter).
For the estimation of total flavones and flavonols, an aliquot of 2 mL of the test solution, 20 mL methanol and 1 mL 5% aluminium chloride in methanol (w v -1 ) were mixed in a volumetric flask, the volume was filed with methanol up to 50 mL and the mixture was left for 30 min. The absorbance was measured at 425 nm (Bonvehi et al., 1994).

Total condensed tannins content
Total tannins content was determined using a colorimetric method of methyl cellulose. An aliquot of the extract was dissolved in methanol 1:10 (Sarneckis et al., 2006). Treatment Sample: To a 10 mL centrifuge tube, 3.00 mL of methyl cellulose solution was added to the required volume of tannin-containing solution (0.25 mL of supernatant), and the tube was capped, inverted several times, and allowed to stand for 2-3 min. Following the addition of 2.00 mL of saturated ammonium 5 sulphate solution, the sample was made up to final volume with deionized water, mixed again, and allowed to stand for 10 min at room temperature before centrifugation, which was performed at 4,000 rpm for 5 min.
Control Sample: The same volume of tannin-containing solution was added as per the treatment sample. Following the addition of 2.00 mL of saturated ammonium sulphate solution, the sample was made up to final volume with deionized water (in place of the methyl cellulose polymer solution), mixed, and allowed to stand for 10 min at room temperature before centrifugation as above.
The concentration of total tannins was estimated from a calibration curve, constructed by plotting known solutions of catechin (12.5-200 μg mL -1 ). The results are expressed as mean (milligrams of catechin equivalent per gram of fresh matter).

Total carotenoids and chlorophyll-a analysis
The protocol of Nagata and Yamashita (1992) was used for the estimation of carotenoids and chlorophyll-a. Samples of pigments were extracted with acetone-hexane (4:6). 1 g of fresh tissue and 4 mL extraction solution were mixed with vortex for 1 min. Samples were centrifuged at 5,000 rpm for 5 min. The supernatant was collected and the extraction procedure was repeated six more times. All fractions were combined. Then the optical density of the supernatant at 663 nm, 645 nm, 505 nm and 453 nm were measured by spectrophotometer at the same time. From this value, the content of chlorophyll-a and carotene was estimated using our proposed equations: Chlorophyl-a (mg 100 mL -1 ) = 0.999A663-0.0989A645. β-Carotene (mg 100 mL -1 ) = 0.216A663-1.22A645-0.304A505 +0.452A453.

Analysis by HPLC
Individual anthocyanins were analysed by the high-performance liquid chromatography (HPLC) method and the identification was based on comparing retention times and on-line spectral data in comparison with original standards. The calibration curves of each standard compound were used for the quantification stage. Calibration curves made by plotting known solutions (1.25-20 μg mL -1 ) were used for the estimation of concentration. The results were expressed as mean (catechin mg equivalent per gram of dry matter). The analyses were performed using an HPLC Shimadzu Nexera comprising a gradient pump, Shimadzu Nexera X2, a ProStar model 410 AutoSampler and a ProStar model 330 Photodiode Array Detector on a reversed-phase Waters C18 x select (250 mm × 4.6 mm, 5 mm) column at a temperature of 25 °C.
To measure anthocyanins by HPLC, 1 mL of the supernatant was evaporated into a sample concentrator, at room temperature, under a stream of nitrogen gas and the pellet was dissolved in 20 mL of 50% methanol in water.
The composition of the eluent was H2O/HCOOH (90:10, A) and CH3OH (100, B) with the flow rate at 1 mL min -1 . The linear gradient program employed for the elution was: 5% B for 0 min, from 5% to 50% B in 25 min, from 50% to 95% B in 30 min, followed by a return to the initial conditions in 10 min and reequilibration of the column. The chromatogram was monitored at 520 nm.
To separate the organic acids (quinic, malic, ascorbic, citric and fumaric acid), homogenized tissue was diluted 1:4 in 5% meta-phosphoric acid and the solution was membrane-filtered (0.22 μm). A 250 × 4.6 mm ID, 5 μm, Waters × select C18 column operating at 20 °C, under isocratic conditions (mobile phase flow rate 0.8 mL min -1 ) was used for the HPLC analysis. The mobile phase was acidified with 0.1% formic acid in water solution and extracted acids were detected at 210 nm with a PDA (Photodiode Array) detector. Calibration curves were constructed by standard solutions (quinic, malic, ascorbic, citric and fumaric acid) at a range of acid concentration from 31.25-1,000 ppm, for the quantification of acids.

