Physicochemical and antioxidant properties of honey across bee species from North Eastern Hill region of India

Abstract

The present investigation focuses on the comprehensive assessment of the physico-chemical and antioxidant potential of honey samples from various bee species, Apis dorsata, A. florea, A. cerana himalaya, A. mellifera, Lepidotrigona arcifera, and Tetragonula sp. from the North Eastern Hill (NEH) region of India. This study reveals for the first time that honey of stingless bees, Lepidotrigona arcifera (LAH) and Tetragonula sp. (TSH) consistently excels in total phenolic content (TPC), total flavonoid content (TFC), proline content (PC) and total antioxidant activity-CUPRAC (TAAC) and total antioxidant activity-DPPH (TAAD). Total flavonoid content in honey was positively correlated with antioxidant activity. Proline content demonstrated strong positive correlations (0.80) with TAAC, indicating its influence on antioxidant potential and stability. Higher levels of total phenolic content were associated with greater antioxidant activity in honey samples. The principal component analysis (PCA) of physicochemical, and antioxidant properties of the honey produced by different species identified two principal components that explain almost 87.3% of the total variance of the data with the first component explaining 74.4%, second PCA explained an additional 12.9%. The PCA highlighted a clear distinction between species based on these biochemical markers, suggesting that honey from different species possesses unique compositional characteristics that could influence their quality and functional properties. Honey production has excellent potential in high-altitude areas, from which the branding of honey and export may be executed with suitable extension activities to promote beekeepers in the NEH region of India.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-98040-w.

Introduction

Honey is the natural sweet ingredient produced by honeybees from the nectar of blossoms or from exudations of living parts of plants or excretions of plant-sucking insects on the alive parts of plants, which honeybees gather, transform and combine with specific constituents of their own, store and leave in the honeycomb to ripen and mature. Honey is deciphered as a semi-liquid product containing a complex mixture of carbohydrates, specifically the monosaccharides glucose and fructose, and other enzymes, pigments, lactones, wax, vitamins, amino acids, organic acids, minerals, and pollen¹. Honey has valuable nourishing, healing, and preventative properties². The composition and quality of honey are variable and depend upon various factors like botanical origin, geographical region, seasonal variation, processing techniques, and storage conditions³. Chemical attributes that make honey nutritious include its high polyphenol content, and the presence of enzymes, minerals, and vitamins⁴. Honey also possesses antimicrobial, anti-inflammatory, antidiabetic, and low glycemic index properties in addition to its nutritional value. The physicochemical properties of honey, such as moisture content, ash content, pH, colour intensity, electrical conductivity, viscosity, specific gravity, glucose, sucrose, fructose, and total protein content, contribute to its overall quality and nutritional value^5,6.

The composition of honey differs between different species of bees. The Apini tribe and the Meliponini tribe are two types of bees that produce different quality of honey. The protein content and SDS-PAGE profile of honey from the Meliponini tribe showed more protein bands compared to honey from the Apini tribe, indicating potential biomarkers for authenticity and quality⁷. However, the sugar content in honey from Apis mellifera and Melipona beecheii bees was similar, with fructose and glucose being the main sugars present⁸. Several factors influence the quality of honey produced by different species of bees. The botanical origin of the honey plays a significant role in determining its composition and quality⁹. Additionally, the harvesting and the level of environmental pollutants can contribute to variations in the quality of honey¹⁰. The amount of honey produced and pollen collected depends not only on the floral abundance of an area but also on the morphological characteristics of worker bees, such as their proboscis length and body size¹¹. For these purposes, the most common determinations are melissopalinology, sensory, biological, and physicochemical methods¹².

The physico-chemical investigation of honey is imperative to the honey industry, as these factors are closely related to storage quality, texture, granulation, flavour, and the nutritional and therapeutic qualities of honey^13,14. A good quality honey possesses low water activity, which inhibits microbial growth, and it possesses antioxidant, antimicrobial, and anti-inflammatory effects¹⁵. Additionally, the physicochemical properties of high-quality honey include low moisture content, high electrical conductivity, and low hydroxymethylfurfural content¹⁶. The superiority of honey is essential for obtaining premium prices in markets and gaining trust among consumers in the present times. Owing to the adulteration reports that tarnished the trust in top honey brands of India, it is the need of the hour to determine the originality of the honey samples by analysing their nutraceutical attributes of physical and chemical properties and their botanical origin.

India is the 7th largest producer and exporter of honey in the world. It exported 74,413.05 MT of natural honey in 2021-22, valued at US$ 163.77 million. The present statistics show that honey production has increased from 76,150 MTs (2013-14) to 1,20,000 MTs (2019-20), which is a 57.58% increase in the country (www.indiastat.com). The honey market in India reached a value of INR 19.2 billion in 2020 (Indian Honey Market Outlook 2021–2026). Prior estimates state that the NEH regions of India produce 500 thousand kg of honey annually, with 40 thousand beekeepers, mainly marginally, with one or two hives. As the National Beekeeping and Honey Mission (NBHM) in India aims to achieve the goal of ‘Sweet Revolution’ through the National Bee Board (NBB) for the overall promotion and development of scientific beekeeping in the country, it also encourages studies on exploring the potential of high-altitude honey.

The Northeastern part, sub-Himalayan forests and cultivated areas with oilseed crops and citrus orchards were listed as one of the potential beekeeping regions at the Proceedings of the 37th International Apicultural Congress held at Durban, Africa. Still, the potentiality of becoming one of the “honey capitals” is yet to be explored. The concept of pollinator farming practices has not yet been adopted in the NEH region, although a few small-scale traditional apiaries in Rangpo, Central Pendam, and Namchi for commercial purposes or household are prevalent. The dynamic climatic conditions in mountain areas result in varying flowering and nectar production rhythms. Hence, the native bee species of this region are ideally acclimatised to ensure the pollination of mountain crops^17,18. The north-eastern region’s tribals are also primarily more proficient in sourcing feral colonies of domesticated species of bees, especially A. cerana cerana and A. cerana himalaya, and stingless bees for honey production^19–21. Furthermore, the non-domesticated Apis dorsata honey is also harvested which serves as additional income to these indigenous local communities.

