An Introduction to the Asian Honey Bee and its Behaviours
To beekeepers, Asian and European honey bees are pretty much identical. They live in social colonies consisting of a single queen, many workers, and a few drones. They make honey, gather nectar, pollen, propolis and water for their hive, and provide pollination services for crops and local flora. It is correct to assume these two species compete with one another for resources. The most contested factors are tree hollows (both are cavity-nesting species) and flower sources (both are generalist foragers).
This article introduces Apis cerana through its swarming habits. Most literature on A. cerana is written in Japanese, Chinese, German or Indonesian languages and only abstracts, if any, are available in English. To any A. cerana beekeepers reading, I appreciate your input.
Apis cerana Swarms Multiple Times a Year
Both swarming and absconding involve the queen and a cluster of the original workers leaving the hive to establish a new colony. In swarming, a large amount of workers and daughter queens remain in the original colony. In absconding, very few (if any,) workers remain, and no queen; the colony has abandoned this location altogether. Brood and resources may remain, depending on the efficiency in preparing for abscondment.
Migration and Absconding
Theoretically, forage is available year-round in tropical environments which suggests that A. cerana will swarm anytime according to colony requirements and nectar flow. However, this is not always the case, as the study by Chinh and colleagues showed that in tropical north Vietnam, A. cerana swarmed twice a year (Chinh et al., 2004). In temperate climates, flowering has seasonal variations, and A. cerana’s reproduction follows pattern of spring & summer, declining in autumn and non-existent in winter (Matsuura & Sakagami, 1973).
A. cerana swarms and absconds far more than A. mellifera (6-8 times a year, depending on your sources, compared to the 1 or 2 swarms of A. mellifera per year). In a survey of Indian hives by Pradeepa & Bhat (2014), they found that absconding was determined by shortage of flowering resources, colony disturbances, and incidence of pests & disease, with the greater wax moths being the most prevalent. In regions with dry and wet seasons, forage has an annual cycle, and it’s likely that bees abscond to follow the nectar flow.
Hive Departure is Temperature Driven
In A. cerana hive temperature begins increasing 40 minutes prior to departure (Zhu et al., 2019). This continues linearly up to ~3°C higher than the average temperature of ~35°C, peaking at 7 minutes post-departure. The temperature increases were not consistent across the whole hive. The central regions where brood rearing take place had the lowest increase in temperature; recall that brood have a narrow temperature tolerance for optimal development. By comparison, regions near the hive entrance increased by 7.6°C.
The authors suggest that real-time remote temperature monitoring will be beneficial for A. cerana monitoring, as it gives beekeepers a ~40 minute window to prepare and even control for either absconding or swarming of bees (Zhu et al., 2019).
Competition with Apis mellifera
In southern China, A. cerana foraged on more diverse plants than A. mellifera (Liu et al., 2013). Where the species overlapped in their foraging range, the European honey bees aggressively take over flowers from A. cerana foragers (Gross et al., 2019). The general pattern for foraging bees: in their native range, the Asian bees are more efficient than introduced A. mellifera, and vice versa (Manila-Fajardo & Cervancia, 2003).
In introduced or agricultural environments, the unfamiliar plants cause both species to rely on their generalist foraging behaviours to determine which flowers are most suitable, and here the more aggressive A. mellifera dominate the preferential resources, forcing A. cerana to forage at second-class sources (Sakagami 1959).
When comparing the waggle dance of A. cerana and other Asian bee species, Dyer & Seeley found no “obvious dialectical differences” – in other words, the waggle dance of the species was quite identical (Dyer & Seeley, 1991). This held true when the cavity-nesting A. mellifera was compared against the open-nesting A. florea (Sen et al., 2003). Fortunately, honey bee espionage is not practiced.
A. cerana Starts and Finishes Forage Earlier than A. mellifera
In Himalayan apple orchards, A. cerana begins foraging earlier in the morning and stops later in the evening than A. mellifera. However,The European bees carry larger pollen loads, spend a longer time on individual flowers and touch more stigmas than A. cerana, making them the superior apple pollinator (Verma & Dulta, 1986).
