Never underestimate honeybees: Their exceptionally diverse adaptive strategies

Graphical abstract

Highlights

• •Honeybees are susceptible to the external environment.
• •Honeybees employ diverse adaptive strategies to their environment.
• •Brain functions and sensory systems help honeybees survive.
• •Communication and division of labor are key to maintaining honeybee population stability.

Abstract

Background

As major pollinators in terrestrial ecosystems, honeybees are highly sensitive to environmental changes. To survive, they have evolved diverse adaptive strategies to cope with complex and fluctuating conditions. Considerable progress has been made in elucidating the mechanisms of these adaptations, which involve brain function, communication, division of labor, and sensory systems.

Aim of review

Although numerous studies have investigated honeybee environmental adaptation, most remain confined to isolated disciplinary perspectives. To date, no systematic review has comprehensively examined the multidimensional synergy network underlying honeybee adaptation. In this sense, an updated synthesis of honeybee adaptive strategies that incorporate different knowledge areas is still lacking. Our review aims to develop a global synthesis of diverse honeybee adaptive strategies, which encompass brain function, communication, division of labor and sensory systems. We highlight the structure, development, function, and regulatory factors of the brain and its pesticide response strategies, and summarize the classification, roles, and key regulatory factors involved in communication, division of labor, and sensory systems in honeybees. They do not function independently, and their synergy is central to the environmental adaptation of honeybees. Finally, we analyze current research limitations and propose future directions.

Key scientific concepts of review

Brain function, communication, division of labor, sensory systems, and their synergistic interactions collectively underpin the adaptive strategies that enable honeybees to survive in complex environments. A comprehensive understanding of these molecular and structural mechanisms is essential to ensure honeybee fitness and to support resilience-oriented breeding programs.

Introduction

Bees (Hymenoptera: Apoidea: Anthophila) constitute about 20,000 described species across seven families: Melittidae, Apidae, Megachilidae, Andrenidae, Halictidae, Stenotritidae, and Colletidae [1,2]. Nearly 21, 200, 2,500, 2,900, 4,000, 4,300, and 5,700 described species have been recorded in Stenotritidae, Melittidae, Colletidae, Andrenidae, Megachilidae, Halictidae, and Apidae, respectively. In addition, Apidae contains the vital pollinator honeybee [2]. The vital role of honeybees in maintaining terrestrial ecosystems and supporting the pollination of both wild plants and agricultural crops is well established. They also produce economically valuable products such as honey, pollen, royal jelly, and propolis [3,4]. Given their ecological and agricultural significance, investigating the effects of the environment on the survival of honeybees and their ability to adapt to the environment has become an important research focus.

The resilience of honeybees in diverse environments stems from adaptive strategies involving advanced brain function, efficient communication, division of labor, and specialized sensory systems. Honeybee brains support sophisticated learning and memory, enabling navigation, foraging, and individual recognition [[5], [6], [7], [8]]. Communication through pheromones and behavioral signals allows rapid and accurate information transfer about resource location [9,10]. The division of labor enables individuals to assume roles that optimize colony efficiency and survival, with colonies capable of dynamically adjusting role allocation to environmental conditions [11]. Sensory systems permit precise perceptions of direction, odor, and temperature, facilitating accurate flower recognition and environmental assessment [12,13]. Together, these traits enhance survival and adaptability.

Although reviews have examined the environmental effects on honeybees [[14], [15], [16], [17], [18], [19], [20]], few have synthesized adaptive strategies from a comprehensive and integrative perspective. This review highlights how honeybee brain function, communication, division of labor, and sensory abilities promote environmental adaptation. We also discuss the molecular mechanisms underlying these traits and outline potential strategies for resilience breeding.

The literature reviewed was identified using NCBI PubMed, Baidu Scholar, and Google Scholar. Keywords included “honeybee brain function”, “communication”, “division of labor”, “sensory system”, “strategies”, and “environmental adaptation”. Relevant publications were downloaded, categorized, critically assessed, and compared to extract overarching themes. Studies with unclear mechanisms were also considered, and their limitations and prospects for future research are discussed.

Honeybee brains and related regulatory mechanisms

The human brain contains about 86 billion neurons [21], whereas the honeybee brain has about 950,000 neurons and a volume of ∼ 1 mm³. Despite its small size, the honeybee brain supports complex behaviors, including foraging, navigation, and decision-making, which are crucial for colony success [[5], [6], [7], [8]]. The key features of honeybee brains include multifunctionality, heterochronic development between workers and queens, caste-specific gene expression, and high learning and memory capacity [[22], [23], [24], [25]]. In the following sections, we summarize the adaptive strategies of honeybee brains and their regulatory mechanisms, with a focus on structure, learning, memory, and resilience to complex environments.

Brain of honeybees

The adult honeybee brain has a complex structure comprising three main regions: the protocerebrum, deutocerebrum, and tritocerebrum. The protocerebrum contains two optic lobes, the central body, and the mushroom bodies (MBs). The MB is the complex center of the nervous system, is responsible for processing olfactory information and is involved in learning, memory, and other cognitive functions. The deutocerebrum includes the antennal lobes, whereas the tritocerebrum consists of bilateral lobes that innervate the digestive tract and the labrum [26,27].

Brain development is heterochronous between workers and queens. During the larval stage, queens develop larger and faster-growing brains than workers [22,28]. In contrast, adult workers have larger brains and more developed cognitive abilities than queens [22,29]. This caste-specific divergence supports behavioral specialization, enabling workers to perform tasks such as nursing, foraging, colony defense, and nest maintenance more effectively.

