Chemical analysis and sensory evaluation of honey produced by honeybee colonies fed with different sugar pastes

Abstract Supplemental feeding of honeybee (Apis mellifera L.) colonies is essential for colony buildup. Honey samples obtained from feeding honey bee colonies with different commercial sugars, including sugar paste, sugar paste + vitamins +amino acids, and sugar paste + vitamins +protein as pollen substitute, were studied to determine the effects of feeding bees on the physicochemical and sensory characteristics of honey, compared with the honey produced by a control group (no supplemental feeding). Analyzed honey samples from the different groups were in accordance with the criteria described in Council directive 2001/110/CE, 2002. Elsewhere, significant differences (p > .05) were detected in color, free acidity, diastase activity, hydroxymethylfurfural, sugar profile, and conductivity between all honey. In terms of mineral content, the honey from hives faded with sugar paste + vitamins +protein and control group had higher values for Na, Mg, P, K, Ca, Mn, Fe, Cu, and Zn. Related to sensory analyses, no differences in flavor and basic taste were found in all analyzed honey (p > .05) independently the type of feeding. For the visual attributes, only differences were found for the color. Supplementary feeding with different sugar pastes and proteins does not affect the physicochemical characteristics of honey. For the sensory analyses, control sample presented significant differences only for color and chemical odor attributes compared with honey from bees receiving supplementation.

identified, although there are several factors, acting in combination or separately. These are thought to include climatologically difficult years with a consequent nutritional impact on colonies, the effect of neonicotinoid insecticides and unsuitable management practices (Pajuelo, Torres, & Bermejo, 2008).
The nutrition of bees is essential in certain periods of the year (Guler et al., 2014). Supplementary feeding can range from a diet that consists almost exclusively of carbohydrates, to others more balanced that carry energy, protein, and lipid. In the first case, the main interest is to accelerate the energetic metabolism of the whole hive, while in the second case it is sought to stimulate breeding and so increase in population density. In general, honeybees feed upon nectar of flowers and pollen. The nectar fulfills their carbohydrate requirements (Brodschneider & Crailsheim, 2010). However, the use of pollen as a protein and lipid source has its contraindications, since pollen can be the vehicle for transmission of pathogens (Durrer & Schmid-Hempel, 1994;Singh & Kundu, 2010).
The carbohydrate needs of honeybee colonies can be provided by sugar cane, sugar beet, and sugar maize due to their low cost (Ruiz-Matute, Weiss, Sammataro, Finely, & Sanz, 2010;Sammataro & Weiss, 2013). Certainly, supplementary diets with increased protein intake have been provided for decades (Standifer, Moeller, Kauffeld, Herbert, & Shimanuki, 1978;Herbert, 1992. Of great importance the adequate feeding during of seasons when food resources are drastically reduced (such as winter or dry periods), with little natural food available to maintain the queen during egg-laying and to generate healthy of spring, renew food stocks and allow honey production (DeGrandi-Hoffman et al., 2016;Morais et al., 2013). The effect on the quality of honey of supplementing the bee diet with protein has not been studied, but it is known that feeding bees with high percentages of sucrose, corn syrup, and high fructose syrup can generate problems of indirect adulteration of the honey obtained (Guler, Bakan, Nisbet, & Yavuz, 2007). The effect of supplementation on the physicochemical and the sensory quality of honey has not studied in detail.
The sensory analysis applied to honey is an important complement to the physico-chemical parameters and pollen analyses. It can confirm defects in fermentation, and the presence of impurities, the odor of smoke, metallic tastes, and other characteristics that common laboratory routine analyses do not access (Piana et al., 2004). In this sense, many studies have been carried out on physicochemical and sensory analyses for use as analytical tools for the quality control of honey in relation to its botanic origin (Anklam, 1998;Galán-Soldevilla, Ruiz-Pérez-Cacho, Serrano, Jodral, & Bentabol, 2005;González-Lorente, De Lorenzo, & Pérez-Martin, 2008).
The present study aimed to evaluate the physicochemical compositions and sensory properties of the honey produced by honeybees fed with supplemental sugars pastes of different compositions, including vitamins, amino acids and/or protein as pollen substitute, compared with the honey produced by a control group (with no supplemental feeding).

