Cultivar and year effects on the chemical composition of elderberry (Sambucus nigra L.) fruits

Due to their outstanding anthocyanin content, elderberries are mainly used in the food industry to produce pigment concentrations. Thanks to the increase in demand, elderberry is grown on ever greater areas in Hungary and in neighbouring countries. Cultivar use, however, is very one-sided, being practically restricted to ‘Haschberg’. As this cultivar has many negative properties, growers have begun to plant and test new cultivars. When determining the commodity value of cultivars, it is important to examine not only the physical traits of the fruit, but also their main chemical parameters. In the present experiment the chemical properties (soluble solids and titratable acid content, total polyphenol and anthocyanin content, antioxidant capacity) of the fruit of 11 elderberry cultivars (‘Haidegg 13’, ‘Haidegg 17’, ‘Haschberg’, K3, ‘Korsör’, ‘Samdal’, ‘Samidan’, ‘Samocco’, ‘Sampo’, ‘Samyl’, ‘Weihenstephan’) were analysed in three consecutive years. In addition to the comparative evaluation of the cultivars, this work also aimed to discover correlations between the components and to study the effect of the year on the chemical composition of the fruit. Significant differences were found between the cultivars for the soluble solids content (F(10;8.74)=9.71; p=0.001), the titratable acid content (F(10;22)=7.91; p<0.001), the polyphenol content (F(10;22)=9.77; p<0.001), the anthocyanin content (F(10;8.52)=36.18; p<0.001) and the antioxidant capacity (F(10;22)=3.61; p=0.006). A year effect was proved for the water-soluble solids content (F(2;30)=4.02; p=0.028) and the antioxidant capacity (F(2;30)=5.21; p=0.011). Among the chemical properties, a significant positive linear correlation was only detected between the polyphenol and anthocyanin contents (r=0.91; p<0.001). Among the cultivars, ‘Sampo’, ‘Samidan’ and ‘Weihenstephan’ exhibited outstanding polyphenol and anthocyanin contents. The soluble solids content and antioxidant capacity of ‘Haidegg 17’ were also promising.


Introduction
As the result of changes in eating habits, attention has increasingly moved to fruit and vegetables with high pigment contents that could be used to replace artificial food colourings (Wissgott and Bortlik, 1996). Elderberry (Sambucus nigra L.) has been used as a medicinal plant for hundreds of years, but its cultivation area has increased greatly over the last few decades, thanks to its diverse uses and easy production (Charlebois et al., 2010). Elderberries have exceptionally high anthocyanin content compared to other fruit species (602.9-1265.3 mg CGE 100 g -1 FW) (Veberic et al., 2009). The dark red pigment concentration extracted from the berries is used as a colouring for squashes, jams and other products of plant origin (Charlebois et al., 2010).
Apart from anthocyanins, the fruit also contain numerous other bioactive compounds, the health benefits of which have been reported by many authors (Netzel et al., 2005;Knudsen and Kaack, 2015;Mlynarczyk et al., 2018). The berries have significant vitamin C content (26-36 mg 100 g -1 ), and among the minerals the quantity of potassium, calcium and magnesium is considerable (Vulic et al., 2008). The carbohydrate content is low compared to that of other fruit species (6.5-18.4 g 100 g -1 ) and consists mainly of simple sugars (glucose and fructose) (Veberic et al., 2009). The titratable acid content is low to medium, and citric acid is the main organic acid (0.6-1.7 mg 100 g -1 ), though there is also a substantial quantity of malic acid (Kaack et al., 2008). Foodstuffs prepared from elderberries have notable antioxidant capacity (5.04-6.37 mmol 100 g -1 ) (Akbulutu et al., 2009;Cejpek et al., 2009).
The 'Haschberg' cultivar is widely grown in the major elderberry-producing countries in Europe (Austria, Denmark, France, Hungary and Germany). In order to eliminate the problems caused by the uneven ripening and disease susceptibility of 'Haschberg' and by the fact that the production of a single cultivar leads to a short harvesting and processing season, increasing emphasis is now laid on the breeding of cultivars specifically for the food industry and on the comparative evaluation of cultivars (Kollányi et al., 2005;Möhler et al., 2009;Matejicek et al., 2015;Thomas et al., 2015).
In addition to papers on the chemical traits of American cultivars (Sambucus canadensis L.) (Özgen et al., 2010;Perkins-Veazie et al., 2015;Wu et al., 2015), many authors have also reported differences in the chemical composition of the fruit of a number of European cultivars originating from Sambucus nigra (Kaack and Austed, 1998;Lee and Finn, 2007;Möhler et al., 2009;Veberic et al., 2009;Fejer et al., 2015). In Hungary the first analyses of chemical components and the determination of optimum harvesting dates were performed on 'Haschberg' (Stéger-Máté et al., 2002;Stefanovitsné, 2004), while later these studies were extended to include the cultivars 'Sampo' and 'Samocco' (Szalóki-Dorkó et al., 2015). Similar comparative evaluations on the chemical parameters of other European cultivars with promise for cultivation have not yet been carried out.
However, the chemical composition of the fruit is not only determined by genetic factors. Environmental and climatic factors also influence fruit composition, which thus exhibits a diverse chemical profile each year (Salvador et al., 2015). The effect of internal and external factors on the fruit composition of American cultivars has been proved by Thomas et al. (2013) for American cultivars and by Ferreria et al. (2020) for the European cultivars 'Sabugueiro', 'Sabugueira' and 'Bastardeira', but to the best of our knowledge no paper has yet dealt with the effect of climatic factors on the chemical components of major European elderberry cultivars.
In the present experiment the chemical parameters of 11 elderberry cultivars were evaluated in three consecutive years. Measurements were made on the soluble solids content, the titratable acidity, the antioxidant capacity and the polyphenol and anthocyanin contents of fruit picked at the optimum harvesting date. The results were statistically analysed to discover whether the year and the genotype influenced the given parameters and whether there were any interactions between individual chemical traits. The three years of data made it possible to compare the cultivars and to select genotypes with better chemical properties than 'Haschberg'.

