The scientific hypothesis of this study was to examine the differences in various types of honeys from Czech and Slovak beekeepers due to their bioactive compounds evaluation measured by polyphenolic content and antioxidant capacity of two methods (DPPH and ABTS tests), and correlations between them.

There were evaluated 11 honey samples (nectars and honeydew honeys) of 6 honey types of Czech (produced in Moravian region, CZ) and Slovak origin (SK). There were 1 sample of acacia honey (A1) sample (SK), 1 rape honey (R1; CZ), 3 samples of floral honey (FL1-FL3; all CZ), 1 multi flower honey (MF1) sample (SK), 1 sample of forest honey (FO1; SK), 4 samples of honeydew honeys (HD1- HD4, all CZ), collected from private beekeepers, seasons 2013 and 2014. The samples were stored in glass jars in the dry dark place at room temperature (approximately 20 °C) and after jar opening they were analyzed up to ten days.

For the determination of honey color a modified spectrometric method of Beretta et al. (2005) was used. The honey samples were diluted to 50% (w/v) solution with distilled water. They were sonicated for 5 min and filtered through a paper filter and used for color analyses. Absorbances of samples were measured at two wavelengths, 450 nm and 720 nm against blank on the spectrometer (Libra S6 Biochrom, GB). The difference in absorbances (ΔA) was expressed. Determinations were made in triplicate.

The polyphenolic (PP) content in honeys was evaluated by a modified spectrophotometric method with Folin- Ciocalteau reagent (Socha et al., 2009). Samples for PP evaluation were prepared with 10 g honey and 40 mL of distilled water. After sonication (5 min) the solutions were filtered through a paper filter and quantitatively transferred into a 50 mL volumetric flask. Afterwards the samples were made up to 50 mL with distilled water and used for the analyses of PP. To extracts (0.1 mL) with 1 mL of distilled water, Folin-Ciocalteau reagent (1 mL; 10% (w/v); Penta Chemicals, CZ) was added and after agitation it was left for 5 min in the dark at room temperature, then 1 mL of sodium carbonate (10% (w/v); Penta Chemicals, CZ) solution was added and mixed again. After 15 min of standing in the dark at lab temperature absorbance of samples was measured at λ = 750 nm against blank using a Libra S6 Biochrom spectrometer (GB). Gallic acid was used as a standard and PP values were expressed as gallic acid equivalents (GAE) in mg.100g^-1 sample. Determinations were made in triplicate.

For the total antioxidant capacity (TAC) determinations modified spectrometric methods using ABTS and DPPH reagents (Škrovánková et al., 2018; Beretta et al., 2005) were used.

The procedures for honey samples extraction were same for both determinations. There were mixed 10 g of honey sample and 40 mL of distilled water. After sonication (5 min) the solutions were filtered through a paper filter and quantitatively transferred into a 50 mL volumetric flask. Afterwards the samples were made up to 50 mL with distilled water and used for the analyses.

ABTS method : To 50 μL of honey extract the reactive radical mixture (4 mL), composed of ABTS (2,2′-azinobis- 3-ethylbenzthiazoline-6-sulphonic acid; Sigma Aldrich, CZ) (12 mL; 3.5 mM) with K₂S₂O₈ (0.06 M; Lukes, CZ) and acetic buffer (pH 4.3), was added. The reaction mixture was shaken vigorously on a Vortex mixer and for 30 min it was left to react without light exposure at room temperature. Honey samples absorbance (A) and absorbance of control samples (AC) were after time limit measured at λ = 734 nm against blank by spectrometer Libra S6 Biochrom (GB). Inactivation (I) was calculated from the decrease of absorbance (%) according to relation (1). Results of TAC (ABTS) were calculated from inactivation using calibration curve with trolox as standard. It was expressed as trolox equivalents (TE) in mg.100g^-1 sample. Average results were obtained from three parallel determinations.

(1) I = A C − A A C .100

DPPH method : To prepared honey extract (0.2 mL) a DPPH (1,1-diphenyl-2-picrylhydrazyl) solution in ethanol (1.9 mL; 0.02 mM; Sigma Aldrich, CZ) and acetate buffer solution (1 mL; pH 5.5) were added. The mixture was shaken vigorously on a Vortex mixer in capped glass and left without light exposure for 1 h at room temperature. Absorbance of samples (A) and absorbance of control samples (AC) was measured at λ = 517 nm against blank on the spectrometer (Libra S6 Biochrom, GB). Their inactivations (I) were also calculated from the decrease of absorbance according to relation (1) and the values were expressed as trolox equivalents (TE) in mg.100g^-1 sample. Average results were obtained from three parallel determinations.

All analyses were provided in triplicate, and the data were expressed as mean values ± standard deviation (SD). Statistic evaluation of the results was made by Statistica program, StatSoft version 9.0 (Dell, USA) using parametric test comparing mean values of two independent assortments (Student t-test). Differences at a 95% confidence level (p <0.05) were considered statistically significant. Correlations between the parameters evaluated were obtained using Pearson’s correlation coefficient (r).

In terms of our measurements and also due to sensorial examination, honey color ranges from white pale, pale yellow to amber and dark brown with the values ΔA from 0.139 to 0.817, the average 0.478 (Table 1). As expected, as the darkest honeys with the highest values were evaluated honeydew honeys (HD samples). Their values (average 0.779) were nearly five times higher than the brightest ones, acacia and rape honeys (average 0.161); and nearly three times to the values of floral honeys (0.268). Multi flower and forest honeys belong to the group of darker honeys. Several compounds, pigments are responsible for honey color. To the most important honey pigments belong water soluble polyphenols, flavonoids, and lipid soluble carotenoids (Isla et al., 2011).