| Supplemental Feeding and honey harvest
The feeding and harvest were carried out in the beekeeping unit of Research Farm of the University of Murcia, Faculty of Veterinary, starting the experimental treatments in December 2017. Twenty four colonies were distributed randomly into four experimental groups with various feeding supplements: M1, control group (no supplemental feeding); M2 sugar paste; M3 sugar paste with vitamins and amino acids; M4, sugar paste with 3% protein (pollen substitute) and vitamins. The colonies received the supplements in plastic trays containing 1 kg of the supplement every 15 days during the experimental period (from December 2017 to April 2018).
The honey was collected in April 2018 by centrifugation and filtered through a sieve. Honey obtained was kept unpasteurized in glass containers at room temperature until physicochemical and sensory analysis.

| Chemical analyses
The water content was determined using a refractometer at 20ºC.
The samples were homogenized at room temperature and directly deposited in the prism of the refractometer. The obtained refractive index in each sample was related to the water content of the honey, according to the relationship between honey water content and refractive index (Bogdanov, Ruoff, & Oddo, 2004;Chataway, 1932).
The pH was measured directly in the water solution of the honey sample using a pH-meter (CRISON, GLP 21) (AOAC, 2012;Bogdanov, 2009). The electrical conductivity was measured in accordance with AOAC (2009). Twenty grams of honey in distilled water were weighed and transferred to a 100 ml flask, completing the volume with water. The solution was transferred to a beaker, and the electrodes were immersed. The reading of the conductance of the solution was in µs/cm. Hydroxymethylfurfural (HMF) was determined by using the AOAC (1990) Official Method 980.23. Five grams of honey was dissolved in 25 ml of distilled water and treated with a clarifying agent (0.5 ml of Carrez I and 0.5 ml of Carrez II solutions), and the volume was made up to 50 ml. The solution was filtered, and the first 10 ml was discarded. The absorbance of the filtered solution was measured at 284 and 336 nm against an aliquot of the filtered solution treated with NaHSO 3 . HMF was determined as follows: HMF/100 g of honey = (Abs 284-Abs 336) × 14.97 × (5/g of the sample).
The diastase activity was measured using the Phadebas method for α-amylase. Phadebas is a synthetic reagent that produces a blue color when it is hydrolyzed by the diastase. Adsorption was determined using a UV/VIS spectrophotometer at 620 nm, the absorbance is directly proportional to the diastase activity in the honey sample. Results were expressed in Gothe units per gram of honey (Bogdanov, 2009).
The sugar content was determined by HPLC with a RI (refractive index) detector and analytical stainless-steel column in polar aminopropylsilane (NH2). Five grams of honey dissolved in water were transferred to a 100 ml volumetric flask containing 25 ml of methanol and topped up with water. The solution was filtered through a syringe filter (Bogdanov, 1997 Technologies, Santa Clara, CA, USA). In a preliminary step, samples were heated and sonicated to facilitate honey homogenization.

| Panel selection
Eight panelists aged from 20 to 52 from the Faculty of Veterinary, University of Murcia, previously selected and trained according to ISO 8,586, were further trained in the appearance, odor, flavor, taste, mouthfeel, and textural parameters of honey (Table 1). All the sensory analyses were carried out following Piana et al. (2004)

| Sample preparation and testing procedure
Six samples from each of the four groups (M1, M2, M3, and M4) were presented to the tasting panel. The samples of honey were presented in transparent 100 ml jars containing 30 g of honey at 20 ± 2ºC. Each sample was given a random three-digit code and presented in a different order for each panelist sample was evaluated in triplicate in different sessions, in which three or four samples were presented.

Visual attributes
Color intensity Degree of amber color (varying from water white to dark ambar)

Crystallization
Phenomenon that causes the loss of fluidity. The size of the crystals must be uniform (for crystallized honey).

Viscosity
Force required to remove honey from a spoon (for liquid honey)

Olfactory and aroma attributes
Overall intensity Strength of the stimuli perceived by the nose or by olfactory receptors via retronasal way.

Associated with different flowers
Fruity Associated with different fruits: acid, ripe and tropical

Vegetal
Associated with gardens, green notes, dry leaves, and wood Warm Associated with foods characterized by their sweet smell and taste.

Chemical
Not associated with food, it is characterized by its aggressiveness (smoked, phenolic, sulfuric, vinegary).

Taste and Mouthfeel
Sweetness Sensation produced by products that contain sugars such as sucrose and fructose.

Sourness
Sensation produced by products that contain acids, such as citrus.

