Introduce participants to passive acoustic monitoring (PAM) and its conservation use cases Discuss the tradeoffs to consider when developing a PAM project and its sampling design Present options for PAM data collection and analysis, with accompanying hands-on demos The rapid decline of fauna worldwide requires creative solutions to improve biodiversity monitoring. Passive acoustic monitoring (PAM) has greatly improved our ability to survey biodiversity efficiently across time and space. PAM utilizes automated recording devices that can remotely record species-specific sounds at large spatial and temporal scales. These sound detections can then be used to estimate species presence, distribution, and occupancy. PAM can also be used to detect anthropogenic noise and survey for sounds indicative of illegal activity (gunshots, chainsaws). PAM is a powerful tool because it is non-invasive, cost- and labor-efficient & able to generate long-term data at scale that can be repeatedly analyzed. Hardware/Data Collection Acoustic monitoring has been revolutionized with the recent advent of new technologies such as miniaturized and affordable hardware (e.g., Audiomoths) which significantly lower the barrier to entry. During this workshop, we will discuss the different considerations you should take into account when deciding what technology, sampling design, and analysis to use. For the hands-on part of this section, we will walk participants through the configuration and deployment of Audiomoths (devices provided by presenters for the purposes of this workshop).

Software/Data Analysis The recent development of user-friendly software that incorporates machine learning and cloud computing has become a vital step in the PAM pipeline. During this workshop, we will guide participants through accessible and intuitive workflows for analyzing data from autonomous recorders using Arbimon. This free, web-based platform offers a comprehensive set of eco-acoustic analytical tools through an intuitive no-code interface. We will walk participants through a variety of workflows using a sample dataset (or participants can use their own data). These workflows will cover how to annotate spectrograms, automate species detection and identification, and holistically analyze soundscapes. We will also explain how sound detections can be used in post-hoc ecological analyses (e.g., occupancy models, SDMs).

Approximate Schedule:

Introductions (15-minutes) Presentation: Intro to PAM (20-min.)

Demo/hands-on practice: hardware (40-min) Audiomoth configuration, deployment Q&A/break (15-min) Presentation: Analyzing PAM data (15-min.)

Demo’s/hands-on practice: software Arbimon platform (10-min) Pattern matching (35-min.)

Event detection & clustering analysis (40-min.)

Convolutional neural networks (CNN; 30-min.)

Insights (20-min.)