Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Physico-Chemical, Enzymatic, Mineral and Colour Characterization of Three Different Varieties of Honeys from Kashmir Valley of India with a Multivariate Approach

Physico-Chemical, Enzymatic, Mineral and Colour Characterization of Three Different Varieties of... Pol. J. Food Nutr. Sci., 2015, Vol. 65, No. 2, pp. 101­108 DOI: 10.1515/pjfns-2015-0022 http://journal.pan.olsztyn.pl Original article Section: Food Quality and Functionality Physico-Chemical, Enzymatic, Mineral and Colour Characterization of Three Different Varieties of Honeys from Kashmir Valley of India with a Multivariate Approach Gulzar Ahmad Nayik*, Vikas Nanda Department of Food Engineering and Technology, Sant Longowal Institute of Engineering and Technology, Longowal 148106 (Punjab) India Key words: Kashmir, honey, acacia, enzymatic, pine honeydew, multifloral The present study was undertaken to determine the physico-chemical properties (moisture content, reducing sugars, proline content, electrical conductivity, ash content, pH, titrable acidity, HMF, water activity, total soluble solids and total solids), enzymatic activity (diastase and invertase), colour characteristics (mmpfund, ABS450 and CIE L*a* b*) and mineral content (Cu, Mn, Fe, Zn, Pb and Cd) of three different varieties of honeys from Kashmir valley of India (acacia honey, pine honeydew and multifloral honey). Of the honey samples analysed, only pine honeydew were grouped in dark category of honey (L*<50) while acacia honey and multifloral honey were confirmed as light coloured honeys (L*>50) and possessed both red and yellow components. The concentrations of mineral content were found highest in pine honeydew followed by multi floral and least in acacia honey. All the physico-chemical properties and enzymatic activity indicated that all the three analysed varieties of honey met the criteria set by the International Honey Commission and revised codex standards for honey. The source of honey had a significant effect (p<0.05) on physico-chemical characteristics, enzymatic activity, mineral content and colour properties. Strong and positive correlations exhibited among minerals, colour (mmpfund) and L* value indicated that dark coloured honey contained high mineral content. Multivariate analysis proved to be an effective tool in classifying the three varieties of honey based on physico-chemical characteristics, enzymatic activity, mineral content and colour properties. INTRODUCTION Honey, the wonderfully rich golden liquid is the miraculous product of honey bees and a naturally delicious alternative to sugar. Chemically, honey is mostly dominated by sugars in the form of fructose and glucose (70­80%), water (10­20%) and other minor constituents such as organic acids (gluconic acid, acetic acid), mineral salts (potassium, calcium, sodium, phosphorus etc.), vitamins (ascorbic acid, niacin), proteins, enzymes (invertase, glucose oxidase, catalase, phosphatases), volatile chemicals, phenolic acids and flavonoids [Bouseta et al., 1996; Terrab et al., 2001; Gheldof et al.,2002; Blasa et al., 2006; Ouchemoukh et al., 2007; Nayik & Nanda, 2015]. The composition and physico-chemical properties of honey are wholly dependent on the plant species visited by the honeybees as well as the processing, storage, regional, and climatic conditions [Saxena et al., 2010]. The colour, which serves as an indicator of floral origin ranges from light to dark often reddish or yellowish [Soria et al., 2004; Terrab et al., 2004]. It is well known fact that there is a strong correlation among colour, antioxidant activity, electrical conductivity and ash content [Marghitas et al., 2009]. Generally dark coloured honey has higher phenolic content and consequently * Corresponding Author: Cell: +91­9478153553; Fax: 01672­280057; E-mail: gulzarnaik@gmail.com (Gulzar Ahmad Nayik) higher antioxidant activity as compared to honey with light colour [Berreta et al., 2005]. Multivariate analysis viz. principal component analysis (PCA) and linear discriminate analysis (LDA) have been extensively used to classify honey based on physico-chemical data, mineral analysis and colour characteristics. Conti et al. [2007] classified three Italian honey samples (acacia, multifloral, honeydew) on the basis of their physico-chemical data and mineral content by applying multivariate statistical analysis. Chudzinska & Baralkiewicz, [2010] analysed fifty-five Polish honey samples to rationalize and interrupt the analytical data for the determination of thirteen metallic and nonmetallic elements by using PCA tool. Similar study was reported by Devillers et al. [2002] who examined French honeys. Rearing of honey bees for production of honey (apiculture) is one of the prevalent agricultural activities done in India. As per the data published by Press Information Bureau [2012­2013], there are more than 1600 honey producing units in Jammu and Kashmir where honey production is near about 2000 metric tons. The physicochemical characterisation of different honeys from different parts of the world has been carried out extensively [Azeredo et al., 2003; Finola et al., 2007; Ouchemoukh et al., 2007]. Honeys produced in India from various unifloral sources as well as from commercial honeys were already studied by various authors [Ahmed et al., 2007; Nanda et al., 2009; Saxena et al., 2010]. After successful studies of characterisation of quality parameters of Eucalyp- © Copyright by Institute of Animal Reproduction and Food Research of the Polish Academy of Sciences © 2015 Author(s). This is an open access article licensed under the Creative Commons Attribution-NonCommercial-NoDerivs License (http://creativecommons.org/licenses/by-nc-nd/3.0/). tus, Brassica, Helianthus, L. chinensis, Z. mauritiana, C. sinensis and P. persica [Nanda et al., 2003, 2006, 2009], the present study was undertaken, not only to study the physico-chemical parameters, mineral content, enzymatic analysis and colour characteristics from three different varieties from Kashmir valley of India viz. acacia honey, pine honeydew and multifloral honey but also to perform pattern recognition methods by using PCA and LDA to classify honey varieties which has never been performed till now. MATERIALS AND METHODS Honey sample The present study was carried out using 30 samples (10 samples from each variety) of three different raw and fresh honey samples (acacia honey, pine honeydew and multifloral honey) collected from bee keepers during September 2012 to May 2014 from different regions (Pulwama, Srinagar, and Budgam) packed and sealed in glass bottles of Kashmir valley of India and stored at 4oC. The honey samples were kept at ambient temperature 25±1oC overnight before the analyses were performed. The origins of each honey sample were confirmed by microscopic pollen analysis (Melissopalynology). Honey samples were classified according to their botanical origin using the method