From an industrial point of view, promoting high-altitude honey production from the NEH region offers a multifaceted approach to growth. It creates a lucrative niche market for premium-priced honey with unique characteristics, expands product portfolios for beekeepers, and fosters a socially responsible image through sustainable practices and local community support. Furthermore, it presents opportunities for job creation, and economic development in these regions, and even contributes to ecosystem health through pollination. However, responsible beekeeping practices, fair trade principles, and robust scientific evidence are essential for this initiative’s long-term success. Considering these the present investigation focuses on the comprehensive assessment of the physico-chemical and antioxidant potential of honey samples from various bee species: Apis dorsata, A. florea, A. cerana himalaya, A. mellifera, Lepidotrigona arcifera, and Tetragonula sp. from the NEH region of India.

Results

This study comprehensively analyzes the physico-chemical and antioxidant attributes of diverse honey samples derived from distinct bee species. The acquired values for specific properties underwent meticulous comparison and assessment against the Food Safety and Standards Authority of India (FSSAI) honey standards, as stipulated in the Food Safety and Standards (Food Product Standards and Food Additive) Amendment Regulations of 2019. This evaluation aims to deepen our scientific comprehension of the quality characteristics inherent in honey from varied honeybee populations.

Physical properties

In exploring honey samples from various bee species, distinct physical properties emerged (Fig. 1; Supplementary Table 1). Tetragonula sp. honey (TSH) and Lepidotrigona arciferal honey (LAH) had the densest honey with a significant higher specific gravity (SG) than all other four species of honey. A. florea honey (AFH) has numerically the least dense honey with the lowest SG of 1.416. Electrical conductivity (EC) variations were evident, A. florea honey (AFH) and A. mellifera honey (AMH) displaying the significant lowest at 0.280 and 0.300 mS/ cm, and TSH registering the significant highest EC at 0.770 mS/ c. A. dorsata honey (ADH) had moderate EC value at 0.445 mS/ cm and all values fell within the FSSAI limit (≤ 0.8 mS/ cm). Moisture content (MC) ranged from the lowest at 15.87% in TSH to the highest at 19.34% in AMH, complying with the FSSAI limit (< 20%). pH levels varied, with A. c. himalaya honey (ACHH) being mildly acidic (formic acid) (FA) at pH 4.33, ADH at pH 4.29; AFH slightly more alkaline at pH 4.41, AMH mildly acidic (formic acid) (FA) at pH 4.01, LAH at pH 3.57 and TSH at pH 3.48 exhibited increased acidity. Distinct total soluble solids (TSS) values in honey from different bee species revealed varied concentrations. LAH exhibited the significant highest TSS content at 79.42%, showcasing a rich composition. ACHH and TSH had slightly lower but comparable TSS values. ADH demonstrated TSS at 71.59. TSH had the highest percentage of suspended solids (TS), at 84.1, and AFH the lowest at 81.88. The refractive index at 20 °C varied, with TSH and LAH having the significant highest value indicating denser honey. Pfund values categorized ACHH as amber (93 mm), ADH slightly darker at 94 mm, AFH as light amber (84 mm), AMH within amber (89 mm), LAH as dark amber (115 mm), and TSH as amber (102 mm) showcasing diverse honey colours (Fig. 2). LAH and TSH emerged as the bee species producing honey with the higher TSS and suspended solids, respectively. AFH presented lower values in TSS and suspended solids, showcasing a distinct composition.

Fig. 1.

Physical properties of honey from different bee species in the NEH region of India. Data points from three biological replicates were analyzed using one-way ANOVA, followed by the Least Significant Difference (LSD) post hoc test. Data points with different letters indicate significant differences of P < 0.05.

Fig. 2.

Colour characteristics of honey from different bee species of the NEH Region of India.

Correlation between physical properties: The correlogram visually represented Pearson’s correlation coefficient matrix for the physical properties of honey across various bee species. It represented relationships within these properties (Fig. 3). A moderate positive correlation of 0.68 existed between the refractive index (RI) and total soluble solids (TSS), indicating a proportional increase. A moderately positive correlation of 0.72 between the refractive index (RI) and specific gravity (SG) implied a tendency for specific gravity to increase with honey colour. A positive correlation of 0.50 between SG and total solids (TS) indicated that higher specific gravity corresponded to higher total solid content. On the contrary, negative correlation of − 0.69 between total solids (TS) and pH suggested an inverse relationship, aligning with the acidic nature of honey as pH decreased with increasing total sugar content. A positive correlation of 0.58 between pH and moisture content revealed that as pH became more acidic, moisture content (MC) tended to decrease, consistent with the expectation of higher acidity associated with lower moisture levels. Despite a non-significant correlation (r = 0.47), the positive association between total soluble solids (TSS) and total solids (TS) suggested a slight rise in total solids (TS) as TSS content increased, possibly due to dissolved solids in honey.

Fig. 3.

Correlogram representing the Pearson’s correlation coefficient (p ≤ 0.01) matrix between physical properties of honey from different bee species.

