Recently when I’ve mentioned Bee Lab to my friends, their first question is “How can you do experiments when you’re remote?” Unfortunately, our ants aren’t too privy to Zoom, but we have the next best thing: video of previous experiments.
Usually, our experiments are set up to investigate how colonies make decisions. Because we couldn’t conduct any in-lab experiments, remote work provided an opportunity to look back at videos of previous setups in finer detail- how do individual ants make choices?
This summer, I’ve used video from our 2019 Summer experiments to learn about how individual ants navigate maze-like bridges. This was definitely not what the original experiment was designed to examine. Its original purpose was to help us understand how our turtle ants, Cephalotes varians, decide which nests to move into. This species has multiple nests per colony, and lives in holes that beetle larvae make in trees.
[1]: Two turtle ants stand next to a nest in a mangrove twig. (A third peeks out from the nest.)
We designed an experimental arena (see Marylin’s blog post) with nests on the top of five towers, connected to a central tower by the maze-like bridges. Some of the nests were easier to travel between (extra simple plastic bridges between them), so we wanted to see if those were the ones the ants seemed to like the best.
[2] Left: An experimental arena, with five nest towers and one central tower.
Each nest tower holds two nests (the red tubes with blue T-shaped bases).
[3] Right: The design of the maze-like bridges that connect each nest tower
to the center tower. Each junction of the “maze” has been given a label.
I was curious how the ants made decisions on these mini mazes. To gain some intuition, I started mapping out the paths that some individuals took. We had recorded video in the Summer 2019 experiments using this arena. I chose a couple promising clips and poked around to see if there was anything interesting going on. Here’s a pair of paths I drew out:
[4] Left: A bird’s eye view of the experimental structure. The bridges one particular ant traversed are marked with purple arrows. Notice that those two nest towers are connected by a thin bridge- there’s a much easier way to get between them! Center: The path this ant took from one nest tower to the central tower. Right: The ant’s slightly-more-complicated journey from the central tower to another nest tower.
Here, an ant had a super efficient journey one way across the bridge. It then climbed up the central tower, then went down another side of the tower. On its trip back across an identical bridge, it took a… very different journey. Was it lost? Did it just really like the view on one particular section? I don’t know what exactly was going on here. Paths like this made me really interested in learning more about how the ants were navigating these mazes.
Over the summer, I wanted to collect some data describing how the ants were walking along these bridges- I settled on recording the series of junctions of the maze that each ant traveled to (along with some other information). But to get this data, I needed to watch videos. A lot of videos. About 72 thirteen-minute videos.
Initially, it felt like this task was just a means to an end. My mindset was that I just needed to get through this repetitive task so that later, I could get to the fun stuff and play with the data. As I collected the data, I started taking some extra notes to keep myself entertained:
Oh my god ant 22
Ant go spinny
An ant falls off of tower 0 in like the last second of the video it's really funny
Ant 52 goes UNDER bridge!!! (shouldn’t be possible?!)
[5] One of the ants falling off of the central tower (oops).
(Don’t worry about the ants falling off the structure- they’re fine! They live in trees after all, so they’re built to survive falling off a branch.)
Initially, there didn’t seem to be any rhyme or reason to the behavior of the ants on the bridges. Some managed to cross the bridge, some ended up where they started, and the list of junctions the ants were traversing just seemed to me like a random series of numbers. If there was some sort of pattern, I didn’t think I’d be gaining much insight into it from watching these isolated moments on video.
However, along the way I began to learn more about how the ants were responding to our experimental setup. I started noticing some behaviors I wasn’t totally expecting. Some of the ants looked to be… intentionally jumping different places? Having read Rajan’s blog post I knew that this species is able to control the direction they glide, but I had never seen it before! Before they launch themselves, the ants lean forward, and maybe wiggle a bit as they align themselves with their target. It reminds me of how the squirrels in my backyard act when they’re about to make a risky jump to the bird feeder. Sometimes, the ants are jumping from the bridges or or nests towards the ground. Gliding is definitely faster than climbing down the towers, which also have the bowtie maze pattern on their sides. Other times though, the ants jumped from nests and aimed for the bridges! Initially I was a bit frustrated that the ants were “cheating,” but I grew to appreciate that witnessing this behavior gave me insight into ways we can improve our future setups.
[6] One instance of an ant jumping from a nest (red cylinder) to a bridge.
For instance, we weren’t expecting that the ants would treat the ground as their primary route between nests- we had assumed they would use the bridges. Another thing that grabbed my attention was that the ants didn’t seem to like the smooth plastic central tower. Often, upon reaching it, they would immediately turn around and return over the bridge they had just crossed. This information is super useful when we’re designing new experiments. It seems like the ants really didn’t like smooth, completely vertical plastic. If we want to make a surface for the ants to climb on, it seems that a smooth, entirely upright one isn’t a good candidate!
Finally, the notes I took about funny moments ended up helping me make sense of weird entries in my data set. For example, some of the journeys I recorded don’t start or end at the ends of the bridges. My records of the times the ants jumped onto or off of the bridges helped me confirm that I hadn’t misrecorded those junctions. In another instance, one of the walks an ant took lasted a long time, even though the ant didn’t pass through many junctions. Returning to my notes, I realized that this was the ant that seemed to take a nap in the middle of its journey, resulting in a very long recorded walk time.
I like to imagine that our ants have some colony-level plan to thwart my hopes for a nice dataset. Fortunately, I suppose, this (probably) isn’t the case. They’re living organisms, they make decisions, and can be unpredictable. In watching individual ants make choices, I’ve been amazed by how, despite individual behaviors not seeming to make much sense, colonies are still capable of making coordinated decisions that do.
[7] An ant spinning in place (video is at 5x speed).
Media Credits
[1]: Photo by Alex Wild. alexanderwild.com/Ants/Taxonomic-List-of-Ant-Genera/Cephalotes/i-PfqGtQQ/A
[2]: Photo by Matina Donaldson-Matasci.
[3]: Graphic by Fletcher Nickerson, design developed by Elena Romero, Marylin Roque, and Fletcher Nickerson.
[4]: Photo by Matina Donaldson-Matasci, graphic by Fletcher Nickerson.
[5], [6], [7]: Video recorded by Tom Fu, setup by Jarred Allen, Matina Donaldson-Matasci, Tom Fu, Kate Lord, and Marylin Roque.
No comments:
Post a Comment