Characterization and Classification of Spanish Honeydew and Blossom Honeys Based on Their Antioxidant Capacity

Abstract

Honey is a very appreciated product for its nutritional characteristics and its benefits for human health, comprising antioxidant, anti-inflammatory, antifungal, and antibacterial activities. These attributes depend on the specific composition of each honey variety, with the botanical origin as one of the distinctive features. Indeed, honeydew and blossom honeys show different physicochemical properties, being the antioxidant capacity, mainly relying on the phenolic compound content, one of the most important. In this work, Folin-Ciocalteu (FC) index, total flavonoid content (TFC), and the antioxidant capacity based on the ferric reducing antioxidant power (FRAP) assay were determined for a total of 73 honeys (50 blossom honeys and 23 honeydew honeys). Mean content of oxidizable species (FC index) ranges from 0.17 to 0.7 mg eq. gallic acid g^-1, with honeydew honeys being the ones with higher values. Regarding TFC, mean values above 1.5 mg eq. quercetin g^-1 (method applied in the absence of NaNO₂) were obtained for honeydew honeys and heather honey. Lower and not discriminatory values (below 0.3 mg eq. epicatechin g^-1) were obtained in the presence of NaNO₂. The maximum antioxidant capacity was observed for thyme honeys (2.2 mg eq. Trolox g^-1) followed by honeydew and heather honeys. Individually, only the FC index was able to discriminate between honeydew and blossom honeys, while the other spectroscopic indexes tested allowed the differentiation of some honey types according to the botanical origin. Thus, a holistic treatment of the results was performed using partial least square discriminant analysis (PLS-DA) for classification purposes using FC, TFC, and FRAP results as data. Honeydew and blossom honey were satisfactorily discriminated (error 5%). In addition, blossom honeys can be perfectly classified according to their botanical origin based on two-class PLS-DA classification models.

Keywords: PLS-DA; blossom honey; ferric reducing antioxidant power; honeydew honey; total flavonoid content; total phenolic content.

Figure 1

Boxplots with the total content of oxidizable species in honey samples according to the botanical origin. For botanical origin code refer to Table 1. Blue boxes correspond to blossom honeys, while orange ones correspond to honeydew honeys. The same letter indicates no significant differences (α = 0.05).

Figure 2

Boxplots with the total flavonoid content in honey samples according to the botanical origin. (A) method I (absence of NaNO₂) and (B) method II (presence of NaNO₂). For botanical origin code refer to Table 1. Blue boxes correspond to blossom honeys, while orange ones correspond to honeydew honeys. The same letter indicates no significant differences (α = 0.05).

Figure 3

Boxplots with the antioxidant activity (FRAP assay) in honey samples according to the botanical origin. For origin code refer to Table 1. Blue boxes correspond to blossom honeys, while orange ones correspond to honeydew honeys. The same letter indicates no significant differences (α = 0.05).

Figure 4

PLS-DA results for the classification of honeys according to both blossom or honeydew origin. Three latent variables were used to build the model.

Figure 5

Two-class classification models obtained by PLS-DA of blossom versus honeydew honeys. The calibration set (filled symbols) comprises the 60% of all samples randomly chosen and the prediction set (empty symbols) comprises the 40% remaining samples.