Determination of soluble solids content and titratable acidity
The soluble solids content (SSC) of the fruit was analysed using an HI 96801 digital refractometer. The titratable acid (TA) content was determined on the basis of the Hungarian standard (MSZ EN 12147:1998). Fruit samples weighing 10 g were made up to 100 ml with distilled water, after which they were titrated with 0.1M sodium hydroxide (NaOH) solution in the presence of dimidium bromide-disulphine blue until the colour changed. The total acid content was expressed as citric acid equivalents and the results were given as w/w%.

Determination of ferric reducing antioxidant power (FRAP)
The antioxidant capacity of elderberries was determined using the FRAP method described by Benzie and Strain (1996). The FRAP method involves the reduction of Fe 3+ ions to Fe 2+ , which in turn form a bluecoloured complex with 2,4,6-tripyridyl-S-triazine (TPTZ). The intensity of the colour depends on the antioxidant concentration. Absorbance was measured at 593 nm. Antioxidant capacity was defined as ascorbic acid equivalents (mmol AAE l00 g -1 FW) based on the ascorbic acid standard calibration curve.
Determination of total polyphenol content (TPC) TPC was measured according to the method of Singleton and Rossi (1965) with slight modifications. Samples weighing 500 μl were placed in test-tubes, to which 2.5 cm³ Folin-Ciocalteu's reagent and 7.5 cm³ sodium carbonate (20%) were added. After 2 hours, the absorbance of each sample was measured at 760 nm. The calibration curve was made using gallic acid. The results were expressed as micrograms of gallic acid equivalents on a fresh weight basis (mg GAE 100 g -1 FW).

Determination of total anthocyanin content (TAC)
Total anthocyanin content (TAC) was determined using ethanol and hydrochloric acid, as described by Füleki and Francis (1968). Samples weighing 0.1 g were taken from the supernatant obtained after centrifugation. After the addition of 0.2 ml conc. HCl, the samples were made up to 10 ml with 96% alcohol. After 30 minutes in the dark, the absorbance was measured at 510 nm. The results were expressed as milligrams of cyanidin-3-glucoside equivalents on a fresh weight basis (mg Cy3G 100 g -1 FW).