Chemical properties

The chemical analysis of honey from different bee species revealed unique traits (Fig. 4; Supplementary Table 2). Lepidotrigona arcifera honey (LAH) had the significant higher total reducing sugar at 79.56 g/ 100 g, followed by Tetragonula sp. honey (TSH) at 78.98 g/ 100 g and each other was non-significant. A. cerana himalaya honey (ACHH) and A. mellifera honey (AMH) also showed non-significant sweetness in each other. For apparent sucrose (AS), Tetragonula sp. honey (TSH) had the highest at 4.77%, while A. florea honey (AFH) had the lowest at 1.95%, all below the FSSAI 5% limit, confirming non-adulteration. The fructose/glucose ratio (FGR) varied; Lepidotrigona arcifera honey (LAH) and Tetragonula sp. honey (TSH) had significant higher values, within the FSSAI range (0.950–1.500). A. florea honey (AFH) had the significant lowest total ash (TA) at 0.180%, while Tetragonula sp. honey (TSH) exhibited the nearest values of around 0.210%. A. cerana himalaya honey (ACHH) had a significant highest total ash at 0.334%. All samples recorded total ash within the maximum FSSAI limit (0.5%). Significant high acidity was observed in Tetragonula sp. honey (TSH), Lepidotrigona arcifera honey (LAH), and A. cerana himalaya honey (ACHH) at 0.20, 0.19 and 0.19, respectively and non-significant in each other. A. mellifera honey (AMH) and A. dorsata honey (ADH) had moderate acidity at 0.15 and 0.15, while A. florea honey (AFH) showed significant lowest acidity (formic acid) (FA) at 0.10. All samples had acidity (formic acid) (FA) below the FSSAI limit (0.2%), indicating higher freshness. The hydroxymethylfurfural (HMF) levels (mg/kg) were significant low in Tetragonula sp. honey (TSH) at 10.88, while A. dorsata honey (ADH) had the significant highest at 34.21. All honey samples recorded HMF levels within the permissible limit of 80 mg/ kg, indicating freshness was preserved by avoiding heating or processing. Tetragonula sp. honey (TSH) had the highest total protein at 3.59 g/ kg, followed by Lepidotrigona arcifera honey (LAH) at 3.42 g/ kg. A. mellifera honey (AMH) and A. cerana himalaya honey (ACHH) had moderate total protein (TP) values at 1.20 g/ kg and 2.95 g/ kg, while A. florea honey (AFH) had lowest TP levels at 0.51 g/ kg.

Fig. 4.

Chemical properties of honey from different bee species in the NEH region of India. Data points from three biological replicates were analyzed using one-way ANOVA, followed by a Least Significant Difference (LSD) post hoc test. Data points with different letters indicate significant differences of P < 0.05.

Correlation between chemical properties: A significant strong negative correlation of -0.80 suggests an inverse relationship between hydroxymethylfurfural (HMF) and the overall acidity (formic acid) (FC) of honey. Higher HMF levels may be associated with lower acidity. Significant negative correlations of -0.55 between HMF and apparent sucrose (AS), and − 0.76 between HMF and F/G Ratio (FGR), imply that as HMF levels increase, there is a significant decrease in both apparent sucrose and the F/G ratio. This suggests that aging or heating processes lead to reduced sucrose content and less efficient sucrose conversion, resulting in higher HMF levels. The significant positive correlation of 0.58 between apparent sucrose and F/G Ratio indicates that sucrose-rich honey may have a more efficient conversion of sucrose to fructose and glucose during ripening. Significant negative correlations of -0.89 between HMF and total protein, and positive relation of 0.87 between F/G Ratio and total protein (TP) suggest that as HMF and apparent sucrose (AS) increase, there might be a strong and moderate decrease in total protein, respectively. Additionally, honey samples with a higher F/G ratio may exhibit higher total protein. The strong significant negative correlation of -0.85 between HMF and total reducing sugars (TRS) reinforce the notion that aging or heating processes might contribute to decreased reducing sugars in honey. Correlations between HMF and freshness indicators (total ash, acidity) suggest that HMF tends to increase with age or exposure to heat, reflecting changes in the chemical composition of honey over time. Positive correlations among apparent sucrose, F/G ratio, total protein, total reducing sugar, and total ash indicate interdependencies among these properties, highlighting the complex nature of honey’s chemical composition (Fig. 5).

Fig. 5.

Correlogram representing the Pearson’s correlation coefficient matrix (p ≤ 0.01) between chemical properties of honey from different bee species in the NEH region of India.

Antioxidant properties

The comparative analysis of total phenolic content (TPC) across various honey varieties reveals distinctive profiles, highlighting their potential as sources of antioxidants (Fig. 6; Supplementary Table 3). In the total phenolic content (TPC), A. cerana himalaya honey (ACHH) and A. dorsata honey (ADH) showed closely aligned values at 785.21 mg GAEs/ kg and 791.42 mg GAEs/ kg, indicating moderately antioxidant potential. A. florea honey (AFH) had a significant lowest but notable phenolic content at 677.60 mg GAEs/ kg, while A. mellifera honey (AMH) exhibited significant phenolic presence at 704.40 mg GAEs/ kg. Among stingless bee honey, Lepidotrigona arcifera honey (LAH) had significant highest TPC at 886.15 mg GAEs/ kg, followed by Tetragonula sp. honey (TSH) at 847.18 mg GAEs/ kg, showcasing rich phenolic content and potent antioxidant capabilities. A. cerana himalaya honey (ACHH) honey exhibited moderately antioxidant activity, recording a total antioxidant activity-CUPRAC (TAAC) of 878.32 mg AAE/ kg. A. florea honey (AFH) honey displayed significant lowest antioxidant activity at 584.58 mg AAE/ kg. While Lepidotrigona arcifera honey (LAH) honey stood out with the significant highest total antioxidant activity-CUPRAC at 968.94 mg AAE/ kg. Tetragonula sp. honey (TSH) also exhibited significant antioxidant activity with a total antioxidant activity-CUPRAC of 921.90 mg AAE/ kg. In total antioxidant activity-DPPH, A. cerana himalaya honey (ACHH) has moderate value at 163.09 mg AAE/ kg, while A. dorsata honey (ADH) has at 149.23 mg AAE/ kg. A. florea honey (AFH) had significant the lowest activity (103.58 mg AAE/ kg). In stingless bee honey, Lepidotrigona arcifera honey (LAH) led with 171.81 mg AAE/ kg, while Tetragonula sp. honey (TSH) showed notable efficacy at 173.24 mg AAE/ kg and non-significant each other. A. dorsata honey (ADH) has 342.63 mg/ kg proline content, indicating good stability. A. florea honey (AFH) maintained 286.13 mg/ kg proline content, contributing to stability. A. mellifera honey (AMH) with 354.26 mg/ kg proline content aligned with A. dorsata honey. Among stingless bee, Lepidotrigona arcifera honey (LAH) stood out with 587.12 mg/ kg proline content, and Tetragonula sp. honey (TSH) showed 649.28 mg/ kg proline content, both significantly surpassing others, indicating enhanced stability and resistance to crystallization. All honey varieties exceeded the FSSAI limit of 180 mg/ kg, emphasizing their resilience against crystallization.

