
Paperis Articles
How Do Ants Find Their Way Home?
Why? · Ants
Your field's papers, as audio
Pick your fields and journals — Paperis summarizes the key papers and turns them into audio. Listen on your commute.
Start Paperis free →How Do Ants Find Their Way Home?
The child listens. The parent reads. This page helps you answer the questions your child may ask after the story. You can see the story script with the script button during play.
What this story is about
Children often see ants come to a crumb in a line. Then the ants carry the crumb back along the line. This story follows how people found out how ants get home. It looks at a test with smell on paper strips. It looks at desert ants with long legs (pig hairs glued on) or short legs (legs cut). It also looks at ants that were carried by other ants. Here is the answer. Ants use many ways together to find home. Which way matters most depends on the kind of ant and where it lives. Two kinds of ants are in this story. Some ants mark a path with smell and follow it. Desert ants in North Africa (Cataglyphis) leave no smell on the path. The "steps measure distance" test was done with desert ants.
How people found out
Ants that follow a smell path. Ants depend a lot on smell to find food and to find their way. Some of these smells are pheromones. A pheromone is a smell an ant puts out to send a message to other ants. Other smells are not pheromones (PMID 22137100). In one study, a team found the path smell of the pavement ant (Tetramorium immigrans) and made it in the lab. In the lab, ants followed this smell about 21 times farther than a path with no smell. Outside, the team put paper strips to an apple bait. One strip had the smell. One did not. The smell strip brought many ants to the apple (PMID 34738201). This smell made the ants' feelers respond (PMID 34738201). In another ant (Pheidole dentata), the nerve of one feeler was cut. These ants still did most of their jobs well. But they had trouble following a smell path. The team thinks ants need both feelers to tell where a smell is (PMID 28505612). The story does not talk about this feeler test. It only says ants smell the path with their feelers.
On a smell path, ants still use their eyes and the sky. One study followed 20 wood ants on long trips home. The ants shared one smell path. But each ant kept using its own way inside it. The team says this shows the ants use their own memory. They think what the ants see is likely the main cue (PMID 37993707). A desert harvester ant (Veromessor pergandei) walked on a smell path. When the team changed the light pattern above, the ants changed their way too (PMID 31422422). For this ant, the smell told the ant what to do, like follow, search, or go back. The sky told the ant which way to go (PMID 33677697). In another test, garden ants (Lasius niger) walked on a smell path on a turning table. The ants made a U-turn and kept walking toward the nest. They stayed on the path. The team thinks the ants stay on the smell path but use their eyes to choose the way along it (PMID 26909169). On a path with two forks, routes that turned left then right had more errors (32%) than routes that turned the same way twice (3%). On these left-right routes, errors dropped by 30% when the smell was there (PMID 22972897).
Desert ants make no smell path. The desert ant Cataglyphis fortis lives on salt flats in Tunisia. It looks for food alone. It can walk more than 100 meters. Then it must find a small nest that is hard to see. It is very hot, and food can be anywhere. So this ant does not make smell paths (DOI: 10.1186/1742-9994-6-5). Instead, it uses path integration (DOI: 10.1186/1742-9994-6-5). Path integration means the ant keeps adding up which way it went and how far. So it always knows which way home is, and how far (DOI: 10.1126/science.1126912). The ant can walk thousands of times its body length. Then it still goes home in an almost straight line (DOI: 10.1142/s0219878906000800).
The sky shows the way. Books used to say this. Bees and ants use the sun as their main compass. They use polarized light only when clouds hide the sun. A test with desert ants showed something else. Their sky compass uses polarized light most. In training, the ants saw only part of the light pattern. Later, they went home in the wrong direction in a set way. This stayed the same even when they could see the sun. With only the sun, the ants chose the right way (PMID 16888039). The top edge of the ant's big eyes has special cells for UV light. These cells read the light pattern (PMID 26138802). A test from 1985 covered the big eyes. The sun and landmarks were hidden too. Three small eyes on top of the head still read the light pattern and helped the ant go home (PMID 17779641). When the ants could not use polarized light, they walked between the sun way and the wind way. The higher the sun was, the more they went with the wind (PMID 17361400).