A comprehensive understanding of the regulatory network underlying brain development is essential for understanding honeybee adaptability. Considerable progress has been made, particularly in uncovering brain dimorphism between workers and queens. Caste-specific gene expression drives this dimorphism, with 53 differentially expressed genes (DEGs) identified between worker and queen brains during the third larval stage (L3) and 1760 DEGs identified during the pupal stage [30,31]. Genes such as Osiris 17, Takeout, Alpha-methyldopa hypersensitive protein (Amd), Forked, Neurotrophin-1 (Nt-1), and Masquerade (Mas) are upregulated in the L3 queen brain, while chymotrypsin inhibitor and members of the hexamerin family (hex110, hex70c, and hex70b) are upregulated in the L3 worker brain (Fig. 1A) [30]. These genes play key roles in caste-specific brain development.

Fig. 1.

Factors regulating heterochronic brain development in honeybees. Caste-specific transcription profiles (A), interactions between miRNAs and genes (B), nutrition (C), and cell death (D) govern heterochronic brain development. L3: third larval stage. MB: mushroom body.

Recent studies have identified additional regulators. The minibrain (mnb) gene is critical: RNA interference (RNAi) in workers reduces brain volume and size, along with smaller MB and optic lobes (Fig. 1A) [32]. Similarly, knockout of the Pax6 gene toy results in worker brain atrophy and microcephaly, underscoring its role in worker brain development. However, whether toy is also implicated in caste-specific dimorphism remains unclear. Despite these advances, the molecular basis of honeybee brain development remains incompletely understood. Emerging methods such as single-cell omics and spatial transcriptomics hold promise for mapping spatiotemporal gene expression and characterizing cell populations involved in brain development, thereby advancing the understanding of regulatory mechanisms underlying brain dimorphism.

MicroRNAs (miRNAs) regulate gene expression by binding to 3′-untranslated regions of mRNAs, thereby influencing brain function and development [33,34]. Studies have reported that miRNA–mRNA interactions are also associated with caste-specific brain development in honeybees. Compared with those of workers, the L3 queen brains presented elevated expression of miRNA-34, miRNA-210, and miRNA-317 Negative regulatory effects have been demonstrated between miRNA-34 and hex70b, hex110, Takeout, and Nt-1; between miRNA-210 and hex70b, hex110, and Mas; and between miRNA-317 and Forked and Takeout (Fig. 1B) [30,35]. These interactions may contribute to asynchronous brain development between workers and queens. However, a complete miRNA–mRNA regulatory network has yet to be established, and further studies are needed to clarify these mechanisms.

Nutrition and programmed cell death also influence brain dimorphism. Queens receive greater nutritional input, particularly royal jelly, during early development, promoting accelerated brain growth (Fig. 1C) [30]. Moreover, extensive cell death occurs in queen brains during late pupal stages, which may also account for the size differences between adult queen and worker brains (Fig. 1D) [28,[36], [37], [38]]. Thus, caste-specific transcription profiles, miRNA–mRNA interactions, nutrition, and cell death collectively govern heterochronic brain development. However, the molecular mechanisms underlying brain development in drones and other bee species remain poorly understood and require further investigation.

Despite its small size, the honeybee brain plays a pivotal role in coping with pesticide stress (Fig. 2 and Table 1). For example, 131 DEGs have been identified in imidacloprid-treated brains, 130 of which are downregulated and associated with chemosensory responses, detoxification, and immunity [39]. Fluvalinate treatment alters the expression of 546 genes enriched in pathways related to the neural response, visual transduction, amino acid metabolism, and lipid metabolism [40]. In addition, 6 circular RNAs (circRNAs) involved in post-transcriptional regulation, potentially affecting neuronal apoptosis through the mTOR, Wnt, Hedgehog, and apoptosis pathways, and 15 miRNAs associated with dorsoventral axis formation, apoptosis, phototransduction, and Hippo and Wnt signaling are differentially expressed following fluvalinate exposure [41,42]. Dinotefuran treatment also leads to DEGs enriched in transforming growth factor-β signaling, and protein metabolism; differentially expressed circRNAs associated with neuroactive ligand–receptor interactions, and Wnt and Hippo signaling; differentially expressed miRNAs involved in immunity, behavior, and development; and differentially expressed long non-coding RNAs (lncRNAs) linked to the immune response and growth [[43], [44], [45]].

Fig. 2.

Response of the honeybee brain to pesticide exposure. Many differentially expressed genes, miRNAs, circRNAs, and lncRNAs play important roles in enhancing the survival ability of honeybees under imidacloprid, fluvalinate, and dinotefuran stress.

Table 1.

Differentially expressed (DE) genes and non-coding RNAs (ncRNA: miRNA, circRNA and lncRNA) play important roles in honeybee brain responses to imidacloprid, fluvalinate, and dinotefuran stress.