Meteorological data in the years of the experiment
The meteorological data used to evaluate the year effect were taken from the www.metnet.hu website for the 23 rd district of Budapest. The precipitation distribution differed greatly in the three years ( Figure 1). The largest quantity of precipitation was recorded in 2018 (710 mm) and the smallest in 2019 (687 mm), with 709 mm in 2017. The mean annual temperature was highest in 2019 (13.5 °C) and lowest in 2017 (12.3 °C).

Statistical methods
Three technical replicates were measured each year for each cultivar (from the same pooled sample), and these were then averaged. The data were analysed using one-way ANOVA models (Dobson, 2002;Hang, 2014), and the model residuals were checked for normality using the Kolmogorov-Smirnov test and graphical representations of distribution (histogram, QQ plot). Pairwise comparisons were made using the Sidak p-value adjustment. Levene's tests were run to check the homogeneity of variances. If there was significant heteroscedasticity, Welch's ANOVA and the Games-Howell post hoc test were applied. Homogeneous subsets were labelled on mean and standard deviation plots for each variable. Linear relationships between pairs of investigated variables were analysed using Pearson's correlation and scatterplots. Statistical analysis was performed with IBM SPSS (Version 25, IBM Corporation, Armonk, NY, USA), where p-values of <0.05 were considered to indicate statistical significance. Plots were generated in Excel (Microsoft Office 365).

Soluble solids content (SSC)
The soluble solids content is one of the most important quality parameters for elderberries, as it determines their market value. The processing industry will only buy up fruit with a soluble solids content of at least 12% (Sidor and Gramza-Michalowska, 2015). This criterion was met for all the genotypes tested, with the exception of 'Sampo', 'Samyl' and 'Weihenstephan' (Figure 2). The SSC values varied between 10.8 and 14.56%, which is within the range previously reported by Csorba et al. (2019) and Kaack et al. (2005). The SSC value was lowest for 'Sampo' in two of the years, confirming the findings of Safránková (2011). Szalóki-Dorkó et al. (2015), however, reported that the SSC level of 'Sampo' was similar to that of the other cultivars tested. The present data showed outstandingly high values for 'Korsör' in two years, somewhat contradicting the findings of Lee and Finn (2007). According to Kaack (1997) the SSC value of 'Haschberg' is considerably lower than that of other cultivars, but this too was not confirmed in the present work. The most promising cultivars were found to be 'Haidegg 17' and 'Korsör', together with the experimental Hungarian genotype K3. The statistical analysis of the three-year data confirmed that the cultivar exerts an influence on the soluble solids content (F(10;8.74)=9.71; p=0.001). Titratable acidity (TA) Quantitative changes in the acids, and especially in their ratio compared to the sugar content, influence the flavour. The combined data for the three years showed the lowest titratable acidity for 'Samocco' and K3, while the highest value was exhibited by 'Samidan' in two years and by 'Sampo' in one year ( Figure 3). The data recorded for 'Sampo' and 'Samocco' are in agreement with those of Kaack (1997). However, Kaack and Knudsen (2015) found lower TA values for 'Samyl' than for 'Samdal', in contradiction to the present findings. According to Szalóki-Dorkó (2016) 'Samocco' had higher acid content than 'Haschberg', which was not confirmed in the present study. The acid content of 'Haschberg' proved to be the lowest of all the cultivars tested by Safránková (2011), which again contradicted the present data. The statistical evaluation of the data proved the significant effect of the cultivar on the mean titratable acidity (F(10;22)=7.91; p<0.001). Ferric reducing antioxidant power (FRAP) Elderberries are a rich source of antioxidants such as phenolic acids, flavonols and anthocyanins (Cejpek et al., 2009). Numerous factors influence the quantity of antioxidant compounds accumulated in the fruit, which also depends greatly on the genotype (Scalzo et al., 2005), as confirmed by the present statistical analysis (F(10;22)=3.61; p=0.006). This was also demonstrated by Hegedűs et al., (2008) for other berry fruit species. The mean FRAP values of the 11 cultivars can be seen in Figure 4. The values found here were slightly higher than those reported by Stéger-Máté et al. (2007). In all three years the cultivar with the highest FRAP value was 'Haidegg 17', followed by 'Haidegg 13' and 'Weihenstephan'. The results of Mlynarczyk et al. (2015), who found that 'Haschberg' had greater antioxidant capacity than 'Sampo' or 'Samyl', were not confirmed in the present study. Matejicek et al. (2015) also reported considerable antioxidant capacity for 'Haschberg', with low values for 'Haidegg 13' and 'Korsör'.

