How the Wise Old Wood Gods Laugh

Who saw the first primeval calf.

One day through the primeval wood
A calf walked home as good calves should;

But made a trail all bent askew,
A crooked trail as all calves do.

Since then three hundred years have fled,
And I infer the calf is dead.

Sam Walter Foss

The pull quote above is from Sam Walter Foss’s The Calf-Path. I looked it, and him, up because I thought it’d be cute to quote it in this article, but then it turns out Sam Walter Foss is awesome so I had to stop writing this and put together a “selected poems” collection of my favorites. I’ll probably be quoting him a lot and linking back to that page. The article ends with a quote from another poem of his. Appropriately for someone with the name FOSS, his work is all in the public domain, so you can find his work online other places too. Okay, now back to

Bee Problems

Suppose you’re a scout bee looking for nectar to bring back to your hive. No other scouts have reported back yet, so you don’t know how much nectar is out there, how it’s distributed, or the best direction to start looking. What search pattern do you use?

The most important rule is that it should be random. You’re one of many in a large hive. If you do something simple, like fly in a gradually widening spiral centered on the hive, your sisters will probably do the same, since they’re very similar to you. That’d be a waste of time, having everybody cover the same ground. You could try to coordinate, but bees don’t get any easier to spot from a distance when you’re also a bee.

Bees tend to follow something like the Lévy flight rule, which is a fun algorithm because it looks really dumb, but it’s been proven to be optimal under some assumptions, and it’s common in nature and technology.

A random Lévy flight a bee might take .

Paths traveled by ten bees in a simulated Lévy flight. Each color is a different bee, each using the same rule.

One hundred bees.

One thousand.

As you scale up, the wisdom in the randomness reveals itself. It all adds up to a dense coverage pattern that gradually thins out the further you get from the hive. It works under a broad variety of conditions.

But any rule has its “adversarial” cases—worst-case scenarios a hypothetical or real enemy might set up so that your rule performs poorly. In this case, if you place most of the nectar in a spot near the limit of their search radius, the bees probably won’t find it, because to find it they’d have to fly in a long straight line. Bees only beeline when heading home, because for most of their time on Earth, that hasn’t been the patterns flowers have grown in. Unfortunately…humans.

Source

The neat strips we like to mow are a problem for bees, sometimes, if they’re wider than their search radius. It’s more noticeable on big farms that get mown using tractors—fewer species of bees are able to find flowers across a larger gap.

In this model, the bees are probably going to widen their search to the left, because everywhere they go they find plants, some of which have nectar. But they might never realize there’s a whole strip to their right, too, on the other side of the vast wasteland.

That’s why some Belgian scientists and farmers have been testing a change to how they mow—instead of neat lines, mow in randomly-generated sine waves. This lets them mow the same total area as before, but distributed in a shape that lets bees find their way across more easily.

This seems to be working! After a couple years of this, you get larger and more diverse bee populations, improving the overall health of the ecosystem and your crops.

Explore/Exploit

Both the bees and the farmers in this story face a classic tradeoff with an unfortunate technical name. When you’ve found something that works, do you exploit that discovery forever, or do you explore for even better ones? The right answer at the individual level tends to be fiendishly complicated, but the right answer for a society is usually pretty simple: diversity and individualism. If different people follow different paths, you win no matter what. Either the explorers find more nectar and alert the rest, or they come home empty-handed but to a hive with the slack to support them, thanks to their parallel work exploiting what they already had.

It’s…not common to describe bees as individualistic, but I think it fits here. Part of what makes the Lévy algorithm optimal is that different bees travel different amounts—most stay close to home, but a few travel further, and then a few of those travel further still, and so on. They’re not coordinating using hive mind telepathy (probably), they’re just all doing their own thing and sharing the rewards.

Similarly, the farmers in Belgium already had a phased mowing system that worked, and no proof that it wasn’t optimal. So almost all of them stuck to that, while a few experimented with wiggly mowing. So long as you have institutions that help you share the risks and benefits, this tends to work the best, which is convenient because some of us are natural risk-taskers and most aren’t. Presumably due to similar evolutionary pressures to the ones that shaped the bees.

And Justin Bloom, if left alone,
Would set the world on fire;
And Gontoseed, and all his breed,
Would stagnate in the mire.
While one would plunge in the abyss,
One saunter on the grass,
One holds back from the precipice,
One leaps the wide morass.

Though one is full of rest and sleep,
And one is full of noise,
They both together work to keep
The world in equipoise.

On this wide planet there is room
For both; and both we need.
Three cheers, three cheers for Justin Bloom!
Three cheers for Gontoseed!