Odor motion detection by an olfactory system aids navigation of complex odor plumes.

ORAL  · Invited

Abstract

For many animals, survival depends on the ability to navigate odor plumes to their sources. This task is complicated by turbulent air motions, which break continuous odor streams emanating sources into disconnected odor patches swept by the wind. Animal studies have revealed a general strategy to navigate odor plumes: reorient upwind when the odor is present, but go crosswind or downwind when signals become sparse to regain contact with the plume. In this strategy, the olfactory system is used to detect the identity, intensity and arrival time of odor packets, while the main directional cue is wind direction. This is because gradients of odors, which can be detected by comparing odor intensity between the two antennae, tend to fluctuate in many directions.

We discovered that besides detecting the identity and intensity of odor packets, the Drosophila olfactory system also detects the direction of motion of odor packets. Simulations and theory show that odor motion provides a secondary directional cue, which points towards the center of the odor plume and therefore is complementary to the wind direction. Using a virtual reality setup to decouple wind from odor signal, we find that flies detect odor motion from the temporal correlations of the odor signal between its two antennae, in a computation similar to motion detection in vision. Manipulating spatio-temporal correlations in the virtual odor signal demonstrates that flies indeed exploit odor motion when navigating odor plumes. Thus, Drosophila can compute the direction of motion of odors independent of the wind, and they use this capability in natural plume navigation. This work suggests a novel role for previously observed bilateral signal processing in the olfactory circuit.

*Funding: Swartz Foundation for Theoretical Neuroscience, NIH F32MH118700, and K99DC019397 (NK), Yale PEB program (MD), by NIH R01GM106189 and R01GM138533 (TE), by NSF IIS-1631864 (BM and MR), and by NIH R01EY026555 and R01NS121773 (DAC).

Publication: Kadakia N, Demir M, Michaelis BT, Reidenbach MA, Clark DA*, Emonet T*, Odor motion sensing enables complex plume navigation, bioRxiv 2021.09.29.462473; doi: https://doi.org/10.1101/2021.09.29.462473 *co-corresponding authors

Presenters

  • Thierry Emonet

    • Yale University

Authors

  • Nirag Kadakia

    • Yale University
  • Mahmut Demir

    • Yale University
  • Brenden T Michaelis

    • University of Virginia
  • Matthew A Reidenbach

    • University of Virginia
  • Damon A Clark

    • Yale University
  • Thierry Emonet

    • Yale University