Statistical analysis
The results were expressed as mean ± SE (Standard Error) of the three replications (50 fruits each) and all determinations were analysed three times. Data were processed by analysis of Variance (ANOVA), and statistical significance was processed using the Student's test at P < 0.05. Statistical analysis and correlations 6 were obtained using JMP v. 11 statistical software (SAS Institute Inc., Cary, NC, USA). Principal Component analysis (PCA) was used in evaluating the measurements and their contribution to the variability of the biotypes studied (Biniari et al., 2020).

Results
Fruit pomological characteristics The fruits of A. × andrachnoides were red-orange, spherical, with slightly granular surface like A. andrachne as opposed to the granular surface of A. unedo, but their taste was sweet, like A. unedo, while A. andrachne fruits were rather astringent. In terms of length, width and weight the hybrid fruits had intermediate characteristics compared to the parental species (in Kalamos) or closer to A. andrachne (in Varympompi), while fruit moisture was similar to A. andrachne in both areas and 10-15% lower than A. unedo (Table 2), which had the largest fruits. Soluble solids, pH and total titratable acidity In Kalamos, the hybrid and A. andrachne fruits showed the highest concentration of total soluble solids (27.20 and 28.00 °Brix) and A. unedo the lowest (20.40 °Brix), but this result was not repeated in Varympompi where all species had similar measurements (Table 3). PH values of all three species were low and showed stability. The highest value of pH (3.85 and 3.87) was recorded in A-K and A-V, and of total acidity in the H-K (0.77) and U-V (0.79) ( Table 3). PH was strongly correlated with both citric and fumaric acid (Figure 1).

Phenolics and antioxidant capacity
A-K fruits had half the concentration of total flavonoids (20.51 mg catechin/g fresh tissue) compared to all other samples ranging from 54.82-56.83, but this biotype had the highest concentrations of orthodiphenols, total flavanols, total flavones, total phenolics, condensed tannins and total antioxidant activity FRAP and DPPH (Table 4). Total-flavanols and condensed tannins did not differ in the two biotypes of all three species. In Kalamos the hybrid had average concentrations of ortho-diphenols, total-flavanols, total flavones and flavonols, total phenolics, condensed tannins, and total antioxidant activity FRAP and DPPH compared to the parental species, while in Varympompi this could only be seen for total-flavanols, condensed tannins and antioxidant activity DPPH (Table 4).  Values are the mean (± SE) of three analyses from three different lots of fruits. Values followed by different lowercase letter within each trait are significantly different at the 5% level, determined by the one-way Anova (Student's t test, P < 0.05). † mg catechin/g fresh tissue, † † mg caffeic acid/g fresh tissue, † † † mg rutin/g fresh tissue, † † † † mg trolox/g fresh tissue Organic acids, anthocyanins, chlorophyll and carotenoids There was variation in the organic acids content in different biotypes (Figure 2). The hybrid, compared to the parental species, presented average concentration of citric acid in both areas. The highest concentration of quinic acid was found in A. unedo in Varympompi, and the highest concentration of fumaric acid in A. andrachne in Kalamos (Table 5). A-V presented the highest concentration of malic acid (2,221.9 μg/g fresh tissue), over double from U-K (1,013.3 μg/g fresh tissue) that presented the lowest concentration. A. andrachne and the hybrid seemed to have higher malic acid concentrations than A. unedo. A. unedo in both areas and the hybrid in Varympompi had a content of 798.3-858.1 μg/g ascorbic acid, significantly higher than the other three biotypes, representing a better source of vitamin C.  (Table 6) being the least abundant (0.03 to 0.11 mg/g fresh tissue). Total and individual anthocyanins concentrations did not seem to follow a pattern as far species and biotypes are concerned, apart from Kalamos where the hybrid exhibited average values of total anthocyanins and delphinidin 3-O-galactoside compared to the parental species (Table 6). The A-K exhibited the highest concentration of total anthocyanins (0.82 mg delphinidin 3-O-galactoside/g fresh tissue) and cyanidin 3-O-glucoside (0.53 mg/g fresh tissue) ( Table 6). A-V presented the highest concentration of delphinidine 3-Ο-galactocide and cyanidin 3-Orutinoside/cyaniding-3-O-glucoside equiv and the second highest cyanidin 3-O-glucoside (Table 6).
The concentration of total chlorophyll was very low in all biotypes, β-carotene was the only carotenoid detected and its highest concentration was recorded for H-K (0.037 mg/g fresh tissue, Table 5). Table 5. Quinic, malic, ascorbic, citric, fumaric acid (Q-a, M-a, A-a, C-a, F-a respectively)   some of the characteristics measured showed an average value between the two parental species. Average of two areas fruit mass of the hybrid was 1.55 g, closer to A. andrachne (1.11 g) and less than half of the A. unedo (3.65 g), which had the largest fruits with the highest moisture content (Table 2). Although the morphology of the hybrid fruits was closer to that of A. andrachne their taste was sweet, similar to A. unedo, while A. andrachne fruits were rather astringent due to their high concentration of tannins (Table 4).
The mass of A. unedo fruits was similar to that in two Spanish forests (Molina, et al., 2011) andin Mersin Lapa-Gülnar N, Turkey (Ozcan andHaciseferogullari, 2007) and A. andrachne fruit mass was in accordance with Serce et al. (2010) data (1.15 g) in Turkey.
The hybrid fruits had similar moisture content to A. andrachne, while A. unedo fruits had the highest moisture, something expected because of their biggest size, as the main fruit components were shown to be moisture and carbohydrates (Ruiz-Rodríguez et al., 2011). Moisture content in A. unedo fruits was of 64.46-67.81 (w/w), a bit higher to that counted by Özcan and Haciseferoğullari (2007) and Barros et al. (2010) (60 and 52% respectively) in fruits collected from N Turkey and Natural Park of Montesinho territory (NE Portugal) respectively. However, A. unedo fruit moisture was lower than the conventional fruits (75-95%), as shown for other wild fruits, e.g. Vaccinium oxycoccus (Souci et al., 2008).
The geographic variation of fruit characteristics revealed by the data is referred also in previous studies of A. unedo fruit production (Herrera, 1998;Ogaya and Peñuelas, 2007;Molina et al., 2011). It could be explained by variation of local soil characteristics and meteorological conditions that effects the extended reproductive cycle of all three species affecting also fruit production.
Soluble solids, pH and total titratable acidity Total soluble solids and acidity in all three species were in the range found by Celikel et al. (2008) and Sulusoglu et al. (2011) for A. unedo in Turkish mountains and similar to other wild fruits, providing the unique flavor and making these fruits suitable in food technology. PH values of all three species were low (3.34 -3.85) and showed stability in agreement with previous studies on A. unedo (Cavaco et al., 2007;Ruiz-Rodríguez et al., 2011), but with remarkable difference with pH values (8.06 to 9.11) reported for eight different genotypes of A. andrachne by Serce et al. (2010). PH was strongly correlated with both citric and fumaric acid so that a simple measurement of pH could reveal biotypes with high concentration of citric or fumaric acid.