Within the same region, the extended foraging times of A. cerana made it a better pollinator of rapeseed than A. mellifera (Pudasaini & Thapa, 2014). These Asian bees persisted in pollinating under unfavourable conditions (foggy, low temperatures). When comparisons were made, the authors found A. cerana to visit a higher number of flowers than Apis mellifera, concluding it to be a better pollinator. The authors draw this conclusion without commenting on rapeseed seed set (i.e. successful pollination) or the time they spent at flowers, unlike the apple example above. The bees are seen visiting flowers, but are they actually pollinating?
Artificially Mixed Apis cerana but not Apis mellifera Colonies
Sakagami reports that a hybrid colony of A. mellifera and A. cerana bees is possible when formed with an A. cerana queen and workers less than three days old (Sakagami 1959). Once established, younger bees from either species can be introduced to bolster the population, so long as the founding colony started with equal numbers of both bees. This method is not always successful, as the author reports cases of A. mellifera turning on their A. cerana nestmates.
When things are going well, the workers and drones from either species are treated as if they are in their own hive. The author also noted that the Asian bee queens can suppress the ovaries of European worker bees, indicating that bees can be influenced by queen pheromones from a different species.
Rejection of A. cerana Bees by A. mellifera colonies
When the above was reversed, and young A. cerana workers were transplanted into A. mellifera colonies, the European bees rejected the Asian workers. Perhaps this one-way relationship is due to the dominance of A. mellifera over A. cerana, and the latter will accept the former as a result of their superiority. This was the same case with A. cerana drones. While A. mellifera tolerates A. mellifera drones from other colonies, they do not extend this to A. cerana drones – perhaps due to their cuticular hydrocarbon profile?
However, the drifting of A. cerana workers (and not beekeeper-mediated introduction), was tolerated by the European colonies. “Tolerated” is not the same as “accepted, as the author reports that even inside the hive, the drifted A. cerana bees were bullied by some of the A. mellifera workers. Drifting was most commonly observed after relocating one hive into the vicinity of another, and was confirmed by the fanning patterns – Asian honey bees fan facing away from the hive entrance, whilst European bees fan facing towards it (Sakagami 1959).
If A. cerana comb is placed inside an A. mellifera hive, then any unsealed brood was removed by the host colony. Capped brood remained until emergence, whereby the young bee was expelled from the hive. Despite the smaller cell size, A. mellifera utilises the comb as normal: resource storage and brood rearing. By contrast, when A. mellifera comb is placed in A. cerana colonies – the Asian workers would “gnaw” at the lower parts of the comb. If any brood was there, it would be gnawed too – otherwise A. mellifera brood was “tolerated”. Unsealed larvae still died, as A. cerana workers do not provide the same brood care as A. mellifera (Tan et al., 2009 & Tan et al., 2012).
Beekeeping with Apis cerana
Chan & Ron report that in rural Malaysia, traditional hives of simple boxes or modified logs are used for A. cerana beekeeping (Chan & Ron 2010). Langstroth hives are deemed too intensive for the rural beekeepers (think comb spacing, queen excluders, wired foundations, frame building and so on). The outcomes from top bar designs were similar to Langstroth hives in both good (honey building, colony health) and bad factors (absconding rates, wax moth prevalence1). In other words, the beekeeping experience was equal between Langstroth and non-Langstroth hives.
In Indonesia, A. cerana inhabits managed forests, such as the Wanagama Teaching Forest on the island of Yogyakarta (Permadi et al., 2021). The villagers hang bee boxes known as “stups” (“glodogan”) from branches. Honey is extracted by heating – the comb is melted in pan for 10-15 minutes at a maximum temperature of 60°C2. The mixture is cooled and filtered, and the honey is sold off for production. As of 2019, there are a total of 460 stups, producing around 895 kilograms of honey.
A modified Langstroth hive, the Newton hive, is used for A. cerana. This variation has the same construction with respect to bee space, but with a lower volume to match the smaller Asian bees. A diagrammatic breakdown is presented below, extracted from the presentation by the International Centre for Integrated Mountain Development in the Hindu Kush region of the Himalayas.