Pesticides	DE gene/ ncRNA	The main pathways or roles that involved	Reference
Imidacloprid	Genes	Chemosensory responses, detoxification, and immunity	[39]
Fluvalinate	Genes	Neural response, visual transduction, amino acid metabolism, and lipid metabolism	[40]
	circRNA	The regulation of brain nerve apoptosis	[41]
	miRNA	Dorsoventral axis formation, apoptosis, phototransduction, and Hippo and Wnt signaling	[42]
Dinotefuran	Genes	Transforming growth factor-β signaling, and protein metabolism	[43]
	circRNA	Neuroactive ligand–receptor interactions and Wnt and Hippo signaling	[45]
	miRNA	Immunity, behavior, and development	[44]
	lncRNA	Immune response and growth	[43]

Functionally, imidacloprid reduces the number of MB synapses and impairs learning [46,47]; fluvalinate negatively affects memory, learning, survival, and sucrose responsiveness [48,49]; and dinotefuran decreases the olfactory learning and homing flight ability of honeybees [49,50]. Further analysis revealed that differentially expressed genes and non-coding RNAs (ncRNAs: miRNAs, circRNAs, lncRNAs) in the brain may mitigate the harmful effects of pesticide stress on honeybees through combined actions involving detoxification, the oxidative stress response, and immune pathways. However, the differential expression of genes and ncRNAs was obtained through RNA sequencing. Their specific roles and links to adaptive outcomes in honeybees require further investigation. Overall, the differential expression of genes and ncRNAs may represent compensatory mechanisms that increase survival under pesticide stress (Fig. 2). However, most studies have focused on a limited set of pesticides, primarily fluvalinate, imidacloprid, and dinotefuran. Broader investigations are needed to elucidate brain defense mechanisms against diverse chemical stressors in the environment.

Learning ability

Honeybee learning is a neural capacity that underpins a wide range of complex behaviors, including navigation, foraging decisions, and individual recognition, all of which contribute to survival and adaptability. The honeybee brain supports multiple forms of learning, including associative, non-associative, social, color, aversive, rule-based, olfactory, gustatory, and appetitive learning [8,[51], [52], [53]].

Learning ability varies significantly among individuals, ages, castes, and species [[54], [55], [56], [57], [58]]. At the individual level, differences in responsiveness to nectar and pollen rewards generate interindividual variation in learning, supporting foraging specialization [55]. At the age level, associative learning in young honeybees can be enhanced by increasing reward sensitivity, whereas older foragers show reduced responsiveness [56]. At the caste level, queens outperform workers in olfactory learning, likely because of their diet of royal jelly (Table 2) [57]. At the species level, Apis cerana exhibits superior learning of grating and color patterns but shows lower learning scores during the acquisition phase than Apis mellifera does, suggesting slower learning [58,59]. Compared with stingless bees (Melipona rufiventris), A. mellifera more rapidly learns random associations between patterns and colors (Table 2) [60]. These findings indicate that individual, age, caste, and species differences each influence learning ability. Further comparative studies across species and castes, particularly those involving specialized or complex tasks, are needed to clarify the evolutionary and adaptive basis of learning in bees.

Table 2.

Caste and species traits related to honeybee learning.

Caste or species	Bees	Learning content	Comparison of learning abilities	Reference
Caste	Young honeybee	Olfactory learning	Better	[57]
	Older foragers		Worse
Species	Apis cerana	Grating and color patterns learning	Better	[58]
	Apis mellifera		Worse
	Apis cerana	Learning during acquisition phase	Worse	[59]
	Apis mellifera		Better
	Apis mellifera	Random relationships between patterns and colors	Better	[60]
	Melipona rufiventris		Worse

Honeybee learning is regulated by genes. During olfactory learning in honeybees, gene expression is asymmetrical, with more genes expressed at lower levels in the left hemisphere than in the right hemisphere, suggesting that the right hemisphere plays a dominant role [25]. In failed honeybee learners, most downregulated DEGs are associated with brain development, neurodegeneration, and sensory processing, potentially exacerbating learning deficits [61]. RNAi experiments have shown that the elav/Hu family gene (elavl2) is essential for learning in honeybees [62]. In addition, DNA methyltransferases (Dnmts) are related to the regulation of inhibitory and excitatory processes in honeybee reversal learning [63]. These findings highlight the critical role of genes in regulating learning. However, additional genetic factors remain to be identified to fully elucidate the molecular basis of honeybee learning.

In addition to gene regulation, the gut microbiota, DNA methylation, dopamine, and octopamine positively affect on honeybee learning (Fig. 3A). A normal gut microbiota is required for olfactory learning, and Gilliamella apicola has been shown to enhance reward-based learning in honeybees [64,65]. Dopamine can primarily regulate the learning process that links scents to toxins, and enhance aversive learning in honeybees [66,67]. Octopamine promotes appetitive learning and enhances learning ability in newly emerged workers [56,68]. Downregulation of octopamine receptors via RNAi impairs honeybee learning, further confirming its role [69]. DNA methylation, which is mediated by Dnmts plays an important role in honeybee relearning [63,70,71]. Despite these advances, further studies are needed to identify additional modulatory factors that contribute to learning and enhancing honeybee resilience.

Fig. 3.

Factors that affect the learning ability of honeybees. (A) The gut microbiota, DNA methylation, dopamine, and octopamine positively affect honeybee learning. (B) Different pesticides, urban air pollution, Nosema ceranae, deformed wing virus (DWV), cadmium, lead, heat, and cold have negative effects on honeybee learning.

Learning ability can be disrupted by environmental stressors, including pesticides, heavy metals, parasites, viruses, urban air pollution, and temperature extremes (Fig. 3B). Thiamethoxam impairs differential and visual learning in foragers [72], whereas deltamethrin significantly reduces olfactory associative learning [73]. Deformed wing virus (DWV) also impairs associative learning [74]. Cadmium, urban air pollution, and cholinergic pesticides adversely affect olfactory learning [[75], [76], [77]], and the ability of cadmium to impair olfactory learning is likely caused by the dysregulation of the expression of genes encoding cytochrome P450 enzymes, olfactory proteins, and chemoreceptors and the generation of oxidative stress [75]. Additional stressors, including Nosema ceranae, lead, heat, and cold, also impair the learning of honeybees [[78], [79], [80], [81]]. Collectively, these findings demonstrate that multiple environmental factors negatively affect honeybee learning. However, the underlying defense mechanisms remain poorly understood and require urgent investigation to mitigate these impacts.

