Each A. dorsata specimen used for morphometric analysis was photographed. The specimens were dissected into separate body sections: head, thorax, abdomen, fore-wing, hind-wing, and hind-leg (Ruttner et al., 1978);(Ruttner, 1988);(Cao et al., 2012). Dissection was performed using a scalpel and forceps. Each dissected body part was mounted on a microscope slide and was subsequently photographed to enable accurate measurement of the corresponding morphometric parameters.

Comparative morphometrics analysis of A. dorsata from three locations revealed both shared and significantly different traits among the populations Table 4. Eleven morphometric characters: body length (BL), head width (HW), compound eye width (CW), abdomen length (AL), abdomen width (AW), forewing width (FWW), hindwing width (HWW), number of hamuli (NH), length of metafemur (LMF), length of metatibia (LMTB), and width of metatarsus (WMTT) did not differ significantly among the three locations. These results indicate a relatively conserved pattern for these traits across populations. The significant average size of these morphometric parameters suggests that the body size of South Sumatra population is larger than in Belitung and West Kalimantan Table 4.

In contrast, twelve morphometric parameters have significantly different variations, namely proboscis length (PL), head length (HL), compound eye length (CL), distance between ocelli (DO), antennal socket distance (AD), thorax length (TL), thorax width (TW), longitudinal diameter of tergite 3 (LT3), longitudinal diameter of tergite 4 (LT4), forewing length (FL), hindwing length (HWL), and the length of the metatarsus (LMTT) Table 4. These findings indicate the presence of distinct morphometric differentiation among populations across sampling locations.

The Belitung giant forest honey bee populations exhibited the highest mean values for proboscis length (5.81±0.53 mm) and antennal socket distance (0.79±0.08 mm), which were significantly greater than those observed in the other locations (P<0.05). Meanwhile, the South Sumatra A. dorsata populations showed the highest mean values for ten morphometric including head length (3.56±0.52 mm), compound eye length (2.76±0.24 mm), distance between ocelli (0.78±0.21 mm), thorax length (4.18±0.44 mm), thorax width (3.57±0.58 mm), longitudinal diameter of tergite 3 (2.81±0.12 mm), longitudinal diameter of tergite 4 (2.74±0.12 mm), forewing length (12.82±0.64 mm), hindwing length (8.71±0.27 mm), and metatarsus length (2.85±0.23 mm). Overall, the results demonstrate that while several morphometric traits are conserved across A. dorsata populations from different locations, a specific subset of characters exhibits significant geographic variations.

Apis dorsata from South Sumatra did not differ significantly nearly half of the measued paramters among A. dorsata colonies, suggesting moderated intra- population homogeneity. The significant variations detected primarily involved overall core body dimensions, wing dimensions, and hind-leg structures. These traits are closely related to flight performance, dispersal capacity, and resource transport. Colony SS1 of this giant forest honey bee consistently exhibited higher mean values Table 3, suggesting localized influences on structural development. Similarly, morphometric variability of A. dorsata in India showed limited differences within the same geographic (Uniyal et al., 2019); (Rajak & Basavarajappa, 2016). Environmental factors such as resource availability, microclimate, and colony nutritional status are important factors in size variations (Hepburn & Radloff, 2004);(Hoiss et al., 2012). Variations in body parts of the bees associated with feeding proboscis, compound eyes ocelli, and abdominal characteristics may reflect differences in ecological conditions across the island. Larger body size is generally associated with longer flight distances and greater resistance to habitat fragmentation (Greenleaf et al., 2007) ;(Cariveau et al., 2016), which could be an advantage in areas with dispersed floral food sources. Furthermore, larger wing size can influence flight efficiency and foraging range, two characteristics often related to environmental conditions and resource distribution (Hoiss et al., 2012). Tropical rainforests in South Sumatra likely provide abundant and diverse food sources. This is reinforced by the positive relationship between vegetation diversity and the variation in pollen collected by bees, supporting the hypothesis that resource-rich environments promote the development of more optimal morphology (Raffiudin et al., 2024).

Apis dorsata from West Kalimantan showed a similar degree of morphological uniformity, with only 10 parameters differing significantly, mainly in head dimensions, abdominal measurements, tergite lengths, hind-wing length, and the number of hamuli Table 3. The relatively high proportion of non-significant parameters suggests greater local homogeneity, possibly facilitated by more continuous forest habitats.

