Session M36: Animal Behavior and Social Interactions I

Cephalopod skin patterns as windows into brain dynamics

Coleoid cephalopods (octopus, cuttlefish and squid) possess a unique skin display system that operates under direct neural control. This allows them to very rapidly change their body color, pattern and texture. They use this system for many functions, most prominently for camouflage. In camouflaging, a cephalopod extracts a statistical summary of the visual environment, and then adopts a matching skin pattern so as to escape detection from predators. Modern experimental and computational tools now make it possible to study this complex process in detail. I will describe our recent work revealing the space of camouflage patterns and dynamics of pattern matching. While sleeping, cephalopods paradoxically show rapid skin pattern changes. We performed behavioral manipulations and neural measurements showing that this is a distinct stage of sleep, strikingly similar to vertebrate REM sleep in its wake-like properties. These include the reactivation of waking skin patterns during sleep, suggestive of functions involving motor control and/or memory consolidation.    

Characterizing the dynamics of syllable emergence in the juvenile songbird

As juveniles, songbirds, such as the Bengalese finch, learn to sing. Their songs are built from sequences of stereotypical acoustic units known as syllables. The dynamics of emergence of these syllables in the process of learning is not understood. During learning, syllables can be created through syllable differentiation, corresponding to an increasing difference over time between the acoustic structure of pairs of pre-existing syllables. Syllables can also emerge de-novo, with seemingly no relation of similarity to other syllables in the repertoire. To characterize the dynamics of emergence, the acoustic structure of songs is usually classified using manual labeling and quantified by computing scalar acoustic measures such as pitch, amplitude and spectral entropy. To go beyond such human-defined features, we use low-dimensional latent representations that can be learned from the syllables' spectrograms to analyze the vocal repertoire of songbirds at different developmental points. We classify syllables in each bird's repertoire using unsupervised clustering methods and analyze the dynamics of the latent representations in the time periods prior to and following syllable differentiations and creations. This allows us to describe the relation between the spectral structure of syllables and the timing of their appearance within different song contexts. This new understanding of the dynamics underlying the emergence of new syllables should facilitate the search for neural correlates of learned motor repertoires in the songbird.   

A Novel High Throughput Experimental Assay for Studying Distance Estimation

Visual distance estimation is a common and important task for survival and reproduction for both vertebrates and invertebrates. Nevertheless, it is primarily studied within vertebrates, particularly in primates, with large brains where neurons can be difficult to access. On the other hand, existing experimental assays used for probing distance estimation in invertebrates have tended to be low throughput. As a result, our understanding of the biological implementation of distance estimation at the level of complete neuronal circuits is limited. Here, we present a novel experimental assay that leverages a simply engineered rig and machine learning tools to provide a high throughput means of probing the neuronal underpinnings of distance estimation in insects. By combining this assay with common tools and techniques in neuroscience, we can dissect the visual circuits that detect distance in the small insect brain. Our results will reveal algorithms that generate distance estimates, their implementation in visual circuits, and what visual features contribute to visual distance estimation.   

Robust gait stability analyses using dynamical machine learning

While animals move through the world with apparent ease and stability, underlying these movements are complicated neuromechanical dynamics that we have had difficulties replicating in simulations or robots. In particular, understanding how gait is stabilized has proven difficult given the complex and intertwined dynamics of muscles controlling locomotion and the neural processes that allow an animal to adapt to perturbations in its environment. Floquet Theory is a promising data-driven methodology for quantifying and understanding gait stability, combining neural and mechanical dynamics into a single framework. However, its application has been limited due to the amount of data it requires and its sensitivity to measurement noise. Here we leverage dynamical system theory and generative machine learning models, combined with finite-size corrections, to enhance the accuracy, and robustness of Floquet calculations starting from realistic data set sizes. We show applications of this method from fly, mouse, and human gait dynamics. This approach opens avenues for the exploration of neuromuscular mechanisms of gait stabilization and can be readily applied to the study of periodic dynamics more broadly.   

Phase oscillator model of inter-limb coordination in free-running mouse locomotion

While locomotion is a basic part of the behavioral repertoire of most animals, it is modulated via a complex consortium of control structures and pathways yet to be fully understood. Mouse studies have revealed important gait motifs and neural control mechanism but have been mostly constrained to a framework of discrete measurements and perturbative experiments. Recent innovations in machine learning and computer vision have allowed for the continuous tracking of body part trajectories in natural settings, which has opened up a new dimension of possibilities for the detailed study of animal behavior in high spatiotemporal resolution. We track mouse limb trajectories in spontaneous, unconstrained locomotion in open arenas and model the associated limb dynamics as a system of phase-coupled oscillators. This linearized Kuramoto description outputs not only traditional gait descriptors such as phase offset and speed, but also novel measures such as inter-limb coupling and underlying noise structures. Using these model parameters, we generate a description and comparison of inter-limb dynamics at various locomotion speeds and turn angles, and investigate the source of naturally occurring variations in locomotion.   

