Passive Acoustic Monitoring within the Northwest Forest Plan Area: 2023 Annual Report

Abstract

Here we document progress in implementing large-scale passive acoustic monitoring across the Northwest Forest Plan area to track population trends of northern spotted owls (Strix occidentalis caurina), barred owls (S. varia), marbled murrelets (Brachyramphus marmoratus), and a broad array of forest-adapted wildlife. In 2023, we deployed 4,012 autonomous recording units across 1,009 randomly selected 5-km{superscript 2} hexagons, generating nearly 2.2 million hours of recordings, representing approximately 1 petabyte of acoustic data. These data were processed with PNW-Cnet v5, the latest version of our convolutional neural network model, trained on 135 sound classes representing over 80 species and environmental sounds. Model performance demonstrated high precision for focal species and many additional taxa, substantially reducing manual review effort while enabling broad-scale biodiversity assessments. Results confirmed northern spotted owl detections in all 20% sample areas, with occupancy varying geographically and declining notably in the Tyee study area. Barred owls were widely detected, with the highest prevalence in Oregon and Washington and comparatively lower occupancy in California. Marbled murrelets were consistently detected in coastal areas, particularly the Olympic Peninsula and Oregon Coast Range. Beyond these focal species, PAM and PNW-Cnet generated robust datasets for a wide range of birds, mammals, and disturbance indicators, underscoring the value of random-site, multi-species monitoring. The 2023 field season marked the first full implementation of the 2% + 20% NWFP sampling design, expanding monitoring coverage while strengthening collaborations with federal and state partners. These efforts provide the foundation for long-term, cost-effective wildlife monitoring and inform conservation strategies in dynamic forest ecosystems.