How do bees solve navigational challenges in 3D? – 3DNaviBee

We shall apply the following methods and technologies to reach the targeted objectives and tovalidate the above-mentioned assumption :
1. Development of a new methodology for recording and analysing the 3D movements of bees over several hundred meters based on an innovative radar application. This approach will allow for the implementation of new kinds of navigation experiments with bees in the lab and in the field and the development of powerful unsupervised statistical analyses of high-resolution 3D trajectories.
2. Running behavioural experiments on 3D navigation in bees across different behavioural contexts and spatial scales as bees learn to orient, search for resources, develop foraging routes between known resources, and return to their nest. This will bring entirely new information about the importance of height control for information sampling and decision-making by bees, that are out of reach with current methodologies.
3. Building of an integrative computational model of bee 3D navigation based on our empirical data. This will constitute the first theoretical framework to investigate computational mechanisms mini-brains use to navigate in 3D in ecologically relevant spatial scales.

For the first time, we have achieved the tracking of untagged flying insects (bumblebees) using the radar system developed in Task 2 and we have reconstructed the 3D flight trajectories of untagged bumblebees flying in a big outside tent.

3DNaviBee may have a strong important impact in fundamental research in electronical engineering by demonstrating that radar technology allows the tracking of untagged small and fast flying targets (bees) in the 3D space and at high spatio-temporal resolution. This methodology will enable to provide new fundamental knowledge on the biology of key pollinators, ask entirely new questions about insect navigation in the field, and explore new grounds into movement ecology. On the longer-term, we expect our methodology to represent a cornerstone in quantitative ethology as it could be easily modified to suit a wide range of animals. Finally, or models may also prove useful for improving pollination practices and conservation plans in the worrying context of generalised pollinator declines.

1. Even N, Bertrand O, Lihoreau M (2020) Navigation by honey bees. In « Encyclopedia of Animal Cognition and Behavior » (Vonk J, Shackelford TK eds). Springer.
2. Brebner J, Makinson J, Bates O, Rossi N, Lim KS, Dubois T, Gomez-Moracho T, Lihoreau M, Chittka L, Woodgate J (In press). Bumblebees strategically use ground-level linear features in navigation. Animal Behaviour.
3. Dore A, Pasquaretta C, Henry D, Ricard E, Bompard JF, Bonneau M, Boissy A, Hazard D, Aubert H, Lihoreau M (2020). A non-invasive radar system for automated behavioural tracking. bioRxiv. doi:10.1101/2020.12.09.418038.

How animals acquire, process and combine information about the world to accurately navigate is a fundamental question in biology. For small-brained insects, successful navigation may not require complex cognitive operations but instead emanate from optimized combination of relatively simple mechanisms involving vision. While almost all knowledge on insect navigation has been deduced from 2D recordings, many insects navigate in 3D. For flying insects, such as bees, gaining altitude may allow to look above trees, give access to additional visual landmarks on the ground, but may also profoundly modify the perception of travel distances.
The overarching aim of 3DNaviBee is to understand how flying insects use visual information to navigate in 3D across motivational contexts and spatial scales. We will focus on the buff-tailed bumblebee (Bombus terrestris), a model organism for insect cognition and navigation research, that shows sophisticated spatial behaviours and can be easily manipulated in experimental setups year-round. When foraging, bumblebees display complex behavioural sequences including phases of orientation, exploration and familiar route following. We hypothesise that bees can adjust their flight altitude to selectively make use of the height-dependent visual information in different behavioural contexts.
To test this hypothesis, electrical engineers, neuroethologists, cognitive ecologists and computational biologists from the universities of Toulouse (France) and Bielefeld (Germany) will cooperate to address three aims:

Aim 1. To develop a new methodology for recording and analysing the 3D movements of bees over several hundred meters based on an innovative radar application. This approach will allow for the implementation of new kinds of navigation experiments with bees in the lab and in the field and the development of powerful unsupervised statistical analyses of high-resolution 3D trajectories.

Aim 2. To run behavioural experiments on 3D navigation in bees across different behavioural contexts and spatial scales as bees learn to orient, search for resources, develop foraging routes between known resources, and return to their nest. This will bring entirely new information about the importance of height control for information sampling and decision-making by bees, that are out of reach with current methodologies.

Aim 3. To build an integrative computational model of bee 3D navigation based on our empirical data. This will constitute the first theoretical framework to investigate computational mechanisms mini-brains use to navigate in 3D in ecologically relevant spatial scales.

3DNaviBee may have a strong important impact in fundamental research in electronical engineering by demonstrating that radar technology allows the tracking of untagged small and fast flying targets (bees) in the 3D space and at high spatio-temporal resolution. This methodology will enable to provide new fundamental knowledge on the biology of key pollinators, ask entirely new questions about insect navigation in the field, and explore new grounds into movement ecology. On the longer-term, we expect our methodology to represent a cornerstone in quantitative ethology as it could be easily modified to suit a wide range of animals. Finally, our models may also prove useful for improving pollination practices and conservation plans in the worrying context of generalised pollinator declines.