Conceived and designed the experiments: GK ES. Performed the experiments: GK ES. Analyzed the data: GK ES. Contributed reagents/materials/analysis tools: GK ES. Wrote the paper: GK ES IK.
The authors have declared that no competing interests exist.
Giant honeybees (
Giant honeybees (
Generally, defence behaviors of honeybees may involve physical contact with aggressors. A prominent example is their capacity to recruit stinging guards
Honeybee colonies also defend themselves without physical contact with their enemies, minimizing the risk for the defending bees. Examples include the aposomatic coloration, which is characteristic of all hymenopterans
This article investigates whether Giant honeybees succeed defending their nests against hornets by shimmering, and how prey and predator interact. The honeybee colony behavior is analyzed concerning the occurrence, strength and repetitiveness of shimmering under the aspects of proximity and velocity of predatory wasps. This proves shimmering as a colony response to approaching wasps. On the other hand, the wasp behavior has been investigated in response of the time course and the strength of shimmering regarding proximity to the honeybee nest. In two different scenarios of experiments, shimmering waves, in particular in their ‘big-scale’ shape, are proved to repel wasps within a specified range around the honeybee nest. ‘Small-scale’ shimmering is effective in preventing wasps from predation by generating ‘confusion’. It is demonstrated that the predation activity of wasps near honeybee nests and the defence responses of Giant honeybees through shimmering base on a reciprocal, mutually adjusted relationship with possibly coevolutionary roots. Lastly, it is discussed how shimmering benefits Giant honeybees, in particular to support their open-nesting habit.
This article investigates the prey-predator interactions between colonies of Giant honeybees (
In both scenarios, a single camera recorded the behaviors of hornets and bees in PAL format enabling a frame-by-frame analysis at a rate of 25 images per second. Prior to these experiments documented in this article, the authors had observed thousands of shimmering waves in hundreds of Giant honeybee colonies on several expeditions in India and Nepal over 15 years. Based on this broad experience, both scenarios of the two experimental nests were recognized to typify the shimmering behavior of giant honeybees and the respective flight behaviors of hornets under the given prey–predator relations.
In
The nest-specific axes were defined by their real-world coordinates: x, the horizontal (sideways) directions in regard to the vertical flat of the
In
The focus of this article is on shimmering behaviors. Shimmering is made up by abdominal movements of quiescent individuals predominantly at the surface of the honeybee nest, displaying wave-like processes. These abdominal movements of surface bees were detected by image analysis (Image-Pro, Flir), assessing the shimmering waving strength (
(A) Continuous assessment of abdominal thrust activity of the experimental Giant honeybee nest, while a predatory wasp was present in front of it (
The waving strength W ( = number of abdomen-shaking bees per frame) depends on the hornets' distances from the nest dxz (A) and on the hornets' flight velocities vxz (B); time zero defines the onset of the waves; (A) five dxz classes (Cdxz = 1–5; coded in yellow to red; for definition, see
Categories of distance of the hornets from the nest before shimmering | Number of episodes | Distance of the hornets from the nest before shimmering | Shimmering activity as response to the presence of the hornets | ||
Cdxz | dxz [cm] | n | dxz [cm] | Wpeak | W600 |
1 | 0–18 | 6 | 12.7±1.8 | 12.6±1.3 | 71.1±4.4 |
2 | 18–30 | 19 | 25.9±0.7 | 12.1±2.0 | 67.1±6.9 |
3 | 30–40 | 49 | 35.8±0.4 | 10.0±0.8 | 55.2±2.9 |
4 | 40–60 | 79 | 49.2±0.7 | 8.2±0.5 | 49.4±2.0 |
5 | >60 | 48 | 71.5±1.5 | 6.5±0.5 | 42.4±2.1 |
201 |
Shimmering activities of 201 episodes in terms of Wpeak and W600 in dependence of the pre-wave distances of the hornets from the nest (definitions, see
Categories of flight velocity of the hornets before shimmering | Number of episodes | Distance of the hornets from the nest before shimmering | Shimmering activity as response to the presence of the hornets | ||
Cvxz | vxz [cm s−1] | n | dxz [cm] | Wpeak | W600 |
1 | 10–20 | 7 | 41.15±6.04 | 6.43±0.72 | 37.08±2.63 |
2 | 20–25 | 25 | 40.27±3.33 | 7.49±0.75 | 43.23±3.51 |
3 | 25–30 | 29 | 41.01±3.38 | 9.28±0.98 | 51.36±3.42 |
4 | 30–40 | 53 | 38.60±2.05 | 8.88±0.90 | 53.08±3.23 |
5 | 40–50 | 30 | 47.23±3.53 | 9.06±0.83 | 55.58±3.97 |
6 | 50–60 | 21 | 47.68±4.71 | 8.86±1.04 | 53.36±5.24 |
7 | 60–80 | 21 | 53.78±4.01 | 9.89±0.95 | 58.28±4.20 |
8 | >80 | 16 | 56.98±5.35 | 11.14±1.26 | 61.88±5.58 |
202 |
Shimmering activities of 202 episodes in terms of Wpeak and W600 in dependence of the hornets' pre-wave flight velocities. The hornet data (vxz) were assessed in the interval 400 ms prior to shimmering. Shimmering responses (means±SEM) were categorized by eight classes (Cvxz = 1 to 8) of flight velocity vxz of the hornets (cf.
