Bee Mimicry: Occurrences & Advantages

The Role Mimics in an Ecosystem

Hoverflies are an example of a “Batesian” mimic, whereby a harmless species evolves to imitate the warning signals of one more harmful. By looking like a bee, a hoverfly signals its defensive behaviours to deter a potential predator: venomous stings. On the hand, robber flies are “aggressive” mimics- some Florida species visually imitate a worker bee to avoid detection and appear harmless… until it attacks.

Beekeeping-relevant mimicry involves chemical manipulation of the intruder’s cuticle, enabling it to avoid detection inside the hive. Mites, Braula flies and the death’s head hawkmoth fall into category. This article describes these chemical mimics, along with examples of the phenomena in bees and its evolutionary niches. I hope that both beekeeping and non-beekeeping audiences can appreciate the diversity of insects beyond asking whether their photo of a hoverfly is “a type of bee”.

Why Mimic a Bee?

In short, insects mimic bees for predator defence, to blend in amongst prey (“wolf in sheep’s clothing”) and for access to the hive, which is a store of food (honey, larvae) and shelter (enclosed, warm space) for the mimicking insect.

Bees defend themselves by stinging, which makes them less appealing to your average predator (e.g. toads; Brower & Brower 1965). Therefore, non-stinging insects like hoverflies gain a survival advantage by mimicking bees. Predators like the aforementioned toad have learned to associate the bee’s appearance with a painful sting and would subsequently avoid attacking a visual hoverfly mimic.

On the other hand, some predatory insects mimic bees to blend in, so that once an unsuspecting forager gets close, the predator captures and consumes it. Robber flies mimic bees and freely approach flowers to catch insects in this manner. So successful are these species that the University of Florida nicknames their three endemic species as “bee killers”. One of them, Mallophora orcina has a diet consisting of 80% honey bees, and over-predation is reported as a contributor to annual colony loses (Bromley, 1950).

There are the integrative mimics, those who mimic bees and live in their colonies undetected. Once inside, these mimics begin to feed on the larvae and stores, exploiting the hive for shelter. An example of “Wasmannian” mimicry is seen in the carrion beetle Necrophila americana who visually mimic cuckoo bumble bee (Psithyrus spp.), enabling them to raid a bumble bee colony (Bombus spp.; Fisher et al., 1986). This is not a standalone phenomenon, as the authors report of other beetles mimicking bumble bee aspects to their advantage.

Chemical Mimicry: Cuticular Hydrocarbons of the Death’s Head Hawkmoth

When invading a hive, the death’s head hawkmoth is ignored by workers. At first this was thought to be due to the sounds made by the moth- its screeches sound similar to a queen bee, causing workers to “freeze”. This was not the case, and nor was it due to the visual similarities between the moth and workers, as shown below… Sight is useless in the dark hive.

Moritz and colleagues showed that the cuticular hydrocarbon produces chemical camouflage (Mortiz et al., 1991). The moth’s hydrocarbon profile is similar to that of honey bees, consisting of palmitoleic acid, oleic acid, hexadecenoic acid and stearic acid. Readers familiar with the honey bees’ cuticular hydrocarbons will recognise these components as being part of their chemical makeup. We know that a foreign bee is rejected from entering the hive, so perhaps a combination of moth and bee chemicals provide the invader with an “invisibility cloak” against the hive guards?

Varroa Avoid Detection by Chemical Mimicry

The less destructive cousin of Varroa destructor, V. jacobsoni, has a cuticular hydrocarbon profile like its host Apis mellifera bees (Nation et al., 1992). Unlike the death’s head hawkmoth above, these mites obtain chemical camouflage either passively from close contact with adults and larvae, or actively via feeding on bee fat bodies (Ramsey et al., 2019). A total of 91 compounds were shared between V. jacobsoni and A. mellifera, with mites having slight differences in their proportions (Martin et al., 2001). This imperfection could be why bees can detect and remove V. jacobsoni more readily, and why it is (relatively) less harmful than V. destructor.

Speaking of V. destructor, this species mimics its host cuticular hydrocarbon profile to such a fine degree that even subtle inter-colony differences are accounted for (Kather et al., 2015). By feeding on fat body tissues, the mites synthesise bee-specific compounds (and by extension, the colony-specific compounds), mimicking the host’s odour. V. destructor is efficient. Chemical changes in the mite’s cuticle begin within 20 minutes after introduction to a new host. Even when a dead mite is placed on a live bee, the parasite passively picks up the host cuticular hydrocarbon profile, camouflaging the corpse.