Phenolics and antioxidant capacity
In the area of Kalamos, A. andrachne fruits had the highest concentrations of ortho-diphenols, total flavanols, total flavones, total phenolics, condensed tannins, and total antioxidant activity FRAP and DPPH, followed by the fruits of the hybrid that had average values between the two parental species. Values of phenolic compounds and antioxidant properties falling between the parental species were also found in seeds but not in fruit skins of grape hybrids (Xu et al., 2010), or in coconut oil from hybrids compared with their parents, where one hybrid was outstanding among the cultivars presenting the highest levels of lauric acid, total phenolic compounds, and antioxidant activity (Arlee et al., 2013).
Ortho-diphenols is an important antioxidant polyphenolic group occurring in fruits, wine and olive, of high impact. Many authors have mentioned ortho-diphenol compounds as the most active antioxidants and of high antimicrobial activity (Bisignano et al., 1999). In the present study, the biotype A-K had the highest concentration of ortho-diphenols with no difference from H-K (1.59 and 1.36 mg caffeic acid/g fresh tissue).
Antioxidant capacity is connected with anthocyanins, flavonoids and flavanols concentration of polyphenolic compounds (Garaguso and Nardini, 2016) and depends on their structure, the number and position of the hydroxyl groups and the nature of the substitutions on the aromatic rings (Balasundram et al., 2006). The biological and nutritional value of fruits can be expressed by the determination of total antioxidant capacity with tests like DPPH (1,1-diphenyl-2picrylhydrazyl free radicals test) and FRAP (ferric reducing antioxidant power) (Koleva et al., 2000;Gazdik et al., 2008). It is difficult to compare the antioxidant capacity of the three Arbutus species with relevant reports in the literature due to different methods of determination used. However strawberry tree fruits ranked fourth between 28 different fruits (Garcia-Alonso et al., 2004).
The values of phenolic content and antioxidant capacity of A. andrachne fruits confirmed previous data by Serçe et al. (2010) for this species in Turkey. There were also similarities between our results on A. unedo with previous studies on the phenolic profile of A. unedo samples from Italy, Portugal and Spain (Pallauf, et al., 2008;Fortalezas et al., 2010;Mendes et al., 2011). Quantification of the comparison of phenolic compounds was not possible either due to the use of different units or lack of reporting the quantities in these studies.
Organic acids, anthocyanins, chlorophyll-a and carotenoids A. unedo in both areas and the hybrid in Varympompi had a content of 498.7-858.1 μg g -1 ascorbic acid, significantly higher than the other three biotypes representing a better source of vitamin C, higher even than citrus fruits (Fatin and Azrina, 2017). Seker and Toplu (2010) who studied A. andrachne and A. unedo in Turkey, also found a higher concentration of ascorbic acid in A. unedo than A. andrachne, with mean values in both species much higher (3-4 times) than those found in the present study. On the other hand, in the present study all biotypes of the tree species had 9-15 times higher concentration of ascorbic acid in comparison to A. unedo fruits (55 μg g -1 fresh tissue) in Salamanca, West Spain (Pallauf et al., 2008).
The anthocyanin profile of all Arbutus biotypes revealed the presence of four anthocyanins, i.e., delphinidin 3-O-galactoside, cyanidin 3-O-rutinoside, cyanidin 3-O-glucoside and peonidin 3-O-glucoside. In previous reports the anthocyanin profile of A. unedo was characterized by the existence of delphinidin and 12 cyanidin (Pallauf et al., 2008) that has been identified as the characteristic red colour-factor of the fruit (Proloiac and Raynaud, 1981). In the present work A. andrachne, particularly in Varympompi, presented the highest concentration of delphinidin and cyanidin and indeed the colour of A. andrachne fruits are usually darker red than the A. unedo and the hybrid. The identification of delphinidin 3-O-galactoside, cyanidin 3-Orutinoside and cyanidin 3-O-glucoside is similar with Pallauf et al. (2008) and Fortalezas et al. (2010) on A. unedo fruits study, although in the present study one more anthocyanin, peonidin 3-O-glucocide, is described in the anthocyanin profile of strawberry tree fruits. The concentration of A-V delphinidin 3-O-galactoside, reached up to 6 times higher than U-K concentration (126.83 and 23.38 mg g -1 fresh tissue respectively) and almost two times higher than U-V (74.64 mg/g fresh tissue), giving to A-V biotype a high impact because of the beneficial effects of anthocyanins both as antioxidants and regulators of adipocytokine gene expression and anti-cancer factors (Vattem and Shetty, 2005;Tsuda et al., 2006).
The very low total chlorophyll content in fruits of all biotypes proved that they were collected at the full maturity stage. Regarding carotenoids analysis, β-carotene was the only carotenoid detected, in accordance to Alarcao-E-Silva et al. (2001) and Pallauf et al. (2008) reports for A. unedo.

