Melophorus bagoti Field Work

Field Work at Simpson's Gap


Landmarks of various kinds are used in the research. This landmark is a beacon right at the nest entrance, where the white lines painted on the ground intersect.



A grid is sometimes painted on the ground. This grid marks out an area between the feeder in the foreground and the nest in the background (where the white bucket is). This grid was made with spray paint, which does not last long. We now mostly use tent pegs stuck in the ground, around which strings can be wound to mark out a grid.

To get an ant to go home, we offer it a crumb of cookie in a container. Most ants are willing to grab a piece in their mandibles, and then they run home.

The Cataglyphis get Tunisian cookies. Melophorus get Australian cookies. Arnott's tea biscuits (Morning Coffee and Milk Coffee) are favourites. Maybe Arnott's should be the official sponsor of our research. In more recent years, mealworm pieces were added as reward, as a good number of ants prefer them.


photo by Barbara Muser

A grid allows us to plot the path of an ant on paper. Here, Daniela Strano of Macquarie University is doing the work (inset).

(In the other photograph, she is holding a Centralian blue tongue lizard at the Reptile Centre in Alice Springs.)

roadside nest

Sometimes, nests at odd locations can be useful for research. This one, indicated by the red flag, is right at the side of the road.

Daniela Strano and I set up a feeder at the side of the road opposite the nest. The road provides a uniform and open surface for tracking the ants. This informal research showed that ants displaced sideways, that is, along the road, adjusted for the displacement and can head towards the nest. They must have used the view of trees and bushes behind the nest somehow. Such piloting presages our pursuit of view-based matching in desert ants.

Field work at the CAT site

When we started at this field site, it belonged to the CSIRO Arid Zone Research. The organinisatin has since moved to a nearby premise, and the field site now belongs to the Centre for Appropriate Technology, who kindly still lets us do research on the premises. We are very grateful, and can say that CAT has been acknowledged in a good number of publications.

We have tested the role of ultraviolet (UV) wavelengths on how it affects the desert ants' navigation using the terrestrial panorama. We trained ants to come to a feeder, both in an artificial arena as shown in the photo, and in the natural scenery (imagine this set up without the circular arena). Our key manipulation was blocking the scene that the ants saw at the feeder with a transparent plastic that cut out most wavelengths under 400 nm. This would greatly reduce the stimulation for the ants' UV receptor, although not completely, a knock-down but not a knock-out manipulation. The surrounding plastic was uniformly tall.

We hypothesised that the biggest difference in UV-green contrast would define where the skyline is for ants. The sky contains lots of UV light, while earthly objects reflect a lot of green. This idea is due to Ralf Moeller. The UV-blocking clear plastic would make the biggest UV-green contrast at the top of the clear plastic, making the skyline uninformative. We tested ants that had gone home with food, were captured again and brought back to the start of the journey for a test. Indeed, these zero-vector ants oriented worse with the clear plastic, giving some support to the hypothesis.

In this figure, ants were tested with their natural scene. The tested ants had travelled to the feeder that they had visited many times, took food, returned home, and were captured just before they entered their nest. These ants are called zero-vector ants because they have run off the vector calculated from path integration. They were returned to the feeder for a test, and the graph represents their heading: where they crossed a circle at 30 cm distance from the release point. On the x-axis, 0 is the nest direction, while 180 and -180 both represent the opposite direction (nest-to-feeder). The graph is thus cylindrical. Ants attempting to head home with the UV-blocking plastic around them (UV block, grey dots) were more scattered in their distribution of headings. The curve fits are splines without theoretical import, meant to help readers to visualise the data.

The story is more complex, with more conditions effected. A paper has appeared: Schultheiss, P., Wystrach, A., Schwarz, S., Tack, A., Delor, J., Nooten, S. S., Bibost, A.-L., Freas, C. A., & Cheng, K. (2016). Crucial role of ultraviolet light for desert ants in determining direction from the terrestrial panorama.  Animal Behaviour, 115, 19-28.

One type of apparatus that we have used consists of long channels into which ants are forced to travel.

The rectangle at the near end goes around a nest entrance. The plastic is too slippery for the ants to climb over, and many of them take to travelling down the plastic channel. In experiments, we would cover the floor with a thin layer of sand to make it more natural and give the ants footing. At the far end would be a feeder, into which some venturesome foragers would eventually drop. We can then train them to come back via an inbound channel, or else transport them to a long test channel for testing. A measuring tape would be placed near the test channel in experiments to tabulate where the ants turn back and forth in searching for the nest.