The wave strength W was originally defined and calibrated as the number of surface bees per frame that were actively ‘shaking’ or ‘lifting’ their abdomens. This ‘wave’ parameter was also applied in a more general sense as one of the ongoing aspects of movement of nest mates, regardless of whether or not shimmering had been identified as the respective process. For explanation, the value Wpeak characterized the intensity of shimmering at its peak, and the value W−400 ms was used to quantify the levels of movement activity 400 ms before the onset of the shimmering waves (which may include residual waving activity).
To estimate the number of bees that participated in a shimmering wave, the integral ∫
The
The flight velocity
The behaviors of both predator and prey have been synchronized to the onset of the shimmering waves. This trigger concept not only associates the shimmering waves of the bee colony to the hornet behaviors considering shimmering as responses to the hornets' behaviors, but also vice versa, that is, it allows considering hornet behaviors as responses to shimmering. The shimmering waves and the flight behaviors of the hornets were monitored from 400 ms prior to until 1000 ms after the onset of shimmering. The question was whether the flight behavior of the hornet in the vicinity of the honeybee nest represented adequate cues for affecting shimmering. In detail, the investigation was on whether the strength of the shimmering waves (W) was dependent on the flight parameters of the hornet (dxz, vdxz, vf) in the pre-wave period, 400 ms before the onset of waving. To the contrary, changes of the hornet behavior in the first 600–1000 ms after the onset of shimmering was tested as obvious responses to shimmering.
An episode between the honeybee as prey and the hornet as predator was defined by the shimmering behavior of the bee colony and by the flight behaviors of the hornet in the front of the bee nest. As detailed in earlier sections, the
In
The waving strength W600 gives the numbers of bees which had shaken their abdomens over 15 frames (definition see inset and text); it depends on the hornet's distance from the nest dxz (A) and on the hornet's flight velocity vxz (B); time zero in the insets defines the start of the shimmering waves (cf.
In
Shimmering is not necessarily a response to wasps, it may also be provoked by homing and departing bees (
While the category of ‘reactive’ hornets is unequivocally defined, the category of ‘nonreactive’ hornets remained to be crucial because of its dependence on gross subjective manual criterions, which did not allow deciding in detail about any residual pattern of reactivity of the hornets to e shimmering. To proof whether ‘nonreactive’ hornets did not respond to shimmering or showed any odd reaction patterns, we tested the time courses of the respective behaviors of hornets categorized as ‘nonreactive’ for the existence of any obscure responsiveness.
In
Gaussian distributed data sequences were compared by parametric tests (t-test). If the normality test failed, the software automatically used nonparametric tests (Chi-square test, Wilcoxon Signed Rank Test). These tests traced differences in behaviors between two experimental states, e.g. how the behaviors of the hornets differ between two time intervals, such as before and after the onset of shimmering. Correlations were characterized by the regressions of the original data values of the respective behavioral classes or facultatively (see below) of their arithmetic means. The regressions were fitted by optimizing their coefficients of determination (R2) and tested by Spearman rank order correlation test.
