The first time that a bumblebee leaves the nest and takes to the air, the stakes are high: lose your way and you could be lost forever. The pressure is on to learn as much as possible about the lay of the land before venturing further afield. ‘Bumble bee nests are hidden in the undergrowth’, says Tom Collett from the University of Sussex, UK, adding, ‘The bees have to learn the exact relationship between objects that define the position of the nest and the nest hole’. So instead of embarking on an epic journey and keeping their limbs crossed for a safe return home, the novices set about exploring the vicinity. First, they fly tiny looping circuits that are centred on the nest, gradually broadening the survey until they have learned enough about their surroundings to guide themselves home at the end of their maiden flight.

‘One thing that they need to know is whether objects are near or far’, says Collett. He explains that bees, and insects in general, use the speed that the image of an object travels across the retina of the eye to estimate their distance to an object: images of nearby objects move much faster than images of distant objects. Insects often simplify the task of estimating distance by making sure that the head does not rotate, compensating for any body rotation by moving the head in the opposite direction to the body's rotation. But bees may use a different strategy to estimate the distance separating two objects, such as the nest entrance and nearby foliage. They could circle around the nest so that objects that are near by move slowly across the eye, while more distant objects move faster. Wondering how bumblebees learn the layout of objects surrounding the nest, Collett and colleagues Olena Riabinina, Natalie Hempel de Ibarra and Andrew Philippides set about filming the insects' looping learning flights (p. 2633).

The team set up a bumblebee nest box, complete with a queen and several dozen workers, on the roof of a building at the University of Exeter, and then filmed the first tentative flights of new bees as they emerged from the nest. The insects initially strayed no more than a few centimetres from the nest entrance, but they eventually flew out of camera view 20–30 s later. Then the team began the painstaking task of analysing how the learners moved their heads and bodies to find out whether they moved their heads to compensate for body rotations to stabilise the image as they circled around the nest. First, they measured the position of the bee's body in each frame of the movie as it circled the nest and then they meticulously measured the head movements by hand.

‘We didn't know what we were going to get’, laughs Collett, but eventually the team was surprised to see that instead of stabilising the image rotation completely by pivoting the head to counteract the body rotation, the learners' head movements under-compensated for the body rotations, allowing their heads to rotate slightly. Calculating how images move across the retina as the head rotated a little, the team could see images of objects close to the entrance of the nest would move slowly, while remote objects would be moving faster, allowing the bee to distinguish between landmarks that were remote from the nest and those that would guide them home safely.

Riabinina
O.
,
Hempel de Ibarra
N.
,
Philippides
A.
,
Collett
T. S.
(
2014
).
Head movements and the optic flow generated during the learning flights of bumblebees
.
J. Exp. Biol.
217
,
2633
-
2642
.