Human-mediated dispersal of Geniotrigona thoracica (Apidae: Meliponini) colonies promotes high genetic diversity and reduces population structuring in managed populations

Introduction

Stingless bees, members of the tribe Meliponini, represent a diverse clade of eusocial Hymenoptera widely distributed across tropical and subtropical regions (Quezada-Euán, 2018; Wongsa et al., 2024). With approximately 600 described species, they exhibit considerable variation in morphological characteristics, colony structure, and foraging ecology (Hrncir & Maia-Silva, 2013; Quezada-Euán, 2018; Rattanawannee & Duangphakdee, 2019). As dominant pollinators in many tropical ecosystems, stingless bees contribute significantly to the reproductive success of both native flora and cultivated crops (Heard, 1999; Wongsa, Duangphakdee & Rattanawannee, 2023). Their ecological effectiveness is underpinned by traits such as floral constancy, perennial colony maintenance, reduced defensive behavior due to non-functional stingers, and efficient worker recruitment, all of which enhance their value as pollinators in natural and agroecosystem (Bartelli, Santos & Nogueira-Ferreira, 2014; Wongsa, Duangphakdee & Rattanawannee, 2023).

During the past two decades, the practice of stingless beekeeping, or meliponiculture, gained increasing traction in Thailand among both commercial and small-scale beekeepers, reflecting its dual role in promoting ecological sustainability and generating supplementary income (Rattanawannee & Duangphakdee, 2019). To date, at least 33 stingless bee species across 10 genera have been reported in Thailand, with Geniotrigona thoracica emerging as one of the most successfully managed species for commercial purposes (Rattanawannee & Duangphakdee, 2019; Wongsa, Duangphakdee & Rattanawannee, 2023). This species is well managed to standard wooden hive boxes and is increasingly utilized for pollination services in open-field cultivation of economically important crops. Beyond its pollination role, G. thoracica is also valued for its high-yield and high-value production of honey and propolis, as well as for the commercial trade of whole colonies (Rattanawannee & Duangphakdee, 2019).

In domestic markets, the honey produced by G. thoracica is typically sold at prices ranging from 1,000 to 1,200 Thai Baht (approximately 30–36 USD) per kilogram, substantially higher than that of honey derived from Apis mellifera, and at least twice the price of honey from native species such as A. cerana, A. dorsata, and A. florea (Rattanawannee & Duangphakdee, 2019). Notably, a premium product known as “emerald honey,” which is produced by G. thoracica from nectar predominantly collected from Melaleuca cajuputi in the southern provinces of Pattani and Narathiwat, can command prices as high as 6,000 Thai Baht (approximately 180 USD) per kilogram (A. Rattanawannee, pers. comm., 2024). Furthermore, fully provisioned G. thoracica colonies maintained in wooden hives are commercially valued between 6,000 and 8,000 Thai Baht (approximately 180–242 USD) per colony (A. Rattanawannee, pers. comm., 2024). With increasing consumer demand for high-quality stingless bee products, meliponiculture, particularly involving G. thoracica, holds substantial promise as a supplementary livelihood strategy for rural communities throughout southern Thailand.

The commercial exchange and relocation of stingless bee colonies facilitate their movement beyond native geographic boundaries (Chapman et al., 2018; Jaffé et al., 2016b). This anthropogenic activity has been associated with ecological concerns, particularly the risk of introducing non-native species into novel environments, which may negatively impact indigenous bee communities and disrupt local biodiversity and ecological functions (Beekman et al., 2008; Byatt et al., 2015; Inoue & Yokoyama, 2010; Kondo et al., 2009; Soland-Reckeweg et al., 2009). Moreover, translocating colonies across regions increases the likelihood of disseminating parasites and pathogens that can threaten both wild and managed bee populations (Byatt et al., 2015; Chapman et al., 2018; Lozier & Zayed, 2017; Meixner, Kryger & Costa, 2015; Oldroyd & Nanork, 2009). Understanding the genetic composition and population structure of stingless bees is therefore critical for developing sustainable management strategies (Koffler et al., 2017; Lozier & Zayed, 2017). Such genetic insights can inform domestication efforts and assist beekeepers in minimizing the risks of inbreeding and genetic erosion, which are common challenges in meliponiculture (Chapman et al., 2018).