A. cerana Disease and Parasite Risks are Lower than A. mellifera
I think that A. cerana is less impacted by Varroa mites due to the bees’ frequent migration/absconding tendencies. If the colony is on the move, then mites emerge from the comb and into an empty hive, dying later.
Further, A. cerana workers clean themselves (“autogrooming”), each other (“allogrooming) and brood cells more thoroughly than A. mellifera (Rosenkranz et al., 1993). When A. cerana brood is infested by Varroa (the endemic V. jacobsoni), the targeted brood was removed before cell capping. The authors attribute this to an “alien scent” in infested cells, along with a possible shift in the brood pheromone profile following Varroa introduction. In infected drone cells, there appeared “closed pores”3; as the worker bees fill the opening with wax to isolate comb containing mites or disease like Sacbrood or European foulbrood disease (Boecking, 1999).
In general, A. cerana is a more robust honey bee and is a better species for beekeeping in the poorer areas of Asia. Less management means less equipment and less treatment. More swarms means more queens, which could means more brood and honey for consumption.
Predator Defence
A variety of predators exist in Asia, and A. cerana has developed adaptations to combat them. Online, the video of a hornet being bee-balled to death is a frequent occurrence (look it up). Vespine wasps are a major predator for A. cerana colonies throughout Southern Asia. Further adaptations to hornet attacks include shimmering and zig-zagging, to absconding (Shah & Shah, 1991; Robinson, 2013).
Evasive Flight Measures
Zig-zags are evasive flight measures that occur when hornets “hawk” outside a hive, i.e. when hornets hover at the hive entrance to catch returning foragers. The variation in flight confuses the hornet, decreasing its chances of predation. The dodging tactic is combined with a “crash landing”- the forager lands at the hive entrance at such high speeds that she collides with the walls or a guard bee (Shah & Shah, 1991).
A. mellifera have no such adaptations, if anything, the foragers slow down and swivel from side as to side (as in a normal returning forager). This makes them easy prey for the hawking hornet. During take-off, the bee flies slowly in wide circles above the hive prior to seeking forage, again easy prey for hawking hornet (Tan et al., 2007).
Bee-balling, or Cooking, Hornets to Death
Consider what happens when a large & (heavily-armoured) solitary hornet enters the nest and begins consuming bees. the response of A. cerana is to send 400-500 workers and surround the hornet, beating their wings and stinging it. This raises the temperature inside the ball to ~48°C and increases carbon dioxide concentrations, suffocating the hornet to death.
The balling method is suitable in defending from species like Vespa orientalis, as they have a lower lethal temperature limit than V. mandarinia (44°C compared to ~46°C respectively; Ken et al., 2005), whereas bees do not start dying until 50°C. Indeed, the temperature inside bee balls varied between 45-46°C in A. cerana (Sugahara & Sakamoto, 2009), but only 44°C in A. mellifera. (Papachristoforou et al., 2011). Within A. cerana bee balls, Asian giant hornets die in less than ten minutes as carbon dioxide concentrations increase to 3.6% (compared to the ~0.04% CO2 in air).
“I See You” Shakes. Buzzes, Arms and Tails.
Shaking & tails are more common in open-nesting species like A. dorsata. Between 20 and 200 bees hang beneath the main cluster forming a “tail” that can launch to pursue and sting an intruder. Tan & colleagues describe this as an “I see you” signal, whereby the guard bees simultaneously vibrate their abdomens from side to side and buzz loudly for a few seconds (Tan et al., 2012). It works, for hornets do not approach the colony. Should a hornet enter anyway, it is balled to death… After all, it was warned.
Unlike in the open nesting species, Robinson reports that the formation of “tails” was rarer in A. cerana (Robinson, 2013). Rather than forming a dedicated attack unit, the tails are decoys to draw the hornet’s attention away from the main cluster to protect the queen and the majority of workers. The author studied Vespa spp. attacks on swarms and found the formation tails (along with “arms” at the side) frequently occurred in response to hornet attacks.
