The Ant SatNav


A new study has provided novel insight into the way ants memorize and navigate routes ‘home’ whilst out foraging.

Foraging Leaf Cutter Ants (c)  A. Bockoven

Foraging Leaf Cutter Ants (c) A. Bockoven


Ants have been subject to numerous studies regarding navigation techniques and information exchange across a social hive. These have shown that ants can navigate across large distances between their nest and various foraging locations thanks to visual cues. More recent studies have started to demonstrate that this ability to navigate is undiminished when travelling backwards – an event that would occur when the foraging ant is returning to the nest with a large piece of food.


Backward Navigating Ant With Large Food Item  (c) Sarah Pfeffer/University of Ulm

Backward Navigating Ant With Large Food Item (c) Sarah Pfeffer/University of Ulm

These findings have challenged the idea that ants use egocentric memories of an area in order to navigate.


Egocentric navigation is a self-to-object spatial coding system. This means that the ant represents the location of objects relative to its own location and transient axes, such as front-back, left-right and up-down.

Egocentric Navigation (c) Maria Kozhevnikov, Harvard Edu 2016

Egocentric Navigation (c) Maria Kozhevnikov, Harvard Edu 2016


Initial thoughts were that if the ant is travelling in the same area but with a different egocentric view – ie. the ant is travelling backwards, rather than when it first navigated the area, in a head-forwards direction – it makes the initial egocentric redundant. The ant shouldn’t be able to navigate. However this is not the case.

This means that the ant must be using other cues to navigate.


This study was conducted near Seville, in the natural environment of the desert ant Cataglyphis velox. The team buried barriers around an active nest to constrain the foragers to navigate along a one-way route that, whilst not affecting the efficiency of ant navigation, allowed correct observation of their activity. This route included 90 degree turns in order to better evaluate orientation capacities. Pieces of cookie were placed at the end of the run in order to encourage forager movement along it.

Close Up of Cataglyphis Velox - Desert Ant (c) Alamy

Close Up of Cataglyphis Velox – Desert Ant (c) Alamy

Once forager ants had run along the route sufficient times to both memorize it and avoid collisions with other foragers, the barriers after the turn were removed and their navigational technique was observed.

In order to return to the nest, it was assumed that ants would either follow visual cues – along the path they had been taught due to the buried barriers – or use their path integrator, their original path back to the nest.

Path integration is a spatial awareness factor associated with animals. It entails a knowledge of the path followed to reach a given point. The path integration factor (PI) is the knowledge of the linear path that needs to be followed in order to return to the initial departure point, even if the initial journey does not follow this linear path.

Path Integraation (c) Wikipedia 2017

Path Integraation (c) Wikipedia 2017


The study aimed to observe the ants choice of direction following a redirection caused by the 90 degree turn. Ants were placed at the end of the run and prompted to return to the nest with their piece of food.

Ants carrying small pieces of cookie – thus able to travel forward facing -proceeded along the ‘learnt’ artificial path back to the nest despite the barriers having been removed thus also the obligation to take this path. This demonstrates that they are exclusively using their visual cues to navigate.

Wood ant with Aphid Nymph (c) Solar 2017

Forward Navigating Wood ant with Aphid Nymph (c) Solar 2017


However, ants with large pieces of cookie, thus ants having to drag them backward facing, once exiting the barriers, travelled back to the nest using their PI – ignoring their visual memory that would have led them on the ‘learnt’ path.

So backward facing ants relied on a different navigational cue than the forward facing ant.

Additionally, ants traveling backwards but released in a novel area (ie. not the ‘learnt’ end of path) were observed as putting down their food and ‘peeking’ at their surroundings to get their bearings. Once they have gotten their bearings, they will continue backwards in a straight line and back to the nest.


So what does this mean?


This new study suggests that ants, when travelling backwards – ie. the ant is dragging the food object –  do not use the relative distance and immediately perceived objects in the area around them to navigate. Instead they use a combination of navigational factors to create its own ‘map’ home.

(c) Shadmia's World. WordPress.

(c) Shadmia’s World. WordPress.

Initially it appears that ants maintain a straight line, whilst travelling backwards, using their celestial compass (ie. navigation via cues in the sky, like sunlight, starlight and moonlight) rather than ground level spatial objects (such as plants and obstacles). It is also suggested that they use this in combination with their path integrator. Indeed, a deliberate modification of the sunlight during the study (using mirrors) caused the backward facing ants to change direction.


However the study has also found that, when travelling backwards, ants will also still ‘peek’ at the forward facing surroundings.

If the food being transported is too large to allow a forward ‘peek’, the ant will put down the food, turn around and move forward for a few steps, apparently to ‘get its bearings’ using the visual egocentric terrestrial cues it used to get to the food source in the first place.

So, having established these two navigational techniques that are not always used in conjunction, it appears that ants can maintain a direction of travel independently of their body orientation, explaining how they can navigate correctly whilst travelling backwards:

Indeed, it appears that whilst ants require visual egocentric alignment for ‘visual scene recognition’ – ie. the ant uses its eyes to analyzes and evaluate its placement within an environment – they can also use this acquired directional information – ie. where the ant came from and which direction it was travelling in – and use it as a holonomic frame of reference.


A holonomic frame of reference means that the coordinates observed – or in this case the visual cues observed – work in function with an observed time – when the ant travelled through the area. This way the any recognizes the distances and space between each visual cue but without itself being the centre of reference.

Using this technique allows the ant to decouple its travel direction – for example, North, South, East, West – from their body orientation – forward facing vs. backfacing – and thus navigate backwards without encountering any hindrance to its navigational capacity.

(c) Dan Sampson. Instagram.

Ant dragging Cheese (c) Dan Sampson. Instagram.

Overall this shows that ants are highly flexible when it comes to navigational information exchange: They can move from terrestrial to celestial cues as well as from egocentric to holonomic directional memories.

The cognitive or physical adaptations that have led to this style of navigation remain largely unknown. Grasping the complexity of ant behaviour remains an essential stepping stone towards a better understanding of the multiplicity of navigational tools observed in the natural world. It will also help understand the evolutionary procedures that led some creatures to maintain multiple tools whilst others didn’t.



Schwarz, S. Mangan, M. Zeil, J. Webb, B. and Wystrache, A. (2017) ‘How Ants Use Vision When Homing Backwards’ Current Biology 27: 1 – 7.

Read the full paper here: 

6,156 total views, 2 views today

The following two tabs change content below.

Lillian Stanton

Lillian is a student currently studying biology and media communications. She is hoping to follow a post graduate degree in scientific journalism.

Latest posts by Lillian Stanton (see all)

You may also like...

Leave a Reply

Your e-mail address will not be published. Required fields are marked *

Blue Captcha Image