A lot of researchers, including me, have investigated how landmarks or the shape of the horizon can provide useful information for animals to learn to navigate. But flying animals like bees can also look downward, seeing the world from above, something that terrestrial animals rarely get the chance to do. Far less is currently known about how they make use of this ability. One of the more striking ways that humans have reshaped the world is our network of roads, and pigeons have famously been shown to follow human highways to get home. Although roads are a human innovation, long, linear features are common in nature in the form of rivers, coastlines, boundaries between different types of habitat, and on a smaller scale, animal tracks. We were excited to try to find out how flying insects make use of these features.

 

As part of her PhD studies, Joanna Brebner used harmonic radar to track the flights of bees in a fascinating landscape in Southern Spain. The ground was so flat that there were no useful cues on the skyline at all, but the land was a rice farm which was divided into a rectangular grid by pathways and access roads. We found that not only did these paths have a profound effect on the bees’ flights but that they appeared to use them in a variety of different ways, depending on context. Bees exploring the world for the first time ever in fields and grassland follow a typical flight pattern in which they make a series of loops around their nest, almost like the petals of a flower; in our Spanish site, their exploration flights looked very different, following roads far more often than you’d expect by chance and even flying parallel to  them at other times, as though their whole conception of space was influenced by the grid structure. Spending so much time around the paths is bound to bias the way bees learn about their environment and, indeed, they discovered a feeder placed by a path intersection much faster and made far more use of it, than one that was the same distance away but placed in the middle of a field.

 

As bees gained experience and transitioned from exploring to foraging at a feeder we provided, they initially followed the roads to the feeder quite faithfully even though this involved an inefficient right-angle turn, but they soon started to cut the corner, gradually straightening their route. We think this happens because, when you are unfamiliar with your surroundings, following a road is much easier than trying to keep track of your location with enough precision to direct yourself straight to a hard-to-spot feeder; essentially, you only need to memorise what to do at intersections, rather than constantly choose the exact right direction of travel. As they learn more about the environment, though, they become ever surer about where they are and can abandon the easy but inefficient pathways for a direct shortcut. Interestingly, even very experienced foragers often seemed to rejoin the pathway around 30m short of the nest and follow it back home, rather than fly there directly. This might be a clever tactic to avoid overshooting the nest: if you fly directly to where you think the nest is but can’t find it, it could be in any direction so you have a long search ahead; but if you join the road to the south of the nest, you just have to follow it north until you see the nest.

 

Speaking of searching, we unexpectedly removed the feeder on occasion, to see how the bees would go about looking for it. A group of bees who expected to find a feeder next to a pathway searched by flying up and down the pathway. As they continued to search for it, they started flying along other pathways, but seldom left them to explore the fields. Another group of bees had been trained to search for feeders that were always kept within a single field. These bees searched in a very different way, flying in loops inside the field. When they did reach the pathways that formed the field boundaries they typically turned back. In both cases the bees seem to use their knowledge of where the feeders are usually found to reduce their search space: in one case they expect to find feeders next to pathways so don’t waste time searching anywhere except pathways; in the other they use the pathways as a boundary and concentrate their search within the area they define.

 

All in all, this work shows that ground level linear features can provide a lot of information for a navigator and that bees are clever enough to make use of different aspects of this information depending on the context.

 

In academia, the first-named author on a paper is typically the person who did the bulk of the hard work – in this case, Joanna Brebner. The last-named author is typically a supervisor or mentor who guided the project – this was my first last-author paper and it’s one I’m particularly proud of. It’s been a great pleasure to watch Joanna flourish as a world-class scientist over the years (although I’m far too modest, of course, to suggest my advice played any role in that).