Skip to main content

Multimodal navigation in insects: Use of olfactory, visual and idiothetic cues in ants

Final Report Summary - ANT NAVIGATION (Multimodal navigation in insects: Use of olfactory, visual and idiothetic cues in ants)

Being able to navigate efficiently through the environment is a requirement for most animal species, including humans. Solitary ant foragers, that don’t rely on social cues such as chemical trails, are specialized for individual navigation and little else, thus their movements are a window onto the spatial computations underpinning navigation. Allied to the practicality of studying ants in the lab or field, ant navigation is a rare opportunity to understand in detail how behaviour emerges from the interaction between brain, body and environment. The classic view is that ants rely on path integration (PI) when in unfamiliar terrain and as they become experienced they rely more on learnt, mainly visual, information (1-3). However, recent behavioural observations show that the ants’ navigational toolkit is more diverse, for instance using olfaction for orientation (4-6). Also we start to see interesting multimodal interactions (e.g. (7, 8)). We know little about the integration of multimodal cues at the behavioural-output level, and even less about the mechanisms behind. Yet such knowledge is necessary in order to understand the computational strategies that ants need for navigation in their environment.
In this Marie Curie IEF ANT NAVIGATION project (Project # 624765), we approached the question of cue integration by studying the fine details of ants’ movements during navigation. Such fine details of the ants’ movements can identify the computational strategies being used for navigation and are key for the understanding of navigational mechanisms. It was our goal to understand if insects need complex cognitive mechanisms to implement multimodal interactions or if simple movement strategies are sufficient for multimodal cue integration (see also:
The aim of our first objective was to look at the fine details of paths from ants navigating by path integration and at interactions with other sensory cues. We find that PI guided desert ants modulate their speed as they travel along their homing route and search for their nest. The walking speed decreases significantly along the ants’ homing path, i.e. the speed is about half as high at the end of a homing path compared to the maximum. More interestingly, we see that ants slow down before reaching their goal. Also, during the subsequent search paths their walking speed is lower. By moving slower, when path integration suggests they are close to the nest, ants allow visual cues to have a higher weight because they can act for a longer period of time. Two further sets of experiments have shown that the response to unfamiliar and familiar sensory cues depends on the position along the homing path. More specifically, we have seen that the further along the path the ants are the more disturbed they are by modifications in the visual scene and the stronger they weight learnt visual cues. Hence, the slower the ants are walking the more receptive they are both for unfamiliar and unfamiliar cues. Thus path integration produces slower walking at the times when other cues might be more important and walking speed could be an indirect mechanism for weighing cues. Based on these exciting findings we performed an additional experiment to investigate in more detail the influence of the walking speed on path accuracy and cue balancing (see also (8)). It seems that the walking speed positively influences the ants’ accuracy in both the use of vision and path integration. The influence on cue balancing is thus more complex, however, having data for different path lengths and different visual environments will allow us to analyse it in fine detail.
Our second objective was to explore the interaction of vision and olfaction in naïve and experienced ant foragers. It has been shown that desert ants pinpoint an inconspicuous nest entrance by following the plume of odour from the nest (6). But we don’t know how the use of olfactory cues depends on the structure of the visual environment. Further, learning walks are known to be performed by naïve foragers learning the visual world around the nest or when the familiar visual scene has changed (9, 10) whereby ants inspect the surroundings in very particular ways such that relevant visual information can be picked up. However, it is not known if learning walks are adapted to take into account the directionality and nature of olfactory information. The analyses of this objective are still ongoing.
The presence of a second sensory cue enhances initially the learning performance of a unimodal cue, but the components of the bimodal cue are fused together after several training trials and ants will no longer respond to either of the components presented alone (11). This demonstrates that there is an interesting multimodal interaction but as yet we do not have any information on whether these interactions might be mediated by simple sensori-motor gain mechanisms. The power of the experiments from our third objective is that we can go a step further and look in detail at the motor output of the ants when navigating in the different situations for what it can tell us about cue integration. Here, wood ants learnt to approach an inconspicuous and scentless feeder that is either marked with a unimodal or bimodal cue and the ants’ paths were video-recorded with a high-resolution camera during the training conditions and in different tests. We were able to replicate the finding of (11) by showing bimodally trained wood ants have a performance deficit when one cue is removed. Interestingly, those ants have slower and more sinuous paths and only a small proportion of the tested ants were able to approach the target. Also, when looking at paths characteristics of ants when they are either guided by vision, olfaction or both we see interesting differences in the walking speed, path straightness or accuracy. In general, it seems that the walking speed is increased in the presence of bimodal cues. Also, paths of ants purely navigating by the use of olfaction are more sinuous and thus longer. The addition of the visual cue makes these paths shorter but they don’t reach a straightness seen in ants purely using vision.
Understanding the mechanisms’ behind multimodal interactions is another step towards a deeper knowledge of how insects, with their small brains, can perform sophisticated navigational behaviours; An endeavour of general interest to the fields of animal behaviour, cognitive psychology, neuroscience and robotics. For instance, gaining a richer understanding of the computation involved in the navigational strategies of insects, can better inform the design of small autonomous robots that might one day match insects’ behavioural performance.

1. M. Collett, L. Chittka, T. Collett. Spatial Memory in Insect Navigation. Curr. Biol. 23, R789-R800 (2013). 2. R. Wehner. Desert ant navigation: how miniature brains solve complex tasks. J. Comp. Physiol. A 189, 579-588 (2003).
3. M. Knaden, P. Graham. The sensory ecology of ant navigation: from natural environments to neural mechanisms. Annual review of entomology, 61, 63-76 (2016). 4. C. Buehlmann, P. Graham, B. S. Hansson, M. Knaden. Desert
ants use olfactory scenes for navigation. Anim. Behav. 106, 99-105 (2015). 5. C. Buehlmann, P. Graham, B. S. Hansson, M. Knaden. Desert ants locate food by combining high sensitivity to food odors with extensive crosswind
runs. Curr. Biol. 24, 960-964 (2014). 6. C. Buehlmann, B. S. Hansson, M. Knaden. Path Integration Controls Nest-Plume Following in Desert Ants. Curr. Biol. 22, 645-649 (2012). 7. M. Collett. How Navigational Guidance Systems
Are Combined in a Desert Ant. Curr. Biol. 22, 927-932 (2012). 8. A. Wystrach, M. Mangan, B. Webb. Optimal cue integration in ants. Proc. R. Soc. B-Biol. Sci. 282, (2015). 9. M. Muller, R. Wehner. Path Integration Provides a
Scaffold for Landmark Learning in Desert Ants. Curr. Biol. 20, 1368-1371 (2010). 10. P. N. Fleischmann, M. Christian, V. L. Muller, W. Rossler, R. Wehner. Ontogeny of learning walks and the acquisition of landmark information
in desert ants, Cataglyphis fortis. J. Exp. Biol. 219, 3137-3145 (2016). 11. K. Steck, B. S. Hansson, M. Knaden. Desert ants benefit from combining visual and olfactory landmarks. J. Exp. Biol. 214, 1307-1312 (2011).