Conclusions
The present study enriches the previous knowledge on the potential use of Arbutus species found in Eastern Mediterranean region, and for the first time provides information on the polyphenolic content and antioxidant capacity of the natural hybrid A. × andrachnoides in comparison with the two parental species A. unedo and A. andrachne. The hybrid fruits had pomological characteristics closer to that of the A. andrache in almost all biotypes, and values of ortho-diphenols, total flavanols, total flavones, total phenolics, condensed tannins, and total antioxidant activity FRAP and DPPH average between the two parental species. The results obtained revealed the significant differences in the polyphenolic content among the various biotypes of the three species, whereas the A. unedo biotypes were closer to each other. Therefore, this study confirms earlier reports for a wide variability in the chemical composition and properties of A. unedo and A. andrachne fruits.
The antioxidant capacity positively correlates with the total flavanols, and the otho-diphenols with the total soluble solids. Four anthocyanidins were identified and quantified and A. andrachne biotype from Kalamos proved to have the highest antioxidant activity and anthocyanin potential. Both A. unedo biotypes and A. × andrachnoides from Varympompi had the highest concentration of ascorbic acid compared to the other biotypes, whereas A. unedo from Varympompi exhibited the highest concentration of quinic acid. With regards to the utilization of the species, the differences revealed within the present study in combination with previous phylogenetic, ecophysiology and in vitro propagation studies will allow both pharmacy and food industry to develop suitable strategies for the best utilization of those species.