described by Louveaux et al. [1978]. The following terms were used for frequency classes: predominant pollen (>45% of pollen grains counted), secondary pollen (16­45%), important minor pollen (3­15%) and minor pollen (<3%). Physico-chemical analysis The samples of honey were analysed according to the methods established by AOAC [2012] and International Honey Commission [2009] for moisture content, ash content, electrical conductivity, diastase and invertase activity, reducing sugars, proline content, titrable acidity, pH, total solids and total soluble solids, hydroxymethylfurfural content and water activity. The colour characteristics of the honeys were assessed according to the CIE L*a*b* method described by Saxena et al. [2010]. Colour (mmpfund) was determined by the method described by White [1984] while colour intensity (ABS450) was done by the method adopted by Beretta et al. [2005]. Mineral elements Cu, Mn, Fe, Zn, Pb and Cd were determined by using air acetylene flame atomic absorption spectrometer (AAS4141). The response from the equipment was periodically checked with known standards. Hollow cathode pump and air-acetylene flame were used for all the samples from three varieties. Calibration curves were constructed for each element by using appropriated standard solutions. Statistical analysis The results were expressed in triplicates, mean, standard deviation and correlation was obtained by using Microsoft Excel 2007. The significant differences were obtained by a one-way analysis of variance (ANOVA) followed by Duncan's multiple range test (DMRT) (p<0.05). Principal component analysis and linear discriminate analysis were performed to classify honey samples from different botanical sources by using XLSTAT. RESULTS AND DISCUSSION The pollen spectra of honey samples studied have been briefly described and the percentages are related to pollen of nectar producing plants. Acacia honey contained 54­60% pollen of R. pseudoacacia sp. The honeydew element/pollen grain (HDE/P) ratio was in a range of 2.79­3.01 in pine honeydew (Pinus wallichiana), which was in good agreement with Louveaux et al. [1978]. The microscopic analysis revealed some fungal spores in pine honeydew which is in good agreement with those found in Greek pine honeys [Karabagias et al., 2014]. Multifloral honey contained 2­5% pollens of Plectranthus rugosus, other pollens found were those from Prunus sp., Brassica sp., Thyme sp. and Ailanthus sp. The mean (±S.D) results obtained from the physico-chemical, enzymatic and colour characteristics of honey samples are presented in Table 1 while mineral content is shown in Table 2. Moisture content was reported as an important parameter and honey with lower moisture content showed longer shelf life [Fredes & Montenegro, 2006]. In this study all the samples from three varieties showed the level of moisture content (18.2 to 19.11) which was lower than the limit prescribed by Codex standard for honey [Codex Almentarius, 2001] and International Honey Commission [2009] and the results were significantly different (p<0.05) for each honey obtained from different sources. Our results were in agreement with previous studies [Yilmaz & Kufrevioglu, 2000; Pan & Ji, 1998; Duman et al., 2008]. The moisture content of honey was affected by various factors such as harvesting time, climatic factors and maturity period [Nanda et al., 2003; Finola et al., 2007]. The oBrix values ranged from 79.16 to 80.03 while the refractive index of the samples ranged from 1.4889 to 1.4908. With the increase in the solid content, there was increase in refractive index. The total solid content of the samples was in a range of 80.89% to 81.8% which were quite similar to the results (77.8­80.4%) reported by Saxena et al. [2010]. The pH values of all the honey varieties ranged from 3.52 to 3.78 which confirmed that all the honey varieties were acidic in nature. The pH values were consistent with the results reported by various authors [Azeredo et al., 2003; Ouchemoukh et al., 2007; Saxena et al., 2010; Kamboj et al., 2013]. Results indicated maximum acidity in the form of formic acid in pine honeydew variety (0.39%) followed by acacia honey and multifloral floral varieties in decreasing order with 0.17% and 0.14%, respectively. The titrable acidity and pH of all the samples analysed was found within the corresponding limits as described by Codex Almentarious Commission [2001]. A high strong positive correlation was found between pH and titrable acidity (r=0.99).The percentage ash content is an indicator of the mineral content present in a given sample. The maximum ash content was found in pine honeydew (0.39%) followed by acacia honey (0.06%) whereas ash content on multifloral floral honey was at 0.05%. Similar results were reported by Kamboj et al. [2013] (0.140­0.211%) and Nanda et al. [2009] (0.13 to 0.35%). As per previous studies, electrical conductivity (EC) was used to distinguish honeydew honeys from blossom honey [Mateo & Bosch-Reig, 1998]. The electrical conductivity of all analysed honey samples ranged from 0.25 to 0.79 mS/cm. As per G.A. Nayik & V . Nanda TABLE 1. Physico-chemical parameters of honey samples with Duncan's multiple range test results. Parameter Ash (%) Electrical Conductivity (mS/cm) Titrable acidity (%) pH Moisture (%) Total soluble solids (°B) Total solids (%) HMF(mg/kg) Water activity ABS450 (mAU) Color (mmpfund) L* a* b* Diastase Number (DN) Invertase Number (IN) Proline (mg/kg) Reducing Sugars (%) Acacia honey (n=10) 0.06±0.03 Pine honeydew (n=10) 0.35±0.02 Multifloral honey (n=10) 0.05±0.01b 0.25±0.04b 0.14±0.02a 3.52±0.02b 19.11±0.20a 79.16±0.25b 80.89±0.55a 22.64±0.30a 0.566±0.00a 148.31±3.07c 39.16±1.45b 54.64±0.81b 3.49±0.29b 9.39±0.30b 14.93±2.10b 12.58±1.94b 168.05±3.66b 72.81±2.97c 0.26±0.05b 0.17±0.02a 3.55±0.0.0ab 18.6±0.08b 79.7±0.10a 81.40±0.08a 5.49±0.07b 0.507±0.00c 188.05±2.30b 43.32±1.36b 61.43±1.60a 4.94±0.08a 12.32±0.07a 15.51±1.40b 9.40±1.27b 292.02±2.65b 66.24±3.05b 0.79±0.06a 0.39±0.02b 3.78±0.11a 18.2±0.12b 80.03±0.15a 81.8±0.12a 6.79±0.10b 0.523±0.00b 525.39±2.98a 143.04±3.10a 15.49±0.3c 2.25±0.08c 5.28±0.10c 25.99±2.79a 15.83±2.82a 570.95±2.15a 60.6±3.94a Results are expressed as mean values±standard deviations. Means in a row with same superscripts are not significantly different (P<0.05), n=number of samples. Codex Alimentarius [2001] electrical conductivity value for the nectar honey must be less than 0.80 mS/cm while according to the EU, minimum electrical conductivity for pine honey should be near 0.80 mS/cm. The results of electrical conductivity were consistent with reported results of Escuredo et al. [2012] in blossom and honeydew honeys. Positive and strong correlation (r=0.90) between ash content and electrical conductivity (Table 3) indicated that higher ash content resulted in higher electrical conductivity. Tosi et al. [2002] reported that hydroxymethylfurfural (HMF) as a quality parameter to check the