Memory ability

Honeybees possess highly developed memory abilities, including short-term, mid-term, and long-term memory (STM, MTM, and LTM, respectively), which are essential for survival and resource acquisition. Notably, the right hemisphere plays a greater role in STM, whereas the left hemisphere is more responsible for LTM during olfactory learning [25]. However, the specific brain regions responsible for STM, MTM, and LTM remain to be fully identified.

The memory ability of honeybees differs across species. For example, A. mellifera workers exhibit superior MTM and memory retrieval capacity but perform less effectively than A. cerana workers do in terms of memory performance involving grating and color patterns [58,82]. Similarly, A. mellifera meda drones exhibit greater odor memory recall than A. cerana indica and A. mellifera carnica drones do [83]. These interspecific differences suggest that memory variation may reflect evolutionary strategies for adaptation to diverse environments. Comparative analyses of species- and caste-specific memory will contribute to clarify the relationships among genes, neural structures, and cognition, thereby deepening our understanding of the evolution of memory in social insects.

Memory types differ in duration and function in honeybees. STM persists for seconds to minutes without requiring transcription or translation. MTM (minutes to hours) and LTM (hours to days) require transcription and translation, although the maximum memory duration remains uncertain [23]. Typically, a single learning trial induces STM and MTM, whereas repeated trials are required for MTM, early LTM (e-LTM, ∼24 h, translation-dependent), and late LTM (l-LTM, days later, dependent on both transcription and translation) [23,84,85]. However, Villar et al. (2020) challenged this paradigm, showing that translation-dependent memory can occur within 4 h of a single trial and persist for up to 3 days [86]. These differences underpin task specialization. STM is well suited for within-patch foraging, which lasts only seconds during flower-to-flower flights. In contrast, MTM and LTM are more suitable for identifying pollen- and nectar-rich floral resources and remembering rewarding flower locations [23]. Thus, distinct memory types represent evolutionary adaptations that increase survival under various environmental conditions.

The formation and consolidation of honeybee memory are tightly regulated by transcription and influenced by beneficial and harmful factors. Several genes play key roles in honeybee memory, including elavl2, calcium/calmodulin-dependent kinase II (CaMKII), protein kinase A (PKA), NMDA glutamate receptor (NR1), and Dnmts (Fig. 4A) [[87], [88], [89]]. The elavl2 is essential for memory consolidation and undergoes alternative splicing as a possible adaptive mechanism in A. mellifera [62]. CaMKII functions as a molecular “memory switch”, critical for LTM formation [87,90]. PKA is necessary for LTM formation during olfactory learning [88]. NR1 contributes to memory formation [89]. Dnmts regulate memory specificity: enhancing memory specificity during high experience is used in multiple-trial training but reducing memory specificity during low experience is employed in one-trial training [63]. Together, these genes highlight the central role of transcriptional control in memory processes.

Fig. 4.

Factors that influence the memory ability of honeybees. Honeybee memory is regulated by genes (A) and miRNAs (B), and influenced by beneficial (C) and harmful factors (D). LTM: long-term memory. STM: short-term memory. JH: juvenile hormone. DWV: deformed wing virus.

The miRNAs also contribute significantly to memory regulation in honeybees (Fig. 4B) [[91], [92], [93]]. For example, miRNA-124 and miRNA-12 are conducive to defining the early phases of transient and long-lasting memory in honeybees [91], whereas miRNA-932 is essential for LTM, with its loss impairing memory formation [92]. Despite progress, the interplay between genes and miRNAs remains incompletely understood. Future research into miRNA–gene regulatory networks will clarify the molecular mechanisms underlying honeybee memory.

Several endogenous and environmental factors enhance honeybee memory (Fig. 4C). The gut microbiota plays a major role: Lactobacillus strains improve honeybee memory by converting tryptophan into indole derivatives that activate the host aryl hydrocarbon receptor [64]. Short neuropeptide F and floral scents improve visual memory [94,95]. Nutritional content and Ca²⁺ are conducive to LTM [96,97]. Juvenile hormone (JH), floral tea polyphenols, protein degradation inhibitors, octopamine, and DNA methylation facilitate STM, memory retention, memory formation, olfactory memory, and stimulus-specific associative LTM, respectively [69,[98], [99], [100], [101]]. Sleep also supports memory consolidation, whereas deprivation impairs extinction memory [[102], [103], [104]]. Collectively, these findings emphasize the variety of beneficial factors that reinforce honeybee memory and thereby enhance adaptability.

In contrast, numerous harmful factors including pesticides, pollutants, pathogens, and toxins, impair honeybee memory (Fig. 4D). For example, pesticides such as thiacloprid, coumaphos, acetamiprid, clothianidin, fipronil and imidacloprid impair memory performance, STM, memory retention of learned tasks, memory processing, olfactory memory, and memory of learned colors of honeybees, respectively [76,[105], [106], [107], [108], [109]]. The effects of different acaricides on honeybee memory vary, and the effects of formic acid (a widely used acaricide) are particularly damaging [110]. The DWV disrupts memory formation [74], whereas Nosema ceranae impairs odor memory [74,79]. In addition, the air pollutants, nicotine and morphine have negative effects on odor memory, associative memory formation, and LTM consolidation, respectively [77,111,112]. These findings demonstrate that honeybee memory is highly vulnerable to diverse environmental stresses. How honeybees resist or mitigate such impairments is a pressing area for future research.