Fig. 6.

Antioxidant properties of honey from different bee species in the NEH region of India. Data points from three biological replicates were analyzed using one-way ANOVA, followed by the Least Significant Difference (LSD) post hoc test. Data points with different letters indicate significant differences of P < 0.05.

The comparative analysis revealed that Lepidotrigona arcifera honey (LAH) consistently excels in total phenolic content (TPC), total flavonoid content (TFC), and total antioxidant activity (CUPRAC and DPPH). A. mellifera and A. dorsata honey also exhibited moderate antioxidant properties, whereas A. florea honey (AFH) showed lower antioxidant activity. These variations in antioxidant properties emphasized the diverse health-promoting potential of honey, influenced by factors such as bee species, floral sources, and environmental conditions.

The correlation analysis indicates significant positive associations (r = 0.76) between total flavonoid content (TFC) and DPPH-assayed antioxidant activity (TAAD). This suggests a strong connection, indicating that honey samples with higher flavonoid content may strongly increase antioxidant activity in the DPPH assay. Additionally, a significant strong positive relationship (r = 0.90) is observed between TAAD and TAAC, suggesting that higher activity in one assay may lead to strong higher activity in the other, influenced by differences in assay methods. Strong positive relationships are identified with a correlation coefficient of (r = 0.84) between TAAD and total phenolic content (TPC), as well as a correlation coefficient of 0.96 between TAAC and TPC. This highlights that honey samples with higher total phenolic content exhibit greater antioxidant activity in both DPPH and CUPRAC assays, emphasizing the significance of phenolic compounds in honey. Furthermore, the significant strong positive relationships (r = 0.78) between TAAD and proline content (PC), and (r = 0.80) between TAAC and proline indicate that honey samples with higher proline content show enhanced antioxidant activity in CUPRAC and DPPH assays. This underscores the potential contribution of stability-related factors, particularly proline content, to the observed antioxidant properties in honey. The significant moderate positive relationship (r = 0.67) between total flavonoid content (TFC) and proline content suggests that honey samples with increased flavonoid content also exhibit higher levels of proline, highlighting the combined influence of flavonoids and proline on the overall antioxidant potential and stability of honey (Fig. 7).

Fig. 7.

Correlogram representing the Pearson’s correlation coefficient matrix (p ≤ 0.01) between antioxidant properties of honey from different bee species in the NEH region of India.

Multivariate analysis

The principal component analysis (PCA) of honey species parameters revealed a distinct clustering of species based on their physico-chemical and antioxidant properties. The first PCA (Dim1) accounted for 74.4% of the total variance, while the second PCA (Dim2) explained an additional 12.9%, collectively capturing 87.3% of the variation in the dataset. Species like Apis florea honey (AFH) and Apis dorsata honey (ADH) displayed distinct parameter profiles, with Apis dorsata honey showing a strong association with moisture content (MC), hydroxymethylfurfural (HMF), and pH (pH), while Apis cerana himalaya honey (ACHH) was linked to traits like total antioxidant status. On the other hand, Tetragonula sp. honey (TSH) exhibited a close relationship with refractive index (RI) and total soluble solids (TSS). The PCA also highlighted a clear distinction between species based on these biochemical markers, suggesting that honey from different species possesses unique compositional characteristics that could influence their quality and functional properties (Fig. 8).

Fig. 8.

Principal component analysis of the physicochemical, and antioxidant properties of honey from different bee species in the NEH region of India.

PC (proline content), pH, SG (specific gravity), RI (Refractive Index), and TS (total solids) are pivotal factors shaping patterns in the honey dataset. PC, highlighted in blue, holds particular significance. SG, also in blue, plays a strong role, reflecting its importance as a physical property influenced by factors such as sugar concentration and water content. TS, another blue variable, is highly influential, indicating that variations in total solids contribute to observed patterns. Apparent sucrose (AS), and total flavonoid content (TFC) moderately influence the observed patterns, with AS highlighted in dark blue, suggesting a moderate representation. While not as impactful as the blue variables, AS contributes to the dataset’s patterns, being an important parameter related to honey composition and purity. TFC, also in dark blue, has a moderate impact, indicating the role of flavonoids in shaping variability in the dataset. TAAD and electrical conductivity (EC) are less represented in the dataset, indicating their minor impact on primary patterns. TAAD, depicted in the dark, suggests its variations have a limited effect compared to blue and dark blue variables. EC, also in the dark, provides information on ion conductivity in honey but exerts less influence on observed patterns in the dataset (Fig. 8).

Notably, proline content (PC), pH, specific gravity (SG), refractive index (RI), and total solids (TS) demonstrated the highest Cos2 scores, making them the top five contributors to both PCA1 and PCA2 (Table 1). These variables influence the characteristics represented by these principal components, as indicated by (Table 1; Fig. 8). The loading values and matrices offer a comprehensive insight into variable relationships within the principal components. For PCA1, strong positive loadings in specific gravity (SG), total solids (TS), refractive index (RI), and proline content (PC) contrast with high negative loadings in moisture, pH, and hydroxymethylfurfural (HMF). This highlights trade-offs, with increasing SG and TS associated with higher PC but lower moisture, pH, and HMF. The second principal component (PCA 2) highlights unique relationships with high negative loadings for total ash (TA) and total antioxidant activity-DPPH (TAAD), suggesting an inverse association between them. Conversely, electrical conductivity (EC) and refractive index (RI) show positive loadings, indicating a positive relationship (Table 1).

Table 1.

Component loadings of physico-chemical, and antioxidant properties of honey from different bee species in the NEH region of India.