honey freshness and high temperature processing. All the three honey varieties showed an HMF level lower than the limit (40 mg/kg), recommended by the Codex Alimentarius [Codex Alimentarius, 2001] and statistical analysis revealed a significant difference (p<0.05) among the varieties. The maximum content of HMF was found in multifloral source (22.64 mg/kg) followed by pine honeydew (6.79 mg/kg) and minimum in acacia honey (5.49 mg/kg) as shown in Table 1. These results suggested that all the samples from three varieties of honey were raw and unprocessed. The HMF values of the analysed samples were consistent with the values reported by Yilmaz & Kufrevioglu [2000] and Duman et al. [2008]. The water activity is an important factor for stability of food by preventing microbial growth. The water activity of the honey samples was in the range of 0.507 to 0.566 and these values were statistically different from each other (p<0.05) (Table 1). Our results were quite similar to those of Greek honeys and Indian honeys [Lazaridou et al., 2004; Saxena et al., 2010]. The acceptability of honey by the consumer and its market price mainly depends on colour of honey. According to the classification of honey based on Pfund mm values set by USDA [1985] colour of honey ranges from water white (<8 mm) to dark amber (>114 mm). Based on this classification, the pine honeydew variety (143.04 mm) was classified as dark amber while multifloral honey (39.16 mm) and acacia honey (43.32 mm) as extra light amber. It was reported that colour of honey was affected by chemical composition, primarily due to the presence of pigments such as chlorophylls, carotenoids, flavonoids and polyphenols [Lazaridou et al., 2004; Finola et al., 2007; Juszczak et al., 2009]. The honeys with light colour usually contained low ash content and low antioxidant activity while those with dark colour showed reverse of these values [Gheldof et al., 2002; Marghitas et al., 2009]. There was high positive correlation found between colour and ash content (r=0.98) and colour values of all samples from different sources were significantly different (p<0.05). Colour intensity (ABS450) is an index for confirming the presence of pigments such as carotenoids and some flavonoids. ABS450 values of the tested samples ranged from 148.31 to 525.39 mAU and were statistically different (P<0.05) (Table 1). All obtained values were in the range as reported by Beretta et al. [2005] in different commercial honey varieties of Italy. TABLE 2. Statistical analysis of mineral content (mg/kg) of honey from different sources. Sources Pine honeydew Acacia honey Multifloral honey Cu 0.12±0.01 Mn 1.08±0.01 Fe 2.11±0.01 Zn 1.04±0.02 Pb 0.21±0.02 Cd 0.19±0.02a 0.07±0.02b 0.08±0.03b 0.12±0.01b 0.38±0.02a 0.95±0.01b 1.01±0.02a 1.42±0.03b 1.26±0.03b 0.06±0.03b 0.07±0.03b 0.04±0.04b 0.09±0.02b Results are expressed as mean values±standard deviations. Means in a column with same superscripts are not significantly different (P<0.05). TABLE 3. Correlation among some physicochemical parameters and minerals (Pearson correlation coefficients, p<0.05). pH pH Ash Color L* a * * Ash ­ 1.00 0.90 -0.89 -0.76 -0.81 0.78 0.77 0.90 0.91 0.74 0.70 0.90 Color ­ ­ 1.00 -0.98 -0.81 -0.87 0.51 0.77 0.99 1.00 0.89 0.87 0.98 L* ­ ­ ­ 1.00 0.87 0.92 -0.80 -0.82 -0.95 -0.99 -0.91 -0.87 -0.97 a* ­ ­ ­ b* ­ ­ ­ ­ Cu ­ Mn ­ ­ Fe ­ ­ ­ ­ ­ ­ ­ ­ 1.00 0.98 0.84 0.85 0.96 Zn ­ ­ ­ ­ ­ ­ ­ ­ ­ 1.00 0.89 0.86 0.98 Pb ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ 1.00 0.92 0.88 Cd ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ 1.00 0.85 EC ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ 1.00 0.99 0.66 -0.65 -0.51 -0.54 0.66 0.59 0.64 0.64 0.64 0.66 0.66 ­ ­ ­ ­ 1.00 -0.08 -0.62 -0.47 -0.25 -0.32 0.88 ­ ­ ­ ­ ­ 1.00 0.69 0.80 0.70 0.58 0.79 1.00 0.95 -0.04 -0.89 -0.72 -0.83 -0.85 -0.75 -0.81 ­ 1.00 0.08 -0.84 -0.81 -0.90 -0.90 -0.82 -0.88 Cu Mn Fe Zn Pb Cd a a EC EC= Electrical Conductivity The L*, a*, b* values obtained for three different varieties of honeys are shown in Table 1. According to L* values, Gonzalez-Miret et al. [2005] classified honey into two groups: light honey with L*>50 and the dark honey with L*<50. On the basis of this classification, the pine honeydew was placed in group of dark category while acacia and multifloral honey in light category. It is evident from the Table 1 that acacia honey (61.43) and multifloral floral (54.64) honeys showed higher L* values than pine honeydew (15.49). It established that acacia honey and multifloral floral were of light colour while pine honeydew was of dark colour. In all the honey samples from three varieties of honey, a* values were ranged from 2.25 to 4.94 while b* values from 5.28 to 12.37. These values were helpful in concluding that all the honey samples possessed both red and yellow components. Similar results were reported for Slovak honeys [Kasperova et al., 2012]. The diastase activity, indicator of high temperature exposure of the analysed samples ranged from 14.93 DN (multifloral honey) to 25.99 DN (pine honeydew) (Table 1). All the honey samples showed the values within the Codex Standard i.e. more than 8 DN. Similar results were published by Meda et al. [2005] for Burkina Fasan honey, and by Saric et al. [2008] for Croatian honeys. The presence of high diastase number in the analysed samples might be due to moderate climate of Kashmir valley. According to the Codex standard, invertase number, indicator of honey freshness, should be more than 4. The invertase number ranged from 9.40 IN (acacia honey) to 15.83 IN (pine honeydew). Thus both IN were >4 and DN>8 which determined that all honey samples were fresh and unprocessed. Our results were consistent with reported results of Hasan et al. [2013] for Iraqi honeys and Dinkov et al. [2014] for acacia, sunflower and tilia honeys. All the honey samples had a high proline content ranging from 268.05 to 570.95 mg/kg (Table 1) thus could be considered as ripened and unadulterated ones. Similar results were published for Algerian honeys, Malaysian honeys and Indian honeys [Ouchemoukh et al., 2007; Saxena et al., 2010]. The total reducing sugars value in the analysed samples ranged from 60.6 to 72.81% (Table 1), which is in agreement with the standards proposed by EU Directive 2001/110, which should be more than 60% for blossom honeys and at least 45% for honeydew honey. Similar results were reported by Khalil et al. [2012] for Algerian blossom honeys. Mineral analysis The minerals found in the analysed honey samples are expressed in Table 2. Statistically significant differences (p<0.05) were observed among all analysed samples. Copper (Cu) content was found highest in multifloral honey (0.38 mg/kg) and lowest in both acacia honey and pine hon- G.A. Nayik & V . Nanda TABLE 4. Principal component analysis Total variance explained PC 1 2 3 Initial Eigen values 15.77 3.90 1.18 % of variance 65.72 16.26 4.93 Cumulative % 65.72 81.98 86.92 