Skip to main content

Navigating the most challenging habitats on earth: unravelling the architecture of a universal compass system

Periodic Reporting for period 2 - UltimateCOMPASS (Navigating the most challenging habitats on earth:unravelling the architecture of a universal compass system)

Reporting period: 2020-12-01 to 2022-05-31

When lost in the desert at night, or in dense forests, people tend to walk in circles. This is because the seemingly simple act of walking in a straight line involves a complex interplay of various sensory modalities, the motor system and cognition. Many insects released in the same type of uncharted territory, does not walk in circles, but rather keeps steadfastly to a chosen bearing. The main goal of this project is to unravel the sensory and neuronal architecture of the newly discovered ‘snapshot compass’ of the beetles, that supports orientation over all continents and terrestrial habitats on earth (except Antarctica). This goal is realized through a fusion of biology, bio-informatics and mathematics.

A quickly growing pool of studies indicates that neuronal networks are modulated in a context-dependent manner. Therefore, to truly understand how the tiny compasses of insects works, and to formulate the core computational algorithms underlying these remarkable systems, I aim to obtain the first ever brain recordings from the compass of a freely orienting insect. Is this possible? Certainly! But only in an animal with a robust orientation behaviour, and that is strong enough to carry a little backpack of electronics. The large dung-beetles, with their easily manipulated orientation behavior, offers a unique opportunity to attain this holy grail of neuroethology.

Three main objectives are currently being pursued over the course of this five-year project:
1) Input to the compass: To identify what celestial compass cues are extracted from different habitats, and how the sensory system is designed to support their extraction.
2) The compass at work: To characterise the neuronal principles, e.g. sensitivity, and connectivity, under-pinning the integration of compass cues in the snapshot compass system, and how these are modulated in different types of habitats and behavioural contexts.
3) A universal compass model: To formulate the core computational algorithms underlying a snapshot compass system that supports orientation in a broad range of light levels and terrestrial habitats.

While the challenge of cue integration has been solved effectively in the compass system of the beetles, it remains a key problem within the field of cognition and perception, as well as for the design of artificial intelligence systems.
Within this project we are asking several questions concerning the integration of different directional cues. Our new sky rooms has allowed us to characterise the directional reliability of an ersatz sun at different elevations and wind at different speeds as guiding cues for a species of ball-rolling dung beetle. We could then demonstrate that the relative reliability of these cues determines which directional cue that will dominate the network when the cues are put in conflict. Or in other words: what cue the insect will primarily steer by. By carefully modeling this system, we have further been able to show that the neural circuitry in the insect brain appears to provide an ideal substrate for this type of integration.

But how do insects remember how far to go and in what direction to move? That has been tricky to address, partly because insects seem to be remembering distance and direction through a strange combination of fast-updating and long-lasting memories. These two are typically considered incompatible aspects of memory, yet insects somehow combine them. We suddenly realized that we needed to freeze the beetles to understand this! This is because when the temperature is reduced to melting ice temperature (0ºC), electrical activity in the brain stops. If the memory state is maintained as an electrical signal, it is wiped out, but survives if it is stored in synapses between neurons. In this part of the project, we found that the cooled down insects still moved in the expected direction, but had forgotten the distance they should travel. Counterintuitively, the best explanation for this is not two separate memories but one common memory that encodes both the direction and distance combined. These findings have significant implications in understanding the role of memory for navigation and opens up for more exciting discoveries about navigational networks.

Cue integration is a relavant topic of study across a range of different species, from humans to insects. In our latest study we provide a novel, in-depth exploration of this topic. Our study goes beyond previous work by first characterising the relative reliability of wind and sun cues in isolation, then using this relationship to direct behavioural experiments along with computer simulations allow us to generate a concrete model of behaviourcue integration for directed behaviours. Furthermore, we describe a potential neural substrate for this model based on known anatomical properties of the insect central complex. This highlights the importance of our work in developing understanding of the insect central complex, which drives orientation and navigation behaviours in all insects studies to date.
So far in the project, we have progressed beyond the state-of-the-art in several areas. Firstly, we have shown that light pollution forces a change in dung beetle orientation behaviour, which is a major discovery. For the many other species of insect, bird, and mammal that rely on the night sky for orientation and migration, similar effects could dramatically hinder their vital night-time journeys.

Secondly, by disrupting the neural activity of beetles we have also shown that memories of distance are lost before memories of direction. Together, these findings present new insights into the functional properties of home vector memories and provide the first empirical evidence that a biological process that can be disrupted by cold-induced anesthesia is essential to support homing by path integration.

Thirdly, we have been able to complete an in-depth exploration of multimodal (sun and wind) cue integration in the dung beetle compass with an indoor setup which completely isolates the cues under study. Our descriptive modelling points to a vector-sum based integration mechanism, for which the neural circuitry in the insect central complex appears to provide an ideal substrate. This again supports the idea that the insect nervous system, despite its small size, is both sophisticated, flexible and remarkable.

Fourthly, we have pioneered new methods for manipulating and tracking orienting insects.

For the remainder of the project, we expect to determine exactly how directional cues are used together in a heading compass network, and how these relationships adapts to the current environment and state of the animal. A major (and very risky) goal of this project is to electrophysiologically record from the compass cells of the beetle while freely navigating on the savannah. This is still our goal during the remainder of the project. As a proof-of-concept, we next aim to run the modelled neural snapshot compass controller on the first complete version of the ‘beetle-bot’. The robot will be released side by side with the beetles on our next field trip in May. This will allow us to evaluate the performance of the model in different types of habitats.
A ball-rolling dung beetle (Kheper lamarcki) in our newly built skyroom (Photo: C Collinridge)