The time courses of the behaviors of honeybees and hornets in three steps were compared, using the One Way Repeated Measures ANOVA (e.g. Friedman test on Ranks, Sigmastat). The time correlations of the original data over a number of discrete time intervals in 40 ms steps were proved (e.g. from time zero to 400 ms after the onset of waving) and was adjusted for ties. In the first step, a test was conducted to see whether the pairs of variables of prey (W) and of predator (vdxz, θxz) tend to increase together during the experimental time after the onset of shimmering in two time intervals (0–400 ms; 400–800 ms). Here, Dunns Method tested the original data per time interval, and the Spearman test their mean values. In a second step, the relative differences among the treatment groups of prey (rel W) and of predator (rel vdxz, rel θxz) were proved. In a third step, if the relative differences were because of random sampling variability, the respective regression of the arithmetical means of the same time intervals as the appropriate description of the correlation was accepted. This three-step statistical procedure estimated the type of interaction between prey and predator, although the original predator data would have been too unfriendly for a straight one-way Repeated Measures ANOVA.
In this study, the conditions under which shimmering waves are produced by giant honeybee colonies were investigated. In
Wave-like processes also occur without the presence of a hovering wasp. Foraging nest mates departing from or arriving at the nest mostly are the source for the generation of shimmering in absence of predators. However, in
This study investigates several flight parameters of predatory hornets in the vicinity of Giant honeybee colonies regarding their significance for eliciting shimmering. In
The second parameter to categorize the shimmering responses in
The arousal state of a Giant honeybee colony when predatory wasps are nearby is also expressed by the repetitiveness of shimmering. A good measure of it is the residual waving strength W−400 (lower inset of
Percentage of hornet episodes (n = 317) observed in the respective five dxz classes (A) and eight vxz classes (B) of hornets' flights; (C) the relationship between the flight velocities vxz (abscissa) of individual hornets and their distances dxz to the nest (ordinate); dashed lines (A,C) give the average hovering distance of the hornets (cf.
Finally, these data of
Long-lasting experiences over more than one decade with Giant honeybee colonies have enabled the authors hypothesize that shimmering waves do have antipredatory goals. If so, it should be possible to observe that shimmering lowers the chances of the hornets to prey on the curtain bees on the surface of the Giant honeybee nests. In the following sections, this surmise is investigated and questioned whether shimmering is able to distract wasps from grabbing bees, whether it is able to repel wasps or is even able to make wasps turn away from the nest.
The primary question for an obvious antipredator impact of shimmering on wasps is whether wasps respond to shimmering. In
Hornet behaviors were categorized in five distance classes Cdxz = 1–5 (see
To quantify this responsiveness of the wasp to shimmering, the change in distance of the hornet within one second after the onset of waving (Δdxz(1s)) was chosen. Taken together the responses at all five proximity categories (Cdxz = 1 to 5) the data correlated linearly with the proximity of the hornets to the nest prior to the shimmering wave (
Summarizing, the authors conclude that at distances greater than the mean hovering distance, hornets tended to approach the nest under the influence of shimmering. This suggests first, that hornets were attracted by shimmering if they were further away from the nest by at least half a meter. Second, at the mean hovering distance, hornets were not affected by shimmering; they stayed neutral with regard to approaching or leaving the nest site. Third, when the hornets were closer than the mean hovering distance, they withdrew from the nest in the course of shimmering. This hornet behavior is indicative of avoidance behavior, suggesting that shimmering plays a role in repelling predatory hornets, but only when close to the nest.
In
The time courses of shimmering (A,D) and of the hornets' flights (B–C,E–F), ‘reactive’ (A–C) and ‘non-reactive’ (D–F) episodes (for definition, see text). Honeybee and hornet behaviors were synchronized to the start of the shimmering waves (abscissas give the time in milliseconds after the start of the shimmering waves). The hornets' behaviors are shown in terms of distance velocity vdxz (B,E) and turning angle θxz (C,F); see inset and text for definition. Two classes of hornets were defined according to their distance to the nest in the 400 ms interval prior to shimmering: dxz<45 cm (red circles, 84 ‘reactive’ episodes; 59 ‘non-reactive’ episodes), and dxz>45 cm (yellow circles, 65 ‘reactive’ episodes; 108 ‘non-reactive’ episodes). Different brown-shaded areas define two test intervals in relation to the time course of shimmering (brown-shaded: 0–400 ms, grey-brown shaded: 400–1000 ms). Circles and bars give arithmetical means±SEM. Big full (red or yellow) circles give significant differences of the data in relation to the starting time of the wave at t = 0 ms (P<0.05; Holm-Sidak test, Friedman Repeated Measures ANOVA on Ranks).
