As humans, we are used to seeing the world in colour, identifying and categorising objects, and memorising their locations. It’s intuitive to us. Can we say the same for insects? In addition to mere categorisation, can they also learn to associate specific objects or object properties with a reward (e.g. a coloured flower-shaped object = food)? Using that example, do they associate any coloured object with food or can they even sort out different colours (blue-coloured flower-shape = toxic, red-coloured flower-shape = food)? These are all questions for which an answer can’t be taken for granted, or we risk being drearily anthropocentric: by automatically saying that they can do all these things without experimental data, we are projecting our own abilities on them; by saying they cannot do these things without experimental data, we are stroking our phylogenetic ego (my new band name, btw).
I will not be exploring all these questions in detail here; suffice it to say that yes, they can. What I do want to explore is navigation in ants, which is visually-mediated (a symplesiomorphic trait), because what they can achieve is really quite remarkable. I’ll be concentrating on the ants for two reasons: they kick ass and they’re model organisms for this research (as are bees, a tradition going back to von Frisch (1914)), since they need to have the ability to navigate fluently in their habitats, often using visual clues, in order to find food and their nests. Also, they’re easy to experiment on (as the myrmecologist [ant specialist] who advised David Attenborough in his Trials of Life documentary series says (paraphrasing): you can observe ants like a god. You know their every move and they don’t even know you exist). In addition, some ants have colour vision (Camlitepe & Aksoy, 2010), can perceive shapes and patterns, as well as high speed motion. Perhaps I’ll make this a series and go on to do the same for other insects; time (and requests) will tell.
In ants, navigation has specifically been studied, especially in desert ants (Wehner, 2003). In the desert, the landscape is relatively barren and the ants go over 100 m away from their nests looking for dead arthropods (i.e. food), and the main question in these studies revolves around what cues they use to find their way back to their nest – laying pheromones is out of the question because they evaporate immediately in the heat and the road to a new food source is too tortuous for a pheromone trail to be followed. It turns out they can use the stars (the sun being the main one, of course), including the moon at night (Klotz & Reid, 1993), as well as landmarks (Graham & Cheng, 2009). Besides the actual position of the star, the polarisation pattern produced by light rays as they are scattered by particles in the atmosphere (Rayleigh scattering; this includes moonlight (Gál et al., 2001)) can be detected by ants, as can the colour of the sky (Wehner, 1997), and both are used as compasses (Wehner & Labhart, 2006) (as an aside, the Vikings used crystals, probably of calcite, to detect polarisation of sunlight and thus as navigational aids (Konnen, 1985)); basically, they measure the intensity of long wavelength light compared to the UV background to see how it changes (Wehner, 2003). Note that the compasses are used independently – the scattered and direct light are both processed and compared, not combined (Wehner & Müller, 2006). This is how they always return to their nests in a more or less straight line, without getting lost. They can also keep time with some circadian clock mechanism (not identified yet) so that their directions are up to date with the constantly changing light patterns.
It must also be said that the ants will not land immediately back at their nests. On their road to the food source, they take in the landmarks and count their steps (compensating for changes in steepness, of course), as well as use their compass, to know where they are. But they don’t follow the same road back to their nest, since when looking for food, they may go in several directions until they locate something. When they want to go back, they integrate all the knowledge accumulated to find their way back in a more or less straight line. But they won’t land exactly at their nest, but at some point close to it. This is where smell may come in handy, since nest entrances tend to be very inconspicuous. What they do from their location is go in a gradually widening spiral – eventually, they will find the nest entrance. Of course, they also have visual snapshots of what the area looks like, taken from some distance away and from several positions. When near the nest, they call up these images to remember where the nest is. This, of course, depends on how well their path integration system worked: if they don’t end up near the nest, they’ll die while looking for the entrance in this way. This path integrator is more or less fixed and doesn’t improve with training (Merkle & Wehner, 2009), most probably because the details of past searches are erased and only the final path integration result is kept.
But what about forest ants? They can’t see the sky or any starlight. But what they can see is the silhouette of the canopy, a canopy that is always changing and never the same in two places. Forest ants in fact memorise the entire canopy structure above them (Ehmer, 1999) and so can navigate through the forest without the aid of a compass. The coolest ant in this respect is the neotropical rainforest ant Gigantiops destructor – the ant species with the largest eyes (Gronenberg & Hölldobler, 1999). An individual worker can travel through 20 m of rainforest – with all the trees and other objects in the scenery – without using any chemicals (Beugnon et al., 2001). Importantly though, these ants do not need their path integrator unless they stray away from a settled foraging route – and even then, they only use it to find their way back to the trail.
Note that these techniques are not applicable to long-distance migrations, but only to short or medium distances to a food source, etc., referred to in the literature as “homing” distances.