Total polyphenol content (TPC)
The mean polyphenol contents of the 11 elderberry cultivars included in the experiment exhibited great variability, as shown in Table 1. The data were not always in agreement with those published in the literature. Lee and Finn (2007) obtained lower values in two years for 'Haschberg' (364 and 510 mg GAE 100 g -1 ) and also for 'Korsör' (387 and 582 mg GAE 100 g -1 ), while Ferreria et al. (2020) reported higher values (820 ± 45 and 1476 ± 282 mg GAE 100 g -1 FW) for the Portuguese cultivars ('Sabugueiro', 'Sabugueira', 'Bastardeira'), though these values depended not only on the genotype but also on the year. The polyphenol contents reported for elderberries by Wu et al. (2004) were as high as those reported here (1950 mg GAE 100 g -1 FW) and were outstanding compared to those of other berry fruit species (red-and blackcurrants, gooseberry).
One-way ANOVA also proved that the genotype had a significant effect on the mean polyphenol content (F(10;22)=9.77; p<0.001). With the exception of 'Haidegg 17', all the cultivars had higher polyphenol content than 'Haschberg'. The best results were recorded for 'Samyl' and 'Weihenstephan', but 'Korsör', 'Samidan' and 'Samocco' also contained substantial quantities of polyphenol. Earlier studies in Hungary (Szalóki-Dorkó, 2016) gave contradictory results, as the Danish 'Samocco', 'Sampo' and 'Samyl' cultivars were found to have polyphenol contents similar to that of 'Haschberg'. Few papers have been published on the polyphenol content of European elderberry cultivars; most research has involved either cultivars derived from Sambucus canadensis L., which is native to America, or wild-growing plants (Özgen et al., 2010;Thomas et al., 2013;Duymus et al., 2014;Wu et al., 2015).

Total anthocyanin content (TAC)
In addition to soluble solids content, the quantity of anthocyanins present in the fruit is also an important trait, since 90% of the harvested berries are used by the food industry to produce pigment concentrates (Kaack, 1990;Charlebois et al., 2010). Like the polyphenol content, the anthocyanin content, shown in Table 1, also exhibited considerable differences between the cultivars. The values recorded in the present work differed from some of those found in the literature. Lee and Finn (2007), for example, reported lower values, depending on the year and cultivar (364 and 582 mg CGE 100 g -1 FW), while higher anthocyanin contents, similar to those found here, were recorded by Veberic et al. (2009)  Statistical analysis also revealed the significant effect of the cultivar on the mean anthocyanin content (F(10;8.52)=36.18; p<0.001). With the exception of 'Haidegg 17' and 'Sampo', all the cultivars had higher anthocyanin contents than 'Haschberg', as also reported by Kaack and Knudsen (2015). In contrast, Mlynarczyk et al. (2020) detected higher anthocyanin content in 'Haschberg' than in 'Samyl' or 'Sampo'. In the present work the highest value was recorded for 'Samyl', which agrees with the results of Möhler et al. (2009) and Kaack (1989). On the other hand, outstanding values were found by Szalóki-Dorkó et al. (2015) for 'Samocco', and by Kaack et al. (2008) and Kaack and Austed (1998) for 'Sampo'.