Distance by steps: the long legs and short legs test. Before this test, it was not yet clear how desert ants measure distance (DOI: 10.1126/science.1126912). Matthias Wittlinger and Harald Wolf from the University of Ulm in Germany worked with Rüdiger Wehner from the University of Zurich in Switzerland. A short news story in the journal (written by an editor, not the team) says this. Wittlinger saw something by chance. Ants with legs cut short before going home started to look for the nest too early. This made him think ants have a step counter inside (DOI: 10.1242/jeb.02688). The team trained ants on a 10-meter path from the nest to food. Then they caught ants at the food. For some ants, they cut the legs short ("stumps"). For other ants, they glued pig hairs to the legs ("stilts"). Then they gave each ant a bit of food. They put it on a test path right next to the training path. They marked where the ant began to search for the nest (DOI: 10.1242/jeb.02688). Ants on stilts thought they went farther and walked past home. Ants on stumps thought they went less far and stopped too soon (DOI: 10.1126/science.1126912). The summary does not say how many meters too far or too short. So the story gives no number.
Steps, not time? The team had to check one thing. Did the ants use the steps, or something else, like time (DOI: 10.1242/jeb.02688)? They filmed the ants with a fast camera. They measured step length, step speed, and walking speed. At the same step speed, ants on stilts took longer steps. Ants on stumps took shorter steps. Ants with longer legs also stepped a little more slowly, on average (DOI: 10.1242/jeb.02688). The ants walked almost the same as before. Their legs worked together very well (DOI: 10.1242/jeb.02657). The team used this data to correct for step size and speed. Then they guessed how far each ant would go. The guesses matched the test results. The team says this gives more support to the step counter idea (DOI: 10.1242/jeb.02657). The paper's summary does not say they tested travel time on its own. It only says the corrected guesses matched. So the story does not say "it was not time." It only says the guesses from step size and speed matched. The paper's words are "strongly indicative" of a step counter (DOI: 10.1242/jeb.02657). The word "proving" is only in the editor's news story (DOI: 10.1242/jeb.02688).
Eyes measure distance too. Some desert ants were carried by other ants. They did not walk. But they still measured the distance by watching the ground move past their eyes. This is called optic flow. This eye information did not pass to the step counter. The team thinks the two work on their own, side by side (PMID 27609893). In an older test from 2006, the testers moved ants without letting them walk. The ants saw the ground move. But they could hardly measure the distance. That team said optic flow helps only a little, and walking is needed (PMID 16823570). In 2006, people moved the ants. In 2016, other ants carried them. So the two tests are not the same. A study from 2018 covered the lower part of the eyes, which see the ground move. It started from a puzzle: how do the two systems join to make one answer? It showed the two distance systems keep separate memories and work together. The team gave two ideas. Maybe each memory counts more or less in different cases. Or maybe one memory wins. They gave a rough number, but the way they join is still not settled (PMID 30304010).
The last few steps: view and smell. Path integration makes small errors that add up. So it cannot bring the ant exactly home. The ant gets there by using views of the nest it learned before it went out for food (PMID 39904309). Desert ants were trained to link their nest door with one smell. In a test, they searched where that smell was. They could find the smell even in a mix of four smells. The team says desert ants use smells near the nest as landmarks (DOI: 10.1186/1742-9994-6-5). This is not the same as a smell path. Desert ants do use smell. They just leave no smell on the path.
Answers you often hear
"Ants follow a smell home." This is true for many ants, but not for desert ants. Ants that walk in a line follow a smell path (PMID 22137100, PMID 34738201). But desert ants in North Africa make no smell path because of the heat and food (DOI: 10.1186/1742-9994-6-5). Also, ants on a smell path still use their eyes and the sky (PMID 37993707, PMID 31422422). The story keeps these two apart.
"Ants use the sun." Half true. With only the sun, desert ants choose the right way. But when they can see both the sun and polarized light, they use polarized light most (PMID 16888039).
"Ants count their steps." There is good support, but add two things. First, this does not mean ants count numbers like people. It means the distance adds up with each step. Even the paper put "step counter" in quotes (DOI: 10.1126/science.1126912). Second, desert ants also measure distance with their eyes (PMID 27609893). Also, the test was done with desert ants. It does not show that all ants do the same.