‘Reactive’ hornets in close vicinity to the nest (dxz<45 cm) showed a strong reaction to shimmering and turned away from the nest with a maximum speed vdxz of around 18 cm s−1 (
The additional analysis in
Abscissas, the waving strength W assessed by the number of abdomen-shaking bees per frame; ordinates, the hornets' behaviors measured by the parameter distance velocity vdxz (A–B;E–F) and turning angle θxz (C–D;G–H) using the data of ‘reactive’(A–D) and ‘non-reactive’ (E-H) episodes (cf
In
(A,B) Mean trajectories of the approaching hornet (for compilation of trajectories, see text); the sectors between the dashed lines define the four divisions of the mean pre-wave flight directions of the hornets for pooling the x- and y-values of the positions of the hornets approaching to the nest in 40 ms intervals. Thick lines, arithmetical means of x- and y-values of the hornet's position; horizontal and vertical bars, SEM. Note, that the trajectories before the onset of the waves (coded by black thick lines) are straighter than after the onset of the waves (coded by red and blue thick lines). (C) Hornet flight behavior under the influence of big-scale (full red circles) and small-scale (open blue circles) shimmering waves; ordinate,
Big-scale and small-scale waves were categorized according to a threshold of 40% of the maximal waving strength (cf.
The data obtained from this study provide new insights into the complex spatial and temporal patterns of interaction between bee-hawking hornets and Giant honeybees under defence. In support of their open-nesting life-style, Giant honeybees have evolved a set of defence strategies that keep predatory animals, birds, and wasps in particular, off the nest. The obviously most spectacular defence action refers to the recruitment and release of flying defenders
It has been the conventional view
Bee-hawking hornets incessantly approach honeybee nests, again and again, to prey on them, without showing the tiniest sign of habituation (which was tested in 335 episodes of shimmering waves). The data obtained do not provide any support for the ‘proximity-avoidance’ hypothesis that would propose that hornets are deterred by the honeybee nests and would avoid its vicinity. Nevertheless, the factor ‘proximity to honeybee nest’ apparently modulates their responsiveness to shimmering, essentially because the honeybee colony also alters its defence response depending upon the distance of the intruder.
However, three aspects have been proved in support of the ‘shimmering-repels-wasps’ hypothesis that assumes that bee-hawking hornets show an avoidance response to shimmering when they come too close to the honeybee nest. First, it was proved, particularly in
Thus, the capacity of shimmering to repel wasps is limited to a distance of around half a meter from the nest, which equals the mean hovering distance. This restriction in the defensive coverage of colony of Giant honeybees is likely to be associated with the obvious ultimate goal of colony defence to generate a safety zone around the nest that should keep predatory wasps away from the nest, preventing them from catching bees directly from the nest surface. If the wasps, nevertheless, succeed in intruding this shelter zone they should not stay long. This is exactly what was observed during shimmering: When wasps approached the nest, the honeybee colony continuously generated shimmering waves that repeatedly repelled the wasp. Shimmering evidently benefits the honeybee colony because it lowers, factually to zero, the hunting success of those predatory wasps that want to seize bees from the nest surface. The chance to observe wasps in trying to seize giant honeybees from the surface of the nests is quite rare. In a total of estimated 30 minutes of own observation over years and of a hundred of trials by the wasp to catch surface bees, we have not observed any successful grasp.