eydew (0.12 mg/kg). The analysed honey varieties showed low concentration of Cu as compared to honey studied by Conti et al. [2007], Nanda et al. [2009] and Kamboj et al. [2013], but similar concentration of copper was found in Brazilian honey [Liberato et al., [2013]. The manganese (Mn) concentration of all analysed samples ranged from 0.95 mg/kg to 1.08 mg/kg with highest concentration in pine honeydew (1.08 mg/kg). Similar findings were reported by Nanda et al. [2009]. Among the other minerals, iron (Fe) content was found in higher concentration in all samples with highest value in pine honeydew (2.11 mg/kg) and lowest in multifloral honey (1.26 mg/kg). The concentrations of zinc (Zn), lead (Pb) and cadmium (Cd) were found very low (<1mg/kg) except zinc in pine honeydew (1.04 mg/kg) (Table 2). The results for Mn, Fe, Zn, Pb and Cd were in agreement with those reported by Liberato et al. [2013] for Brazilian honey of Apis mellifera from different floral origins. The mineral content of the honey depends on geographical area, climatic conditions and soil type of the floral source. It was reported that soil in Kashmir valley of India did not contain high amount of minerals like Cu, Fe, Zn and Co [Bhat et al., 2011; Yatoo et al., 2011] which might be the reason of low level of mineral content in the studied samples of honey varieties. Strong and positive correlations were found between all minerals and colour (pfund) while strong negative correlations with L* value, which indicated that honey with dark colour contained high amount of ions (Table 3). All minerals also showed strong positive correlation with electrical conductivity and ash, which demonstrated that higher mineral content, resulted in higher ash content and electrical conductivity. A positive correlation was also found among all minerals, electrical conductivity and pH (Table 3), which showed that both pH and conductivity were dependent on the amount of ions in honey. Similar correlations were observed by Acquarone et al. [2006] between total mineral content and electrical conductivity. Multivariate analysis In our study, the PCA was applied to achieve a reduction of original data matrix while retaining the maximum amount of variability present in data. The factor loading obtained for the first three components (PCs) and the percentage of variance along with cumulative variance is shown in Table 4. The first three principal components accounted for 86.92% of the variance in the honey samples analysed. The first, second and third principal components (PC1, PC2, and PC3) explained 65.72%, 16.26% and 4.93% of the variance, respectively (Table 4). According to loading matrix (Table 5), it was observed that 65.72% variability explained by PC1 was positively correlated with variables viz. colour intensity, colour (mmpfund), iron, zinc, proline, electrical conductivity and negatively with colour L*, a* and b* values (Figure 1) which could be specified as indicators of colour and minerals. Thus such variables could be used to distinguish pine honeydew from acacia honey. Figure 1 also illustrated that all acacia samples located on the left side of PC1 were linked to L*, a* and b* values while all pine honeydew samples positioned on its upright were linked to minerals and electrical conductivity. The second component (PC2=16.26%) was positively corre- TABLE 5. Principal component analysis. Loading of the first three components. Principal components Factor loading pH Moisture Total solids Color Intensity Total soluble solids HMF Ash Titrable acidity Color (mmpfund) Water activity Color (L*) Color (a*) Color (b ) 1 0.660 -0.657 0.657 0.994 0.333 -0.479 0.903 0.923 0.993 -0.279 -0.978 -0.746 -0.872 -0.524 0.746 0.985 0.990 0.901 0.880 0.934 0.674 0.994 0.658 0.978 2 -0.014 0.481 -0.481 -0.027 -0.317 0.840 0.042 -0.082 0.034 0.912 -0.159 -0.577 -0.440 0.782 0.490 -0.102 0.078 0.258 0.130 0.000 0.435 -0.002 -0.307 0.072 3 0.424 -0.163 0.163 -0.048 0.700 0.111 -0.137 0.092 -0.043 0.192 0.025 0.026 -0.008 0.104 -0.245 -0.063 -0.075 0.208 0.250 -0.080 -0.069 -0.053 -0.420 -0.036 Cu Mn Fe Zn Pb Cd Diastase number Invertase number Proline Reducing sugar Electrical conductivity lated with water activity and HMF, which could be interpreted as an indicator of the maturity of honey. Such variables could be used to distinguish multifloral honey from acacia honey and pine honeydew. The third component with less contribution (PC3=4.93%) was positively correlated with TSS and pH but negatively with reducing sugar. Similar results were reported by Kadar et al. [2010] for colour (mmpfund), CIE L* a* b* and conductivity, while Saric et al. [2008] published simi- PC 2: 16.26% - 0.2 -2 - 0.4 - 0.6 -4 0 PC 1: 65.72% 5 FIGURE 1. Projections of the variables on the factor plane for the three honeys (Botanical origins: A: Acacia, PH: Pine Honeydew, MF: Multifloral). TABLE 6. Classification result of LDA of variables in three types of honeys. From/to Acacia honey Pine honeydew Multifloral honey Total Acacia honey 10 0 0 10 Pine honeydew 0 10 0 10 Multifloral honey 0 0 10 10 Total 10 10 10 30 % Correct 100.00 100.00 100.00 100.00 lar results on Croatian honey varieties from three harvesting seasons in which he concluded that in all seasons PC1 was mainly dominated by conductivity, proline and diastase number. Similar results for HMF were obtained by Isopescu et al. [2014] for Romanian honeys. As shown in Figure 1, a natural separation between honeys of different botanical origin was obtained. All the three analysed honey samples from different botanical origins were correctly classified (100%) by using LDA (Table 6). CONCLUSION The analysed three honey varieties from Kashmir valley of India revealed that floral origin significantly affects all the physico-chemical parameters, enzymatic properties, mineral content and colour characteristics except total solids. The concentration of minerals found in all three honey varieties was low as compared to other varieties of honey of Indian origin. Application of multivariate techniques confirmed the validity of physico-chemical analysis as a tool for classifi- cation and characterization of honey obtained from different botanical sources. PCA revealed 86.92% of the variance with the first three principal components with minerals, colour and electrical conductivity dominating variables. LDA proved to be an effective tool which classified the all honey samples 100% correctly. ACKNOWLEDGEMENTS The first author is very much thankful UGC New Delhi India for providing financial assistance in the form of MANF2013­14 and also to local beekeepers of Kashmir valley of India for providing raw honey samples. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Polish Journal of Food and Nutrition Sciences de Gruyter

Physico-Chemical, Enzymatic, Mineral and Colour Characterization of Three Different Varieties of Honeys from Kashmir Valley of India with a Multivariate Approach

Loading next page...
