Geometric morphometric wing venation of Indonesian Apis cerana and Central Sulawesi Apis nigrocincta
Date
2022-07-14Author
Nisa, Nisfia Rakhmatun
Raffiudin, Rika
Juliandi, Berry
Metadata
Show full item recordAbstract
The Asian honey bee, Apis cerana, is widely distributed in almost all Asian regions, including Indonesia. The distributions of A. cerana in Indonesia arrange from Sumatra, Java, and Sulawesi to Papua. Apis cerana sympatrically lives with A. nigrocincta on Sulawesi Island. In contrast with A. cerana, the distribution of A. nigrocincta is limited to Sulawesi Island and the surrounded islands.
Several islands in the Indonesian archipelagos belong to the Sundaland (Sunda Shelf), and others are located in the Wallacea region; both regions have different environmental conditions and species distribution. The differences in a specific geographic area also affected the morphological characters of species, including bees.
Honeybee wing venations show highly heritability that can determined the relationship and distribution of honey bees. However, the wing landmarks venation of A. cerana in Sumatra, as well as A. cerana and A. nigrocincta that live sympatrically in Sulawesi, has not been studied. Therefore, this study aimed to analyse: 1) intraspecies landmark variations of A. cerana from Sumatra at different altitudes, 2) intraspecies landmark variations of A. cerana from the islands of Sumatra, Belitung, Java, Bali, Sulawesi, and the Moluccas, 3) interspecies landmark variations between A. cerana and A. nigrocincta, and 4) interspecies relationship between A. cerana and A. nigrocincta based on wing venation landmarks.
This study used 55 bee colonies, i.e., 50 colonies of A. cerana, from Sumatra, Belitung, Java, Bali, Sulawesi, and the Moluccas, and five colonies of A. nigrocincta from Central Sulawesi. The right wings of ten individuals from each colony of A. cerana and A. nigrocincta were excised using a scalpel and were digitized on 19 wing venation landmarks. The digitization results were analyzed using geometric morphometric methods with Thin Plate Splin (TPS) software. The analysis of landmark visualization (grid deformation) was performed on the intraspecies of A. cerana from Sunda Shelf and Wallacea, as well as on the interspecies of A. cerana and A. nigrocincta. Principal Component Analysis (PCA) and relationship analysis were performed using R software with rooted and unrooted Neighbor-Joining (NJ) methods.
The intraspecies landmark variations of A. cerana Sumatra resulted variations in the deformation grid. The deformation grid displacement of A. cerana from Aceh and West Sumatra were similar to the reference landmark (300 individuals of A. cerana Sumatra). This study revealed the high different deformation grids in A. cerana North Sumatra, Jambi, South Sumatra, and Lampung compared to reference landmarks. Based on the bending energy values, the lowlands of North Sumatra and Jambi have higher bending energy than the highlands. In contrast, the A. cerana from the lowland West Sumatra has lower bending energy values than those from the highlands.
The comparison of A. cerana from the six islands to the reference landmark resulted the variated displacement of landmarks. The reference landmark is a landmark configuration that were aligned from 50 colonies of A. cerana from six islands. The variations of landmark displacement (deformation grid) between A. cerana on the six islands from Sunda Shelf, i.e., A. cerana Sumatra, Belitung, Java, and Bali, resulted a different deformation grid to the reference landmark. However, A. cerana Sulawesi and Moluccas (Wallacea) showed a similar deformation grid to the reference landmark. This deformation grid analysis revealed the same pattern with the geographical differences of Sunda Shelf and Wallacea. Landmarks 11, 16, and 17 showed the highest contribution values among the 19 landmarks. The last two landmarks are known as cubital index characters in traditional morphometrics. Thus, both landmarks can be used as intraspecies markers of A. cerana based on geometric morphometrics.
The analysis of interspecies landmarks showed that the variations of A. nigrocincta were higher than A. cerana. The deformation grid results showed that the A. cerana landmark displacement led to the inferior of the mean A. nigrocincta landmark. On the other hand, A. nigrocincta led to the superior of the mean A. cerana landmark. The most contributing landmark displacement are generated by landmarks 16 and 17. Thus, these landmarks can be used as interspecies markers of A. cerana and A. nigrocincta. The displacement of landmarks affected the bending energy values in both species. Furthermore, A. nigrocincta produced higher bending energy values than A. cerana.
PCA analysis shows A. cerana from Sumatra, Java, and Sulawesi were clustered at the centre of the ordination plot, while A. nigrocincta separated at quadrants one and four of the plot. Thus, the PCA results can differentiate A. cerana and A. nigrocincta. The differences in these two species are supported by an unrooted NJ tree that placed A. cerana from Sunda Shelf and Wallacea at different branches from A. nigrocincta.
The results of NJ rooted were based on the relative warps value of A. cerana from the six islands and are clustered in three groups, namely: (1) the Java group, (2) the Sumatran group, Sulawesi, Moluccas, and (3) the Belitung group. In the second group, A. cerana from Musi Rawas (South Sumatra) revealed a much longer branch than other locations. Thus, for future study, morphometric analyses of different body parts are needed to clarify the results of A. cerana from Musi Rawas.
Based on the results of this study, the geometric morphometric method using wing venation landmarks revealed the variations in the intraspecies level of A. cerana among different altitudes and many islands. Moreover, the geometric morphometric approach show the differences at the interspecies level of A. cerana and A. nigrocincta. The output of this research is expected as the basic method for a fast and accurate honey bee species determination.