"Ants smell with their feelers." True. The path smell makes the feelers respond (PMID 34738201). When the nerve of one feeler is cut, following the smell path gets worse (PMID 28505612).
Questions your child may ask
"Did it hurt the ants when their legs were cut?" The summaries behind this story do not say if the ants felt pain. They also do not say what happened to the ants after the test. We only know that the ants with changed legs walked well and did not trip (DOI: 10.1242/jeb.02688, DOI: 10.1242/jeb.02657). The story says only one fact: "they cut the legs to make them short." It does not say the ants felt pain, or that they did not. If your child asks, the honest answer is, "The study did not write it down, so we don't know."
"Do ants know numbers?" The test showed one thing. When step length changes, the ant gets the distance wrong (DOI: 10.1126/science.1126912). It does not show that ants know numbers. We also do not know well how the step counter works in the ant brain. Ant brains have not been studied much. So ideas about the brain map borrow from other insects, like fruit flies, silk moths, and bees (PMID 39904309).
"Do ants have a map in their heads?" Studies of desert ants point more to an adding machine than a map. In one test, ants walked a known path from food to home again and again, even after they reached home. Their path counter kept adding. In the end, the ants ran past home and away from it. The team says the ants do not use a map of where the nest and food are. They just follow what their path counter says (DOI: 10.1098/rspb.2004.2749). A paper about the desert ant's tool kit also says these tools are not made to give a map (DOI: 10.25849/myrmecol.news_012:085).
"Do the ants near us count steps too?" We can't tell. The step test was done with desert ants. An Australian desert ant (Melophorus bagoti) lives where there are many landmarks. In the same tests, it switched to using the view more easily than the desert ant from open flats (DOI: 10.1242/jeb.054601). Which cues an ant uses most depends on where it lives and what kind of ant it is.
What is known, and what is not yet known
What is known. Many ants follow smell paths (PMID 22137100, PMID 34738201). On those paths, they also use what they see and the sky (PMID 37993707, PMID 31422422). Desert ants find home with path integration and no smell path (DOI: 10.1186/1742-9994-6-5). They get direction mostly from polarized light (PMID 16888039). The leg tests show they get distance from their steps (DOI: 10.1126/science.1126912, DOI: 10.1242/jeb.02657). When other ants carry them, optic flow also gives distance (PMID 27609893). They find the nest door with the view and with smells as landmarks (PMID 39904309, DOI: 10.1186/1742-9994-6-5).
What is not known yet. How the two distance systems, steps and optic flow, join together is still a puzzle (PMID 30304010). One result from 2006 says ants moved by people, without walking, can hardly measure distance (PMID 16823570). Another from 2016 says ants carried by other ants can measure it with their eyes (PMID 27609893). The two tests were done in different ways. How the ant brain does this math is also not well studied in ants (PMID 39904309).
About the evidence. The step test comes from one research group. It has two papers (Science, 2006, and Journal of Experimental Biology, 2007) and one editor's news story. We could not check, from the summaries we collected, if another group did the same test again.
Things to try together
- Count your steps from the front door to the sofa, once with big steps and once with small steps. The same distance takes a different number of steps. Then walk with small steps, but only as many as the big steps. See where you stop. This shows what it means that "ants with short legs stopped too soon" (DOI: 10.1126/science.1126912). This is not a test of people's steps. Share it as a game that shows the idea of the ant test.
- When you see ants outside, watch together. Do they walk in a line, or alone? Ants in a line may be a kind that uses smell paths (PMID 22137100). But you cannot tell the kind of ant or its way just by looking.
- Help your child watch without touching the ant path or the ants. Right after the story praises the leg test, it says, "We never touch an ant's body. We can try it with our own steps instead." This is so the child does not hear it as an invitation to copy the leg test.
Sources
- "The Ant Odometer: Stepping on Stilts and Stumps" (2006) — DOI: 10.1126/science.1126912 — Desert ants on stilts thought they went farther, and ants on stumps thought they went less far. This supports the step counter idea.
- "The desert ant odometer: a stride integrator that accounts for stride length and walking speed" (2007) — DOI: 10.1242/jeb.02657 — Fast video showed the ants walked almost the same. The guessed distances matched the results.
