PHYSICOCHEMICAL CHARACTERISTICS OF SERBIAN HONEYDEW HONEY

Th e aim of this study was to investigate the composition and quality of Serbian honeydew honey. For this purpose, the physicochemical characteristics of 14 honeydew samples were analyzed. Th e physicochemical characteristics of all honeydew honeys from Serbia analyzed in this research can be considered to be within the parameters prescribed for honeydew in general. Th e sum value of glucose and fructose, the content of sucrose, water, hydroxymethylfurfural, acidity, and diastase activity were in line with European and national regulations for honey, for all investigated honeydew samples. Out of a total of 14 tested honey samples, 1 sample did not comply with the national regulations for honey regarding electrical conductivity. According to our results, in most of investigated samples the fructose/glucose (F/G) ratio was greater than 1.11 and glucose/water (G/W) ratio was close to 2. Th is means that they can be categorized as medium-crystallizing honeys. Th e results obtained in this study indicate excellent quality, absence of undesirable fermentation, acceptable freshness and proper manipulation of Serbian honeydews.


Introduction
Th e interest in honeydew has increased because of its nutritional, sensorial and potential therapeutic properties . Honeydew is a distinctive honey compared to blossom honey since it usually has higher electrical conductivity, pH values, ashes, higher content of disaccharides, trisaccharides, and lower content of monosaccharides Živkov Baloš et al., , 2019a. Darker color and diff erent sensory features compared to blossom honey are also the characteristics of this type of honey (Escuerdo et al., 2013;Flores et al., 2015;Živkov Baloš et al., 2019b). Numerous studies have established that honeydew contains a number of bioactive compounds, such as proteins, amino acids and phenolics (Bogdanov et al., 2004;Escuerdo et al., 2013;Silva et al., 2018;Vasić et al., 2019), compared to other types of honey, which classifi es it as health-promoting food.
Basically, honey is a concentrated water solution of fructose and glucose, with small amounts of various complex sugars (Escuredo et al., 2014;Valdés-Silverio et al., 2018). Fructose and glucose are present in nectar or in the carbohydrate excretion of insects that suck the fl uid from the phloem (Lazarević et al., 2017). As a result of high concentration of sugar, honey crystallizes over time. Some components of honey, such as other carbohydrates, pollen grains, air bubbles, and particles can have an impact on the crystallization of honey. Fructose/glucose (G/F) ratio and glucose/water (G/W) ratio can be used to estimate the rate of honey crystallization. Generally, honeys with low G/W and F/G ratio do not crystallize easily (Escuredo et al., 2014).
Crystallization of honey is an undesirable process. During the crystallization changes occur in the textural properties, making honey less appealing to consumers, who prefer liquid and transparent product (Kabbani et al., 2011). Crystallization of honey aff ects the processing of honey during extraction, fi ltration, mixing and bottling (Dobre et al., 2012;Laos et al., 2011).
Honeydew honey is a natural product with complex composition, which depends on bee species, geographical region, available fl oral source and storage conditions (Karabagias et al., 2014). Considering all the factors mentioned above and the number of possible fl oral sources in particular, it is understandable that no two honeys are the same . Honeydew physicochemical quality criteria are well specifi ed by the European Legislation (European Commission, 2002) and the regulation concerning the quality of honey in the Republic of Serbia (Offi cial Gazette RS, 101/2015). Th e major criteria for honey in both of those standards are sugar content (sucrose and sum of fructose and glucose), moisture content, water-insoluble content, electrical conductivity, free acidity, diastase activity and hydroxymethylfurfural (HMF) content.
Th e aim of this study was to investigate the composition of Serbian honeydew honey, in order to obtain the information about the honey quality, and gain an insight into its nutritional suitability.

MATERIAL AND METHODS
A total of 14 honeydew samples were collected from beekeepers at diff erent regions of Serbia. All collected samples were in their original packaging and transferred to the laboratory of Scientifi c Veterinary Institute "Novi Sad" for examination. Manufacturers used fi eld observations for botanical origin determination. Our research included only the samples with confi rmed botanical origin stated on the manufacturing specifi cation label. All the selected samples were produced by Apis mellifera. Honey analyses were carried out immediately aft er sampling. All samples were analyzed in duplicate by methods prescribed in Harmonised methods of the International Honey Commission Methods (2009).

Water content analysis
Water content was determined by refractometry, measuring the refractive index (RI) using a standard model Abbetype refractometer at 20° C. Water content (%) was then obtained from the Chataway table.

Electrical conductivity
Electrical conductivity was measured at 20° C in solutions of honey samples (20.0 g dry matter of honey in volume solution in 100 ml distilled water) using a conductometer Crison (Type Basic 30).

Free acidity
Th e acidity of honey was determined by volumetric method. Ten grams of honey was dissolved in 75 ml of water and solution was titrated with 0.1 M NaOH to pH 8.30. Acidity is expressed in milliequivalents/kg honey (mEq/kg).

Water-insoluble matter
Insoluble matter was determined by gravimetric method. Th e insoluble matter was collected on fi lter of specifi ed pore size by rinsing with warm water. Dried residues (135° C) were weighed until constant weight was obtained.