I mentioned in my insect flight post that some tree-living ants can glide in a directed aerial descent. Here, their visual system is of utmost importance, since they can’t really follow pheromones while falling through the air. Instead, the way they find their tree is by looking at the colours: they actively orient themselves towards white columns (Yanoviak & Dudley, 2006) – in a forest, this is bound to be a tree (covered by lichens).
By the way, the important structure for visual navigation in ants isn’t the compound eye, it’s the individual ocelli found on the head (Mote & Wehner, 1980). This may seem surprising, given that ocelli produce a blurred and astigmatic picture (the light gets focused far behind the retina) – but the ant doesn’t need an exact 10 megapixel picture, it needs a quick spatial summary of its locations, and the ocelli are far more suitable for this than the compound eyes, since they are much faster at producing images. The polarisation detectors, however, are in the dorsal upper rim of the compound eyes (Fent, 1985) – the picture above (Labhart & Meyer, 1999) shows in which insect orders this area has also been discovered. In ants, the ocelli are also sensitive to polarisation (Mote & Wehner, 1980).
Of course, ants have another major advantage: eusociality and the associated group intelligence. This is best demonstrated when they make the move to a new nest or even when group foraging. All the ants have to reach the same destination, but each individual ant makes small mistakes on the trajectory – but if you look at all the individuals in the colony (as the ants themselves do – they can keep track of each other at all times), and take the average line from all their mistakes, you get a straight line to the destination. This is called the many wrongs principle (Simons, 2004). The accuracy of this ‘technique’ increases with group size.
This brings up a related problem, one that collectively-moving humans (in cars or on busy streets) also encounter: traffic. If there’s too much traffic on a specific trail, the overall foraging efficiency is reduced as workers move more slowly (Burd, 1996). The best ant to study this in is the leaf-cutting ant genus Atta, which forms extremely dense nests (10⁶ workers! [Fowler et al., 1986]) and depends on a highly-organised and equally dense system of foraging trails (Kost et al., 2005). These trails can be up to 300 m long and are clear of debris – on an ant scale, they are the equivalent of highways through the forest, leading directly from the residence to the workplace. Of course, like on all highways, traffic can very easily occur, especially when several thousand individuals are on the same tail going in both directions – there are no segregated lanes. If traffic becomes too high, there are three strategies: divert flow by clearing a new trail (Dussutour et al., 2004), create lanes in the trail for each direction (Couzin & Franks, 2003) or change the timing of the in- and outgoing flow so that everything is better synchronised (Dussutour et al., 2009). The importance of vision in this system should not be underestimated: in one experiment where the ants were forced to go through unidirectional trails (one to the food, one to the nest) despite being told by pheromones to go the wrong way, the ants successfully used visual cues to override any directions given by magnetic or chemical cues (Ribeiro et al., 2009).
However, ants do not always move in collectives. For example, when the ant Temnothorax albipennis needs to change nest or find a new food source, the scouts are sent out individually. The scout finds the new site, comes back to the nest and recruits only one individual. They go on what is called a ‘tandem run’ (Möglich, 1978), in which the scout teaches the worker the route to the new site – incidentally, the discovery of this process was the first example of teaching in non-human animals (Franks & Richardson, 2006). However, it isn’t extremely effective, as the follower ant gets lost halfway most of the time. In this case, the leader stays still – most of the time, the follower will return after a while, tapping the leader on the hind legs or the gaster with her antennae to signal that she’s ready to continue. If she doesn’t return, the leader will then look for the follower and stray out of the learned path.
Why go through all this bother? Surely the scout can simply carry the follower to the new site to show her the site is good (in fact, this would be three times faster!). The reason is the same reason why when you walk to a place, you have a much easier chance of learning the road than simply taking the bus: the follower, on her way to the new site, frequently looks around and sometimes even turns in a circle. She does this simply to have a look at her surroundings and memorise the landmarks. The leader knows this – that’s why she waits patiently. Ultimately, when the tandem run is complete, the follower can then become a leader herself by recruiting another nestmate and showing her the way. However, this system would not be possible at all if the ants were incapable of learning their surroundings.
