Scale Invariance in Lateral Head Scans During Spatial Exploration

Universality connects various natural phenomena through physical principles governing their dynamics, and has provided broadly accepted answers to many complex questions, including information processing in neuronal systems. However, its significance in behavioral systems is still elusive. Lateral head scanning (LHS) behavior in rodents might contribute to spatial navigation by actively managing (optimizing) the available sensory information. Our findings of scale invariant distributions in LHS lifetimes, interevent intervals and event magnitudes, provide evidence for the first time that the optimization takes place at a critical point in LHS dynamics. We propose that the LHS behavior is responsible for preprocessing of the spatial information content, critical for subsequent foolproof encoding by the respective downstream neural networks.

Figure 1

Spatiotemporal distribution of LHS behavior. [(a)–(d)] LHS profile during open field exploration. (a) A representative LHS during the pause along with its movement before and after the scan. Note the forward shift in the animal’s body position ($+$) before and after the LHS; there is a directional change in nose trajectory (trailing white line) only during the LHS. (b) Movement trajectories (in grey) and temporal occupancy maps (units in seconds, color coded in log scale) of individual rats, superimposed with LHS (in black) during 90 minutes of exploration. (c) Correlation between LHS counts and respective occupancies of all the rats with fit line for linear regression. (d) Distribution of LHS over time, and LHS counts for each rat. (e) Scale invariant distribution in LHS variables. Probability distribution functions of LHS lifetimes, interevent intervals, and magnitudes [sum of absolute instantaneous angular deviations ($\sum |\mathrm{\Delta }\theta |$) from start ($ss$) to end ($se$) of each LHS], shown for a representative rat with four different step sizes, plotted in log-log scale. Black dots represent power law scaled probabilities, while grey dots represent out scaled probabilities. The values next to each panel are the power law exponent $\pm$ standard error, coefficient of determination ($R^2$) and significance of the exponent ($p$ value).

Figure 2

Comparison of exploratory patterns in different behavioral states. The movement trajectory of an example rat each under different behavioral states (full duration of 90 min) revealing transition through the critical phase of exploration under normal conditions, with unstructured (random) or highly structured (ordered) explorations under the influence of partial anesthesia and amphetamine.