Hydroxymethylfurfural
HMF was determined by an HPLC Dionex UltiMate 3000 Series system with UV detection (Th ermo Scientifi c, Germany). Honey sample (1 g) was dissolved in 25 mL of water, fi ltered through a 0.45 μm nylon fi lter and injected (10 μl) into the HPLC system. Th e HPLC column was 150x3 mm Hyperilsil GOLD, with particle size of 3 μm. Th e mobile phase was methanol: water (10:90, v/v), at fl ow rate of 1 mL/min. All measurements were conducted at room temperature. Th e system was controlled by Chromeleon® 7 soft ware (Th ermo Scientifi c, Germany).
Th e external calibration curves produced by standard solutions were used to quantify the amount of HMF in the samples.

Sugar Composition determination
Th e sugar composition (fructose, glucose, sucrose) was determined by an HPLC Dionex UltiMate 3000 Series system (Th ermo Scientifi c, Germany) equipped with a refractive index detector RefractoMax521 (ERC Inc, Japan) at 35ºC. Honey sample (1 g) was dissolved in 25 mL 25% methanol, fi ltered through a 0.22 μm nylon fi lter and injected (5 μl) into the HPLC system. Th e HPLC column was Hypersil GOLD Amino 150x3 mm (particle size 3 μm), fi tted with a guard column Hypersil GOLD Amino 10x3 mm (particle size 3 μm). Th e mobile phase was acetonitrile: water (8:2, v/v) fi ltered through 0.22 μm membrane fi lter, at a fl ow rate of 1 mL/min. All measurements were performed at room temperature. Th e system was controlled by Chromeleon® 7 soft ware (Th ermo Scientifi c, Germany). Th e external calibration curves produced by standard solutions were used to quantify the amount of sugars in the samples. Honey sugars were identifi ed and quantifi ed by comparing their retention times and peak areas with those of standard sugar solutions.

Diastase activity
Diastase activity was determined by spectrophotometric method (Megazyme International Ireland, 2014). Two grams (2.00 g) of sample were dissolved in sodium maleate buff er and adjusted volume to the mark of volumetric fl ask with water. Amylazyme tablets (Megazyme International, Ireland) were added to buff er solution. In the presence of α-amylase, the substrate is hydrolysed and soluble dyed products are released. Th e reaction was terminated and, following the fi ltration, the absorbance of the fi ltrate is measured at 590 nm. Th e absorbance is directly proportional to the diastase activity of the sample. Th e diastase activity is calculated as diastase number (DN).

Statistical analysis
Statistical analysis was performed by the PAST soft ware package, version 2.12, Oslo, Norway. Th e data were grouped according to the samples of honeydew and presented as mean, standard deviation, minimum, maximum values, and coeffi cient of variation.

Results
Th e results of physicochemical analysis of Serbian honeydew are shown in Tables 1 and 2. Th e water content in all investigated honeydew samples was below 20%, which is the maximum permissible level set by national regulations for honey (Offi cial Gazette, 101/2015).
Th e results of sugar profi le analysis by HPLC-RI are shown in Table 1. Our study revealed that in all examined honeydew samples, the percentage of fructose and glucose ranged from 28.51 to 38.83 and from 24.92 to 34.09%, respectively. Th e sum value of glucose and fructose was in line with European and national regulations (EU, 2002; Offi cial Gazette RS, 101/2015) with the value of over 45 g/100 g for all honeydew samples. Sucrose content in all investigated honey samples was below 5 g/kg honey, which is the maximum permissible level set by the European Legislation (EU, 2002) and national regulations for honey (Offi cial Gazette RS, 101/2015). In 13 out of the total of 14 investigated samples (93%), sucrose content was below the detection limit of the applied method. Detection limit of the applied method is 0.25 %. Th e F/G ratio and G/W ratios were calculated for all samples of honeydew. Th e mean F/G ratio was 1.22 and ranged from 1.06 to 1.50. Th e mean G/W ratio was 1.89 and ranged from 1.50 to 2.32.  4 -unquantifi able value (less than detection limit); STDEV 5 -standard deviation; CV 6 -coeffi cient of variation In accordance with the regulation concerning the quality of honey in the Republic of Serbia (Offi cial Gazette, 101/2015), minimum electrical conductivity in honeydew put in the market is fi xed to 0.8 mS/cm. Th e values of electrical conductivity in the investigated honeydew samples were between 0.61 and 1.99 ms/cm. Out of a total of 14 tested honey samples, 1 sample (No. 11) did not comply with the national regulations for honey.
Maximum value of free acidity in all types of honey (except in baker's honey) was 50 mEq/kg, as is set by regulation (Offi cial Gazette, 101/2015). Free acidity in all tested honeydew samples was below 50 mEq/kg. Th ese data indicate the absence of undesirable fermentation. Th e mean acidity value in the investigated samples was 21.29 mEq/kg. Th e regulation concerning the quality of honey in the Republic of Serbia established the maximum 5-HMF content (40 mg/kg) and minimum diastase activity (8 DN). Generally, all tested samples were in compliance with the provisions of the Regulations regarding the content of 5-HMF and diastase activity. Mean 5-HMF content in investigated honeydew samples was 6.47 mg/kg, with the range 0.50 to 26.01 mg/kg. Minimum value of diastase activity was 12.08 DN and mean activity was 18.43 DN ( Table 2).

