Eric Feferberg/AFP/Getty Images
This question originally appeared on Quora, the knowledge-sharing network where compelling questions are answered by people with unique insights. You can follow Quora on Twitter, Facebook, and Google Plus.
Answer by Ted Pavlic, research scientist in social-insect lab:
Although olfaction (sense of smell) is certainly important to ants, the memory story involves much more than just chemosensory recall and recognition. Different species of ants have different memory capabilities for smell, vision, and even things like the distance and direction of their home nest based on feedback from their step count.
First, it is important to address what is not clearly memory. It’s not clear how much of the “memory” discussed in Matan Shelomi’s answer can definitely be attributed to central nervous system mechanisms over peripheral nervous system mechanisms, let alone neurology over genetics and physiology. For example, does an ant “remember” the smell of its colony, or is there a genetically derived colony odor? Or do workers use the smell of their own cuticles for reference? Regardless, how much recognition goes on in the sensory centers of the brain opposed to odor receptors in the antennae? There is evidence for different mechanisms in different ants, and some of those mechanisms are difficult to interpret as memory. In general, this is a very active area of research right now.
Having said that, the olfaction of honeybees (Apis mellifera) is often used in laboratory studies of learning and memory, and honeybees are very closely related to ants. In particular, researchers make use of the proboscis extension reflex, where a honeybee extends her “tongue” when she detects certain odors. By pairing a responsive stimulus with a novel stimulus, the honeybee can then be trained to respond to that novel stimulus as well (Hebbian learning, those that fire together wire together). Thus, she learned the smell of the novel stimulus; arguably, she remembered it. Relating back to ants, an analog of the PER for ants has been done in at least one case. Ants can also be trained to respond to novel odors after pairing them with new odors.
The mechanisms and behaviors associated with recognition of colony odor (as opposed to odors associated with foods, as in PER) are more complicated. It is true that ant colonies in the laboratory will often tolerate ants from other laboratory colonies over time, but it is not clear that they are becoming more tolerant of different colonies. For example, their nests could simply be gaining the same scent due to tightly controlled laboratory conditions. So it could become difficult for ants to synthesize and/or recognize unique cuticular compounds of other ants; everyone starts to smell the same. You don’t need any learning to drive this latter phenomenon. However, a more interesting case involves observed aggression in the wild between ants from neighboring colonies but not from colonies far away. In these cases, ants behave as if they are learning and remembering the scent of colonies near to them and associating those scents with aggression. Ants with unfamiliar (unlearned) colony scents do not lead to aggressive responses. Again, cuticular recognition is a hot area of research right now. So a lot of these ideas are speculative or isolated in only a handful of studies.
In regard to ant visual capabilities, ants such as Gigantiops and Harpegnathos have very good vision. They use it for hunting and can actually jump onto their prey to make the kill. Harpegnathos in our lab will follow your finger as you move it around their heads (which are equipped with very large eyes). So although some ants, like army ants, are entirely blind, other ants have fantastic vision.
It is true that ants like Cataglyphis (which are also large-eyed hunters) have been shown to respond to polarized light in the sky to help navigate home (which involves memorizing the position of “home” within this landscape). This sensitivity to polarized light has been shown in other insects as well, like honeybees (which are also able to use optical flow and can communicate the polar coordinates of a nest to nestmates who can then “remember” those coordinates and navigate back and forth). However, the memory of Cataglyphis isn’t only for visual cues. They have also been shown to count their strides to update a memorized “home vector” that maintains distance and direction from their current position to home. Although the outward trip of a lone Cataglyphis is circuitous, her return path is straight along this home vector. By artificially lengthening or shortening a forager’s legs for the return portion of her trip, she will overshoot or undershoot her nest entrance by a predictable amount. Thus, she is not simply responding to external cues about her nest entrance (which would require memory on their own). She is navigating based on an internal memorized map that she generated by counting her strides. It may be worth noting that Cataglyphis cannot use pheromone trails as they are native to deserts where it is so hot that pheromone trails would decay too quickly to be useful without very frequent reinforcement. These ants can forage in isolation. Thus, they are unable to maintain a pheromone trail and instead have to depend on vision and local memory for navigation.
And then there are the tandem runners, like Temnothorax, some Camponotus, some Pachycondyla, and even more. The details can vary in the tandem running use across these ant taxa, but a general description makes the point about memory here. Ants like Temnothorax can lead a follower to a new location without the use of a trail. The leader walks a short distance and waits to be touched by the follower. Meanwhile, the follower encodes the visual landscape as she sweeps her head from side to side. After learning the entire route, she can return on her own and even take novel paths—showing the destination as opposed to the entire route was actually communicated, and she memorized that location. There are variants of tandem running, like in Pachycondyla where the follower learns the path while being carried (without any chemicals being used along the trail). In general, these tandem behaviors show the use of both ant memory as well as visual capability.
There are many other examples, including whole colony nest-choice preference shifting consistently after being exposed to different intermediate stimuli. The mechanisms behind all of them aren’t always well understood, but there is enough known to prevent boiling down all ant memory to just following scents.