Intraspecific communication among honeybees

Honeybees are highly social insects, and colony survival depends on sophisticated communication networks that ensure coordination and organization. Effective communication is thus critical to colony prosperity. The communication of honeybees is remarkably rich and diverse, relying primarily on chemical and behavioral signals [9,10,113].

Chemical signal

In honeybees, chemical communication—primarily through pheromones and volatile disease markers (VDMs)—enables the transfer of information essential for coordinating tasks such as brood care, nest defense, and foraging [9,113]. Honeybee pheromones include brood, worker, queen, and drone pheromones. While brood, worker, and queen pheromones are relatively well studied, drone pheromones remain largely unexplored. VDMs, on the other hand, are produced mainly by diseased honeybees.

During the larval stage, honeybee larvae release a blend of fatty acid esters known as brood pheromones, which regulate the behavior and physiology of nearby worker to support brood rearing (Fig. 5A). Brood pheromones can inhibit ovary development, regulate behavioral development, influence foraging, and attract workers to nursing cells [[114], [115], [116]]. Their volatile components are critical for function: starved larvae release volatiles that increase worker visitation to brood cells [117]. Experiments have identified E-β-ocimene as a key volatile component, with gene expression analyses linking aatc-like, longitudinals lacking protein-like, and farnesyl pyrophosphate synthase to its production under starvation conditions [117,118]. Thus, E-β-ocimene appears to be a crucial brood pheromone component that helps larvae signal hunger and increases worker visits to brood cells. However, other volatile components that are involved in ovary inhibition, behavioral regulation, and foraging modulation remain to be characterized.

Fig. 5.

Intraspecific communication among honeybees relies on chemical signal (A) and behavioral signal (B). (A) The chemical signal of honeybees includes pheromones (brood pheromones, worker pheromones, and queen pheromones) and volatile disease markers (VDMs). 9-ODA: 9-oxo-2-decenoic acid. HOB: methyl p-hydroxybenzoate. QRP: queen retinue pheromone. QMP: queen mandibular pheromone. (B) The behavior signal of honeybees contains tremble dances, waggle dances, stop signals, and shaking signals.

In adults, pheromones regulate communication and colony organization. The queen produces complex pheromones, including the queen mandibular pheromone (QMP) and queen retinue pheromone (QRP) (Fig. 5A), both of which vary across species and with queen age [9,[119], [120], [121], [122]]. These pheromones enable queens to regulate colony behavior, suppress worker reproduction, and attract drones without direct aggression [[123], [124], [125]]. Worker messaging and queen movement are critical for queen pheromone dispersal [126]. The QMP component, 9-oxo-2-decenoic acid (9-ODA), released by virgin queens, attracts drones [127,128]. However, in mated queens, methyl p-hydroxybenzoate (HOB), another QMP component, reduces drone attraction to 9-ODA, thereby preventing additional mating [129]. In contrast, the QRP mainly recruits workers for grooming and feeding. Collectively, these findings indicate that queen pheromones are vital for colony regulation and stability.

Worker pheromones, including ethyl oleate, the Nasonov pheromone, 2-heptanone, alarm pheromones, pentacosane, tricosane, and Z-(9)-tricosene, influence worker activity and intraspecific interactions (Fig. 5A). The roles of different worker pheromones differ. For example, ethyl oleate provides a foraging inhibition signal [130], whereas the Nasonov pheromone appears to be an attractive signal for entry into a new nest, colony recognition or social cohesion inside the hive. In contrast to the Nasonov pheromone, 2-heptanone causes a defensive response and repulsive behavior. Alarm pheromones can be used to communicate danger to honeybee colonies, allowing workers to prepare for further attack and defense. Furthermore, pentacosane, tricosane and Z-(9)-tricosene can clearly increase the number of foraging flights, thereby increasing foraging activity [131,132]. These pheromones collectively sustain colony coordination and task organization. Additionally, as yet unidentified, worker pheromones may further expand our understanding of communication within colonies.

VDMs also serve as chemical signals that facilitate disease management. Workers detect VDMs to identify sick broods and trigger hygienic behaviors, such as uncapping and removing diseased larvae, thereby reducing pathogen spread (Fig. 5A) [113,133,134]. For example, ethyl phenethyl acetate is a key compound of VDMs that induces hygienic behavior when workers suffer from chalkbrood [134]. The 2,5-dimethylpyrazine is the most sensitive VDM, and it triggers hygienic behavior when workers sustain American foulbrood [133]. Multiple diseases are significant factors that threaten the survival of honeybees [15,20,135], and thus, identifying VDMs that are associated with other diseases would provide valuable tools for improving honeybee health and colony survival.

Behavioral signal

In addition to chemical communication, behavioral communication—including tremble dances, waggle dances, shaking signals, and stop signals—plays a critical role in transmitting information about food sources, direction, distance, and colony needs (Fig. 5B) [136]. Notably, honeybee dances occur in complete darkness within the hive and rely on multimodal cues such as olfaction, taste, tactile cues, vibrations, and sound [[137], [138], [139], [140]]. By following dances, recruits acquire both a location vector and directional instruction, enabling efficient exploitation of resources within familiar territory [141].

Tremble dance. In foragers, tremble dances involve rapid body oscillations along the central axis while moving slowly through the hive, often changing direction [142]. They regulate nectar processing and convey foraging conditions [10]. Foragers may perform this dance after a long unloading delay [142,143]. Various factors, such as the collection of salty nectar, the experience of crowded feeders, capture at a food source, and the inability of honeybees to find unloading nectar upon arrival, can trigger unloading delays [143,144]. In addition to long unloading delays, crowded conditions at a food source also lead to tremble dancing in honeybees [144].