 
/lp/de-gruyter/physico-chemical-enzymatic-mineral-and-colour-characterization-of-AG9iIXmy0c

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
de Gruyter
Copyright
Copyright © 2015 by the
ISSN
1230-0322
eISSN
2083-6007
DOI
10.1515/pjfns-2015-0022
Publisher site
See Article on Publisher Site

Abstract

Pol. J. Food Nutr. Sci., 2015, Vol. 65, No. 2, pp. 101­108 DOI: 10.1515/pjfns-2015-0022 http://journal.pan.olsztyn.pl Original article Section: Food Quality and Functionality Physico-Chemical, Enzymatic, Mineral and Colour Characterization of Three Different Varieties of Honeys from Kashmir Valley of India with a Multivariate Approach Gulzar Ahmad Nayik*, Vikas Nanda Department of Food Engineering and Technology, Sant Longowal Institute of Engineering and Technology, Longowal 148106 (Punjab) India Key words: Kashmir, honey, acacia, enzymatic, pine honeydew, multifloral The present study was undertaken to determine the physico-chemical properties (moisture content, reducing sugars, proline content, electrical conductivity, ash content, pH, titrable acidity, HMF, water activity, total soluble solids and total solids), enzymatic activity (diastase and invertase), colour characteristics (mmpfund, ABS450 and CIE L*a* b*) and mineral content (Cu, Mn, Fe, Zn, Pb and Cd) of three different varieties of honeys from Kashmir valley of India (acacia honey, pine honeydew and multifloral honey). Of the honey samples analysed, only pine honeydew were grouped in dark category of honey (L*<50) while acacia honey and multifloral honey were confirmed as light coloured honeys (L*>50) and possessed both red and yellow components. The concentrations of mineral content were found highest in pine honeydew followed by multi floral and least in acacia honey. All the physico-chemical properties and enzymatic activity indicated that all the three analysed varieties of honey met the criteria set by the International Honey Commission and revised codex standards for honey. The source of honey had a significant effect (p<0.05) on physico-chemical characteristics, enzymatic activity, mineral content and colour properties. Strong and positive correlations exhibited among minerals, colour (mmpfund) and L* value indicated that dark coloured honey contained high mineral content. Multivariate analysis proved to be an effective tool in classifying the three varieties of honey based on physico-chemical characteristics, enzymatic activity, mineral content and colour properties. INTRODUCTION Honey, the wonderfully rich golden liquid is the miraculous product of honey bees and a naturally delicious alternative to sugar. Chemically, honey is mostly dominated by sugars in the form of fructose and glucose (70­80%), water (10­20%) and other minor constituents such as organic acids (gluconic acid, acetic acid), mineral salts (potassium, calcium, sodium, phosphorus etc.), vitamins (ascorbic acid, niacin), proteins, enzymes (invertase, glucose oxidase, catalase, phosphatases), volatile chemicals, phenolic acids and flavonoids [Bouseta et al., 1996; Terrab et al., 2001; Gheldof et al.,2002; Blasa et al., 2006; Ouchemoukh et al., 2007; Nayik & Nanda, 2015]. The composition and physico-chemical properties of honey are wholly dependent on the plant species visited by the honeybees as well as the processing, storage, regional, and climatic conditions [Saxena et al., 2010]. The colour, which serves as an indicator of floral origin ranges from light to dark often reddish or yellowish [Soria et al., 2004; Terrab et al., 2004]. It is well known fact that there is a strong correlation among colour, antioxidant activity, electrical conductivity and ash content [Marghitas et al., 2009]. Generally dark coloured honey has higher phenolic content and consequently * Corresponding Author: Cell: +91­9478153553; Fax: 01672­280057; E-mail: gulzarnaik@gmail.com (Gulzar Ahmad Nayik) higher antioxidant activity as compared to honey with light colour [Berreta et al., 2005]. Multivariate analysis viz. principal component analysis (PCA) and linear discriminate analysis (LDA) have been extensively used to classify honey based on physico-chemical data, mineral analysis and colour characteristics. Conti et al. [2007] classified three Italian honey samples (acacia, multifloral, honeydew) on the basis of their physico-chemical data and mineral content by applying multivariate statistical analysis. Chudzinska & Baralkiewicz, [2010] analysed fifty-five Polish honey samples to rationalize and interrupt the analytical data for the determination of thirteen metallic and nonmetallic elements by using PCA tool. Similar study was reported by Devillers et al. [2002] who examined French honeys. Rearing of honey bees for production of honey (apiculture) is one of the prevalent agricultural activities done in India. As per the data published by Press Information Bureau [2012­2013], there are more than 1600 honey producing units in Jammu and Kashmir where honey production is near about 2000 metric tons. The physicochemical characterisation of different honeys from different parts of the world has been carried out extensively [Azeredo et al., 2003; Finola et al., 2007; Ouchemoukh et al., 2007]. Honeys produced in India from various unifloral sources as well as from commercial honeys were already studied by various authors [Ahmed et al., 2007; Nanda et al., 2009; Saxena et al., 2010]. After successful studies of characterisation of quality parameters of Eucalyp- © Copyright by Institute of Animal Reproduction and Food Research of the Polish Academy of Sciences © 2015 Author(s). This is an open access article licensed under the Creative Commons Attribution-NonCommercial-NoDerivs License (http://creativecommons.org/licenses/by-nc-nd/3.0/). tus, Brassica, Helianthus, L. chinensis, Z. mauritiana, C. sinensis and P. persica [Nanda et al., 2003, 2006, 2009], the present study was undertaken, not only to study the physico-chemical parameters, mineral content, enzymatic analysis and colour characteristics from three different varieties from Kashmir valley of India viz. acacia honey, pine honeydew and multifloral honey but also to perform pattern recognition methods by using PCA and LDA to classify honey varieties which has never been performed till now. MATERIALS AND METHODS Honey sample The present study was carried out using 30 samples (10 samples from each variety) of three different raw and fresh honey samples (acacia honey, pine honeydew and multifloral honey) collected from bee keepers during September 2012 to May 2014 from different regions (Pulwama, Srinagar, and Budgam) packed and sealed in glass bottles of Kashmir valley of India and stored at 4oC. The honey samples were kept at ambient temperature 25±1oC overnight before the analyses were performed. The origins of each honey sample were confirmed by microscopic pollen analysis (Melissopalynology). Honey samples were classified according