Did chemical mimicry emerge in response to Varroa-sensitive hygienic behaviours? If the mites are perceived to be a part of a bee, they are less likely to be groomed off the adult’s body like pollen or sugar. Le Conte and colleagues found the chemical mimicry to be more effective in mites being transferred between the same species, here the Asian honey bees A. cerana (Le Conte et al., 2015). Presumably, this is due to a longer coevolutionary period between A. cerana and Varroa with them. So, it is a combination of passive and active processes that enables Varroa to avoid detection while parasitising their host.

Braula Flies also Avoid Detection by Cuticular Hydrocarbon Mimicry

Like mites, the bee louse, Braula coeca also mimics the cuticular hydrocarbon profile of bees through close proximity (Martin & Bayfield 2014). The fly prefers inhabiting the head and thorax of the queen, avoiding detection despite being the most attended and most groomed member in the colony; Braula intercepts workers during trophallaxis, consuming Her Majesty’s share.

I could dive into the chemistry and tell you that the “alkene chemical groups that are most strongly mimicked by Braula, which suggestes that these compounds are use by bees for nestmate recognition” however I defer the results as an exercise for the reader.

Begging for Food: The Mimicry of Small Hive Beetles

Small hive beetles are effective mimics of honey bee larvae, in that the beetle’s begging behaviour triggers nursing behaviour in worker bees (Ellis et al., 2002). Cape honey bees A.m. capensis build beetle jails whereby they attempt to “starve” beetles to death. Previous work showed that starved beetles lived up to a fortnight (Ellis et al., 2001), but when incarcerated they lived for up to two months. How?

The authors found beetles mimicking the trophallaxis response by stroking their jailor’s antennae. This was not a foolproof mimic, as initial attempts were dismissed, and only after multiple (>5) attempts did guards regurgitate honey as well as royal jelly for the beetles to feed on. They speculate that there is also an element of chemical mimicry, and I agree, for it is unlikely that the guard bees suddenly decided to “take sympathy” on their prisoners after multiple stroking attempts (Ellis et al., 2002 & Pirk, 2022).

Cape Honey Bees as Chemical Mimics

Adapted from Cape Honey Bees

Honey bee workers are sterile. If they lay eggs, they emerge into unfertilised, drone bees. This does not usually occur, for worker bees police one another and prevent worker-laid brood from emerging. Cape honey bees are the exceptions, as workers can lay both male and female eggs that go undetected by the host colony, presumably due to chemical mimicry of queen-laid eggs.

Martin and colleagues, they found that when A.m. capensis workers laid eggs in A.m. scutellata colonies, the eggs were allowed to hatch and develop into workers (Martin et al., 2002). In contrast, when an A.m. scutellata worker laid eggs in her colony, they were removed by other workers within 24 hours. Is it possible that both fertilized and unfertilized eggs possess a scent, and when the full chromosomal count is present in fertilized eggs, the resulting scent is recognise as a “normal egg”?

Müllerian Mimicry, or How Bees Got Their Stripes

Many bees and stinging wasps have bright and obvious colouration, with the bees’ black and yellow being most memorable example. The literature suggests that colourations are for thermoregulation purposes (i.e. darker colour patterns are associated with longer cool periods in spring, and may help a bee absorb energy from the sun and warm up quicker in these cool conditions) or as a form of camouflage (“a green bee can blend in the tree leaves”; Williams 2007).

Müllerian mimicry is when two harmful species evolve to look alike, reinforcing avoidance behaviours in predators as their pattern recognition systems inform them that eating this animal is a one-way ticket to pain.

Because the females possess a venomous sting, the bright colourations are interpreted as a warning signal to advertise pain and distastefulness to a potential predator (“aposematic” in biological terms). Williams suggests that colouration in bumble bees (and I dare say honey bees) is because of unpredictable colours found in their natural landscapes (Williams 2007).  A variety of flower colours means bumble bees cannot hide from predators by matching with the background vegetation, so they evolved into the striking bee pattern of black and yellow stripes to signal “eat me at your own risk”.

Predators Learn to Avoid Brightly Coloured Prey

The success of the bee colouration (and distinct colouration in general,) is that predators learn to avoid brightly-coloured prey, but only after they’ve been hurt once (reviewed in Chatelain et al., 2023). Chicks stung by bumble bees would not eat a mimic or a stingless bumble bee on account of learned experienced, whereas a chick that has not been stung would eat them anyway.

Some wasp species are predators of bees, but owing to the similar colouration of the two, wasps can reduce predation from another animals due to Müllerian mimicry. A bird may have trouble differentiating between a yellowjacket and a honey bee in flight, so it chooses to avoid both, thus sparing both species.