The movement of bee colonies beyond their native hybrid zones and natural geographic boundaries can lead to genetic consequences such as hybridization and mating interference (Byatt et al., 2015). Hybridization involves the genetic exchange between previously reproductively isolated populations (Byatt et al., 2015), potentially resulting in the erosion of unique genetic lineages and the homogenization of distinct ecotypes (Frankham, Ballou & Briscoe, 2010). This process may ultimately lead to the loss of locally adapted genotypes through genomic swamping (Frankham, Ballou & Briscoe, 2010). Mating interference, on the other hand, arises from interspecific reproductive interactions that negatively impact reproductive success (Byatt et al., 2015; De La Rúa et al., 2009), such as reduced fertility (Byatt et al., 2015; Chapman et al., 2018; Koeniger & Koeniger, 2000; Remnant et al., 2014) or unsuccessful mating attempts, thereby diminishing the overall fitness of native bee populations (Groening & Hochkirch, 2008).

Previous research has shown that the extent of inbreeding and genetic differentiation among wild and managed stingless bee populations varies considerably across species and geographic regions (Chapman et al., 2018; Landaverde-González et al., 2017; Rattanawannee et al., 2020; Santiago et al., 2016). This variability underscores the importance of species and context-specific assessments (Chapman et al., 2018), as findings from one system may not be applicable to another. Importantly, empirical evidence suggests that human-mediated management practices may exert a stronger influence on population genetic structure than natural factors such as dispersal ability, habitat loss, elevation gradients, or climatic conditions (Greenleaf et al., 2007; Kükrer, Kence & Kence, 2021). These observations highlight the need for careful consideration of beekeeping interventions to avoid compromising genetic integrity and local adaptation in stingless bee populations (Jaffé et al., 2016b).

The genetic structure of a population is shaped by the dynamic interplay of evolutionary processes, including natural selection, genetic drift, gene flow, and mutation, all of which influence the distribution and frequency of genetic variation within and among populations (Bluher, Miller & Sheehan, 2020; Hartl & Clark, 1997). In the context of managed pollinators, these forces can be further modified by anthropogenic factors such as selective breeding, artificial colony propagation, and human-mediated dispersal. In this study, we examined the genetic architecture of G. thoracica (Apidae: Meliponini), a stingless bee species of economic and ecological importance that is extensively cultivated for honey production across various regions of Thailand. Our objectives were to determine whether managed populations of G. thoracica exhibit significant genetic differentiation from one another, and to evaluate whether they display elevated levels of inbreeding. By analyzing mitochondrial and microsatellite markers, this research provides critical insights into the extent to which meliponicultural practices influence the genetic diversity and structure of G. thoracica. These findings have direct implications for the sustainable management and conservation of stingless bees, particularly in the context of colony trade, domestication, and the maintenance of genetically healthy populations for long-term apicultural success.

Materials and Methods

Sampling and DNA extraction

Seventy adult worker bees were sampled from 17 meliponaries situated in the southern region of Thailand (Table S1 ; Fig. 1). Each of the bees represented one individual per colony. Collections were performed at the entrance tubes of each nest to ensure colony-specific sampling. All specimens were immediately preserved in absolute ethanol and stored at −20° C prior to laboratory analysis. The geographic coordinates of each sampling site were recorded using a GPS handheld device (Garmin eTrex 20X Handheld GPS). The thorax was used for genomic DNA extraction using a DNeasy^® Blood & Tissue kit (Qiagen, Hilden, Germany) following to the instructions of manufacturer.