Relationship between chemical parameters
Pearson's correlation analysis was used to detect statistical relationships between the chemical parameters, taking the results recorded in the individual years as a single data set (n=33). As also reported by Özgen et al. (2010), a close linear correlation was found between the polyphenol and anthocyanin contents (r=0.91, p<0.001).
In the present study the ratio between the polyphenol and anthocyanin contents (TAC/TPC ratio) ranged from 0.43-0.59 (Table 1), in agreement with the values reported by Jakobek et al. (2007) and Wu et al. (2015). In contrast, Szalóki-Dorkó (2016) found a much lower ratio (0.07-0.18), with the highest anthocyanin ratio for 'Haidegg 13' and the lowest for 'Samdal'. The TAC/TPC ratios reported by Lee and Finn (2007) for 'Haschberg' and 'Korsör' were in agreement with those found in the present work.
No other correlations were detected between the chemical parameters. Although phenolic compounds, including anthocyanins, exhibit intense antioxidant activity and many papers have reported that polyphenols and anthocyanins are strongly correlated with antioxidant capacity (Wang and Lin, 2000;Moyer et al., 2002;Özgen et al., 2006;Mathieu et al., 2015), this correlation could not be detected for the present data. As also reported by Ramaiya et al. (2012), TPC and TAC were not found to correlate with the FRAP parameter.
Influence of the year on the chemical composition of elderberry When examining the effect of the year on the chemical properties of the cultivars, the mean data were evaluated separately for each year (n=11). The statistical analysis (ANOVA) revealed that the year had a significant effect on SSC (F(2;30)=4.02; p=0.028) and FRAP (F(2;30)=5.21; p=0.011) ( Table 2), but not on the other parameters (TA: F(2;30)=1.69; p=0.202; TPC: F(2;30)=0.82; p=0.452; TAC: F(2;30)=1.32; p=0.283). Changes in the chemical traits of the fruit in different years were also reported by Lee and Finn (2007), Thomas et al. (2013) and Ferreria et al. (2020). The fact that the soluble solids content of elderberries changed in response to different climatic factors was also demonstrated by the findings of Tolic et al. (2017). The distribution and quantity of rainfall during the ripening period (July and August) was not consistent in any of the years (Table 3). While the monthly precipitation quantity fell on a number of occasions in 2017 and 2019, in 2018 almost the whole quantity fell on three days. In the case of sour cherries, it was observed by Szabó et al. (2010) that greater rainfall quantities during ripening led to a significant drop in the soluble solids content, but this was not confirmed in the present work, since the highest mean SSC value was recorded in 2019, the wettest year. The quantity of rainfall during ripening may also be decisive for the acid content, with higher quantities resulting in greater acid content and lower quantities in reduced acid content, as seen in 2019 (Mills et al., 1996). The mean temperature during ripening was the highest in 2018, with similar lower values in 2017 and 2019. According to Szabó et al. (2010) the fluctuation between day and night temperatures had the greatest influence on the soluble solids content, with larger temperature differences resulting in higher SSC and smaller differences in lower SSC. This was confirmed in the present work, as the highest mean SSC values were recorded in 2019, when the temperature fluctuation was most pronounced.
The antioxidant compounds in the fruit were influenced by the conditions that preceded harvest, the climate, the temperature, the light intensity, the soil type and mineral fertiliser (Wang, 2006). Some researchers found changes in antioxidant capacity in different years (Connor et al., 2002;Goncalves et al., 2007), as in the present work, while this was not observed in other cases (van der Sulis et al., 2001;Bolling et al., 2010). In studies on strawberry, Wang and Zheng (2001) concluded that higher temperatures (25-30 °C) resulted in considerably higher antioxidant activity and anthocyanin and polyphenol contents in cultivated fruit, while these values were lower under cooler conditions. However, this was not borne out by the present results, since the highest FRAP, TPC and TAC values were recorded in the two coolest harvesting seasons (2017,2018). According to Remberger et al. (2014) the quantity of antioxidant compounds in berries depends less on the temperature than on the rainfall. This was confirmed by the present data, as higher FRAP, TPC and TAC values were found in 2017 and 2019, when the rainfall quantity was also greater.

Conclusions
Due to the end-use of elderberries, the chemical properties of the fruit are just as important as the physical parameters. The present work investigated the effect of cultivar and year on the chemical components of elderberries and the existence of correlations between these components. The results showed that the chemical parameters were determined to a decisive extent by the genotype, though the climatic conditions also influenced annual changes in these properties. A statistical relationship between the chemical traits could only be detected for TPC and TAC. A comparative evaluation of the cultivars led to the conclusion that several cultivars were promising for individual quality traits. 'Haidegg 17' had extremely high soluble solids content and antioxidant capacity, while 'Sampo', 'Weihenstephan' and 'Samidan' were outstanding for their polyphenol and anthocyanin contents. The results of this study provide new information on the chemical components of European elderberry cultivars, which could be of use both for the industrial processing of the cultivars and for the selection of new breeding stocks.