to their botanical origin using the method described by Louveaux et al. [1978]. The following terms were used for frequency classes: predominant pollen (>45% of pollen grains counted), secondary pollen (16­45%), important minor pollen (3­15%) and minor pollen (<3%). Physico-chemical analysis The samples of honey were analysed according to the methods established by AOAC [2012] and International Honey Commission [2009] for moisture content, ash content, electrical conductivity, diastase and invertase activity, reducing sugars, proline content, titrable acidity, pH, total solids and total soluble solids, hydroxymethylfurfural content and water activity. The colour characteristics of the honeys were assessed according to the CIE L*a*b* method described by Saxena et al. [2010]. Colour (mmpfund) was determined by the method described by White [1984] while colour intensity (ABS450) was done by the method adopted by Beretta et al. [2005]. Mineral elements Cu, Mn, Fe, Zn, Pb and Cd were determined by using air acetylene flame atomic absorption spectrometer (AAS4141). The response from the equipment was periodically checked with known standards. Hollow cathode pump and air-acetylene flame were used for all the samples from three varieties. Calibration curves were constructed for each element by using appropriated standard solutions. Statistical analysis The results were expressed in triplicates, mean, standard deviation and correlation was obtained by using Microsoft Excel 2007. The significant differences were obtained by a one-way analysis of variance (ANOVA) followed by Duncan's multiple range test (DMRT) (p<0.05). Principal component analysis and linear discriminate analysis were performed to classify honey samples from different botanical sources by using XLSTAT. RESULTS AND DISCUSSION The pollen spectra of honey samples studied have been briefly described and the percentages are related to pollen of nectar producing plants. Acacia honey contained 54­60% pollen of R. pseudoacacia sp. The honeydew element/pollen grain (HDE/P) ratio was in a range of 2.79­3.01 in pine honeydew (Pinus wallichiana), which was in good agreement with Louveaux et al. [1978]. The microscopic analysis revealed some fungal spores in pine honeydew which is in good agreement with those found in Greek pine honeys [Karabagias et al., 2014]. Multifloral honey contained 2­5% pollens of Plectranthus rugosus, other pollens found were those from Prunus sp., Brassica sp., Thyme sp. and Ailanthus sp. The mean (±S.D) results obtained from the physico-chemical, enzymatic and colour characteristics of honey samples are presented in Table 1 while mineral content is shown in Table 2. Moisture content was reported as an important parameter and honey with lower moisture content showed longer shelf life [Fredes & Montenegro, 2006]. In this study all the samples from three varieties showed the level of moisture content (18.2 to 19.11) which was lower than the limit prescribed by Codex standard for honey [Codex Almentarius, 2001] and International Honey Commission [2009] and the results were significantly different (p<0.05) for each honey obtained from different sources. Our results were in agreement with previous studies [Yilmaz & Kufrevioglu, 2000; Pan & Ji, 1998; Duman et al., 2008]. The moisture content of honey was affected by various factors such as harvesting time, climatic factors and maturity period [Nanda et al., 2003; Finola et al., 2007]. The oBrix values ranged from 79.16 to 80.03 while the refractive index of the samples ranged from 1.4889 to 1.4908. With the increase in the solid content, there was increase in refractive index. The total solid content of the samples was in a range of 80.89% to 81.8% which were quite similar to the results (77.8­80.4%) reported by Saxena et al. [2010]. The pH values of all the honey varieties ranged from 3.52 to 3.78 which confirmed that all the honey varieties were acidic in nature. The pH values were consistent with the results reported by various authors [Azeredo et al., 2003; Ouchemoukh et al., 2007; Saxena et al., 2010; Kamboj et al., 2013]. Results indicated maximum acidity in the form of formic acid in pine honeydew variety (0.39%) followed by acacia honey and multifloral floral varieties in decreasing order with 0.17% and 0.14%, respectively. The titrable acidity and pH of all the samples analysed was found within the corresponding limits as described by Codex Almentarious Commission [2001]. A high strong positive correlation was found between pH and titrable acidity (r=0.99).The percentage ash content is an indicator of the mineral content present in a given sample. The maximum ash content was found in pine honeydew (0.39%) followed by acacia honey (0.06%) whereas ash content on multifloral floral honey was at 0.05%. Similar results were reported by Kamboj et al. [2013] (0.140­0.211%) and Nanda et al. [2009] (0.13 to 0.35%). As per previous studies, electrical conductivity (EC) was used to distinguish honeydew honeys from blossom honey [Mateo & Bosch-Reig, 1998]. The electrical conductivity of all analysed honey samples ranged from 0.25 to 0.79 mS/cm. As per G.A. Nayik & V . Nanda TABLE 1. Physico-chemical parameters of honey samples with Duncan's multiple range test results. Parameter Ash (%) Electrical Conductivity (mS/cm) Titrable acidity (%) pH Moisture (%) Total soluble solids (°B) Total solids (%) HMF(mg/kg) Water activity ABS450 (mAU) Color (mmpfund) L* a* b* Diastase Number (DN) Invertase Number (IN) Proline (mg/kg) Reducing Sugars (%) Acacia honey (n=10) 0.06±0.03 Pine honeydew (n=10) 0.35±0.02 Multifloral honey (n=10) 0.05±0.01b 0.25±0.04b 0.14±0.02a 3.52±0.02b 19.11±0.20a 79.16±0.25b 80.89±0.55a 22.64±0.30a 0.566±0.00a 148.31±3.07c 39.16±1.45b 54.64±0.81b 3.49±0.29b 9.39±0.30b 14.93±2.10b 12.58±1.94b 168.05±3.66b 72.81±2.97c 0.26±0.05b 0.17±0.02a 3.55±0.0.0ab 18.6±0.08b 79.7±0.10a 81.40±0.08a 5.49±0.07b 0.507±0.00c 188.05±2.30b 43.32±1.36b 61.43±1.60a 4.94±0.08a 12.32±0.07a 15.51±1.40b 9.40±1.27b 292.02±2.65b 66.24±3.05b 0.79±0.06a 0.39±0.02b 3.78±0.11a 18.2±0.12b 80.03±0.15a 81.8±0.12a 6.79±0.10b 0.523±0.00b 525.39±2.98a 143.04±3.10a 15.49±0.3c 2.25±0.08c 5.28±0.10c 25.99±2.79a 15.83±2.82a 570.95±2.15a 60.6±3.94a Results are expressed as mean values±standard deviations. Means in a row with same superscripts are not significantly different (P<0.05), n=number of samples. Codex Alimentarius [2001] electrical conductivity value for the nectar honey must be less than 0.80 mS/cm while according to the EU, minimum electrical conductivity for pine honey should be near 0.80 mS/cm. The results of electrical conductivity were consistent with reported results of Escuredo et al. [2012] in blossom and honeydew honeys. Positive and strong correlation (r=0.90) between ash content and electrical conductivity (Table 3) indicated that higher ash content resulted in higher electrical conductivity. Tosi