Complex social waves of giant honeybees provoked by a dummy wasp support the special-agent hypothesis

Abstract

The social waves in giant honeybees termed as shimmering are more complex than mexican waves. it has been demonstrated¹ that shimmering is triggered by special agents at the nest surface. in this paper, we have used a nest that originated by amalgamation of two previously separated nests and stimulated waves by a dummy wasp moved on a miniature cable car. we illustrate the plausibility of the special-agent hypothesis¹ also for complex shimmering processes.

Giant honeybees produce shimmering waves for defense purposes, including against predatory wasps.^2,3 It was recently demonstrated that these social waves are triggered by special agents on the nest surface.¹ These special agents are the first to flip their abdomens in response to a threatening cue. Shimmering is rarely a single-wave event, but more often produced repetitively. It eventually comprises a set of smaller waves elicited in parallel. In other words, shimmering results from an interference of spatially and temporally complex patterns of waves of varying strength levels and repetition rates. This complexity discerns shimmering from the analogue “Mexicanwave”⁴ in football stadiums which are usually uni-dimensional in terms of spatial and temporal distribution. Nevertheless, shimmering still follows simple rules, which elevate individual activities to swarm intelligence,⁵ with the proximate goal to shape visual cues for external addressees.³

In this Addendum we present an experiment with a dummy wasp that was moved by a miniature cable car with constant velocity along a horizontal line 10 cm in front of the experimental nest near its upper attachment zone (Fig. 1), rather than using a tethered-wasp as Schmelzer and Kastberger1 did. Unfortunately, both stimulation methods confine the interpretability of the shimmering process regarding co-evolutionary aspects of predator-prey interactions.³ However, compared to observations of free-flying wasps they have the advantage that shimmering can be initiated under more controlled conditions with respect to position, direction and velocity of the wasp in front of the bee nest. The sample nest (Fig. 1) was formed from merging of two neighboring nests after massive disturbance by a honey hunter. The thicker nest parts on the left had honey cells and were converted after the honey hunting into the mouth zone.³ This is the factual interface between outside and inside the nest and was visible throughout the day. The concave region on the right derived from a formerly separate nest.

Figure 1.

Snapshots from Movie S1, recorded in Chitwan, Nepal in Jan 2009. The bees produced shimmering waves in response to a moving dummy wasp (the black and white-striped cuboid above the nest, which was moved by a miniature cable car). The sequence starts at frame 694 and ends with frame 747; in the last frame (775) the dummy wasp had already been moved back again to the right side. The stimulation started at frame 0 and continued over 3,500 frames. The frames were recorded with a HDT V camera at a frame rate of 50 Hz. See text for details.

We previously noted that a typical shimmering wave lasts shorter than one second,³ irrespective of whether the colony is provoked by a wasp or by other visual cues. At the first sight, the shimmering wave in Figure 1 (see also Movie S1) seems to contradict these principles. We concentrated only on a single movement of the dummy wasp from the right to the left, starting this truncated analysis at frame 694 in the mid of the concave zone of the nest and ended at the left side of the nest at frame 747; in total, this partial process took 53 frames or roughly one second. While the dummy was moved to the left, the waves originated perpendicular, but slightly behind the dummy, in the middle of the concave zone. The waving process was not homogenous (frame 702) but built up as a compound wave with several smaller waves. While their right sides merged, forming a wider shimmering area (frame 704), the left side of the initial wave accompanied the dummy to the left with a lag of some milliseconds. At this moment (frame 709), a new wave started slightly in front of the dummy, its left side merging with the residual left part of the initial wave (frames 713,716). Consequently, the new core area of the resulting waving activity fell behind the dummy wasp. From this time onwards, all successive waves were reinforced regarding their left parts, while their right parts, which were more distant to the dummy, tended to vanish immediately after their emergence (frames 720–723). Thus, the remaining left portions of the successive waves (730–747) added up to form an overall moving shimmering wave, which strikingly accompanied the dummy to the left side of the nest like a shadow.

Real wasps in front of giant honeybee nests would strongly differ in their movements from that of the dummy. Shimmering has been proved to repel free-flying wasps, and the wasps respond with avoidance reactions.³ The wasps alter their flight routes and are mostly driven off the nest. Moreover, the special-agent attribute1 causes that shimmering is not triggered apparently directly at the precise position of the moving wasp but slightly behind, and not ahead of the wasp. This kind of interaction between wasps and honeybees is the very potential of shimmering bees to “wipe away” free-flying wasps from the nest area (Movie S4).³

Nevertheless, some predators neglect the shimmering cues while approaching the nest. In particular specialised birds such as the bee eater Nyctornis athertoni or the honey buzzard Pernis ptilorhynchus find their own ways to get their prey.^6,7 For the bees this is a disturbing situation although frequently occurring. Against birds the giant honeybees have evolved defence behaviours using flying guards6 where hundreds of bees are successively released from the nest within a fraction of a second. In this case, it is possible that the absolute size and the characteristic movement patterns of threatening predators are the cues that trigger mass attacks (flying guards are even released when the honey buzzard is 10 m and more apart from the nest). However, we provide evidence (Movie S2), using the same miniature cable car-driven dummy wasp as shown in Figure 1 that the mass release of flying guards⁶ could also be triggered by a wasp-sized predator that “refuses” to leave the nest. The question whether such an “ignoring” of shimmering by the predator would be an adaptive cue that triggers a mass attack is subject of further investigations.

Addendum to: Schmelzer E, Kastberger G. ‘Special agents’ trigger social waves in Giant honeybees (Apis dorsata) Naturwissenschaften. 2009;96:1431–1441. doi: 10.1007/s00114-009-0605-y.