We propose that homing in honeybees (Apis mellifera) is based on at least two forms of memory, a general landscape memory (GLM) formed during orientation flights and a vector-memory learned while shuttling between particular feeding sites and the hive.
We used the harmonic radar technique developed at the NRIRU by Joe Riley's team. Tiny radar transponders attached to the backs of honeybees allowed us to track the flight paths of the the insects while foraging. The dipole based transponder reflects the first harmonic of the radar transmitter frequency, which can be easily discriminated from typical background radar reflections. The left hand photo shows the radar setup with the transmitting dish antenna. The receiving dish antenna on top is tuned to the first harmonic. Both antennae are rotated by a synchronous motor every three seconds.
General Landscape Memory Flights:
A hive was placed in the experimental area which lacked natural landmarks. Bees were trained to collect sucrose solution from a feeder which was rotated around the hive at a distance of 10m. Once foraging flights between hive and the rotating feeder were well established, we captured bees after they had ingested sugar solution, and quickly transported them individually in a closed box to one of three release points (R1-R3), about 250m from the hive.The bees were then fitted with a transponder, and released.
Initially the bees search for the hive and return to the release site often. After finding landmarks they switch to direct flights towards the hive. We estimate that the optical cues of the hive are visible at distances of not more than 30m . This demonstrates that the bees were able to form a general landscape memory during orientation flights Track 99 058 shows a typical example.
Bees were trained to collect sucrose solution from a feeder (F) east of the hive at a distance of 200m. Once foraging flights between hive and feeder were well established, we captured bees after they had ingested sugar solution, and quickly transported them individually in a closed box to one of eight release points, about 250m from the hive. The bees were then fitted with a transponder, and released. Track 99 205 shows a typical example where a bee released 250m to the SSE of the hive (R7) maintained a straight flight to the west for 200m, before beginning circling flight. This flight was almost exactly along the feeder-to-hive direction for its entire length.
Many bees completed the full length of these flights before starting to search for their hive, and almost all maintained accurate compensation for lateral wind drift. This demonstrates that the bees do not refer to landscape cues during these return flights, but instead use the ‘optic flow’ of the ground beneath them, and their sun compass, to judge both direction and distance.