It’s worthwhile to take a look at the neurological basis of this. The key brain part involved is the mushroom bodies (a.k.a. corpora pedunculata, but this term should be phased out), a connection first made by Bernstein & Bernstein (1969) for the wood ant Formica rufa. The mushroom bodies are the centers of sensory integration and memory, and they are plastic, their structure depending on life experiences added on to a ground plan. As an example, a transition from nest work to foraging work in an ant will lead to an increase in size of the neuropil and a decrease in size of the Kenyon cell body in the mushroom bodies (Kühn-Bühlmann & Wehner, 2006), because when foraging, the environments are much more complex than in the nest (there’s rocks, trees, leaves, bushes, puddles, etc.). It’s worth noting that it is only in the social Hymenoptera (ants, bees) that the mushroom body is so integrated with vision. Another brain part involved is the central complex (Heinze & Homberg, 2007), which takes care of locomotion and is linked to the protocerebrum; in fruit flies and locusts, it is where the polarisation of the light is interpreted. In fact, some brain parts will not really develop until the ant has extensive foraging experience (Wehner et al., 2004) – this is most noticeable in ants exhibiting age-dependent polyethism, i.e. different work tasks depending on age (such as the model desert ant, Cataglyphis bicolor; generally, most ants have this though).
Note that it’s only specific brain regions, not the entire brain that gets enlarged due to experience. Overall brain size in ants is more a function of colony size – the larger the colony, the larger the brain, because of the increased number of social interactions that need to be handled.
Proof that navigational behaviour originates in the brain – I only add this note because some people still believe in that Cartesian mind/brain duality bullshit, and some apply it to non-human animals (at least it’s more consistent than saying humans are special, so they get points for that. They’re still morons, though) – can be seen from our old friends, the behaviour-affecting parasites. The way Cordyceps manipulates the ant to go to its place of death is purely chemical, specifically chemicals in the brain (Mains, 1949); it’s the exact same situation with ants infected by Dicrocoelium dendriticum – if you don’t believe me, sometimes the larvae will settle in the brain (Romig et al., 1980).
And finally, I would like to add another note: you may notice a name appearing quite a lot of times in the literature list below, “Wehner, R.” It’s not an overstatement to say that Rüdiger Wehner single-handedly revolutionised the whole field of research on how ants navigate (especially with his model cataglyphid desert ant experiments that go back to the 1970s!), and his contributions to how animals use polarised light also deserve special mention. The dude’s work is amazing.
Oh, and one last thing: this path integration business and just generally how ants navigate is fundamentally the same as in humans – since both ants and humans evolved as central place foragers (i.e. they go from their nest to a single food source location, not multiple ones), they have converged on the same kind of navigation system.
Bernstein, S. & Bernstein, R. A. 1969. Relationship between foraging efficiency and the size of the head and component brain and sensory structures in the red wood ant. Brain Research 16, 85-104.
Beugnon, G., Chagne, P. & Dejean, A. 2001. Colony structure and foraging behavior in the tropical formicine ant, Gigantiops destructor. Insectes Sociaux 48, 347-351.
Burd, M. 1996. Ecological consequences of traffic organization in ant societies. Physica A 372, 124-131.
Camlitepe, Y. & Aksoy, V. 2010. First evidence of fine colour discrimination ability in ants (Hymenoptera, Formicidae). Journal of Experimental Biology 213, 72-77.
Couzin, L. D. & Franks, N. R. 2003. Self-organized lane formation and optimized traffic flow in army ants. Proceedings of the Royal Society B 270, 139-146.
Dussutour, A., Fourcassié, V., Helbing, D. & Denebourg, J.-L. 2004. Optimal traffic organization in ants under crowded conditions. Nature 428, 70-73.
Dussutour, A., Beshers, S., Deneubourg, J.-L. & Fourcassé, V. 2009. Priority rules govern the organization of traffic on foraging trails under crowding conditions in the leaf-cutting ant Atta colombica. Journal of Experimental Biology 212, 499-505.
Ehmer, B. 1999. Orientation in the ant Paraponera clavata. Journal of Insect Behaviour 12, 711-722.
Fent, K. 1985. Himmelsorientierung bei der Wüstenameise Cataglyphis bicolor: Bedeutung von Komplexaugen und Ocellen. PhD Thesis, University of Zürich.
Fowler, H., Pereira-da-Silva, V., Forti, L. & Saes, N. 1986. Population dynamics of leaf-cutting ants: a brief review. In: Logfren, C. S. & Vander Meer, R. K. (eds.). Fire Ants and Leaf-Cutting Ants: Biology and Management.
Franks, N. R. & Richardson, T. 2006. Teaching in tandem-running ants. Nature 439, 153.
von Frisch, K. 1914. Der Farbensinn und Formensinn der Biene. Zoologisches Jahrbuch der Physiologie 37, 1-238.
Gál, J., Horváth, G., Barta, A. & Wehner, R. 2001. Polarization of the moonlit clear night sky measured by full-sky imaging polarimetry at full moon: comparison of the polarization of moonlit and sunlit skies. Journal of Geophysical Research 106, 22647-22653.
Graham, P. & Cheng, K. 2009. Which portion of the natural panorama is used for view-based navigation in the Australian desert ant? Journal of Comparative Physiology A 195, 681-689.