Parameters	PCA 1	PCA 2
Specific Gravity (SG)	0.260	0.092
Electrical Conductivity (EC)	-0.117	0.380
Moisture Content (MC)	-0.260	-0.060
Pfund	0.234	-0.077
pH	-0.266	-0.125
Total Soluble Solids (TSS)	0.231	0.101
Total Solids (TS)	0.260	0.060
Refractive Index (RI)	0.206	0.321
Total Reducing Sugars (TRS)	0.242	0.072
Apparent Sucrose (AS)	0.241	-0.127
F/G Ratio (FGR)	0.240	-0.084
Total Ash (TA)	0.002	-0.540
Acidity (Formic Acid) (FA)	0.194	-0.244
Hydroxymethylfurfural (HMF)	-0.245	-0.151
Total Protein (TP)	0.240	-0.086
Total Phenolic Content (TPC)	0.234	-0.126
Total Flavonoid Content (TFC)	0.195	-0.260
Total Antioxidant Activity-CUPRAC (TAAC)	0.222	-0.196
Total Antioxidant Activity-DPPH (TAAD)	-0.108	-0.397
Proline Content (PC)	0.271	0.116

Proline content (PC) emerged as a pivotal variable, indicating a strong alignment with patterns observed in both principal components. The high Cos2 score for PC emphasized its significance in capturing the underlying structure of the dataset. Similarly, pH, SG, and TS also exhibited high Cos2 scores, highlighting their crucial roles in shaping both PCA1 and PCA2. These variables (PC, pH, SG, and TS) were key factors that strongly influenced characteristics encapsulated by the principal components. Their notable Cos2 scores quantitatively assessed their impact on the identified principal components, suggesting their instrumental role in delineating major trends and variations within the dataset. The loading matrix provided detailed insights into variable contributions to each principal component. While PCA1 was crucial in capturing dominant patterns, PCA2, though explaining a smaller variance, provided additional perspectives on variable relationships, contributing to a comprehensive understanding of the physicochemical composition that drove variability in the honey samples (Table 1).

Hierarchical clustering

The clustering analysis sought to reveal patterns and similarities among diverse honey samples, categorizing them into distinct clusters based on their chemical composition. These samples disclosed specific associations and dissimilarities, enabling an interpretation of their physico-chemical profiles (Fig. 9). In the hierarchical cluster analysis, the honey samples in the study formed five clusters based on their physico-chemical and antioxidant profiles. Notably, A. cerana himalaya honey (ACHH) and A. dorsata honey (ADH) exhibited similarities in Cluster 1, while A. florea honey (AFH) and A. mellifera honey (AMH) shared resemblances in Cluster 2. Tetragonula sp. honey (TSH) showed similar biochemical makeup with Lepidotrigona arcifera honey (LAH) in Cluster 3. The cluster analysis effectively classified honey samples based on physicochemical properties, revealing distinct characteristics within each cluster. The chosen number of clusters facilitated a comprehensive understanding of relationships among samples, emphasizing shared and distinctive physicochemical profiles (Fig. 9).

Fig. 9.

Hierarchical clustering from the multivariate analysis of honey samples of different bee species in the NEH region of India using ward linkage with rescaled distance clustering.

Discussion

This research focused on deciphering the uniqueness of honey analysed with superior chemical properties, including the highest antioxidant potential, proline, phenolic, and flavonoid contents in honey samples obtained from various bee species of India’s NEH region.

The diverse honey samples from different bee species in the NEH region of India exhibit varying specific gravity (SG) and moisture content (MC), reflecting unique characteristics. Tetragonula sp. honey (TSH) and Lepidotrigona arciferal honey (LAH) show slightly higher SG, indicating denser concentration, while A. florea honey (AFH) displays a slightly lower SG^18,22,23. Tetragonula sp. honey (TSH) stands out with the lowest moisture content at 15.87%²⁴ indicating greater resistance to spoilage. In contrast, A. florea honey (AFH) has the highest moisture content at 18.12%²⁵, nearing the upper acceptability limit. Other honey samples maintain acceptable moisture levels, reflecting the influence of factors like nectar origin, climate, and bee species^26–30. Electrical conductivity varies, with A. florea honey (AFH) having the lowest at 0.28 mS/cm and Tetragonula sp. honey (TSH) having the highest at 0.77 mS/cm^25,31–33, suggesting differences in purity.

Honey from various bee species in the NEH region of India has pH levels ranging from 3.48 to 4.41, reflecting its acidic nature. This pH range differs from Nigerian honey (4.3–6.0)³⁴ but aligns with Jordanian multi-floral honey (3.5–3.7)²⁷ honey from Trinidad and Tobago³⁵, and honey from the West Bank of Palestine (3.03–5.98)²⁶. Similar pH values were reported in different parts of India, including Kerala (3.49–4.45)³⁶, Jammu and Kashmir (3.52–3.78)³⁷, and Meghalaya^38,39. Stingless bee honey exhibited the lowest pH, consistent with Malaysian findings (pH range: 2.65–3.58, average: 3.31 ± 0.26), while Apis sp. honey had higher pH values (3.04–4.28, average: 3.85 ± 0.35) ^[25]. Variations in pH levels may stem from nectar sources, environmental conditions, and bee foraging behaviour²² influencing taste, texture, and preservation of honey⁴⁰. Pfund values quantitatively measure honey colour characteristics across bee species. Honey colour varies due to factors like beekeeper practices, comb handling, minerals, heavy metals, and environmental exposure^18,41,42. The diverse colour in honey reflects botanical sources and regional characteristics, vital for commercial purposes, especially in monofloral honey^20,43,44. The refractive index at 20 °C is an optical parameter reflecting the honey’s transparency and concentration. Tetragonula sp. honey (TSH) exhibited the highest refractive index, signifying denser and more concentrated honey. Conversely, A. c. himalaya honey (ACHH) displayed a slightly lower refractive index, indicating variations in optical properties. This observation aligns with Tiwari’s study on A. c. indica honey, where a lower refractive index was linked to moisture content^12,45.

Honey samples from the NEH region of India bee species exhibited variations in sugar content, reflecting studies by Chou (1994) and Escuredo (2014)^46,47. Lepidotrigona arcifera and Tetragonula sp. honey showed high apparent sucrose (AS), influenced by sucrose esters impacting foraging behaviour^48–51. It also recorded higher acidity, while A. c. himalaya and A. mellifera honey showed moderate acidity. A. dorsata and A. florea honey had lower acidity (Fig. 4)^52,53, indicating varying acidity levels influenced by foraging behaviour and nectar sources. In the HMF assessment, Tetragonula sp. and Lepidotrigona arcifera honey showed lower levels, while A. mellifera and A. florea honey exhibited higher values. A. dorsata honey had the highest HMF content (Fig. 4), suggesting variations due to processing and storage conditions. Lepidotrigona arcifera and Tetragonula sp. honey had the highest total protein (TP), linked to stored pollen, showing significant variation^46,54.

