A look back and a look forward at the turtle ant project

Cephalotes varians soldier (left) 

and two workers (center and right)

It’s hard to believe that it has been 6 months since we received our first Cephalotes varians turtle ants! In that time, we have learned a lot about them.

Our collaborator Dr. Scott Powell and his lab at George Washington University sent us our first queenright colony in February, in a cozy little bamboo nest with a tiny hole just big enough for one soldier head. This first colony taught us a lot about keeping these turtle ants in the lab and helped us formulate our later plans.

Our new turtle ant colonies on their 

first day in our lab

It was June before we received the colonies that we would actually do our experiment on. We were so excited to get four queenright colonies and an additional five queenless fragments with which we could do more pre-experiment trials. We then had to think about what exactly we wanted our experiment to be and what we wanted to know about these ants. Turtle ants are especially interesting to study because of their physically different soldiers that block the entrances to their nests with their enlarged heads. They also live in multiple nest communities made by occupying cavities in branches made by beetle larva. Researchers like Powell and Dornhaus 2013 are particularly interested in investigating these ants’ defense strategies, particularly how colonies distribute their soldiers between their many nests, which has similarities to and possible implications for other defense systems like the immune system and plant defense systems. We wanted to take a step back and look at how the colony and its individuals choose the nests to occupy in the first place.  Because these ants must find cavities to occupy, careful investigation and selection is essential for these ant colonies to be able to grow and defend themselves. Powell 2009 suggests that workers may be the primary investigators and decision makers so we decided to put this to the test. To determine which individuals were making decisions and how they were doing it, we marked every single worker and soldier in our queenright colonies with individualized color patterns.

Painted turtle ants on a water feeder. Workers have three 

colors: one on their head and two on their abdomen. 

Soldiers have two colors on their enlarged head plate.

Forelius mccooki outside turtle ant 

nest with turtle ant soldier guarding 

Forelius mccooki identification picture

We also wanted to see if these ants’ nest decisions are affected by higher threat levels like the presence of other species of ants that could harm their brood or compete for resources. We hypothesized that higher threat levels may lead them to make more hurried and less ideal nest choices than if they have the time to investigate and consider all of their options. Alternatively, higher threat could make the ants pickier about which nests to occupy and prefer easily defensible ones. Thus we set out to find an ant species here in Southern California that we could use to induce this threat. The Robert J. Bernard Biological Field Station (BFS) generously allowed us to collect ants there. We then exposed our turtle ant fragment colonies to these ants to find the one that would produce a nest guarding response and found Forelius mccooki to do the job.

Layout of the experimental arena with 

destroyed boxes in the upper left box

It was then time to start our experiment! We put each colony in a four-box arena with bridges to get from box to box. In each box we put two nests, one large entrance nest that had an entrance the area of 4 soldier heads so it was hard to defend, and one small entrance nest that had an entrance the area of 1 soldier head so it was easy to defend. We decided to destroy their original nests to force the ants to move, since our observations of the first colony indicated that they might not move at all or would move very slowly if left alone. We also used continuous video recording as previous observations suggested that the ants move around all day and night and we didn’t want to miss anything. We also checked on the ants three times a day to get broad patterns and preferences and determine when to stop the experiment. After a week in the arena, the turtle ants had chosen which nests to occupy, so we stopped the experiment and started analyzing the video. Another student in our lab, Jingnan Shi, created computer code in MATLAB for us to screen through the many hours of video and pick out when ants were entering or leaving the nests.

Based on our daily checks, we found that 1) workers rapidly investigated all three new boxes after their original nests were destroyed, 2) larger colonies occupied more nests, 3) all colonies preferred large entrance nests, and 4) workers spread out between nests more than soldiers, who tended to group together. We expected the first two results, but the third was surprising. We expected them to prefer small entrance nests that would be easy to defend. One possible explanation of this preference for large entrances could be that the ants felt extremely exposed in their destroyed nest and time pressured to move their brood to a safe space and could more easily get their brood into large entrance nests, rather than needing to maneuver them through small entrances. The soldiers’ distribution is not surprising in light of this, as they would need to group together to effectively defend a large entrance nest, and thus could not spread out as much as the workers. So far we have not seen much difference between colonies with high versus low threat, but there is still a lot to do on this project, as much of the video still needs to be analyzed. We collected over 500 hours of video that we will use to track individual behaviors and see how they make choices about which nests to defend.