Neural Correlates of Flight Acceleration in Pigeons: Gamma-Band Activity and Local Functional Network Dynamics in the AId Region

Suchen Li, Zhuo Tang, Mengmeng Li, Lifang Yang, Zhigang ShangAnimals 2025 · DOI

TL;DR

This study investigates how pigeon brains encode flight acceleration, a key aspect of dynamic aerial navigation. By analyzing neural activity in the dorsal intermediate arcopallium (AId) during free flight, researchers found that gamma-band brain activity and local functional network dynamics are closely tied to acceleration levels, suggesting the AId plays a crucial role in adapting to changing flight demands.

What Problem Does This Paper Address?

Understanding how animals control complex movements, especially in dynamic environments like flight, is a fundamental challenge in neuroscience. While previous research has explored neural correlates of flight velocity and spatial navigation, the specific neural mechanisms underlying flight acceleration—the rate of change in velocity—remain largely unknown. Acceleration is critical for modulating flight intensity, speed, and direction, especially during maneuvers like takeoff, landing, or obstacle avoidance. This paper aims to bridge this gap by investigating how the avian brain, specifically the AId region (considered analogous to the mammalian motor cortex), represents and processes flight acceleration during natural, three-dimensional flight.

Methods

The researchers employed a multi-modal approach to capture both behavioral and neural data during outdoor free-flight experiments with homing pigeons.

  • Data Acquisition: Pigeons were equipped with a lightweight, wearable system that simultaneously recorded:
  • GPS data: For tracking flight trajectory and speed.
  • Inertial Measurement Unit (IMU): To capture tri-axial acceleration and angular velocity.
  • Local Field Potentials (LFPs): From eight channels implanted in the dorsal intermediate arcopallium (AId) region of the brain, recorded at 1000 Hz.
  • Flight State Classification: Based on the IMU and GPS data, flight was categorized into three states:
  • Acceleration: Speed > 2 m/s and acceleration > 0.5 m/s².
  • Steady Flight: Speed > 2 m/s and acceleration within ±0.5 m/s².
  • Deceleration: Speed > 2 m/s and acceleration < -0.5 m/s².
  • Signal Preprocessing: A key challenge was removing wingbeat artifacts from the neural signals. This was achieved using Variational Mode Decomposition (VMD) and Dynamic Time Warping (DTW) to identify and attenuate noise-contaminated components.
  • Neural Analysis:
    • Time-Frequency Analysis: Power Spectral Density (PSD) proportions and Continuous Wavelet Transform (CWT) were used to analyze neural activity across different frequency bands (Delta, Theta, Alpha, Beta, Gamma) and their temporal dynamics during various flight states.
  • Functional Network Analysis: Inter-channel coherence was calculated to construct dynamic functional brain networks. Topological properties, specifically the clustering coefficient, were analyzed to quantify network connectivity.
  • Machine Learning Decoding: Five distinct frequency bands of LFP signals were used as input for machine learning models (SVM, DNN, CNN) to decode flight acceleration states, validating the role of specific frequency bands.

Datasets

The study utilized data collected from six healthy homing pigeons during outdoor free-flight experiments. The dataset comprises synchronized GPS, IMU, and multi-channel LFP recordings. No pre-existing public datasets were used for training or evaluation; the data collected in this study represents a novel benchmark for analyzing neural correlates of flight acceleration in birds.

Key Insights and Ideas

  • Gamma Band as a Key Indicator: The study strongly suggests that gamma-band activity (30-80 Hz) in the AId region is a primary neural signature for encoding flight acceleration. This band showed significant differences across acceleration states and was the most effective for decoding flight states.
  • Dynamic Network Reorganization: The functional connectivity within the AId region isn't static. It dynamically reconfigures based on flight acceleration, with increased connectivity during demanding phases (deceleration and acceleration) and reduced connectivity during stable flight.
  • "U-Shaped" Relationship for Network Dynamics: The clustering coefficient of the gamma-band network exhibited a non-linear, U-shaped relationship with flight acceleration. This indicates that the brain's local network integration is highest during deceleration, moderate during acceleration, and lowest during steady flight, reflecting varying cognitive and sensorimotor demands.
  • Deceleration as a High-Demand State: Deceleration, often associated with critical maneuvers like landing or obstacle avoidance, showed the most prominent gamma-band activity and the strongest functional network connectivity. This highlights the increased neural processing required for precise control and adaptation during these transitions.
  • Implications for Bio-Inspired Robotics: The findings offer valuable insights for developing more adaptive and intelligent flight control systems for robots, by mimicking how biological systems dynamically adjust to changing motor demands.