Results

Discussion

Previous research has frequently linked artificial selection in managed breeding systems to elevated inbreeding and diminished genetic variation relative to wild progenitors (Bruford, Bradley & Luikart, 2003; Muir et al., 2008; Wang et al., 2014). In contrast, our data show that managed G. thoracica populations in Thailand sustain substantial mitochondrial and nuclear genetic diversity, despite exhibiting pronounced genetic differentiation. This pattern indicates that prevailing stingless bee colony management practices in Thailand exert negligible effects on the overall genetic variability of G. thoracica.

Over the last century, apicultural management has significantly shaped the distribution and genetic structure of social bees worldwide, including various species of honey bee, bumble bee, and stingless bee (Bryant & Krosch, 2016; Chahbar et al., 2013; Chapman et al., 2018; Francisco et al., 2014; Jaffé et al., 2016b; Jensen et al., 2005; Rangel et al., 2016). The stingless bee G. thoracica is a particularly valuable focal species for assessing these impacts, given the rapid expansion of hive trading within Southeast Asia, especially in Thailand and Malaysia (Rattanawannee & Duangphakdee, 2019). Anthropogenic hive translocation may provide genetic and adaptive benefits by increasing allelic diversity and facilitating responses to environmental pressures (Chapman et al., 2018; Todesco et al., 2016; Wongsa et al., 2024); however, it can also generate maladaptive hybrids when reproductive barriers exist, and may result in the loss of regionally adapted genotypes (Byatt et al., 2015; Todesco et al., 2016; Wongsa et al., 2024). In southern Thailand, a major center of stingless beekeeping for over two decades, G. thoracica populations now exhibit genetic patterns indicative of admixture from multiple geographic origins. While the genetic changes are apparent, their phenotypic implications remain uncertain. Notably, in stingless bees, male attendance at mating aggregations is not always linked to hybridization (Law et al., 2024), as illustrated by Tetragonula carbonaria males that visit T. hockingsi aggregations without exhibiting short-range attraction to the latter’s queens (Paul et al., 2023).

This study found no evidence that geographical or physical barriers, such as mountain ranges, urban or agricultural landscapes, or forest cover, significantly influence the population structure of Thai G. thoracica. Similar patterns, gene flow occurs when the absence of dispersal barriers, have been documented in Trigona nigerrima, Trigona corvina, and Scaptotrigona mexicana in Mexico (Rodríguez et al., 2024; Solórzano-Gordillo et al., 2015), Tetragonula carbonaria and Tetragonula hockingsi in Australia (Brito et al., 2014; Law et al., 2024), Trigona spinipes in Brazil (Jaffé et al., 2016a) and Heterotrigona itama in Thailand (Wongsa et al., 2024). The observed structuring of G. thoracica populations is more likely driven by the inherently low dispersal capacity of virgin queens, together with ecological variation among local habitats.

Analyses of population genetic structure indicated that several G. thoracica populations exhibited limited differentiation from geographically distant groups. Such genetic homogeneity is likely maintained through ongoing gene flow, potentially consistent with a stepping-stone dispersal process (Kimura & Weiss, 1964). Anthropogenic factors, particularly the deliberate relocation of colonies by beekeepers, appear to further reinforce interpopulation connectivity. This inference is supported by Bayesian phylogenetic analyses of concatenated mitochondrial COI and 16S rRNA sequences, which grouped 68 of the sampled colonies into a single well-supported clade, with only two colonies forming a separate lineage (Fig. 1A). Moreover, the most common haplotypes, C5 (COI) and S3 (16S rRNA), were shared across all sampling locations (Figs. 1C and 1D), suggesting extensive haplotype mixing among regions. These results suggest that both natural dispersal and anthropogenic colony translocation are important drivers of population structure in this stingless bee species. The elevated occurrence of unique haplotypes in Yala and Chumphon further suggests that colonies may have been introduced from other regions, artificially enhancing local genetic diversity.