Discussion
Honey is a highly viscous solution of sugars, dominantly glucose and fructose in about equal concentrations (Venir et al., 2010). Th e most prevalent sugar in honeydew is fructose, followed by glucose and sucrose. Th ese results are in accordance with other literature data (Kirs et  F/G and G/W ratios are important parameters for predicting the crystallization tendency of honey. Fructose/glucose ratio shows the degree of honey crystallization, because glucose is less water soluble than fructose . Honey samples in which F/G ratio is greater than 1.33 do not crystallize for a long period of time. If the F/G ratio is less than 1.11, honey crystallizes quickly (Escuerdo et al., 2014). Crystallization process is slower or null when G/W ratio is less than 1.7, and it is faster when the ratio is higher than 2 (Dobre et al., 2012) or 2.10 (Venir et al., 2010). According to our results (Table 1), in most of investigated samples, F/G ratio was greater than 1.11 and G/W ratio was close to 2, so they can be categorized as medium-crystallizing honey.
Water is the second largest component of honeydew. Honey moisture depends on the production season, fl oral source, abundance of nectar fl ow, soil, ventilation of beehives, colony strength, meteorological conditions in the area of honey production (primarily air humidity), and maturation and honey harvest time (Escuerdo et al., 2014;Kirs et al., 2011;Lazarević et al., 2017;Sousa et al., 2016;Živkov Baloš et al., 2019b). Moisture signifi cantly aff ects the physical properties of honey such as crystallization, viscosity, and rheological behaviour. Although there are diff erences in moisture content, it can be assumed that the tested honeydews had adequate maturity since moisture was generally lower than the maximum permissible value (20%). As shown in Table 1., moisture of honeydew samples is fairly homogeneous, and this parameter is characterized by a low coeffi cient of variation.
Th e electrical conductivity is oft en used in routine quality control of honey. Th e conductivity is related to the concentration of soluble minerals, organic acids and proteins. It is a useful tool for distinguishing honeys of different botanical origin. Storage time can also aff ect the electrical conductivity of honey. Th e most adequate parameters for distinguishing honeys of diff erent geographical origin are those which described the patterns of pH and electrical conductivity with changes of honey concentration (Acquarone et al., 2007). Honeydew honey is characterized by higher electrical conductivity than blossom honeys, which is a good parameter to distinguish between the two types of honey. Th e values lower than 0.8 mS/cm may indicate adulteration or mixtures with other types of honey .
Free acidity in the examined samples ranged between 4.6 and 48.4 mEq/ kg (CV = 68.2%) ( Table 2). Th e acidity of honey is caused by organic acids (tartaric, citric, oxalic, acetic, etc.), nectar or bee secretions (Yadata, 2014). Th e acidity value varies depending on the fl oral source and bee species (Sousa et al., 2016). Our earlier results (Živkov  demonstrated high acidity of forest honey, as compared with other honey types. Th ese results were also reported by Bergamo et al. (2019) and Primorac et al. (2009). Th e natural acidity of honey can be increased by the storage and ripening of honey, as well as during honey fermentation. Honey adulterated with sugar syrup has acidity lower than 1, while honey that is adulterated with invert sugar has a pronounced high acidity (Yadata, 2014).
5-Hydroxymethylfurfural is formed as an intermediate product in the Maillard reaction from the direct dehydration of sugars under acidic conditions during thermal processing of foods (Pasias et al., 2017). Th is compound is formed slowly and naturally during honey storage. Th us, it is considered an indicator of honey freshness (Serglaio et al., 2019). Diastase activity is closely related to 5-HMF. Th is parameter is sensitive to heat and long storage period. Diastase activity signifi es a possible overheating of honey, above 60° C, as well as prolonged storage . Considering the 5-HMF content and diastase activity found for investigated honeydew honeys, all samples were in accordance with the established limits. Th ese results indicate acceptable freshness and proper manipulation of honeydews.

Conclusion
Th e physicochemical characteristics of all honeydew honeys from Serbia analysed in this research are within the parameters expected for honeydew in general. Additionally, it can be concluded that Serbian honeydew is characterized by good quality as honey samples were within limits established by the European and national Legislation.
Th erefore, further research on honeydew physicochemical and therapeutic properties is required to confi rm the quality and authenticity of this product and for better understanding of the value of this honey.

Author's Contribution:
M.Ž.B. draft ed the manuscript and made substantial contributions to the basic idea; S.J. carried out the HPLC analysis and was involved in draft ing of the manuscript; N.P. carried out other physicochemical analysis and performed statistical analysis; D.LJ.P and S.V.K. were involved in draft ing of the manuscript; M.P. collected the data; V.P. and D.M. revised the manuscript.