In contrast, Belitung showed slightly greater intracolony differentiation, especially in proboscis length, compound eye length, ocellar distance, thorax, abdomen, and wings Table 3. Variations was concentrated in feeding-related (proboscis length) and sensory (compound eye length, ocellar distance). The significant differences in feeding and sensory structures may reflect ecological heterogeneity across the island landscape. Belitung is characterized by fragmented heath forests, grasslands, mangroves, and coastal vegetation mosaics on acidic sandy soils (Mannion, 2011);(Oktavia et al., 2024). Habitat heterogeneity initiates microenvironmental variations that can influences developmental, such habitat heterogeneity can generate microenvironmental variation that influences developmental pathways and enhances phenotypic plasticity in social bees (Hepburn & Radloff, 2004);(Carré et al., 2009). This ecology is consistent with the history of Belitung as part of the Sundaland savanna corridor during the Quaternary period (Bird et al., 2005);(Sarr et al., 2019), which may have contributed to long-term ecological differentiation relative to the main Sundaland regions. Belitung populations exhibited significantly greater proboscis length and antennal socket distance than A. dorsata in South Sumatra and West Kalimantan. Proboscis length is closely linked to nectar foraging efficiency and floral specialization (Hepburn & Radloff, 2004), suggesting potential ecological adaptation and local floral resources composition. The similar pattern of environmental-morphometric variations has been documented in A. dorsata populations in Karnataka (Dhananjaya et al., 2021) and Lombok (Herlambang et al., 2025), as well as differences in landmark fore- wing venation of A. dorsata across Indonesian islands (Zahara et al., 2022).

A comparison of average morphometric values across A. dorsata populations from South Sumatra, Belitung, and West Kalimantan shows that some traits remain consistent, while others vary depending on geographic location. Eleven characters, including body length, head width, compound eye width, abdominal dimensions, wing width, number of hamuli, and hind- leg measurements Table 4, did not differ significantly among locations. These results suggest that several structural traits remain relatively conserved across regional populations.

Conversely, twelve parameters–including proboscis length, head length, compound eye length, ocellar distance, antennal socket distance, thorax dimensions, tergite lengths, forewing length, hindwing length, and metatarsus length–exhibited significant geographic variations. These differences indicate measurable morphometric differentiation among populations from the three islands/regions.

Across the three regions, ocellar distance, abdominal dimensions, tergite length, and hindwing length repeatedly showed significant variations, indicating that these traits may be particularly responsive to environmental or developmental factors. Colonies of A. dorsata from India in different ecological conditions showed differences in thoracic and abdominal characteristics (Rajak & Basavarajappa, 2016). A range of environmental factors, including altitude (Redhead et al., 2016), habitat type, and the availability of resources (Hoiss et al., 2012), are known to influence morphometric traits in bees. Moreover, variations in habitat structure and landscape features can also contribute to differences in insect morphology (Hepburn & Radloff, 2004); (Carré et al., 2009). Overall, these results suggest that morphological variation in A. dorsata is not consistent across all traits, but is mainly associated with certain characteristics that are sensitive to environmental conditions and differ across regions.

Although overall body size was consistent, A. dorsata populations from Belitung exhibited significantly longer proboscises and wider distances between the antennal sockets than those from two other regions. This extended proboscis length may enhance nectar extraction efficiency (Waddington & Herbast, 1987), particularly in habitats with diverse or morphologically restrictive floral resources. The extended proboscis length may further enhance nectar extraction efficiency, particularly in habitats with diverse or structurally complex of flowers.

The distinct morphometric characteristics of the Belitung populations may also be influenced by the island’s geological and ecological history. Belitung formed part of the Sundaland savanna corridor during the late Quaternary and became isolated following the Holocene sea-level rise (Bird et al., 2005);(Sarr et al., 2019);(Meltzner et al., 2017). Present-day Belitung is dominated by heath forests on acidic sandy soils, interspersed with savanna and mangrove habitats (Mannion, 2011);(Oktavia et al., 2024). Soil properties and plant community composition in such savanna-like ecosystems are known to influence insect community structure, including morphological traits in bees (Haddad et al., 2009);(Rodrigues et al., 2018).

Proboscis lenght is an important factor in bees utilize of floral resources, as it influences theis ability to reach nectar in flowers with different corolla depths (Inouye, 1980) ;(Nicolson, 2011). Bees with longer proboscides are often able to exploit a wider range of flowers and esctract netar more efficiently (Harder, 1985);(Cariveau et al., 2016). Palynological data further suggest that A. dorsata in Belitung collects resources from diverse plant species across varied habitats (Bramasta et al., 2023), suggesting that variations in proboscis length represents a functional to local floral resource composition and spatial distribution, likely driven by selective pressures in this environment.

However, multivariate analysis using NMDS ordination revealed considerable overlap among A. dorsata individuals from Belitung, South Sumatra, and West Kalimantan Figure 3, indicating overall morphometric separation among populations is limited. Although ANOSIM showed statistically significant differences (R = 0.06233, p < 0.0005), the low R-value indicates weak differentiation between A. dorsata populations relative to within-population variations. Thus, while A. dorsata Belitung populations exhibit distinct feeding-adaptation, such as a longer proboscis, these may correlate broadly with a conserved morphological framework rather than forming discrete another morphology.

The extended proboscis observed in Belitung populations represents population-level morphological variation associated with local environmental conditions, habitat heterogeneity, and historical island isolation. The NMDS results also support the interpretation that the variations is subtle and quantitative, reflecting geograhical and ecological adaptation or phenotypic plasticity rather than discrete taxonomic divergence.