Classification of Asian Honey Bees
Apis cerana is native to South, Southeast, and East Asia. It is the sister species of Apis koschevnikovi and both are in the same subgenus as the European honey bee, Apis mellifera. Radloff and colleagues propose to alter the taxonomy from the traditional Latin names into morphoclusters- grouping the bees based on their morphological (physical) similarities (Radloff et al., 2010):
- Morphocluster I: Northern cerana. This is the largest-ranging cluster, extending from northern Afghanistan and Pakistan through northwest India, across southern Tibet, northern Myanmar, China and then northeasterly into Korea, far eastern Russia and Japan.
- Morphocluster II: The Himalayan cerana, links the bees of northern India, Southern Tibet & Nepal – they are split into the “Hills” group in the northwest, and “Ganges” group in the northeast, for the river Ganges.
- Morphocluster III: Indian Plains cerana (or just “Plains cerana“) occurs across the plains of central and southern India and Sri Lanka. These were previously the A.c. indica subspecies.
- Morphocluster IV: Indo-Chinese cerana, stems from Myanmar, northern Thailand, Laos, Cambodia and the southern parts of Vietnam.
- Morphocluster V: The Philippine cerana, restricted to the islands of Philippines, except for Palawan Island, which contains Morphocluster VI.
- Morphocluster VI: Indo-Malayan cerana, extend from southern Thailand, through Malaysia and Indonesia.
Note that this is more advanced than the scientific naming of the Western honey bees subspecies, who are still referred to by their Latin names.
Why Morphoclusters and not Species?
Within these morphoclusters, there are subclusters, which is akin to saying “within these species, there are subspecies”. The use of the Latinised “trinomial epithets” (triple names) can cause issues. They are supposed to describe a single variation, rather than a cluster of variations, and can confuse future readers. After all, how did this Latinised name come about, and is it biologically meaningful? For example, Apis cerana indica’s native range extends past India, so why is it called “a.c. indica“?
The morphoclusters suggested by Radloff & colleagues are not binding, but rather a biologically meaningful clarification on the formal naming of A. cerana. The authors themselves speak of limitations in their identifications, such as limited sampling size, and variations in the data from other research (Radloff et al., 2010). I wonder how morphoclusters categorisation can compare something like the Nepalese bees (A. c. indica) with the Philippine bees (A. c. cerana), whereby the former subspecies entombs Varroa-infested drone brood, whereas the latter does not (Boecking, 1999).
Swarming, Absconding & Migration in Apis cerana
This introduction to Apis cerana highlighted their swarming habits (more frequent than A. mellifera) predator defensiveness (far better than A. mellifera) and competitive abilities of the Asian honey bee (outmatched by A. mellifera). Notes on beekeeping with A. mellifera were made, as well as its classification history, and why morphoclusters triumphed over names. If you want more on these Eastern honey bees, do let me know.
Read the whole post and found puns, typos mistakes? Voice your displeasure with a strongly-worded message to admin@tgbee.com.au, or by filling this form.
Footnotes
- I disagree with the authors, and perhaps you do too. Absconding can be a result of excess predation or resources, which changes each year. The authors had done this study over eight weeks from July to September. If this was repeated next year, it is likely they’ll find different results. ↩︎
- † For those wondering if this reduces the benefits of honey, the locals report this relatively short heating time also melts the pollen and propolis into the mixture, increasing the nutritional benefits. It further reduces the water levels, slowing fermentation. Only at higher temperatures (>80°C) do the nutritional benefits break down, and quality is compromised (Shapla et al., 2018). ↩︎
- This “pore” is a cap on drone brood cells that differentiate A. cerana from A. mellifera. When the drone cell is capped, and the drone larvae begins spinning its cocoon, it uses the wax from the capping, forming the pore structure. This pore facilitates gas exchange (“respiration”) with the pupa. Blocking it off hinders drone development completely (Boecking, 1999 citing Rath, 1992 in the American Bee Journal). I wonder if A. mellifera were to have this pore in its drone cells, would it make for a better detection of Varroa-infested combs? ↩︎