The study evaluated TP, TFC, TAAD, and PC in various honey varieties from the NEH region of India. Lepidotrigona arcifera honey displayed the highest TP while A. florea honey had a lower TP attributed to floral sources, geography, and bee species^32,55. Lepidotrigona arcifera honey exhibited the highest TFC, whereas A. florea honey had the lowest^32,55. In the DPPH assay, Tetragonula sp. honey had the highest TAA, and A. florea honey showed the lowest activity. In the CUPRAC assay, Lepidotrigona arcifera and Tetragonula sp. honey demonstrated strong antioxidant capabilities with TAA values, while A. florea honey exhibited the lowest activity⁵⁶. The results emphasize variations in antioxidant potential among different honey varieties.

Honey’s antioxidative processes are strongly linked to total flavonoid content, and water-soluble vitamins, notably vitamin B3, contribute to free radical scavenging^3,46,57,58. Adding bee products like beebread and propolis enhances total phenolic and flavonoid content, and antioxidative activity⁵⁹. The antioxidant properties of honey are influenced by its floral source, with phenolic compounds playing a crucial role³¹. Floral source influences honey’s antioxidant properties with varying polyphenols in types like heather and honeydew⁶⁰. Proline content in honey reflects stability; Tetragonula sp. honey has the highest at 649.28 mg/ kg, A. florea honey has the lowest at 286.13 mg/ kg (Silva, 2016). Proline’s variation depends on flower origin, processing, and added sugars^52,61.

Physico-chemical analysis of diverse honey samples revealed significant correlations. Positive links included RI-TSS, TSS-Pfund Index, and SG-TS, while negative correlations involved TS-pH, pH-MC, and HMF-acidity. HMF strongly correlated with aging indicators, and apparent sucrose positively linked with the F/G ratio. Total phenolic content exhibited a strong positive correlation with antioxidant activity. Proline content showed strong positive correlations with antioxidant potential (TAAC) and stability (TAAD) (Figures 2, 3 and 4).

Samardzic & Ibragic (2022) found a correlation between Pfund colour characteristics in Bosnia and Herzegovina and Turkey honey with both total soluble solids (TSS) and specific gravity (SG)⁶². Darker honey types naturally exhibit higher specific gravity due to denser compounds. Sonawane and Varpe (2023) discovered a robust positive correlation between total solids (TS) and specific gravity in Maharashtra honey²⁴. Meena et al. (2021) reported significant positive correlations between pH and electrical conductivity (EC), EC and moisture, and colour and pollen density in Himachal Pradesh honey⁶³. In Sohra, Meghalaya, a strong association (correlation coefficient of 0.7) was found between specific gravity and moisture content, while a moderate association (correlation coefficient of 0.6) existed between hydroxymethylfurfural (HMF) and pH values⁶⁴. Additionally, Dobrinas et al. (2022) highlighted a positive correlation between total reducing sugars (TRS) and total soluble solids (TSS), providing insights into the interdependencies of various chemical and physical characteristics in honey samples⁹.

In the investigation conducted by Samardzic and Ibragic (2022) on honey from Bosnia and Herzegovina and Turkey, a correlation was established between antioxidant activity and various parameters, including total phenolic content (TPC), total flavonoid content (TFC), proline content, and the colour of honey. Similarly, in Malaysian monofloral honey samples produced by A. cerana and A. dorsata bees, a strong correlation was pragmatic between the colour intensity of the honey and various antioxidant parameters^32,62. These findings align with the current research in honey, where stingless bee honey (Lepidotrigona arcifera and Tetragonula sp.) recorded the highest Pfund values (115- dark amber and 102- amber) with the highest antioxidant values.

Differences in antibacterial and antioxidant activity result from floral sources and regional traits, aiding honey authentication. Electrical conductivity variations indicate diverse floral sources and geographical origins influenced by mineral content. Environmental factors during honey production, like soil composition and beekeeping practices, further contribute to the physicochemical profile.

Principal component analysis (PCA) and hierarchical cluster analysis effectively classify honey based on physico-chemical attributes and pollen profiles⁶⁵. In our study on NEH region honey, PCA revealed two key components. PCA1, representing 74.4% of the total variance, highlighted variables like specific gravity (SG), total solids (TS), and proline content (PC) with high positive values, while PCA2, explaining an additional 12.9% of the variance, contrasted acidity (formic acid) (FA) and DPPH assay-based antioxidant activity (TAAD) with electrical conductivity (EC) and refractive index (RI). Together, PCA1 and PCA2 covered almost 87.3% of the total variance, offering a comprehensive dataset overview.

A physico-chemical study by Wu et al. (2023) on multi-floral honey produced by A. cerana cerana, A. dorsata and Lepidotrigona flavibasis in the Chinese province produced three main principal components with variances of PCA1, PCA2 and PCA3 were 40.9%,31.8% and 19.2%, respectively⁶⁶. A. dorsata honey was designed in PCA1 positive, while the groups with other honey samples were designed in a negative PCA1. The loadings plot showed that compounds had positive values on PCA1, primarily responsible for discriminating A. dorsata honey from other honey samples in the Chinese province.

Through hierarchical cluster analysis, honey samples of the NEH region of India in the present study were distinctly categorized into five clusters based on their physico-chemical and antioxidant profiles. These distinct clusters proposition valuable insights into the distinct characteristics of each honey sample and species. The application of cluster analysis and multivariate data analysis systems have proven to be instrumental in characterizing and classifying honey samples based on their physicochemical properties. This approach demonstrates efficacy in discerning geographical origin, with specific physicochemical attributes playing a significant role in differentiation, as suggested by Escuredo Olga et al.⁶⁵.