When Does Mimicry Fail?

I preface this with a statement that the judgement of a “poor mimic” is biased by human perception. Predators perceive prey differently than we do. For example, we only see a hoverfly at rest, whereas a bird may see it mid-flight, where its appearance may be more convincing. Further, the mimics we see today cannot be deemed failures, for the species had survived and reproduced to present day. Perhaps those species that did not survive the natural selection of urbanisation are the true “failed” species… they no longer exist.

Howarth and colleagues showed that the most successful mimics are those that not only look the part, but also play the part (Howarth et al., 2004). In this case, being active around the same vegetation and times as their model organism. This pattern was easily seen when the mimicked species is common (e.g. a honey bee) and unobserved in rare, solitary species.

Some hoverfly wasp mimics (S. ribesii and E. balteatus) outnumber their wasp models, making predators more likely to attack them in place of the stinging wasp (Azmeh et al., 1998). This results leads to an inverse lesson: those waspy-looking patterns are good to eat! Owing to their abundance, the mimics would be consumed more than the model, and so this mimic is deemed less successful.

The general pattern between mimic and model is that the more common a model, the more closely a mimic should match it. Deviations from expected behaviours are detected by predators who are familiar with its prey’s habits on account of the model’s abundance. Mimics of common species must try harder to produce a believable deception.

The Issue with Batesian Mimicry

Batesian mimicry is detrimental to the modelled species because it slows down predator learning. Like the avian example above, if a predator was to eat a hoverfly rather than a bee, it learns that these “flying black and yellow creatures” are an easy meal and will continue seeking them. This increases the likelihood of a bee being killed by a predator, which isn’t a problem if it’s a social honey bee with tens of thousands sterile worker, but when it’s a solitary species, that’s an issue.

If a predator is stung the first time eating a new species, it takes one attempt to conclude that it is inedible. What if they’re stung after the 10^th or 100^th successful attempts? Would they still be dissuaded after one sting, or will it take multiple stings from that model species before it stops seeking them? Questions for Plato.

The Evolution of Mimicry

In the viewpoint of evolution via natural selection, mimicry is the result of variations in the environment becoming more dominant over time¹. To take the hoverfly example above, suppose that in a population of hoverflies, there exists genetic variation regarding size, colouration, and behaviours. This variation is natural and arises randomly through mutations, genetic drift, and recombination.

Enter the selective pressure – in this case, predation. Hoverflies are preyed upon by species like birds, reptiles, and larger insects. In areas where bees are common, predators usually learn to avoid the lone forager due to her stinging capabilities that hurt them. So predators that encounter such colourations recognise and avoid them based on specific visual cues like yellow and black striping, size or flight patterns (Brower & Brower 1965).

The traits give hoverflies a survival advantage. Those flies that look slightly more bee-like (vivid striping, bee-shaped body) are less likely to be preyed upon than their typical-fly-looking counterparts. This advantage makes the hoverfly more likely to survive and reproduce, passing on these traits their offspring. Over generations, these traits become more pronounced in the population, being refined to include other bee aspects like the aforementioned size, body shape, and bee-haviour.

The above is a simplified example of Batesian mimicry arising from evolution via natural selection. The harmless species mimicking harmful ones to gain protection from predators. The general method above can be copied for aggressive mimics (robber flies) or even chemical mimics (Varroa mites).

Ripple Effects of Mimic Emergence on the Ecosystem

Evolution is not an isolated process. As mimicking characteristics emerge, there is continual feedback from the ecosystem, since the increases in mimic populations impact the behaviour of local predators. Even the real bees are impacted as mimics compete for flowering plants. So long as the presence of actual bees continues to reinforce the predators’ avoidance behaviour (“small yellow flying thing hurts”), then hoverfly mimicry remains evolutionarily stable and can be maintained within the ecosystem.

Bee Mimicry and Beekeeping

Beekeepers constantly encounter chemical honey bee mimics: Varroa and other mites, small hive beetles, even unlikely parasites like the death’s head hawkmoth or Braula flies. The mites and beetles warrant attention, for they’ll overrun weak hives, and you already know the impacts of Varroa. The rarer parasites can be ignored. In terms of visual mimics, the robber flies are predators of concern, and I wish beekeepers in areas with bee killers the very best in managing a wild fly species…

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

1. Occasionally, random mutations may produce changes that are “just right”, slightly resembling those of local bee populations. These changes are not initially intended for the purposes of mimicry, but over time that population will adapt, survive and reproduce, leading to the hoverflies we see today. Sometimes Lady Luck lets species skip the lengthy process of natural selection a fact that is beyond the scope of this article. ↩︎