Saltiness
Sensation produced by products that contain salts, such as sodium chloride

Bitterness
Sensations produced by products such as caffeine.

Persistence
Feeling similar to what is perceived while the product was in the mouth and while continuing over a period of time measurable.

Astringency
Organoleptic property of pure substances or mixture which produce an astringent sensation.

Freshing sensation
Sensation of freshness in the oral cavity (similar to that produced by mint)

Texture attributes
Adhesiveness Ability of honey to stick to the teeth and oral cavity.

Granularity
Geometric attribute of texture relative to the perception of the size and shape of the particles in crystalline honey.
Assessors made a descriptive quantitative analysis of kinds of honey using an unstructured scale (10 cm) by evaluating the appearance (color, fluidity, and the crystallization) and olfactory analysis (odor intensity). The first odor contact could be reinforced by extending the honey sample on the container walls with a spatula.
Besides, each taster had a bottle with coffee beans in order to relax the smell. For basic tastes and aroma evaluation, a small amount of honey was placed on the tongue with a disposable spatula. The sample was allowed to dissolve for a few seconds while the subject did not inhale. Air was released through the nose, keeping the mouth closed, so that the aromas stimulate the olfactory receptors. Water and low salt bread were provided to clean the palate between samples. Finally, texture attributes (adhesiveness and crystallization) were determined.

| Statistical analysis
The statistical analysis used IBM SPSS STATISTIC (Version 24). An analysis of variance (ANOVA) and Tukey's HSD multiple comparison tests (p < .05) was carried out to establish the difference between pairs of groups, according to the physicochemical and sensory parameters.

| Chemical analyses
The water content is one of the most important characteristics influencing the physical properties of honey (Escuredo, Míguez, Fernández-González, & Seijo, 2013). The samples from the present study did not present significant differences (p > .05) concerning the type of supplement administered to bees ( Table 2). The water content values are within the range found in Greek honey (10.50% −20.50%) (Karabagias, Badeka, Kontakos, Karabournioti, & Kontominas, 2014) and are lower than those obtained in blossom honey from Spain (15.50%-18.90%) (Bentabol, Hernández, Rodríguez, Rodríguez, & Díaz, 2014). In general, the values for all the honey samples were within the established legal requirements (Council Directive, 2001).
A pH level of between 3.2 and 4.5 and the natural acidity of honey inhibit the growth of microorganisms (Karabagias et al., 2014).
The type of supplementation used did not alter the pH value, and the pH range was within that accepted for honey . Values for free acidity ranged from 14.17 to 16.83 meq/kg.
Differences between acidity values may be the result of floral origin, the presence of organic acids, or the geographical origin (Isla et al., 2011). In our analysis, there were significant differences (p > .05) between the honey from the control group and all kinds of honey from supplemented groups, regardless of the type of supple- there are no studies that analyze the effect of supplementation on the color of honey. In general, color values depend on the mineral content and floral origin (Piana et al., 2004), but the qualitative pollen analysis of our samples showed that the diversity of pollen types was similar between honey groups. Hence, the difference in color between the honey samples would have been mainly due to the supplementation administered, since the botanical origin was the same for all the groups.