Conclusion

All the honey samples generally met the specified standards of FSSAI in specific gravity, electrical conductivity, moisture content, pH, total soluble solids, total solids, and refractive index. The Lepidotrigona arcifera, Tetragonula sp., and A. cerana himalaya honey samples exhibited desirable chemical properties used to assess the freshness and sweetness. The highest antioxidant potential, phenolic and flavonoid content offering greater health-promoting properties, was observed in L. arcifera honey. This study first time reveals that honey of stingless bees, L. arcifera honey (LAH) and Tetragonula sp. honey (TSH) consistently excel in total phenolic content (TPC), total flavonoid content (TFC), proline and total antioxidant activity (CUPRAC and DPPH). A positive correlation with flavonoid content might slightly increase antioxidant activity, while a negative correlation indicates potential variations between different antioxidant assays. Strong relationships were observed, with a negative correlation between TAAD and a robust positive correlation (0.97) with TAAC, showcasing that higher total phenolic content was associated with greater antioxidant activity in honey samples. Proline content demonstrated strong positive correlations (0.80) with TAAC, indicating its influence on antioxidant potential and stability. A. c. himalaya and L. arcifera honey exhibited lower HMF levels, thus justifying the freshness of the honey from the domesticated species of the region.

This study reveals the exceptional potential of NEH region honey, particularly from stingless bees in India. The honey’s superior antioxidant properties, freshness, and unique characteristics present a compelling opportunity for regional development. By establishing a distinct brand identity, focusing on export potential, and incentivizing beekeeping, particularly of stingless bee species, the region can create a thriving honey industry. Further economic benefits can be achieved through value-added products like nutraceuticals and cosmetics. To fully realize this potential, capacity building for beekeepers, infrastructure development, and establishing market linkages are crucial. This study paves the way for transforming beekeeping into a driver of economic and social growth in the NEH region of India.

Materials and methods

Collection of honey samples

The honey samples from the North Eastern Hill region of India were collected from three sealed cells bee colonies, representing different bee species such as A. dorsata, A. florea, A. cerana himalaya, A. mellifera, L. arcifera, and Tetragonula sp. These six bee’s species raw honey samples as A. dorsata honey (ADH), A. florea honey (AFH), honey (ADH), A. cerana himalaya honey (ACHH), A. mellifera honey (AMH), L. arciferal honey (LAH) and Tetragonula sp. honey (TSH), weighing between 100 and 200 g, were stored under standard laboratory environments. They were later used (within 1 month) for physical, chemical, and antioxidant analyses in the current study.

Physical properties

Refractive Index (RI) and Moisture Content (MC): The refractometric method, recommended by the International Honey Commission (IHC) and Association of Official Analytical Chemists (AOAC’s) 969.38, is preferred for its simplicity and reproducibility¹². Precise temperature control is crucial; honey’s sugar crystals are dissolved at 50 °C before measuring the refractive index at 20–40 °C using an Abbe or digital refractometer. Moisture percentage is determined through an empirical formula or a relative conversion Table⁴¹. Experiments are conducted in triplicate, and results are reported as the mean average⁶⁷.

Electrical conductivity (EC): The procedure entails dissolving a 10 g honey sample in 25 ml of distilled water in a beaker. The electric conductivity meter’s EC cell is then immersed in the solution to measure conductivity, with results reported in milli Siemens per meter (mS/m) following the methods of the International Honey Commission (2009).

Colour (Pfund): The Pfund scale colour of honey was measured using the Hanna Instruments Honey Colour Grade Portable Photometer (H196785). This device, employing an LED and narrow band interference filter, determines colour values in millimeters by measuring light wavelengths passing through the honey. The instrument’s cuvette locking system ensures a consistent 10 mm path length, and calibration involves using glycerol as a blank sample. Readings, compared to an analytical grade glycerol standard reference, are directly taken from the instrument display and expressed in millimeters (mm) Pfund¹.

pH: The pH measurement following the IHC protocol was conducted using a pH meter. This involves dissolving 10 g of honey samples in 75 ml of distilled water. The results, obtained in triplicate, will be presented as the average.

Total Soluble solids (TSS): The total soluble solids (TSS) of the honey samples were measured using a refractometric method. A small drop of diluted honey was placed on the instrument’s prism and illuminated, revealing a distinct interface between light and dark zones. This interface corresponded to the measured refractive index, which could be directly converted to degrees Brix (Bx).

Total Solids (TS): The Total Solids (TS) in the honey samples were calculated by the procedure outlined by Saxena et al. (2010)⁵⁶, subtract the Moisture content from 100%, as expressed in the formula: TS = 100 − Moisture content.

Specific Gravity (SG) @ 27 °C: The purpose of the specific gravity of honey involved using a specific gravity bottle and recording the weights of the empty bottle (A), the bottle with water (B), and the bottle with the honey sample (C). The specific gravity was then computed using the following formula:

Where, A = Mass of empty specific gravity bottle (g); B = Mass of specific gravity bottle with water (g); C = Mass of specific gravity bottle with honey sample (g).

Chemical properties

Total Reducing Sugars (TRS): One gram of the prepared honey sample was correctly weighed and placed into a 250 ml volumetric flask, diluted with 150 ml of water, and mixed. The capacity was adjusted to 250 ml with water. In a ceramic dish, 5 ml each of Solution A and Solution B were taken. Then, 12 ml of honey sample was added from a burette, and the mixture was boiled. Methylene blue indicator (1 ml) was added, and titration was completed within three minutes, with the endpoint marked by a colour change from blue to red.

Where, S = strength of copper sulphate solution; H = volume, in ml, of honey solution required for titration; and M = mass, in g, of honey.

Apparent Sucrose (AS): A one-gram honey sample was liquified in water (250 ml). Then, 100 ml of the solution was assorted with 1 ml of hydrochloric acid, heated, and left overnight. After frustration with sodium carbonate, total reducing sugars were determined following AOAC (2012) guidelines.

F/G Ratio (FGR): The F/G Ratio was determined using HPLC. The sample was dissolved in water, diluted with Acetonitrile, and injected into HPLC-RID for sugar separation and quantification, following AOAC 977.20 (2019) guidelines.