et al. [2002] reported that hydroxymethylfurfural (HMF) as a quality parameter to check the honey freshness and high temperature processing. All the three honey varieties showed an HMF level lower than the limit (40 mg/kg), recommended by the Codex Alimentarius [Codex Alimentarius, 2001] and statistical analysis revealed a significant difference (p<0.05) among the varieties. The maximum content of HMF was found in multifloral source (22.64 mg/kg) followed by pine honeydew (6.79 mg/kg) and minimum in acacia honey (5.49 mg/kg) as shown in Table 1. These results suggested that all the samples from three varieties of honey were raw and unprocessed. The HMF values of the analysed samples were consistent with the values reported by Yilmaz & Kufrevioglu [2000] and Duman et al. [2008]. The water activity is an important factor for stability of food by preventing microbial growth. The water activity of the honey samples was in the range of 0.507 to 0.566 and these values were statistically different from each other (p<0.05) (Table 1). Our results were quite similar to those of Greek honeys and Indian honeys [Lazaridou et al., 2004; Saxena et al., 2010]. The acceptability of honey by the consumer and its market price mainly depends on colour of honey. According to the classification of honey based on Pfund mm values set by USDA [1985] colour of honey ranges from water white (<8 mm) to dark amber (>114 mm). Based on this classification, the pine honeydew variety (143.04 mm) was classified as dark amber while multifloral honey (39.16 mm) and acacia honey (43.32 mm) as extra light amber. It was reported that colour of honey was affected by chemical composition, primarily due to the presence of pigments such as chlorophylls, carotenoids, flavonoids and polyphenols [Lazaridou et al., 2004; Finola et al., 2007; Juszczak et al., 2009]. The honeys with light colour usually contained low ash content and low antioxidant activity while those with dark colour showed reverse of these values [Gheldof et al., 2002; Marghitas et al., 2009]. There was high positive correlation found between colour and ash content (r=0.98) and colour values of all samples from different sources were significantly different (p<0.05). Colour intensity (ABS450) is an index for confirming the presence of pigments such as carotenoids and some flavonoids. ABS450 values of the tested samples ranged from 148.31 to 525.39 mAU and were statistically different (P<0.05) (Table 1). All obtained values were in the range as reported by Beretta et al. [2005] in different commercial honey varieties of Italy. TABLE 2. Statistical analysis of mineral content (mg/kg) of honey from different sources. Sources Pine honeydew Acacia honey Multifloral honey Cu 0.12±0.01 Mn 1.08±0.01 Fe 2.11±0.01 Zn 1.04±0.02 Pb 0.21±0.02 Cd 0.19±0.02a 0.07±0.02b 0.08±0.03b 0.12±0.01b 0.38±0.02a 0.95±0.01b 1.01±0.02a 1.42±0.03b 1.26±0.03b 0.06±0.03b 0.07±0.03b 0.04±0.04b 0.09±0.02b Results are expressed as mean values±standard deviations. Means in a column with same superscripts are not significantly different (P<0.05). TABLE 3. Correlation among some physicochemical parameters and minerals (Pearson correlation coefficients, p<0.05). pH pH Ash Color L* a * * Ash ­ 1.00 0.90 -0.89 -0.76 -0.81 0.78 0.77 0.90 0.91 0.74 0.70 0.90 Color ­ ­ 1.00 -0.98 -0.81 -0.87 0.51 0.77 0.99 1.00 0.89 0.87 0.98 L* ­ ­ ­ 1.00 0.87 0.92 -0.80 -0.82 -0.95 -0.99 -0.91 -0.87 -0.97 a* ­ ­ ­ b* ­ ­ ­ ­ Cu ­ Mn ­ ­ Fe ­ ­ ­ ­ ­ ­ ­ ­ 1.00 0.98 0.84 0.85 0.96 Zn ­ ­ ­ ­ ­ ­ ­ ­ ­ 1.00 0.89 0.86 0.98 Pb ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ 1.00 0.92 0.88 Cd ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ 1.00 0.85 EC ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ 1.00 0.99 0.66 -0.65 -0.51 -0.54 0.66 0.59 0.64 0.64 0.64 0.66 0.66 ­ ­ ­ ­ 1.00 -0.08 -0.62 -0.47 -0.25 -0.32 0.88 ­ ­ ­ ­ ­ 1.00 0.69 0.80 0.70 0.58 0.79 1.00 0.95 -0.04 -0.89 -0.72 -0.83 -0.85 -0.75 -0.81 ­ 1.00 0.08 -0.84 -0.81 -0.90 -0.90 -0.82 -0.88 Cu Mn Fe Zn Pb Cd a a EC EC= Electrical Conductivity The L*, a*, b* values obtained for three different varieties of honeys are shown in Table 1. According to L* values, Gonzalez-Miret et al. [2005] classified honey into two groups: light honey with L*>50 and the dark honey with L*<50. On the basis of this classification, the pine honeydew was placed in group of dark category while acacia and multifloral honey in light category. It is evident from the Table 1 that acacia honey (61.43) and multifloral floral (54.64) honeys showed higher L* values than pine honeydew (15.49). It established that acacia honey and multifloral floral were of light colour while pine honeydew was of dark colour. In all the honey samples from three varieties of honey, a* values were ranged from 2.25 to 4.94 while b* values from 5.28 to 12.37. These values were helpful in concluding that all the honey samples possessed both red and yellow components. Similar results were reported for Slovak honeys [Kasperova et al., 2012]. The diastase activity, indicator of high temperature exposure of the analysed samples ranged from 14.93 DN (multifloral honey) to 25.99 DN (pine honeydew) (Table 1). All the honey samples showed the values within the Codex Standard i.e. more than 8 DN. Similar results were published by Meda et al. [2005] for Burkina Fasan honey, and by Saric et al. [2008] for Croatian honeys. The presence of high diastase number in the analysed samples might be due to moderate climate of Kashmir valley. According to the Codex standard, invertase number, indicator of honey freshness, should be more than 4. The invertase number ranged from 9.40 IN (acacia honey) to 15.83 IN (pine honeydew). Thus both IN were >4 and DN>8 which determined that all honey samples were fresh and unprocessed. Our results were consistent with reported results of Hasan et al. [2013] for Iraqi honeys and Dinkov et al. [2014] for acacia, sunflower and tilia honeys. All the honey samples had a high proline content ranging from 268.05 to 570.95 mg/kg (Table 1) thus could be considered as ripened and unadulterated ones. Similar results were published for Algerian honeys, Malaysian honeys and Indian honeys [Ouchemoukh et al., 2007; Saxena et al., 2010]. The total reducing sugars value in the analysed samples ranged from 60.6 to 72.81% (Table 1), which is in agreement with the standards proposed by EU Directive 2001/110, which should be more than 60% for blossom honeys and at least 45% for honeydew honey. Similar results were reported by Khalil et al. [2012] for Algerian blossom honeys. Mineral analysis The minerals found in the analysed honey samples are expressed in Table 2. Statistically significant differences (p<0.05) were observed among all analysed samples. Copper (Cu) content was found highest in multifloral honey (0.38 mg/kg) and lowest in both acacia honey and pine hon- G.A. Nayik & V . Nanda TABLE 4. Principal component