Also mineral composition is important for honey color. As González-Miret et al. (2005) evaluated lightness is significantly related with minerals such as S, Ca, Fe, As, Pb, and for the dark honey types (chestnut, and honeydew honeys) also Cd is considerable. Due to these facts, there are therefore expectations of higher content of polyphenols and probably also higher antioxidant values for darker honeys (Bertoncelj et al., 2007; Kuś et al., 2014).

In the determination with Folin-Ciocalteau reagent electron-donating antioxidants such as polyphenols, ascorbic acid, and vitamin E are evaluated. The total polyphenols (PP) contents of the honey samples (Table 1) range from 54 to 254.2 mg GAE.100g^-1 with the average 150 mg GAE.100g^-1 honey. There were marked differences between honeys (p <0.05, Student t-test). Also Bertoncelj et al. (2007) mentioned that total phenolic content differs widely among different honey types. As it was expected due to other scientist researches (Moniruzzaman et al., 2014; Alvarez-Suarez et al., 2010) brighter, pale honeys had lower polyphenol values. PP results for honeys were in agreement with literature sources; lowest for acacia honey, where also Beretta et al. (2005) evaluated the lowest content (5.5 mg GAE.100g^-1). It was followed by rape honey and floral ones, forest and multi flower honey, whereas the highest PP content had the darkest honeydew honeys with the nearly five times higher values in comparison to the lowest ones. These findings are in agreement with the values previously reported for other European honeys, as Kuś et al. (2014) determined in Polish honeys total phenolic content in the range 12.2 – 117.4 mg GAE.100g^-1. The composition of honeys is dependent on the botanical origin, floral source, and also seasonal and environmental factors, as well as processing (Kıvrak and Kıvrak, 2017; Cavazza et al., 2013; Dimitrova et al., 2007).

Honey samples exhibited similar order of samples for color and PP values. To characterize the relationships between color and polyphenolic content the correlation (Figure 1) was evaluated. They are strongly related with a correlation factor r = 0.8294. Sant'ana et al. (2014) also discovered that the lowest total phenolic content corresponded to light-colored honey and the highest values to dark honeys. Positive correlation between color and PP determined also other scientists (Alvarez-Suarez et al., 2010; Kuś et al., 2014).

Honey color thus seems to be a relatively reliable parameter to indicate high PP content in honey.

The antioxidant potential that means overall hydrogen/electron-donating activity of present antioxidants, was measured by two methods, with ABTS and DPPH test. The total antioxidant capacity (TAC) results for honey samples are demonstrated in Table 1. The TAC values of ABTS method were in the range from 15.5 to 89.6 mg of trolox equivalents per 100 grams of honey sample with the average 54.5 mg TE.100g^-1; and from 11.3 to 118.7 mg TE.100g^-1 for DPPH method with the average 58.4 mg TE.100g^-1, respectively. There were marked differences between honeys (p <0.05, Student t-test). Also, Frankel et al. (1998) showed in their study that great variations exist in the chemical nature of honey from different floral sources as they found in best honey source 20.3 times higher concentration of antioxidants in comparison with that of the lowest one. The least active honeys in our research were, similarly like in PP evaluation, the brightest honeys (Beretta et al., 2005; Kuś et al., 2014), pale acacia (Bertoncelj et al., 2007) and rape honeys, for both methods.

Higher potency in scavenging of DPPH free radical and also ABTS test showed floral honeys (average 34.8 and 40.2 mg TE.100g^-1, respectively), forest and multi flower honey. Honeys with the best antioxidant potency were evaluated honeydew honeys (average 98.1 mg and 80.6 mg TE.100g^-1, respectively). The highest results for TAC (DPPH and ABTS test) were nearly eleven and nearly six times higher, respectively, than the lowest TAC value. Also, Ferreira et al. (2009) obtained the highest antioxidant values in the dark honeys.

Similar descending order of TAC values was observed also for honey color. The correlations to examine the relationships between them (Figure 1) were evaluated. Antioxidant capacity showed a strong relationship comparing both assays and color intensity, for ABTS evaluation r = 0.8836, and for DPPH test r = 0.8937. High correlation between color and antioxidant capacity was determined also by other researchers (Sant'ana et al., 2014; Pontis et al., 2014; Beretta et al., 2005; Bertoncelj et al., 2007). Honey color thus seems to be a relatively reliable parameter to indicate not only PP content but also antioxidant potential in honey.

Antioxidant capacity determined by both analyses was also strongly positively associated with the polyphenolic content (r = 0.9005 for ABTS; r = 0.8687 for DPPH), as shown in Figure 1. Therefore high PP contents predicate high TAC values in analyzed honey samples. Also Wilczyńska (2014), Sant'ana et al. (2014), Pontis et al. (2014), Beretta et al. (2005), and Bertoncelj et al. (2007) found positive correlation between PP and TAC for honey samples. Although Pontis et al. (2014) determined high flavonone and dihydroflavonol content in some honeys, no correlations between them and antioxidant potential they observed. Generally, our findings and data in the literature have shown a linear relationship between honey color, phenolic compounds and antioxidant capacity.