Gronenberg, W. & Hölldober, B. 1999. <a href=”http://dx.doi.org/10.1002/(SICI)1096-9861(19990920)412:23.0.CO;2-E”>Morphologic representation of visual and antennal information in the ant brain. Journal of Comparative Neurology 412, 229-240.
Heinze, S. & Homberg, U. 2007. Map-like representation of celestial e-vector orientations in the brain of an insect. Science 315, 995-997.
Klotz, J. H. & Reid, B. L. 1993. Nocturnal orientation in the black carpenter ant Camponotus pennsylvanicus (de Greer) (Hymenoptera: Formicidae). Insectes Sociaux 40, 95-106.
Konnen, G. P. 1985. Polarized light in nature.
Kost, C., Gama de Oliveira, G., Knoch, T. & Wirth, R. 2005. Spatio-temporal permanence and plasticity of foraging trails in young and mature leaf-cutting ant colonies (Atta spp.) Journal of Tropical Ecology 21, 677-688.
Kühn-Bühlmann, S. & Wehner, R. 2006. Age-dependent and task-related volume changes in the mushroom bodies of visually-guided desert ants, Cataglyphis bicolor. Journal of Neurobiology 66, 511-521.
Labhart, T. & Meyer, E. P. 1999. <a href=”http://dx.doi.org/10.1002/(SICI)1097-0029(19991215)47:63.0.CO;2-Q”>Detectors for Polarized Skylight in Insects: A Survey of Ommatidial Specializations in the Dorsal Rim Area of the Compound Eye. Microscopy Research and Technique 47, 368-379.
Mains, E. B. 1949. Cordyceps bicephala Berk. and C. australis (Speg.) Sacc. Bulletin of the Torry Botany Club 76, 24-30.
Merkle, T. & Wehner, R. 2009. Repeated training does not improve the path integrator in desert ants. Behavioural Ecology & Sociobiology 63, 391-402.
Möglich, M. 1978. Social organization of nest emigration in Leptothorax. Insectes Sociaux 25, 205-225.
Mote, M. I. & Wehner, R. 1980. Functional characteristics of photoreceptors in the compound eye and ocellus of the desert ant, Cataglyphis bicolor. Journal of Comparative Physiology 137, 63-71.
Ribeiro, P. L., Helene, A. F., Xavier, G., Navas, C. & Ribeiro, F. L. 2009. Ants can Learn to Forage on One-Way Trails. PLoS ONE 4, e5024.
Romig, T., Lucius, R. & Frank, W. 1980. Cerebral larvae in the second intermediate host of Dicrocoelium dendriticum (Rudolphi, 1819) and Dicrocoelium hospes Looss, 1907 (Trematodes, Dicrocoeliidae). Zeitschrift für Parasitenkunde 63, 277-286.
Simons, A. M. 2004. Many wrongs: the advantage of group navigation. Trends in Ecology & Evolution 19, 453-455.
Wehner, R. 1997. The ant’s celestial compass system: spectral and polarization channels. In: Lehrer, M. (ed.). Orientation and Communication in Arthropods.
Wehner, R. 2003. Desert ant navigation: how miniature brains solve complex tasks. Journal of Comparative Physiology A 189, 579-588.
Wehner, R., Meier, C. & Zollikofer, C. 2004. The ontogeny of foraging behaviour in desert ants, Cataglyphis bicolor. Ecological Entomology 29, 240-250.
Wehner, R. & Labhart T. 2006. Polarisation vision. In: Jeffery, K. J. (ed.). Invertebrate Vision.
Wehner, R. & Müller, M. 2006. The significance of direct sunlight and polarized skylight in the ant’s celestial system of navigation. PNAS 103, 12575-12579.
Yanoviak, S. P. & Dudley, R. 2006. The role of visual cues in directed aerial descent of Cephalotes atratus workers (Hymenoptera: Formicidae). The Journal of Experimental Biology 209, 1777-1783.
Research Blogging necessities :)
Wehner, R. (2003). Desert ant navigation: how miniature brains solve complex tasks Journal of Comparative Physiology A: Sensory, Neural, and Behavioral Physiology, 189 (8), 579-588 DOI: 10.1007/s00359-003-0431-1
Ehmer, B. (1999). Orientation in the ant Paraponera clavata Journal of Insect Behavior, 12 (5), 711-722 DOI: 10.1023/A:1020987922344
Mote, M., & Wehner, R. (1980). Functional characteristics of photoreceptors in the compound eye and ocellus of the desert ant,Cataglyphis bicolor Journal of Comparative Physiology ? A, 137 (1), 63-71 DOI: 10.1007/BF00656918