analysis Total variance explained PC 1 2 3 Initial Eigen values 15.77 3.90 1.18 % of variance 65.72 16.26 4.93 Cumulative % 65.72 81.98 86.92 eydew (0.12 mg/kg). The analysed honey varieties showed low concentration of Cu as compared to honey studied by Conti et al. [2007], Nanda et al. [2009] and Kamboj et al. [2013], but similar concentration of copper was found in Brazilian honey [Liberato et al., [2013]. The manganese (Mn) concentration of all analysed samples ranged from 0.95 mg/kg to 1.08 mg/kg with highest concentration in pine honeydew (1.08 mg/kg). Similar findings were reported by Nanda et al. [2009]. Among the other minerals, iron (Fe) content was found in higher concentration in all samples with highest value in pine honeydew (2.11 mg/kg) and lowest in multifloral honey (1.26 mg/kg). The concentrations of zinc (Zn), lead (Pb) and cadmium (Cd) were found very low (<1mg/kg) except zinc in pine honeydew (1.04 mg/kg) (Table 2). The results for Mn, Fe, Zn, Pb and Cd were in agreement with those reported by Liberato et al. [2013] for Brazilian honey of Apis mellifera from different floral origins. The mineral content of the honey depends on geographical area, climatic conditions and soil type of the floral source. It was reported that soil in Kashmir valley of India did not contain high amount of minerals like Cu, Fe, Zn and Co [Bhat et al., 2011; Yatoo et al., 2011] which might be the reason of low level of mineral content in the studied samples of honey varieties. Strong and positive correlations were found between all minerals and colour (pfund) while strong negative correlations with L* value, which indicated that honey with dark colour contained high amount of ions (Table 3). All minerals also showed strong positive correlation with electrical conductivity and ash, which demonstrated that higher mineral content, resulted in higher ash content and electrical conductivity. A positive correlation was also found among all minerals, electrical conductivity and pH (Table 3), which showed that both pH and conductivity were dependent on the amount of ions in honey. Similar correlations were observed by Acquarone et al. [2006] between total mineral content and electrical conductivity. Multivariate analysis In our study, the PCA was applied to achieve a reduction of original data matrix while retaining the maximum amount of variability present in data. The factor loading obtained for the first three components (PCs) and the percentage of variance along with cumulative variance is shown in Table 4. The first three principal components accounted for 86.92% of the variance in the honey samples analysed. The first, second and third principal components (PC1, PC2, and PC3) explained 65.72%, 16.26% and 4.93% of the variance, respectively (Table 4). According to loading matrix (Table 5), it was observed that 65.72% variability explained by PC1 was positively correlated with variables viz. colour intensity, colour (mmpfund), iron, zinc, proline, electrical conductivity and negatively with colour L*, a* and b* values (Figure 1) which could be specified as indicators of colour and minerals. Thus such variables could be used to distinguish pine honeydew from acacia honey. Figure 1 also illustrated that all acacia samples located on the left side of PC1 were linked to L*, a* and b* values while all pine honeydew samples positioned on its upright were linked to minerals and electrical conductivity. The second component (PC2=16.26%) was positively corre- TABLE 5. Principal component analysis. Loading of the first three components. Principal components Factor loading pH Moisture Total solids Color Intensity Total soluble solids HMF Ash Titrable acidity Color (mmpfund) Water activity Color (L*) Color (a*) Color (b ) 1 0.660 -0.657 0.657 0.994 0.333 -0.479 0.903 0.923 0.993 -0.279 -0.978 -0.746 -0.872 -0.524 0.746 0.985 0.990 0.901 0.880 0.934 0.674 0.994 0.658 0.978 2 -0.014 0.481 -0.481 -0.027 -0.317 0.840 0.042 -0.082 0.034 0.912 -0.159 -0.577 -0.440 0.782 0.490 -0.102 0.078 0.258 0.130 0.000 0.435 -0.002 -0.307 0.072 3 0.424 -0.163 0.163 -0.048 0.700 0.111 -0.137 0.092 -0.043 0.192 0.025 0.026 -0.008 0.104 -0.245 -0.063 -0.075 0.208 0.250 -0.080 -0.069 -0.053 -0.420 -0.036 Cu Mn Fe Zn Pb Cd Diastase number Invertase number Proline Reducing sugar Electrical conductivity lated with water activity and HMF, which could be interpreted as an indicator of the maturity of honey. Such variables could be used to distinguish multifloral honey from acacia honey and pine honeydew. The third component with less contribution (PC3=4.93%) was positively correlated with TSS and pH but negatively with reducing sugar. Similar results were reported by Kadar et al. [2010] for colour (mmpfund), CIE L* a* b* and conductivity, while Saric et al. [2008] published simi- PC 2: 16.26% - 0.2 -2 - 0.4 - 0.6 -4 0 PC 1: 65.72% 5 FIGURE 1. Projections of the variables on the factor plane for the three honeys (Botanical origins: A: Acacia, PH: Pine Honeydew, MF: Multifloral). TABLE 6. Classification result of LDA of variables in three types of honeys. From/to Acacia honey Pine honeydew Multifloral honey Total Acacia honey 10 0 0 10 Pine honeydew 0 10 0 10 Multifloral honey 0 0 10 10 Total 10 10 10 30 % Correct 100.00 100.00 100.00 100.00 lar results on Croatian honey varieties from three harvesting seasons in which he concluded that in all seasons PC1 was mainly dominated by conductivity, proline and diastase number. Similar results for HMF were obtained by Isopescu et al. [2014] for Romanian honeys. As shown in Figure 1, a natural separation between honeys of different botanical origin was obtained. All the three analysed honey samples from different botanical origins were correctly classified (100%) by using LDA (Table 6). CONCLUSION The analysed three honey varieties from Kashmir valley of India revealed that floral origin significantly affects all the physico-chemical parameters, enzymatic properties, mineral content and colour characteristics except total solids. The concentration of minerals found in all three honey varieties was low as compared to other varieties of honey of Indian origin. Application of multivariate techniques confirmed the validity of physico-chemical analysis as a tool for classifi- cation and characterization of honey obtained from different botanical sources. PCA revealed 86.92% of the variance with the first three principal components with minerals, colour and electrical conductivity dominating variables. LDA proved to be an effective tool which classified the all honey samples 100% correctly. ACKNOWLEDGEMENTS The first author is very much thankful UGC New Delhi India for providing financial assistance in the form of MANF2013­14 and also to local beekeepers of Kashmir valley of India for providing raw honey samples.

Journal

Polish Journal of Food and Nutrition Sciencesde Gruyter

Published: Jun 1, 2015

References