Phylogeography of the giant honey bees based on mitochondrial gene sequences

 Phylogeography of the giant honey bees based on mitochondrial gene sequences

Our goal was to resolve phylogenetic relationships among Apis laboriosa, and the Apis dorsata subspecies A. d. dorsata, A. d. binghami, and A. d. breviligula, the last two of which have been proposed as full species by several authors. We carried out a phylogenetic analysis of the giant honey bees using mitochondrial cox1 and cox2 gene sequences analyzed with maximum likelihood methods. We obtained strong support for four clades within A. dorsata in the broad sense: the three subspecies or species mentioned above, and a fourth lineage from south India. However, our analysis did not resolve the phylogenetic relationships among the four lineages. The presence of two genetically distinguishable groups of “A. dorsata” in India parallels the presence there of two cavity-nesting honey bees, A. cerana cerana and A. c. indica (the black hill bees and yellow plains bees, respectively). This suggests that past climatic or geological events may have temporarily isolated Indian populations from populations of the Asian mainland, leading to divergence and possibly speciation of Indian giant and cavity-nesting bees, followed by recolonization of India by eastern Asian forms. Recognition of these distinct lineages is important for conservation planning, so that their individual distributions, ecologies, and migration patterns can be considered, and so that the genetic diversity they represent can be maintained.

Introduction

The giant honey bees have a geographic range centered on south and southeast Asia, extending northwest into Pakistan, eastwards through India, Bangladesh, Nepal, Bhutan, Myanmar, Thailand, southern China, and southeast Asia, and through the islands of Malaysia, Indonesia, and the Philippines (Otis, 1996; Kitnya et al., 2020; Huang et al., 2022; Kitnya et al., 2024; Otis et al., 2024; Voraphab et al., 2024; Warrit et al., 2024). Several earlier writers including Maa (1953) and Ruttner (1988) pointed out diversity among giant honey bee populations based on morphological and morphometric data. In particular they noted that the giant honey bees of the Himalayan region, the Indonesian island of Sulawesi, and the oceanic Philippine islands (i.e., those islands never connected to the Asian mainland) differed from one another and from the more widespread form found elsewhere. Maa divided honey bees into three genera—Micrapis, the dwarf honey bees, Megapis, the giant honey bees, and Apis, the cavity-nesting honey bees—and recognized four giant bee species: Megapis breviligula from the Philippines, M. binghami from Sulawesi and smaller nearby islands, M. laboriosa from high altitude Himalayan regions, and the more widespread M. dorsata. Ruttner, like most subsequent authors, recognized just one genus, Apis, and only one species of giant honey bee, Apis dorsata. He and many subsequent authors (e.g., Engel, 1999, 2002) considered the Himalayan form a subspecies, A. d. laboriosa, but noted that additional information might confirm it as a distinct species.

The taxonomic status of A. laboriosa remained contentious for many years despite numerous studies. Sakagami et al. (1980) made detailed morphological comparisons of A. laboriosa from Nepal and A. dorsata collected from many parts of its range, documenting “distinct and stable differences between them” supporting species status of A. laboriosa. McEvoy and Underwood (1988) reported that they could find no morphological differences between male genitalia (the everted endophallus) of A. laboriosa and A. dorsata, but nonetheless supported species status of A. laboriosa on the basis of other morphological differences, habitat, the presence of two species of braulid parasites (Diptera: Bruaulidae, Megabraula) in nests of A. laboriosa but (apparently) not those of A. dorsata, and genetic differences revealed by allozyme electrophoresis.

However, some authors argued that the characters used to support species status of A. laboriosa—including habitat, color patterns, and morphometric characters—could represent intraspecific variation and adaptation to different habitats, and thus took the conservative position that more data were needed, particularly concerning reproductive isolation of populations occurring in sympatry (e.g., Ruttner, 1988; Engel, 1999). Cao et al. (2012a) carried out morphometric comparisons of A. laboriosa and A. dorsata collected from Yunnan, Guangxi and Hainan provinces in China and again found significant differences between them. Collection sites for the two were in relatively close proximity (on the order of 200-300 km) but not strictly sympatric, and they were found at different elevations (A. laboriosa 1500 m and above, A. dorsata 1300 m and lower, though all but one collection was made at 700 m or lower).

