First Flight

Deconstructing A Small Miracle

This morning I observed honey bees in my backyard flying every which way but not alighting on anything.  They could not be foraging for nectar because flowers were still waterlogged from two days of rain. My first thought was that I had a colony ready to swarm. That would be weird because it was so early in the morning. But I went outside anyway and confirmed that my hives were not preparing to swarm. 

Then I put two and two together:

• Bees observed flying in ever larger circles near their hive.

• Bees were not showing signs of swarming, nor could they forage for nectar profitably (because of waterlogged flowers).

• The bees had been hive-bound for two days because of the rain. You would not go out in the rain either if drops were the size of your body.

• Queens lay from 1000 to 1500 eggs a day. In a worker bee's short, six-week lifespan, she will transition through many jobs: feeding the queen, feeding the larvae, packing pollen, accepting nectar from foragers, carrying out the dead, and finally, foraging for nectar and pollen. Each hive may have a thousand or more bees that need to learn to fly every single day.

These were young bees taking their first orientation flights in order to recognize their own home.

Bees don't need to learn the mechanics of flying any more than a fish needs to learn the mechanics of swimming. Flying is an innate skill. It takes about 2-4 weeks before a bee’s wings are stiff enough to fly efficiently. She is then ready to leave the hive for the first time in her life. 

Leaving the hive is easy. Finding her way back to her home is much harder. In the wild, a colony may live behind a small hole in a hollow tree in a forest of similar looking trees. When you consider honey bees are "legally blind" (by Department of Transportation standards), finding one's home seems miraculous. A honey bee needs to get within a few feet before she can even recognize her hive's entrance. Considering that a honey bee can fly 5-7 miles from its home in search of pollen and nectar—without maps or GPS—honey bee navigation is quite the trick.

The photo below gives you a rough idea of a honey bee's poor visual acuity. It is a scene taken at a high school showing a sidewalk, light pole, trees, sky, and buildings. What I’ve learned—from honey bee researchers and from computer simulation—is that this level of acuity is good enough.

What I observed this morning was a backlog of young bees taking their first flights—so called "orientation flights". These young bees fly in ever larger circles and regularly gaze back at their home. A bee may not recognize her own home from more than a few feet away, but she can learn to recognize her neighborhood and her nest's location in that neighborhood. My hives are at one end of a green yard next to a red brick house with a concrete sidewalk and driveway. Once a bee recognizes this unique area, she will know where her home is. 

To these bees, my yard is their first memorized "catchment area": this is an area that, once entered, triggers a visual memory of where the bee is and to what extent it is off-course. Bees (and their hymenopteran cousins, ants) navigate from catchment area to catchment area. You can see this illustrated below. 

I’ve used this clever algorithm for multirotor navigation. Below is an aerial view of a high school near my home. In order to develop a Python algorithm for visual navigation, I used a digital camera with a fish-eye (wide-angle) lens to take four shots—north, south, east, and west—at each numbered location (in red) on this aerial view. 

Below, I show each photograph and its label.

Honey bee researchers have done a pretty good job at defining what features honey bees use in their visual system. I won’t go into detail here, but a data structure known as a Linearly Sensitive Hash table (LSH) can store these feature sets. Given a perceived set of features, the LSH returns an ordered list of locations most like the perceived set. Not only can you retrieve the correct "catchment area," but you can also determine necessary course correction. 

And how do bees navigate from one catchment area to another? They do it by path integration, better known as "dead reckoning". But that is a topic for another time. There is much more to this story.

Later today, when it warms up and dries out a little, the experienced foragers will begin their work. Their flight paths will be "beelines" - straight and purposeful. There will also be the occasional first fliers with their circling orientation flights. 

---

There is so much one can learn by studying the flight paths of honey bees: 

• Swarming is where a colony divides and large contingent flies off to a new home. A much smaller cohort of scout bees that have been to the new home literally herd a cloud of their sisters to their new home.

• Foragers leave their hives with empty honey crops and fly in straight and speedy beelines. 

• Bees arriving full of nectar are much heavier, slower, and less coordinated. Their landings resemble controlled crashes.

• In the Fall when food is scarce, bees may rob weak neighboring colonies. Here, robbing bees leave the robbed hive full of robbed honey. They may climb up the front of the hive to "gain altitude" before taking off. Arriving robbers weave and bob sideways one foot in front of a hive’s landing board before landing…as if they were on the lookout for defending guard bees. 

• Honey bees can get lost. They can be blown off-course from strong winds. When they do get lost, they engage in a similar random search observed in albatrosses, amoebas, deer, jackals, monkeys, sharks, and other foraging animals. This random distribution is called a Levy Walk or Levy Flight. Unlike the path of Browning motion, a Levy Walk looks like someone was searching randomly in a local area and figured out “It’s not here. I need to look somewhere else!”

There are many reasons to love honey bees: honey, wax, medicines, and pollination of our most colorful fruits and vegetables. I like honey bees most for their outsized cognitive ability. I want to understand how the human brain works but I suspect that understanding a simpler, homologous brain first will be more productive. I hope you found this interesting.

---

My book (in progress) addresses the details of navigation in honey bees, robots, and even humans. I show how the faculty of navigation is a necessary precursor and computational foundation for communication, innovation, and reasoning. Non-human animals that navigate long distances also excel in communication, innovation (tool use), and reasoning. Please register (free or sponsored) for Intelligence Evolved so that you can be among the first to know about the publication of this book.