Waggle dance. The waggle dance is a type of dance in which a forager vigorously shakes its abdomen back and forth as it shifts in a straight line across the hive and circles around it to reach the place where it starts running [10]. This dance conveys information concerning the location, direction, and distance of the nectar source to increase foraging efficiency and obtain more food quickly. Accurate waggle dances require social learning in honeybees [[145], [146], [147]]. Two lncRNAs (XR_003305156.1 and MSTRG.6803.3) have been associated with waggle dance behavior [148], although the underlying molecular mechanisms remain unclear.

Shaking signal. Typically, this is directed at younger or less active colony members. To transmit the shaking signal, the sender honeybee grabs another honeybee with its meso- and pro-thoracic legs and then vibrates its body dorso-ventrally dorsoventrally at a frequency of 16 Hz for 1–2 s [149,150]. Foragers typically perform this signal after their first successful trip of the day or following a period of poor weather [151]. Shaking signals increase recipient activity, stimulate interactions, gather additional information related to colony tasks, and mediate collective foraging activity [10,151].

Stop signal. The stop signal consists of a bee head-butting another while producing a high-pitched buzz through thoracic muscle vibration [10]. This signal suppresses waggle dancing, decreases forager dispatch, and restrains departures [[152], [153], [154], [155]]. Stop signaling intensifies in the presence of predators, alarm pheromones, or aggressive encounters at food sources [[155], [156], [157], [158]]. In conclusion, behavior communication in honeybees is highly sophisticated, ensuring precise and efficient colony coordination. However, the molecular regulation and environmental influences underlying these behaviors remain underexplored and warrant further study.

Division of labor in honeybees

Compared with solitary species, honeybee colonies exhibit a sophisticated and complex division of labor, enabling members to specialize in specific tasks, which thereby confer greater competitive advantages. This specialized division of labor is primarily reflected among workers, with seasonal variations, in addition to the fundamental distinction between reproductive and nonreproductive castes [159,160].

In spring and summer, worker honeybees are sequentially categorized by developmental stage as cell cleaners, nurses, middle-aged workers, and foragers, with each group performing distinct tasks to accelerate colony growth and optimize resource accumulation (Fig. 6A). In general, younger workers undertake tasks within the hive, such as cleaning cells, feeding queens and larvae, and converting nectar into honey. Older workers guard the hive entrance, whereas the oldest individuals forage outside the hive to collect propolis, water, nectar, or pollen until death [11,159,161]. Importantly, this division of labor is not fixed but rather highly plastic. For example, when a colony experiences a significant loss of foragers, younger workers can accelerate their transition to foraging and quickly replace the missing members [162]. Such plasticity in labor division enhances colony survival, particularly under unfavorable environmental conditions. In winter, when food resources are limited and the climate is harsh, colony survival becomes a priority. In such situations, honeybees have adopted a comprehensive tactic that involves a large investment for better capacity and longevity when performing any assignment; thus, they become generalists (Fig. 6A) [11]. These findings indicate that the flexibility of honeybee division of labor maximizes efficiency and enables colonies to rapidly adjust their strategies and dynamically respond to environmental change.

Fig. 6.

Division of labor in honeybees and regulatory factors. (A) In the summer and spring, workers are sequentially categorized on the basis of their developmental stage as follows: cell cleaners, nurses, middle-aged honeybees, and foragers. In the winter, honeybees become generalists. (B) Genes such as MRJP2, MRJP3, IIS, Lkr, TRP, TRP2, and TRP3 are essential for honeybee division of labor.

The regulation of the division of labor involves multiple factors, including gene expression, the gut microbiota, JH, vitellogenin, octopamine, and nutritional status. Single-nucleus and spatial transcriptomics reveal that gene regulatory networks orchestrate brain cell activity during the behavioral transition from nursing to foraging [163]. Key genes such as major royal jelly proteins (MRJPs), insulin/insulin-like signaling (IIS), leucokinin receptor (Lkr), A. mellifera foraging gene (Amfor), and tachykinin-related peptides (TRPs) have been identified as central to this process (Fig. 6B) [[164], [165], [166], [167]]. For example, downregulation of MRJP3 and MRJP1 promotes the transition from nursing to foraging, likely by inducing neuronal changes suited to the distinct challenges associated with these roles [166]. Compared with nurses, foragers exhibit elevated IIS and TRP expression but reduced Lkr expression [164,165]. These opposite expression patterns likely constitute a molecular switch governing task transition. Moreover, pollen foragers show higher Lkr but lower TRP2 and TRP3 expression than nectar foragers do [165,167], suggesting fine-tuned regulation that allows adaptation to distinct ecological niches. Further studies demonstrate that TRPs contribute to task specialization by binding to tachykinin receptors (TRPRs), triggering intracellular cAMP accumulation, Ca²⁺ influx, and extracellular signal-regulated kinase (ERK) phosphorylation via dual Gαs and Gαq signaling pathways [167]. Collectively, these findings emphasize that differential gene expression is central to the division of labor, enabling workers of identical genotypes to assume divergent social roles.

The gut microbiota also influences labor division. Compared with nurses, foragers exhibit reduced microbial community stability, lower bacterial loads, and fewer gut microbes, particularly those of gram-positive fermentative genera [168,169]. This microbial divergence contributes to task differentiation. Additionally, JH, vitellogenin, octopamine and nutrients play important roles in regulating changes in the division of labor in honeybees [[170], [171], [172]]. These observations highlight the multifactorial regulation of the division of labor, wherein genetics, physiology, and microbial communities interact to generate phenotypic diversity within colonies. Future studies should integrate laboratory and field data with advanced methodologies—such as single-cell and spatial transcriptomics, metabolomics, proteomics, and computational modeling—to construct a comprehensive framework of the regulatory networks governing labor division and behavioral plasticity under environmental stressors.