Key Results

  • PSD Proportion: The proportion of power in the gamma band significantly decreased as flight acceleration increased from deceleration to acceleration.
  • Time-Frequency Analysis: Gamma-band energy was markedly higher and sustained during deceleration compared to steady or accelerating flight.
  • Functional Network Connectivity: The AId region exhibited the strongest functional connectivity during deceleration, followed by acceleration, and the weakest during steady flight.
  • Clustering Coefficient: A U-shaped relationship was observed between the clustering coefficient and flight acceleration, with the highest values during deceleration and the lowest during steady flight.
  • Machine Learning Decoding: CNN models achieved the highest classification accuracy for decoding flight acceleration states, with gamma-band LFP signals yielding the best performance across all models (up to 89.60% accuracy).

Limitations and Future Work

  • Focus on AId: The study exclusively focused on LFP recordings from the AId region. However, flight acceleration is influenced by multisensory integration (visual, vestibular, proprioceptive). Future work should explore how other brain regions and sensory modalities contribute to this process.
  • Wingbeat Artifacts: While VMD was effective, the authors acknowledge that higher-order harmonics and nonlinear cross-frequency interactions might still contaminate higher frequency bands. Further refinement of artifact removal techniques could be beneficial.
  • LFP vs. Spiking Activity: The study relied on LFP signals, which represent population-level activity. Incorporating spike-based analyses could provide a more granular understanding of the precise spiking dynamics of individual neurons involved in motor processing.
  • Visual Input: The influence of dynamic visual cues, such as optic flow, on AId activity during flight acceleration was not fully explored. Future research could investigate these multisensory interactions.

Summary updated

Abstract

Flight behavior in pigeons is governed by intricate neural mechanisms that regulate movement patterns and flight dynamics. Among various kinematic parameters, flight acceleration provides critical information for the brain to modulate movement intensity, speed, and direction. However, the neural representation mechanisms underlying flight acceleration remain insufficiently understood. To address this, we conducted outdoor free-flight experiments in homing pigeons, during which GPS data, flight posture, and eight-channel local field potentials (LFPs) were synchronously recorded. Our analysis revealed that gamma-band activity in the dorsal intermediate arcopallium (AId) region was more prominent during behaviorally demanding phases of flight. In parallel, local functional network analysis showed that the clustering coefficient of gamma-band activity in the AId followed a nonlinear, U-shaped relationship with flight acceleration-exhibiting the strongest and most widespread connectivity during deceleration, moderate connectivity during acceleration, and the weakest network coupling during steady flight. This pattern likely reflects the increased neural demands associated with flight phase transitions, where greater cognitive and sensorimotor integration is required. Furthermore, using LFP signals from five distinct frequency bands as input, machine learning models were developed to decode flight acceleration, further confirming the role of gamma-band dynamics in motor regulation during natural flight. This study provides the first evidence that gamma-band activity in the avian AId region encodes flight acceleration, offering new insights into the neural representation of motor states in natural flight and implications for bio-inspired flight control systems.
BibTeX
@inproceedings{li2025neural,
  author = {Suchen Li and Zhuo Tang and Mengmeng Li and Lifang Yang and Zhigang Shang},
  title = {{Neural Correlates of Flight Acceleration in Pigeons: Gamma-Band Activity and Local Functional Network Dynamics in the AId Region}},
  booktitle = {Animals},
  year = {2025},
  doi = {10.3390/ani15131851},
}
Figure