We have gathered a large pile of paths by now, so that we have a paper characterizing their search patterns in channels. We have tabulated the distribution of lengths of search segments:

The story is quite complex, but suffice it to say that one model of optimal searching is supported by the data. The full story can be found in the published paper (Narendra et al., 2008, J Comp Physiol A, 194, 929-938).

But we have lots more to explore and understand about the ants' searching behaviour, in channels and over natural terrain. While in the Narendra et al. (2008) paper, the ants had a tendency to drift forward while searching in the channel, we have now found other conditions of searching in the channel in which this drift is not found. The work is in preparation.


We have also been examining a range of topics on how these ants navigate.


The ants will come inside an arena to forage. This gives us the chance to manipulate visual cues on the wall.

Eric Legge (in the picture below), a visiting student from the University of Alberta, came to do some arena work in 2009, along with his principal supervisor Marcia Spetch. Eric was here again in 2010 and 2011. A paper came out in 2010: Legge et al., Animal Cognition, 13, 849-806. Another of Eric's experiments came out in 2014: Legge, E. L. G., Wystrach, A., Spetch, M. L., & Cheng, K. (2014). Combining sky and Earth: Desert ants (Melophorus bagoti) show weighted integration of celestial and terrestrial cues.  Journal of Experimental Biology, 217, 4159-4166.

We have set up artificial obstacles to study the routes that the ants take to navigate.

Such 'route marks' may be displaced to pit the use of these landmarks vs. the broader panorama beyond.

This team effort resulted in: Wystrach, A., Schwarz, S., Schultheiss, P., Beugnon, G., & Cheng, K. (2011).  Views, landmarks, and routes: how do desert ants negotiate an obstacle course?  Journal of Comparative Physiology A, 197, 167-179.


Huge beacons may be set up with some effort. Antoine Wystrach, standing by the beacon, conducted a set of experiments with this huge and obvious beacon (to us anyway). But it was not obvious to the ants.

This work is featured in a paper published in an online journal in 2011: Wystrach, A., Beugnon, G., & Cheng, K. (2011).  Landmarks or panoramas: what do navigating ants attend to for guidance?  Frontiers in Zoology, 8, 21. pdf

From his thesis, Antoine Wystrach, currently at theToulouse Mind & Brain Institute, has accumulated an opus of several papers on M. bagoti elucidating how the ant uses the terrestrial panorama and other cues to navigate:

Wystrach, A., Beugnon, G., & Cheng, K. (2012).  Ants might use different view-matching strategies on and off the route.  Journal of Experimental Biology, 215, 44-55.

Wystrach, A., and Schwarz, S. (2013).  Ants use a predictive mechanism to compensate for passive displacements by wind.  Current Biology23, R1083-R1085. 

Wystrach, A., Schwarz, S., Baniel, A., & Cheng, K. (2013).  Backtracking behaviour in lost ants: an additional strategy in their navigational toolkit.  Proceedings of the Royal Society B: Biological Sciences, 280, 20131677.

Wystrach, A., Schwarz, S., Schultheiss, P., Baniel, A., & Cheng, K. (2014).  Multiple sources of celestial compass information in the Central Australian desert ant Melophorus bagotiJournal of Comparative Physiology A, 200, 591-601.

Wystrach, A., Philippides, A., Aurejac, A., Cheng, K., & Graham, P. (2014).  Visual scanning behaviours and their role in the navigation of the Australian desert ant Melophorus bagotiJournal of Comparative Physiology A, 200, 615-626.

Antoine has had a lot more publications since.

Other designs of channels have also been used. Here is a sophisticated set up created by Sebastian Schwarz to study discrimination learning in the ants.

The paper that has come out: Schwarz, S., & Cheng, K. (2010).  Visual associative learning in two desert ant species.  Behavioral Ecology and Sociobiology, 64, 2033-2041, DOI 10.1007/s00265-010-1016-y.

Other work on M. bagoti headed by Sebastian Schwarz:

Schwarz, S., Albert, L., Wystrach, A., & Cheng, K. (2011).  Ocelli contribute to the encoding of celestial compass information in the Australian desert ant Melophorus bagotiJournal of Experimental Biology, 214, 901-906. Doi: 10.1242/jeb.049262

Schwarz, S., & Cheng, K. (2011).  Visual discrimination, sequential learning and memory retrieval in the Australian desert ant Melophorus bagotiAnimal Cognition, 14, 861-870. DOI 10.1007/s10071-011-0419-0

Schwarz, S., Narendra, A., & Zeil, J. (2011).  The properties of the visual system in the Australian desert ant Melophorus bagotiArthropod Structure & Development, 40, 128-134.