On the other hand, the hornets hardly elicited shimmering when they were more than 50 cm away from the nest. The question here is whether and why giant honeybees do not recognize hornets outside the distance of 50 cm as a threatening peril. It is known that
Another question associated to the concept of a shelter zone of Giant honey bee nests arises here: does shimmering deliver only visual cues to the wasps or does it also utilize pheromone channels? It is known that shimmering is linked to chemical scenting
If wasps should be hindered to feed directly from the honeybee nest it would be sufficient to organize local groups of surface bees to confuse or misguide them
As demonstrated earlier (in
Therefore, the authors propose that the wave-like character of shimmering has been evolved obviously not primarily to confuse wasps, but, as shown above, to repel wasps. A possible explanation for this striking capacity probably has two aspects. First, shimmering may reinforce in wasps innate and not habituating fixed action patterns of avoidance. Second, waving possesses a further sophisticated and strikingly ‘convenient’ effect that may also enforce the innate avoidance of the addressee: When the wave of abdomen-thrusting bees spreads over the nest, the wave front stays indeed ‘behind’ the wasp, it factually press-gangs the wasp away from the place it originally wanted to prey on (see
Although hornets are continuously attracted to the honeybee nests (by their rich resources of protein and sugar), shimmering effectively prevents the potential predators from collecting bees from the nest surface. Hunting episodes in which hornets continuously attempted to ambush flying bees in front of the honeybee nest were recorded. In thousands of wasp episodes in several honeybee colonies, a single case of successful hunt of a hornet having caught a bee from the nest curtain was not observed. However, hornets do have another kind of hunting success if they focus on ingoing and outgoing bees that cross their hovering range. Bees threatened by the bee-hawking wasp are unprotected by the colony-bound collective defence, but are still able to escape by dodging and fast flight (vxz = 2.25±0.03 ms−1, n = 1855 images, n = 107 flights of bees;
The
Theoretically, there is a fundamental problem for a prey–predator relation if a defence action of a potential prey, such as shimmering, does not lead to any physical contact with the enemy. Of course, such traits are less risky for the defenders, but they are obviously less dangerous for the predators, which may learn to ignore ‘unperilous’ signals of the potential prey. However, observations clearly demonstrate that repetitious shimmering efficaciously repels the same hornet again and again. Any habituation effect in hornets can be excluded; obviously, they cannot ignore shimmering, although they repeatedly try, without showing any sign of habituation, to hunt their prey from the honeybee nest. Although the wasps decelerate and approach the nest, shimmering interrupts their landing operations, and elicits avoidance reactions, which take the wasps away from the spot of prey. Mostly they are repelled off the nest, at least half a meter or more, from where they start the next hunting episode.
Because of their persisting and nonhabituating bee-hawking quirks, it is assumed that wasps envisage honeybee nests as a prey of extraordinary attractiveness. Obviously to avoid widespread wasp predation, honeybees have acquired cavity-nesting abilities (in Southeast Asia:
Hence, it has been extremely important for the open-nesting Giant honeybees during their five million years
Regressions of the correlations between shimmering behavior and the hornets' behaviors (see
(0.04 MB DOC)
This video shows two 130 cm-wide nests of the Asian Giant Honeybee
(7.94 MB MOV)
This video shows a nest attached to the ceiling of a water tower in Chitwan, Nepal. A typical, unsuccessful, hunting episode is shown in which a hornet chases a flying bee, but the bee escapes and lands on the nest. The landing of the bee and the manoeuvre of the hornet provokes shimmering, which makes the hornet turn off the nest. The film documents this in original speed (QuickTime; 1.6 MB).
(1.59 MB MOV)
This video shows the same scene as ‘Movie 2’, but in slow motion and contains explanatory text and arrows and the trajectories of the hornet and the flying bee. Light green represents the flying bee, and the shimmering (shaking) nest bees are shown in dark green. The hunting hornet is indicated in red, and in violet when repelled by shimmering. Dark green arrows point out that shimmering repels the hornet. The violet arrow shows the new flight course of the hornet in response to shimmering (QuickTime; 0.9 MB).
(0.90 MB MOV)
This video shows the nest of
(3.27 MB MPG)
We thank Dr. S.M. Man, Dr. R. Thapa, and Dr. M. B. Gewali from the Tribhuvan University, Kathmandu, Dr. Farooq Ahmed and Dr. S.R Joshi from the UNESCO organization ICIMOD in Kathmandu, and Dr. D.K. Sharma from the University Gauhati for their support in Chitwan, Nepal and in Assam, India.