Honeybee sensory systems

As social insects, honeybees have evolved sophisticated sensory systems, including olfaction, gustation, vision, audition, tactile perception, and magnetoreception, all of which play vital roles in their life activities and tasks. While many studies have concentrated on the olfactory, gustatory, and visual systems, research on magnetoreception, especially the auditory and tactile systems, remains limited and requires further investigation [14,[181], [182], [183], [184], [185],[14], [181], [182], [183], [184], [185]. Honeybees serve as a model for vision research, with numerous studies describing the structure and function of their visual system, including their roles in navigation and flight control. Several reviews provide comprehensive summaries of honeybee vision [173], [174], [175], [176]. Below, we focus on the organs and mechanisms underlying olfaction, gustation, and magnetoreception.

Olfactory system

The olfactory system plays a crucial role in maintaining honeybee colony cohesiveness and adapting to complex environments [177]. The honeybee brain possesses a dual olfactory pathway, in which the outputs of the two antennal lobes (the primary olfactory centres) are clearly segregated to higher-order centres in the lateral horn and the MB [12], [178], [179]. Notably, the honeybee olfactory system exhibits pronounced sex- and caste-specific dimorphism, with the olfactory systems of queens and drones differing both functionally and anatomically from those of workers. This variation may reflect their ability to adapt to performing fewer and less diverse behavioral tasks than workers do [180].

The antennae serve as the peripheral olfactory organs of honeybees and house all olfactory sensory neurons (OSNs). Honeybees display stronger odor responses when the right antenna is used than when the left antenna is used [181]. The antennae receive odorants, which enter the sensillar pores and diffuse to the dendritic membrane of the OSNs through the sensillar lymph. Odorants then interact with chemosensory receptors, including ionotropic receptors (IRs) and odorant receptors (ORs) [177], [182], [183], [184], [185], [186]. In addition to IRs and ORs, other olfaction-related proteins include odor-degrading enzymes (ODEs), odorant-binding proteins (OBPs), gustatory receptors (Grs), and sensory neuron membrane proteins (SNMPs) [187]. Their abundance may vary across bee species and castes, reflecting ecological and social specialization [185], [187], [188], [189], [190], [191].

Each olfaction-related protein group serves distinct functions. IRs detect multiple taste qualities, amines, volatile acids, and environmental stimuli such as humidity and temperature [192]. Grs mediate responses to non-volatile pheromones and sweet or bitter compounds [193]. ORs, which are derived from the Grs family, detect semi-volatile and volatile molecules [194]. ODEs degrade odorants [195]. OBPs shuttle odorants to receptors and facilitate the transition from gaseous to aqueous phases [196], [197]. SNMPs support odor detection [196], [197]. These findings indicate that IRs, Grs, ORs, ODEs, OBPs, and SNMPs form the molecular basis of honeybee olfaction. However, the precise mechanisms underlying their integration in odor perception and colony-level behavior remain to be fully elucidated.

Gustatory system

Taste is crucial for selecting profitable resins, water sources, and food resources for honeybees [13]. Gustatory sensilla, located on the forelegs, mouthparts, and antennae, enable taste perception [198]. Grs are key molecular components in this system. Only 10 Grs have been identified in the honeybee genome [191], far fewer than in Anopheles gambiae (52 Grs) [199], Drosophila melanogaster (60 Grs) [200], and Linepithema humile (116 Grs) [201]. This reduced number may reflect the specialization of honeybees on floral products, which narrows tastant diversity [191], [202].

Specific Grs in honeybees recognize distinct substances. AmGr1, AmGr2, and AmGr3 act as sugar receptors. AmGr1 is sensitive to maltose, sucrose, trehalose, and glucose, with mutations reducing responsiveness to glucose and sucrose [202], [203]. AmGr2 appears to function as a coreceptor, broadening the detection range of AmGr1, although AmGr2 mutants still display wild-type-like sugar responses, suggesting that it is not essential [202], [203]. AmGr3 is associated with fructose detection, but deletion of AmGr3 unexpectedly abolishes sucrose perception rather than fructose [202], [203], raising intriguing questions about its precise role. AmGr10 functions as an amino acid receptor, detecting glutamine, lysine, arginine, asparagine, aspartate, and glutamate [204]. These findings highlight AmGr1, AmGr2, AmGr3, and AmGr10 as central to sugar and amino acid detection, whereas the specificities of other honeybee Grs remain largely unresolved.

In addition to sugars and amino acids, honeybees exhibit distinct electrophysiological and behavioral responses to acids, bitter compounds, vitamins, lipids, minerals, heavy metals, and salts throughout their life cycle. Importantly, widespread colony losses and population declines in recent decades have been attributed in part to pesticide exposure [15]. Elucidating how honeybees detect pesticides and other harmful compounds is therefore essential for safeguarding their health.

Magnetoreception system

Magnetoreception allows animals to sense the Earth’s magnetic field and construct magnetic maps for orientation and navigation. Honeybees exhibit magnetoreception, and magnetized individuals show no response to magnetic anomalies [205]. Identifying the location of magnetite and clarifying its role are critical for understanding this sensory modality.