Across most comparisons, genetic diversity metrics did not differ significantly among groups (Tables 2 and 5). Consistent with our expectations, relatively high levels of genetic diversity, as measured by expected heterozygosity, were detected in all provinces of southern Thailand, where G. thoracica colonies are predominantly managed. This pattern was further supported by the elevated values of both expected heterozygosity (H_e) and allelic richness (N_e) observed in all managed apiaries (Table 2). The enhanced diversity in managed colonies is likely the consequence of admixture over time, driven by the exchange of colonies among beekeepers from different localities, which introduces novel alleles into populations (Carvalho-Zilse et al., 2009; Chapman et al., 2018; Wongsa et al., 2024). This inference is reinforced by pairwise per-generation migration rate (Nm) estimates, all of which exceeded three (ranging from 3.60 to 207.83), indicating substantial queen dispersal among populations (Table 3). Specifically, the data suggest that more than three reproductive queens per generation are exchanged between each pair of populations. In line with these findings, low genetic differentiation (F_st) values were observed among geographic localities, and AMOVA results for mitochondrial markers revealed no clear geographic partitioning of genetic variation in G. thoracica (Table 5). This contrasts with the study of Rattanawannee et al. (2017), which identified two distinct genetic groups of the stingless bee Tetragonilla collina in Thailand using geometric morphometric and mitochondrial COI sequence analyses. They proposed that, for this subterranean-nesting species, present-day ecological factors, such as seasonal flooding, exert a stronger influence on spatial distribution than historical biogeography.

Although comparative studies between wild and managed stingless bee populations remain scarce, previous work on Tetragonisca angustula (Santiago et al., 2016) and Heterotrigona itama (Wongsa et al., 2024) reported no detectable differences in genetic diversity between the two management types. In the present study, nearly all F_is values across the defined genetic groups were positive yet statistically non-significant (Table 2). Notably, the absence of significant F_is values in managed colonies was unexpected, as such conditions could be indicative of elevated relatedness among colonies within an apiary, a pattern that may arise from colony propagation practices by beekeepers, as previously proposed (Santiago et al., 2016). In contrast, the Yala population exhibited positive and significant F_is values (Table 2), suggestive of inbreeding, potentially attributable to habitat loss and landscape alterations that may reduce effective population sizes, increase genetic relatedness, and diminish genetic diversity (Lozier & Zayed, 2017). To improve understanding of genetic diversity and inbreeding dynamics in G. thoracica, broader sampling efforts are required, both in terms of the number of colonies and the range of localities represented. In certain localities, only a single colony was sampled, limiting the precision of diversity estimates. Expanding sample sizes would enable more robust statistical inferences and facilitate the assessment of whether habitat degradation is exerting a negative influence on the genetic diversity of G. thoracica.

Bayesian clustering of nuclear genotypes in STRUCTURE supported a single genetic cluster (K = 1), whereas AMOVA of concatenated mitochondrial sequences revealed moderate-to-high population differentiation (F_st = 0.619). This discrepancy exemplifies mito-nuclear discordance, a pattern frequently associated with sex-biased dispersal. In stingless bees, colony founding by queens typically occurs through short-range budding events, averaging approximately 700 m from the natal nest, while males may disperse up to 20 km prior to mating (Bueno et al., 2022). Such asymmetry in dispersal capacity can generate stronger genetic structuring in maternally inherited mitochondrial DNA compared to biparentally inherited nuclear loci (Law et al., 2024; Peters et al., 1999; Quezada-Euán, 2018; Quezada-Euán et al., 2022). Comparable trends have been reported in S. mexicana, where male-biased dispersal has been invoked to explain genetic admixture within drone aggregations (Rodríguez et al., 2024). In stingless bee, drones depart their natal colonies to join “drone congregations” situated near nests with virgin queens (Quezada-Euán, 2018), which may contain several hundred individuals originating from multiple and often geographically distant colonies (Dos Santos et al., 2016; Kraus, Weinhold & Moritz, 2008; Mueller, Moritz & Kraus, 2012). Although meliponine drones generally exhibit shorter effective dispersal ranges than their honey bee (Apis spp.) counterparts (Kraus et al., 2005; Oldroyd et al., 1998; Oldroyd & Wongsiri, 2006), the low genetic differentiation and minimal pairwise genetic distances observed between Chumphon and Narathiwat suggest that ongoing male-mediated dispersal likely contributes to gene flow between these populations (Table 3; Fig. 1). These results indicate that the dispersal distance of stingless-bee drones may be greater than assumed and should be re-evaluated.