Schwarz, S., Wystrach, A., & Cheng, K. (2011).  A new navigational mechanism mediated by ant ocelli.  Biology Letters, 7, 856-858, doi:10.1098/rsbl.2011.0489

Schwarz, S., Schultheiss, P., & Cheng, K. (2012).  Visual cue learning and odometry in guiding the search behavior of desert ants Melophorus bagoti in artificial channels.  Behavioural Processes, 91, 298-303.

Schwarz, S., Julle-Daniere, E., Morin, L., Schultheiss, P., Wystrach, A., Ives, J., & Cheng, K. (2014).  Desert ants (Melophorus bagoti) navigating with robustness to distortions of the natural panorama.  Insectes Sociaux, 61, 371-383. DOI 10.1007/s00040-014-0364-4

Ants use the skyline for navigation

Paul Graham of the University of Sussex visited the field site in 2008 and 2009. His experiments showed that the ants use the skyline to determine a direction for navigation. The skyline is some record of where the top of panoramic terrestrial scene meets the sky.

The ants were trained to come to a feeder for food. If a skyline mimicking what the the skyline from the feeder was made, the ants would follow that. The arena was at some distance from the training ground. Ants displaced to the middle of the arena followed the artificial skyline, even when the arena was rotated so that the skyline and the compass directions (available from the ant's sky compass) conflicted. A paper is published in Current Biology. pdf

Supplemental materials

We later found that when one side of an artificial skyline was higher than what the ants were used to in training, they veered away from that side a bit. Large changes in raising and lowering the skyline all around, on the other hand, had no adverse effects on orientation. We still have a lot to learn about how desert ants use the skyline.

Relevant paper:

Julle-Daniere, E., Schultheiss, P., Wystrach, A., Schwarz, S., Nooten, S. S., Bibost, A.-L., & Cheng, K. (2014).  Visual matching in the orientation of desert ants (Melophorus bagoti): The effect of changing skyline height.  Ethology, 120, 783-792.

Searching in M. bagoti

The red honey ant is by no means perfect navigating home with path integration and the terrestrial panorama. We have often seen foragers miss the nest on routine runs home from a feeder. Missing by just a few centimetres will require them do loop and search systematically. Like the search pattern of other ants, they start out with small loops at the start of search, and these spirals increase in size as the search goes on. The ant in the example path was searching for her nest.

Example search path recorded by Patrick Schultheiss

In one study, additional landmarks were set up near the nest, in the form of tall cylinders. The nest was covered by a board for the duration of the search -- this worked well enough to fool the ants. The team found that the ants' search loops were on average smaller when trained and tested with the landmarks compared with those of ants trained and tested without the cylinders.

Set up used by Patrick Schultheiss et al. in a study on search patterns

Some papers from this line of work:

Schultheiss, P., & Cheng, K. (2011).  Finding the nest: inbound searching behaviour in the Australian desert ant Melophorus bagotiAnimal Behaviour, 81, 1031-1038.

Schultheiss, P., & Cheng, K. (2013).  Finding food: outbound searcing behaviour in the Australian desert ant Melophorus bagoti.  Behavioral Ecology, 24, 128-135.

Schultheiss, P., Wystrach, A., Legge, E. L. G., & Cheng, K. (2013).  Information content of visual scenes influences systematic search of desert ants.  Journal of Experimental Biology, 216, 742-749.  doi: 10.1242/jeb.075077.

Reynolds, A. M., Schultheiss, P., & Cheng, K. (2013).  Are LÚvy flight patterns derived from the Weber–Fechner law in distance estimation?  Behavioral Ecology and Sociobiology, 67, 1219-1226 (DOI 10.1007/s00265-013-1549-y).

Reynolds, A. M., Schultheiss, P., & Cheng, K. (2014).  Does the Australian desert ant Melophorus bagoti approximate a Levy search by an intrinsic bi-modal walk?  Journal of Theoretical Biology, 340, 17-22.

A review:

Schultheiss, P., Reynolds, A. M., & Cheng, K. (2015).  Searching behavior in social Hymenoptera.  Learning and Motivation, 50, 59-67.

Other work on Melophorus headed by Patrick Schultheiss:

Schultheiss, P., Schwarz, & Wystrach, A. (2010).  Nest Relocation and Colony Founding in the Australian Desert Ant, Melophorus bagoti Lubbock (Hymenoptera: Formicidae).  Psyche, Article ID 435838, doi:10.1155/2010/435838.

Schultheiss, P., Schwarz, S., Cheng, K., & Wehner, R. (2012). Foraging ecology of an Australian salt-pan desert ant (genus Melophorus).  Australian Journal of Zoology, 60, 311-319.

Schultheiss, P., & Nooten, S. S. (2013).  Foraging patterns and strategies in an Australian desert ant.  Austral Ecology, 38, 942–951.


Simpson's Gap field site


back to Ken's home page