Hsu and Li proposed that magnetite particles reside in abdominal fat body trophocytes, enabling magnetic field detection, with iron granules restricted to the fat body [206]. This view has been contested [207]. Critics argue that (1) iron granules are present in both the fat body and midgut, (2) the role of the fat body in iron metabolism and ferritin function has been overlooked, (3) no evidence supports trophocytes as magnetoreceptors, and (4) electron diffraction patterns have not been demonstrated despite the presence of observed lattice fringes [207].

Recent findings have indicated the presence of magnetite in both the antennae and abdomens of honeybees [208]. The magnetite crystals in the abdomen are critical for magnetoreception [205], [209], but the contribution of antennal magnetite remains unclear. Overall, research on magnetoreception in honeybees lags behind studies on olfaction and gustation. Future work should investigate the regulatory genes, molecular pathways, and environmental influences underlying this sensory capacity.

Conclusion and future perspectives

Through long-term evolution, honeybees have developed a wide range of adaptive strategies to thrive in diverse environments. This review provides an integrated perspective on these strategies, encompassing brain functions, communication, division of labor, and the sensory system. A deeper understanding of these mechanisms can inform conservation strategies and breeding programs. Nonetheless, many open questions remain, and research can be advanced in the following directions:

• Climate Change and Cognition: While elevated temperatures are known to impair cognition in vertebrates [210], little is understood about invertebrates. Recent work has demonstrated that heat stress significantly impairs memory and learning in bumblebees [211], likely by reducing pollination efficiency. Similar studies in honeybees are urgently needed to assess the impacts of climate on their cognitive functions.

• Regulation of Gene Expressions: Current research focuses on individual genes or transcriptome-wide patterns, but cell type-specific expression dynamics remain unmapped. Emerging tools such as spatial transcriptomics, single-cell sequencing, and high-throughput chromosome conformation capture should be applied to elucidate fine-scale regulatory mechanisms that are involved in honeybee adaptation.

• Multifactorial Adaptation: Honeybee adaptation is shaped by multiple interacting environmental pressures rather than single factors. However, the mechanisms of integration remain poorly understood. Future studies should simulate complex environments to reveal how adaptive pathways interact under realistic ecological conditions.

• Epigenetic Mechanisms: DNA methylation is implicated in honeybee learning and memory [63], [70]; however, the broader role of epigenetic modifications in environmental adaptation remains unclear. Further research could provide valuable insights into the molecular flexibility that enables ecological resilience.

• Neural Plasticity: There is a dynamic and mutually shaping relationship between the sensory system and neural plasticity, but the related mechanisms are largely unknown in honeybees. Future research should address these gaps to better understand honeybee sensory ability.

• Genetic Engineering: Advances in genomics and gene editing offer potential for developing stable transformation systems in honeybees. Such tools could enable targeted breeding strategies and improved adaptation to environmental challenges.

• Beyond A. mellifera: Research is disproportionately focused on A. mellifera. Expanding studies to other bee species will generate novel insights and strengthen conservation efforts. Additionally, protection of honeybee populations and collection of wild germplasm are essential for sustaining ecological functions and supporting human development.

CRediT authorship contribution statement

Guilin Li: Formal analysis, Writing – original draft, Writing – review & editing, Funding acquisition. Shuwei Wang: Writing – original draft, Writing – review & editing. Liu Wang: Investigation, Writing – review & editing. Xiaomeng Wang: Supervision. Yan Li: Writing – review & editing, Visualization. Xingqi Guo: Supervision, Project administration. Hang Zhao: Supervision. Baohua Xu: Writing – original draft, Writing – review & editing, Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Biographies

Guilin Li received her Ph.D. (2020) from the College of Life Sciences at Shandong Agricultural University, China. She is currently a faculty member at the College of Life Sciences, Qufu Normal University, China. Her research focuses on how honeybees defend against and adapt to stressors, and on breeding honeybees for enhanced stress tolerance. Dr. Li has published many papers in national and international journals, such as Cellular and Molecular Life Sciences and Ecotoxicology and Environmental Safety.

Baohua Xu is a Professor of the College of Animal Science and Technology at Shandong Agricultural University, China https://dkdy.sdau.edu.cn/2022/0304/c1555a202017/page.psp. He also holds the positions of a scientist in the National Apiculture Technology System and a Taishan Industry Leader, and formerly served as a Vice President of the Chinese Apiculture Society. With over 30 years of experience in honeybee scientific research and teaching, his long-term research interests focus on honeybee health management, nutrition, and anti-stress biology. As a corresponding author, he has published more than 110 papers on honeybee biology, holds 11 national patents, has developed 9 apiculture standards, and authored an English academic monograph titled Molecular Response Mechanisms of Apis cerana cerana to Abiotic Stress.

Xingqi Guo has served as a professor at the College of Life Sciences, Shandong Agricultural University, and later at the School of Chemical and Biological Engineering, Qilu Institute of Technology (https://www.qlit.edu.cn/hxswxy/xyjj/xyld/). He is a scientist with the Shandong Province Apiculture Technology System, and was formerly an associate editor for the journals Gene and Plant Gene. For over 30 years, his research has focused on pest control and stress resistance in Apis cerana cerana. He has published more than 60 papers as a corresponding author and holds 6 national invention patents.

Hang Zhao is a faculty member at the College of Life Sciences, Qufu Normal University, China. He holds several editorial positions, including Associate Editor for Frontiers in Plant Science, and Youth Editorial Board Member for aBIOTECH and Agriculture Communications. His research focuses on the biology of stress resistance and the regulation of agronomic traits. Dr. Zhao has published more than 40 papers in national and international journals, such as Nature Communications, Advanced Science, Pesticide Biochemistry and Physiology, Trends in Plant Science, New Phytologist, and Plant Biotechnology Journal.