A blog post highlighting the article by H.F. McCreery in Insectes Sociaux

By Helen McCreery (@HelenMcCreery)

Ants are famous for their ability to carry heavy things. But some food items are too large and heavy for an ant to carry alone. In some species, ants that come across these large food items will return to the nest to recruit help to carry it home.

Ant species differ dramatically in how effective groups are at cooperatively transporting these items. A small number of species – such as weaver ants and crazy ants – can rapidly carry extremely heavy objects many thousands of times the mass of each individual. On the other hand, most ant species are relatively uncoordinated, and may not succeed at all.

What makes some ant species excel at cooperative transport? One major challenge of cooperative transport is that the group must agree on a travel direction. I hypothesized a behavioral trait possessed by individual ants – persistence – affects groups’ ability to coordinate their efforts in a particular travel direction, and thus, to succeed.

Persistence is a measure of how likely an individual is to change the direction she’s trying to move an object – or even to give up – if she’s unsuccessful. Highly persistent ants will keep pulling in the same direction for a long time, even if it’s not working. If a group is full of highly persistent individuals, they may fail because they pull in opposing directions for a long time. On the other hand, if a group is full of low-persistence individuals, they may change direction too frequently or simply not try for long enough to get going. Some previous work (McCreery et al. 2016) showed that in most cases high persistence improves coordination, at least in a theoretical context. I tested this hypothesis experimentally, by comparing cooperative transport ability in species that differ in persistence, and by manipulating persistence in groups of one species.

For the first part of this project, I worked with four ant species that ranged from impressively successful at cooperative transport (crazy ants: Paratrechina longicornis) to poor cooperative transporters (Formica pallidefulva), with Formica podzolica and Formica obscuripes making up the middle ground – groups in these two species seem to be moderately effective at carrying objects together (see video below). I measured persistence for dozens of individuals in each of these species by watching them complete an impossible task – trying to move a dead cricket I pinned to the ground. Separately, I measured the degree of coordination of groups in each of these species by recording groups engaged in cooperative transport. I measured two aspects of their coordination: success fraction (in what proportion of attempts did groups move at least 10 cm?) and path directness (did they move in a straight line or did they change direction frequently?). If species with highly persistent individuals were also more successful, it would support my hypothesis.

Video: H. McCreery

In the four species that I compared, I found just that pattern. Importantly, I found that there was substantial variation among species in both persistence and in coordination. P. longicornis individuals were the most persistent by far, and formed groups that were the most coordinated, succeeding in over 97% of attempts. On the other hand, F. pallidefulva individuals had dramatically lower persistence and were the least coordinated – F. pallidefulva groups only succeeded in about a quarter of attempts. The two other species, F. podzolica and F. obscuripes, both had individuals that were moderately persistent, and formed groups that were moderately coordinated – these groups succeeded most of the time, but with relatively indirect paths, they tended to travel about twice the distance they needed to because they frequently changed direction. High individual persistence is correlated with high group coordination in these four species.

In the second part of this project, I wanted to test whether coordination could be improved by increasing persistence so I manipulated persistence in groups of F. podzolica. I did this by measuring the path-directness of groups, and then adding one or two infinitely persistent “fake ants,” and measuring path-directness again. I added fake ants by mimicking the pulling force that such ants would apply on an object. I gave groups of ants heavy baits attached to string. Using a pulley system, I could add a pulling force on the baits by adding weight on the end of the string (see above figure). Check out the video below to see how I measured the pulling force of these ants.

I measured the pulling force of F. podzolica workers by recording them pulling on coiled chains. They enthusiastically pulled on these chains without bait!
Video: H. McCreery

The fake ants that I added to these transport groups were infinitely persistent, because the string would never stop pulling and never change direction. While the force of one ant made no difference, adding the pulling force of two ants going the same direction seemed to moderately improve coordination. These groups tended to move with more direct paths after I added the fake ants. I was surprised to see even a moderate effect, given that I only mimicked a single possible cue (the pulling force); my fake ants did not mimic anything else about real ants. Together with the results from the first part of this project (species comparison), these results suggest an important role of individual persistence in coordination during cooperative transport.

I found that species with more persistent individuals formed more coordinated transport groups, and that artificially increasing persistence moderately improved coordination. It can be difficult to draw strong conclusions when comparing across species, and indeed, the four species I evaluated differ in many more traits than just persistence. It is possible that another difference among these species, instead of persistence, has a more important effect on coordination. Despite this difficulty, comparing cooperative transport across species that dramatically differ is an important step in understanding the mechanisms of coordination. I would be interested to see if other species, beyond these four, fit a similar pattern with respect to persistence and coordination. The results of this study further suggest additional, targeted experiments that could get at the specifics a causal relationship between persistence and coordination.

Reference

McCreery HF, Correll N, Breed MD, Flaxman S (2016) Consensus or Deadlock? Consequences of Simple Behavioral Rules for Coordination in Group Decisions. PLOS ONE 11:e0162768. DOI: 10.1371/journal.pone.0162768