Total Ash (TA): The 5 g to 10 g honey sample was weighed in a silica or platinum dish. After adding a few drops of olive oil to avoid spattering, the assortment was heated until swelling ended and ignited at 600 ± 20^oC in a muffle furnace until white ash. After cooling in a desiccator, the constant weight was recorded. Total Ash (TA), expressed as a percentage by mass, was calculated using the formula: Total Ash (TA), per cent by mass = , Where, M = mass, in g, of the dish with the ash; of the empty dish; and Ml = mass, in g, of the dish with the substantial taken for the test (IS 4941: 1994, Reaffirmed year: 2014).

Acidity (Formic Acid) (FC): To assess, 10 g of the honey sample was liquified in 75 ml of carbon dioxide-free water. The solution was titrated against standard sodium hydroxide solution using 4–6 drops of deactivated phenolphthalein as an indicator, ensuring the pink colour kept on for at least 10 s. The blank was determined with water and the indicator, correcting the volume of standard sodium hydroxide solution based on this determination. Acidity (as formic acid per cent by mass) = 0.23 x V/ M, where V = corrected volume of 0.05 N sodium hydroxide solution required for titration; and M = mass, in g, of the sample taken for the test (IS 4941: 1994, Reaffirmed year: 2014).

Hydroxymethyl furfural (HMF): To analyse HMF content, 5 g of honey were dissolved in 25 ml of distilled water and treated with a clarifying agent (0.5 ml Carrez I and 0.5 ml Carrez II). After increasing the volume to 50 ml, filtration was done, discarding the first 10 ml. Absorbance measurements at 284 and 336 nm were taken, and the HMF content was calculated using the formula in 100 g of honey = (Abs284 – Abs336) × 14.97 × (5 g of sample) (2), Where Abs284 and Abs336 are absorptions at the wavelength of 284 and 336 nm, respectively (AOAC Official Method 980.2, 1979).

Total Protein (TP): The total protein in honey following Bogdanov et al. (2008) was assessed by diluting honey at 1:10 with Milli-Q water or a suitable buffer. Reaction mixtures of 1 mL honey dilution, BSA standard, and 5 mL Bradford reagent were incubated for 5–10 min. Absorbance at 595 nm was measured against a blank, and protein concentration was determined using a calibration curve. TP (mg/kg) = (Protein concentration * Dilution factor * 1000 mg/g)/1000 g/kg, with the dilution factor typically set at 10.

Antioxidant properties

Total Phenolic Content (TPC) (mg GAEs/kg Honey): The Folin-Ciocalteu method for total phenolic content involved extracting 250 µL honey solution with 2.5 ml ethanol, adding 0.5 mL FC reagent, 2 ml sodium carbonate, and incubating for 30 min. Absorbance at 760 nm was recorded, and TPC in mg GAEs/kg honey was calculated using a gallic acid calibration curve, considering a dilution factor (usually 10 for a 10% solution) and molar mass of gallic acid (170.11 mg/mol)⁶⁸.

Total Flavonoid Content (TFC) (mg CEQ/kg Honey): The total flavonoid content was measured using the aluminum nitrate colorimetric method with absorbance at 415 nm⁶⁹. The calculation considered a dilution factor and the molar mass of quercetin.

TFC (mg CEQ/kg) = (Quercetin equivalent concentration (µg/mL) / Dilution factor) * (1000 µg/mg) * (Molar mass quercetin mg/mol) * (1000 ml/L), where, the dilution factor (usually 10 for a 10% solution), the molar mass of quercetin (302.44 mg/mol), and conversion factors (1000).

Total Antioxidant Activity-CUPRAC (TAAC) (mg AAE/kg Honey): The total antioxidant activity was assessed using the CUPRAC assay (mg AAE/kg), involving incubation with CUPRAC reagent and recording the absorbance at 450 nm. Trolox equivalent concentration was calculated based on a Trolox calibration curve, and TAA in mg AAE/kg honey was determined considering a dilution factor and molar mass (176 mg/mol).

Total Antioxidant Activity-DPPH (TAAD) (mg AAE /kg Honey): Total antioxidant activity in honey (mg AAE/kg) was determined with AOAC Method 985.24 (2012), measuring absorbance at 517 nm after a 30-minute incubation with 0.1 mM DPPH solution. TAA-DPPH (mg AAE/kg) was calculated based on absorbance readings and solution parameters, including molar mass (176 mg/mol) and a dilution factor (usually 10 for a 10% solution).

Proline content (PC) (mg/kg Honey): In a 50 mL volumetric flask, 2.5 g of honey was liquified in H2O. A 0.5 mL aliquot was pipetted into three reaction tubes containing HCOOH and Ninhydrin solution. Following boiling, cooling, and the addition of 5 mL of isopropanol (1 + 1), absorbance at 520 nm was measured against a water blank. After colour correction for honey, the resulting value was subtracted. The proline content in mg/100 g of honey was then calculated using a calibration curve with a proline standard solution (AOAC 920.180, 21st edition-2019).

Statistical analysis

Physico-chemical data were analysed in triplicate, and results are presented as means ± standard deviations. All data were checked for normal distribution before statistical analysis. Pearson’s Correlations (p ≤ 0.01) among honey’s physical, chemical, and antioxidant properties from different bee species were determined using R software version 4.3.1. Principal Component Analysis (PCA) was performed on mean-centered and scaled honey parameter data using prcomp. PCA visualization was enhanced with ggplot2, ggrepel, and factoextra including biplots and contribution-based color coding. Hierarchical clustering of honey samples was performed using Ward’s method (ward.D2) on PCA-transformed data, and the dendrogram was visualized with customized cluster highlighting.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

C.N. Nidhi: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation. S.M. Haldhar: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation. K.I. Singh: Supervision, Conceptualization. N.O. Singh: Statistical analysis. B. Sinha: Review & editing. R.N. Kencharddi: Review & editing. L.K. Mishra: Review & editing. M.K. Jat: review & editing. M Choudhary: Statistical analysis.

Funding

This research received no external funding.

Data availability

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.