Diastase is a natural honey enzyme and its activity is used in
Europe as a determinant of freshness.  Table 2). The electrical conductivity of honey is related to the ash content and acidity (Yücel & Sultanog, 2013), and should not surpass 800 μS/cm in blossom honey, from a quality control point of view (Kaškonienė, Venskutonis, & Čeksterytė, 2010).
In terms of mineral content, significant differences (p < .05) were found in our samples (Table 3)
The analysis of the sugar composition of studied honey is shown in Table 4. Significant statistical differences were found for the sugar composition between all the honey groups evaluated (p < .05). The major carbohydrates in the honey were fructose and glucose.
The honey from bees receiving only sugar paste (M2) and from those receiving sugar paste + 3% of crude protein (M4) had a higher fructose concentration than the no supplemented honey (M1) and the sugar paste + vitamins +free amino acids honey (M3).
In terms of glucose content, M2 had a higher mean value (31.41g/100g) than the other honey (p < .05), the lowest mean value corresponding to M4 (29.71g/100g). Honey with a glucose content lower than 30% shows a slow granulation phenomenon over time (Manikis & Thrasivoulou, 2001) so that M2 would show quicker granulation in our study.
Honey from bees receiving sugar paste + vitamins+free amino acids (M3) had the lowest mean sucrose content (2.24 g/100 g), a level significantly different from all the other honey. All the honey samples analyzed in this study did not exceed the limit of 5g/ 100g established by the Council of the European Union. The highest maltose concentrations (1.91%) were found in honey from bees receiving sugar paste alone.
The time honey will take to granulate depends on the Fructose/ Glucose (F/G) and the Glucose/Water content (G/W) ratios. Honey with glucose to water ratio of 1.7 or less are considered nongranulating, while honey with ratios of 2.1 or more predicts rapid granulation (Dobre et al., 2012;Kaakeh & Gadelhak, 2005). Similarly, a glucose-water to fructose ratio higher than 0.50 predicted rapid granulation and a ratio lower than 0.20 predicted slow granulation.
The F/G ratio, therefore, is an important parameter to explain honey granulation because glucose is less water-soluble than fructose, and so induces a tendency to granulation .  Venir, Spaziani, and Maltini (2010), honey fructose to glucose ratio of 1.14 indicates a tendency to granulate more rapidly than honey with a ratio significantly above 1.58. As observed in previous studies that values above 1.3 had a slower granulation tendency (Dobre, Escuredo, Rodriguez-Flores, & Seijo, 2014). The F/G ratio of around 1.2 found in our samples was within the range reported in other studies (Bentabol, García, Galdón, Rodríguez, & Romero, 2011;Dobre et al., 2012).

TA B L E 3 Mineral composition (mg/l) (average ± SD) of the different types of honeys
In terms of the glucose to water ratio M1 (2.36), M2 (2.38), M3 (2.29), and elsewhere M4 (2.21) some researchers have indicated that the G/W ratio can be a better indicator for the prediction of honey granulation (Dobre et al., 2012;Manikis & Thrasivoulou, 2001).
According to Dobre et al., honey granulation is slow when the G/W ratio is less than 1.7 and is complete and rapid when the ratio is greater than 2.
All the samples had fructose to glucose proportions greater than 1 (Table 4) regardless of the type of supplementation administrated, which indicates a greater tendency to granulate in these honeys. In terms of the (glucose-water) to fructose ratio, M1, M2, M3, and M4 indicated a greater tendency to granulation with ratios of 0.47, 0.45, 0.45, and 0.40, respectively. When granulation is incomplete, the crystalline layer is overlaid by a layer of liquid honey with a water content that is higher than that in the original honey. This creates a favorable environment for yeast growth and may lead to fermentation (Escuredo et al., 2013;Tornuk et al., 2013).

| Sensory evaluation
In the present study, all honey analyzed were found to have a similar sensory mean value for "Overall odor intensity," although there was a tendency for this attribute to be higher in the control honey and M3 showed the lowest values (Table 5b). By contrast, there were no significant differences in the floral attribute between honey, although the highest value was observed in the control, honey. Also, all the honey was found to be characterized by fruity, warm, and vegetal attributes.
The control honey was distinguished by its aromatic attribute, while this attribute was not perceived in honey from the honey involving supplementation. All honey samples were characterized by the complete absence of the animal attribute. No significant statistical differences in the basic taste were found (p < .05) (Table 5a). There were no significant ness attribute was not perceived in any honey, and neither were chemical and animal attributes. In general, the honey from the control group was sweeter than the honey from supplemented groups. Similarly, the control honey scored better for color, the intensity of odor, and the attributes of smell (Table 5a and 5b). This result agrees with the result of a study of Turkish honey (Guler et al., 2007)  Finally, no significant differences (p < .05) were found for texture attributes (adhesiveness and granularity) for samples M1, M3, and M4 regarding the type of supplementation.

| CON CLUS IONS
The results of the present study indicated that, regardless of any type of supplementation, all the honey obtained were in accordance with the international legislation in terms of physicochemical properties.
An analysis of the sensory characteristics showed that the control sample only presented significant differences for color and chemical odor attributes compared with honey from bees receiving supplementation.
Honey samples from bees receiving sugar paste + 3% proteins were classified first for the odor attributes fruity and warm. They also had higher values for aroma attributes like the intensity of aroma, warm, aromatic, and vegetal. The result reported in this study will be found useful by apiarists to help them understand the impact of supplementation of honeybee diets with sugar and proteins.

CO N FLI C T O F I NTE R E S T
There is no conflict of interest to declare.

E TH I C A L A PPROVA L
The article does not contain any study with animals or human subjects.