More recently, new distributional records for A. laboriosa (Kitnya et al., 2020) reported A. dorsata and A. laboriosa foraging sympatrically at sites in Arunachal Pradesh, India and in northern Vietnam. Kitnya et al. (2022), found distinct morphological, morphometric, and genetic differences between Indian populations of A. dorsata and A. laboriosa, both in sympatry and in allopatry, providing convincing support for the species status of A. laboriosa.

Until recently the species status of A. d. breviligula and A. d. binghami have received much less attention. Arias and Sheppard (2005) included A. laboriosa, A. dorsata from Thailand and Sri Lanka, and A. binghami in a larger phylogenetic analysis of Apis species using both nuclear (EF-1α intron) and mitochondrial (ND2) sequence data. The giant honey bees were recovered as a monophyletic group and A. laboriosa was consistently recovered as a clade distinct from A. dorsata and A. d. binghami; however, A. dorsata and A. d. binghami were not consistently resolved as separate lineages. Raffiudin and Crozier (2007) used both mitochondrial (cox2, ND2, and the large (16S) ribosomal subunit or rrnL) and nuclear (inositol 1,4,5-triphosphate receptor or itpr) gene sequences in their phylogenetic analysis of Apis taxa, also including the giant honey bees A. dorsata from Sabah, Malaysia, A. d. binghami, and A. laboriosa. Their analyses consistently recovered A. laboriosa as sister to A. dorsata and A. d. binghami. Lo et al. (2010) carried out a more comprehensive coverage of giant honey bees, including A. laboriosa, A. dorsata from Sabah, Malaysia and Palawan Island, the Philippines, A. d. binghami and A. d. breviligula in their phylogenentic analysis of Apis species, using the same set of genes as Raffiudin and Crozier (2007) minus the mitochondrial ND2. Their results strongly supported the species status of A. d. breviligula from the Philippines, though the placement of A. d. binghami remained unresolved.

Kitnya et al. (2024) carried out a taxonomic study of giant honey bees using morphological characters. Their study included A. laboriosa, A. dorsata, and the island lineages A. d. binghami and A. d. breviligula. They found that A. dorsata from mainland Asia differs morphologically from A. d. binghami and A. d. breviligula but concluded that the latter two represent a single morphological species, A. binghami, with two subspecies, A. b. binghami and A. b. breviligula.

In this paper, we accept the species status of A. laboriosa. We use the names A. dorsata dorsata, A. d. breviligula and A. d. binghami for the other distinctive populations of giant honey bees because the species status of the latter two is still subject to investigation. We use the name “A. d. SouthIndia” to refer to a population that appears to be a cryptic unnamed species or subspecies (Smith, 1991; Kitnya et al., 2022). “Apis dorsata in the broad sense” will refer to all giant honey bees excluding A. laboriosa.

The objective of this study is to carry out a phylogenetic analysis for populations of giant honey bees, including representatives from as much of their range as we could obtain, to test whether the lineages within A. dorsata in the broad sense are monophyletic, and to determine relationships among them. Samples include A. laboriosa [Nepal], A. d. dorsata [multiple populations], and the distinctive island populations A. d. binghami [Sulawesi and smaller nearby islands] and A. d. breviligula [the oceanic islands of the Philippines]. We also include the dwarf honey bees, A. florea and A. andreniformis, and the cavity-nesting honey bees A. mellifera and A. cerana as outgroups. We generated partial sequences of the mitochondrial cytochrome c oxidase subunit 1 (cox1) and cytochrome c oxidase subunit 2 (cox2) genes and used Maximum Likelihood methods in MEGA7 to construct phylogenetic trees.

Methods

Field methods

Samples used in this study were collected by multiple researchers from 1989 to 2018 using a variety of collection and preservation techniques. Table 1 gives locality and collection information, and sample IDs corresponding to those used in Figure 1. Most specimens were collected directly from colonies, though some bees were collected while they were foraging. Most specimens are adult worker bees, while a few are pupae collected directly from nests. Individual bees or bee thoraces were preserved in the field in liquid nitrogen (1988–1990) or in 95% ethanol (1991 onwards). Frozen specimens were later stored at −80°C. Ethanol-preserved specimens were stored at 4° to −20°C.