Embryonic Exposure to VPA Influences Chick Vocalisations: A Computational Study

TL;DR

A computational framework for the automated detection, acoustic feature extraction, and unsupervised learning of chick vocalisations is introduced and systematic differences between groups are revealed.

Abstract

In young animals like poultry chicks (Gallus gallus), vocalisations convey information about affective and behavioural states. Traditional approaches to vocalisation analysis, relying on manual annotation and predefined categories, introduce biases, limit scalability, and fail to capture the full complexity of vocal repertoires. We introduce a computational framework for the automated detection, acoustic feature extraction, and unsupervised learning of chick vocalisations. Applying this framework to a dataset of newly hatched chicks, we identified two primary vocal clusters. We then tested our computational framework on an independent dataset of chicks exposed during embryonic development to vehicle or Valproic Acid (VPA), a compound that disrupts neural development and is linked to autistic-like symptoms. Clustering analysis on the experimental dataset confirmed two primary vocal clusters and revealed systematic differences between groups. VPA-exposed chicks showed an altered repertoire, with a relative increase in softer calls. VPA differentially affected call clusters, modulating temporal, frequency, and energy domain features. Overall, VPA-exposed chicks produced vocalisations with shorter duration, reduced pitch variability, and modified energy profiles, with the strongest alterations observed in louder calls. This study provides a computational framework for analysing animal vocalisations, advancing knowledge of early-life communication in typical and atypical vocal development.

Figures (19)

• Figure 1: Computational (top), analytical (middle), and biological outcome (bottom) framework for call-level analysis of chick vocalisations. Horizontal arrows indicate processing steps, and vertical arrows information flow across layers.
• Figure 2: A) Dendrogram of the development dataset showing the clustering structure and optimal cut points, and spectrograms of representative calls extracted from cluster 0 and cluster 1. Within the main clusters, we observed further branching; B) UMAP projection divided into $K = 2$ clusters using HAC.
• Figure 3: A) Dendrogram for VPA-treated chicks showing the clustering structure and optimal cut points, and spectrograms of representative calls from cluster 0 and cluster 1. Within the main clusters, we observed further branching; B) UMAP projection for VPA-treated chicks with $K = 2$ clusters.
• Figure 4: A) Dendrogram for control chicks showing the clustering structure and optimal cut points, and spectrograms of representative calls from cluster 0 and cluster 1. Within the main clusters, we observed further branching; B) UMAP projection for control chicks with $K = 2$ clusters.
• Figure 5: Mean number of calls across 6 time bins by condition and cluster membership. Mean ± SEM.