Anesthesia disrupts distance, but not direction, of path integration memory

Abstract

Highlights d Anesthetizing ants affects their path integration memory d Anesthesia degrades the distance, but not the direction, memory of the ants d The same effect is replicated by a path integration model of the central complex d The results argue against recurrently maintained activity as the memory substrate

Vector navigation in walking bumblebees

“…As in previous studies of insect path integration following chill-coma anesthesia, the orientations of homeward paths following anesthesia were maintained. Although in ants 52 and dung beetles 53 the distances of home vectors were diminished after cooling, these distances were approximately correct across all manipulations in our experiments, suggesting very stable vector memories, potentially due to a long-term storage mechanism of home vector memories in bumblebees. However, experiments with ants revealed that these insects are able to correctly home after anesthetization, but only if they are in a familiar environment containing landmarks.…”

Can the Insect Path Integration Memory be a Bump Attractor?

“…The motor noise level was modelled as a decaying exponential function of the memory amplitude, as in Pisokas et al (2021). The memory amplitude a m was defined as the difference between the maximum and minimum value along the sinusoidal pattern of memory unit values (Pisokas et al, 2021). where y max is the maximum motor noise level (corresponding to zero vector ant paths), and s m is the slope of the exponential function.…”

“…The translational velocity of the simulated agents during homing was iteratively updated using the formula where v ( t ) is the agent’s velocity at time step t, a = 0.08 m / s 2 is an acceleration constant, Δ t is the simulated time step duration, and µ f = 0.15 is a friction constant. The velocity was reset to 0 m / s whenever the agent performed a turn larger than 20 ° simulating a stop on turning as observed in ants (Pisokas et al, 2021).…”

“…Unlike the ants, the bump's location drifts with a constant rate. Wittmann and Schwegler, 1995;Kim and Hallam, 2000;Vickerstaff and Di Paolo, 2005;Haferlach et al, 2007;Stone et al, 2017;Wolff and Rubin, 2018;Hulse and Jayaraman, 2020;Pisokas et al, 2021). Modelling has shown that the columnar organisation of the central complex provides a neural basis for potentially encoding and storing the coordinate values along eight columns of neurons, with each column corresponding to a cardinal axis (Haferlach et al, 2007;Stone et al, 2017).…”

“…To reproduce agent paths resembling those of ants we added motor noise that is parametric to the home vector length, as in Pisokas et al (2021). This motor noise was modelled as a Gaussian noise factor applied to the steering commands of the agent…”