In commercial meliponiculture, artificial colony division is commonly employed to increase colony numbers within an apiary. This practice involves transferring a combination of young and old brood combs, together with honey and pollen pots, from a strong donor colony into a new hive box, thereby establishing a daughter colony (Quezada-Euán, 2018; Santiago et al., 2016). Such management interventions can alter the distribution of mitochondrial haplotypes within a population, with some haplotypes increasing in frequency while others decline or disappear entirely (Santiago et al., 2016). In the present study, this pattern was evident in the Chumporn and Pattani populations, which exhibited pronounced population structuring and a reduced number of haplotypes dominated by a few high-frequency variants (Table 4; Fig. 1). Because mitochondrial haplotypes are maternally inherited, they may be transferred between populations if a colony from one source successfully establishes as a daughter colony within another population (Chapman et al., 2018), thereby contributing to mitochondrial structuring (Francisco et al., 2014).

In natural populations, elevated mitochondrial structure has often been attributed to the short dispersal range of reproductive swarms, a phenomenon reflecting female queen philopatry. This behavior arises because daughter colonies require immediate access to resources, such as propolis and food, provided by the maternal nest to initiate construction of a new hive (Inoue et al., 1984). As a result, the limited dispersal of queens constrains gene flow and reinforces population structure (Santiago et al., 2016). In managed settings as in our study, repeated colony division from a restricted pool of source colonies within an apiary can produce a genetic pattern analogous to that generated by queen philopatry (Santiago et al., 2016). High levels of mitochondrial structuring have similarly been documented in wild populations of multiple stingless bee species, including Melipona beecheii (Quezada-Euán, 2018; Quezada-Euán et al., 2007), Partamona helleri (Brito & Arias, 2010), Plebeia remota (Francisco & Arias, 2010; Francisco, Santiago & Arias, 2013), Tetragonula pagdeni (Thummajitsakul, Klinbunga & Sittipraneed, 2011), Scaptotrigona hellwegeri (Quezada-Euán et al., 2012), Partamona mulata (Brito et al., 2013), Melipona subnitida (Bonatti et al., 2014), and Tetragonisca angustula (Francisco et al., 2017). Additionally, the Australasian clade of stingless bees was recently found to possess an unusually complex mitochondrial genome structure characterized by gene duplications, which appear to contribute to an elevated substitution rate. This phenomenon was first reported in two Australian species (Françoso et al., 2023) and was later identified as a shared feature, to varying degrees, across the entire clade (Li et al., 2024). The implications of this atypical mitochondrial architecture for population genetic analyses remain poorly understood; however, an accelerated rate of mitochondrial evolution could potentially intensify patterns of population structuring observed in mtDNA.

Conclusion

In conclusion, our results indicate that most G. thoracica populations in Thailand exhibit substantial genetic differentiation. While current levels of colony trade and translocation appear not to have disrupted population structure, an escalation of such practices among genetically distinct populations could pose adverse genetic consequences. Preserving the integrity of local gene pools thus requires minimizing genetic admixture. We therefore recommend conducting targeted genetic assessments prior to the introduction of new ecotypes, and ensuring that colony transfers are restricted to populations with demonstrable genetic similarity.

Supplemental Information