Oral: Hydrodynamic study on the water-hopping behavior of mudskipper

Mudskippers, amphibious fish capable of living in water and on land, exhibit an intriguing behavior of hopping on the water's surface without submerging most of their bodies. Although their speed and momentum have been analyzed, questions regarding body-fin movement during splashes and energy efficiency remain unanswered. This study aims to explore the hydrodynamics of water-hopping in mudskippers through experimental setups and theoretical analysis. Using a saltwater tank and high-speed imaging, we capture the fin and body movements during the fish's splash. Our observations reveal that mudskippers extend and maintain their pectoral fins during the impact and then flap their caudal fins to hop when the water splash reaches its maximum radius. The cost of transport (CoT) for water-hopping is found to be smaller and more efficient compared to swimming and crawling. To simulate this behavior, we set up an impact experiment in a water tank with a pneumatic actuator. Two acrylic rods, one with artificial pectoral fins and one without, are thrown into the water surface at the same impact angle and angle of attack as the mudskipper. Our findings demonstrate that the unique anatomical feature of the pectoral fin, which can be bent downward to maximize form drag during water impact, prevents mudskippers from full submersion and allows them to gain momentum from the rebound flow generated by their impact.   

Preventing ankle twists: the role of the human foot arch in enhancing lateral dynamic stability

Maintaining dynamic stability poses a significant challenge during the evolution from quadrupedal to bipedal locomotion in humans. Unlike our ancestors and primate relatives with flat feet, humans have evolved a unique foot arch structure which is known for providing foot stiffness, thereby facilitating activities like running and jumping. Nevertheless, its impact on dynamic stability, particularly lateral stability and ankle twists, remains uncertain. We show that modern humans possess an optimized foot arch height, which provides enhanced lateral dynamic stability when compared to individuals with either flat or high-arch feet. To quantify this lateral stability, we introduce a novel dimensionless stability indicator for single-legged landings, utilizing a nonlinear multi-Degree-of-Freedom mathematical model and an energy-based approach. The typical human foot arch adapts optimally to combing factors such as landing speed, inclination angle, and external disturbances. By contrast, high-arch foot is inept in handling inclinations; whereas flat foot resists inclination but lacks stability at high landing speeds, making it more suitable for foraging primates rather than human hunters. The generalizable nonlinear mathematical model and precise stability indicator for the lower limb have the potential to assist in identifying individuals at a high risk of ankle injuries, informing the design of footwear and robotic feet, and inspiring further research into postural stability.   

The effects of surface feature geometry on the propulsive locomotion of tree-climbing snakes

Being limbless, snakes face unique challenges when climbing trees, sometimes resorting to  wrapping their bodies around the trunk to pull themselves up. However, corn snakes exhibit an alternative climbing technique that allows them to zig-zag up and down trees without wrapping. We model a large tree using a flat, vertical wall that utilizes a single vertical column of force sensors to record horizontal and vertical propulsive-force measurements as the snakes ascend or descend. This study investigates the force output over the body of the snakes through the combination of 3D-kinematic tracking data as well as time-resolved force data. Our findings reveal that the geometry and length of the pegs alters the snakes’ ability to climb. When the pegs were either shorter or narrower on the tip, snakes were less successful, meaning they were unable to complete the climb. In these scenarios, we saw reduced vertical forces applied to the pegs, indicating difficulty gripping the features, resulting in a higher chance of falling off the wall. In addition, we saw less horizontal force, crucial for stabilization, applied by the middle of the body. Conversely, snakes were more successful when the ends of the pegs were wider, as they could be used to generate greater stabilizing and supporting forces. In successful downward climbs, snakes anchored themselves by exerting more vertical forces with the tail. Future work will investigate how different snake species manage similar scenarios with different surface geometry.    

Climbing without feet: Effect of surface-feature size on the physics of vertical snake locomotion

Arboreal environments present numerous physical constraints that are particularly challenging for animals like snakes which lack appendages that many animals use for gripping, propulsion, and stabilization. Tree trunks and branches of varying diameter, flexibility, length, and surface roughness provide these animals with varying “footholds” which they can use to support their weight. Here we created a simplified model of a large-diameter tree consisting of a flat, smooth, vertical surface with force-sensitive pegs acting as surface features. Using 3D kinematics acquired from IR-marker tracking combined with time-resolved force data, we analyze how the distribution of forces over the snakes’ bodies changes with varying peg length. Ordinarily, in corn snakes, strong vertical forces over the forward and midsections of the body support much of the snakes’ weight while strong lateral forces are applied over the body for stability during ascents. With shorter surface features, these lateral forces are reduced, and the snakes rely on comparatively greater vertical components of force over the forward and midsection. In descents, the rear third of the body anchors the animal by applying strong vertical forces, however, these forces are diminished as feature length decreases, inhibiting the snakes’ ability to move. Further work will investigate how species specialization and surface textures on the snake affect physical interactions and overall success of locomotion.   

Unifying framework of vortex interactions from air to water and in between

We explore how organisms from jellyfish to fruit flies harness vortex interactions to boost locomotion. While previous studies focus on bulk fluids, we examine vortex re-energization at fluid interfaces, using water-walking insects as a case study. These insects employ an alternating tripod gait, where hind legs re-energize vortices shed by middle legs, enhancing efficiency. We show how leg positioning either re-energizes or dissipates these vortices. Using dimensionless analysis, we compare this phenomenon across organisms in air, water, and at fluid interfaces, offering a universal framework for understanding energy recapture from vortices.