The identification and classification of bird species using audio recordings have gained prominence in ecological research, biodiversity conservation, and environmental monitoring. This survey provides a comprehensive overview of audio descriptors used in bird species identification, emphasizing the role and effectiveness of various descriptors in capturing distinctive acoustic features of bird vocalizations. Key audio descriptors reviewed include time-domain features like Zero-Crossing Rate (ZCR), frequency-domain descriptors such as Spectral Centroid and Mel-frequency cepstral coefficients (MFCCs), and time-frequency analysis techniques like Short-Time Fourier Transform (STFT) and wavelet transform. Each descriptor’s strengths and limitations are examined, particularly in terms of their ability to characterize the complex sounds within bird vocalizations. Additionally, the integration of these descriptors into machine learning and deep learning frameworks, including k-Nearest Neighbors (KNN), Support Vector Machines (SVM), and neural network models, is reviewed to assess their effectiveness in bird species classification. The paper also discusses methodologies for bird sound detection, segmentation, and feature extraction, providing a holistic view of the audio processing pipeline for bird identification. In examining recent advancements, this survey highlights future directions for enhancing audio descriptor accuracy, improving real-time classification capabilities, and scaling these methods for large datasets. The application of audio descriptor-based identification systems is underscored as a powerful tool in bird conservation, enabling non-invasive, automated monitoring of bird populations across diverse ecosystems. This survey aims to support researchers and practitioners by presenting a detailed, structured review of audio descriptors and offering insights into their application in advancing bird species conservation.