- "PEDOMETER HELPS ANTS GET HOME" (2007) — DOI: 10.1242/jeb.02688 — An editor's summary: a 10 m training path, pig-hair stilts, a test path, a chance finding, and a wish to tell leg movements apart from other information, like travel time.
- "Smells like home: Desert ants, Cataglyphis fortis, use olfactory landmarks to pinpoint the nest" (2009) — DOI: 10.1186/1742-9994-6-5 — Desert ants make no smell paths in the heat, but they learn a smell at the nest door and use it to find home.
- "INFORMATION ACQUISITION IN DESERT ANT NAVIGATION" (2006) — DOI: 10.1142/s0219878906000800 — Desert ants walk thousands of body lengths and come home in an almost straight line, using a light compass and a distance counter.
- "Path integration in desert ants, Cataglyphis: how to make a homing ant run away from home" (2004) — DOI: 10.1098/rspb.2004.2749 — The path counter kept running, so ants going home ran past home.
- "The architecture of the desert ant's navigational toolkit (Hymenoptera: Formicidae)" (2009) — DOI: 10.25849/myrmecol.news_012:085 — An idea that the desert ant's tools are not made to give a map.
- "Vector-based and landmark-guided navigation in desert ants inhabiting landmark-free and landmark-rich environments" (2011) — DOI: 10.1242/jeb.054601 — Desert ants from places with many landmarks switched to the view more easily.
- "The neuroethology of ant navigation." (2025) — PMID 39904309 — Path errors add up, so ants use views near the nest. Ant brains are not well studied yet.
- "Trail using ants follow idiosyncratic routes in complex landscapes." (2024) — PMID 37993707 — Twenty wood ants each used their own way inside a shared smell path.
- "Not just going with the flow: foraging ants attend to polarised light even while on the pheromone trail." (2019) — PMID 31422422 — Desert harvester ants on a smell path changed their way when the light pattern changed.
- "Role of the pheromone for navigation in the group foraging ant, Veromessor pergandei." (2021) — PMID 33677697 — The smell told the ant what to do. The sky compass told it which way.
- "Rotating panoramic view: interaction between visual and olfactory cues in ants." (2016) — PMID 26909169 — Garden ants on a turning table stayed on the smell path and used their eyes for the way.
- "Ant foraging on complex trails: route learning and the role of trail pheromones in Lasius niger." (2013) — PMID 22972897 — On left-right routes, errors dropped by 30% with the smell.
- "Just follow your nose: homing by olfactory cues in ants." (2012) — PMID 22137100 — Ants depend a lot on pheromones and other smells to find food and their way.
- "Identification of the Trail Pheromone of the Pavement Ant Tetramorium immigrans (Hymenoptera: Formicidae)." (2022) — PMID 34738201 — Paper strips with the made smell brought ants to an apple bait.
- "Behavioral Performance and Neural Systems Are Robust to Sensory Injury in Workers of the Ant Pheidole dentata." (2017) — PMID 28505612 — When one feeler nerve was cut, only smell path following got worse.
- "The significance of direct sunlight and polarized skylight in the ant's celestial system of navigation." (2006) — PMID 16888039 — The desert ant's sky compass uses polarized light more than the sun.
- "Experience-related reorganization of giant synapses in the lateral complex: Potential role in plasticity of the sky-compass pathway in the desert ant Cataglyphis fortis." (2016) — PMID 26138802 — The light pattern is read by UV cells at the top edge of the eye.
- "Oceili: a celestial compass in the desert ant cataglyphis." (1985) — PMID 17779641 — With the big eyes covered, the small eyes still read the light pattern.
- "Wind and sky as compass cues in desert ant navigation." (2007) — PMID 17361400 — Without polarized light, ants walk between the sun way and the wind way.
- "Optic flow odometry operates independently of stride integration in carried ants." (2016) — PMID 27609893 — Carried ants measured distance with their eyes only.
- "Desert ants: is active locomotion a prerequisite for path integration?" (2006) — PMID 16823570 — Ants moved by people without walking could hardly measure distance.
- "Two distance memories in desert ants-Modes of interaction." (2018) — PMID 30304010 — The two distance systems keep separate memories. How they join is not settled yet.