11.2.4. Field studies

The optimal area for tracking bees with the harmonic radar is an open, flat and horizontal landscape without any radar reflecting obstacles (trees, houses). The reason for these requirements is that the radar beam should not be reflected either from the ground or from an object because the reflected radar pulse with its energy multiple times stronger than the second harmonic signal radiated off from the transponder would interfere with this small signal. In such an ideal area, transponder-carrying bees can be detected within a radius of up to 3 km (see section?). Examples are given in Fig. 25. The transponder is detected every 3 s defined by the revolution of the radar unit. The circular map resulting from the radar scans is transformed into a Cartesian map with a custom-written program. Bees travelling with 5 – 20 km/h are therefore detected about every 5 – 15 m. The length and width of the radar paint (the technical term for a radar signal) depends strongly on the strength of the harmonic signal, but usually lies in the range of about 15 m at a distance of 600 m. Additional processing of the radar signals allows to substantially improve spatial/temporal resolution. Signal strength of the harmonic radar depends also on the animal's flight height over ground. The unit described above allows for detecting bees reliably between 70 cm and 9 m height for distances up to 1,200 m straight away from the radar. Bees learn quickly to fly with the transponder. Usually, they first land after being released but then start and fly equally fast and reliably as those without a transponder. However, they appear to be more sensitive to wind speeds above > 20 km/h. At such wind speeds they also fly low, making it more difficult to detect them.

The harmonic radar system allowed proving von Frisch's proposal (von Frisch, 1967) that bees communicate distance and direction in their waggle dances (Riley et al., 2005; see section 9). In addition, their navigation strategies could be characterized (Menzel et al., 2005, 2011b). The experiments performed led to the conclusion that navigating bees integrate multiple learning strategies which provide them with the capacity to localize themselves and several goals in such a way that they can perform novel short-cutting flights between them (Menzel et al., 2011a). Such a capacity has been related to a memory structure best conceptualized as a cognitive map (Tolman, 1948).

Fig. 25. Nine examples of flight tracks. In the subfigures A – F three phases of the homing flights are marked in colour. Vector flight component in red, search flight component in blue, straight homing component in green. H marks the location of the hive, R2 and R9 two different release sites. These bees followed a waggle dance that indicated a location 200 m to the east. The thin blue line in C marks a portion of the flight which could not be detected most likely because the bee flew very close to the ground. G – I: Each of these three bees was first trained to a location indicated as FT30° or FT60°. Then feeding at this site was terminated, motivating the bees to stay inside the hive and attend dances. A single dancing bee indicated a location FD30'' or FD60°. Each bee flew first to the place indicated by the dance and then flew directly to the learned place along a novel shortcut. The green area around the place indicated by the dance marks the visual catchment area as defined by the person sitting there and the spatial resolution of the bee eye. Since the spatial resolution of the bee eye is 1.5° it can be calculated at which distance the bee will see the person sitting in front of the hive. There was no visual mark at the trained place during the test (after Menzel et al., 2011b).

1293PN revised Fig 25