Browsing by Author "Hines, Jeff"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    A comparison between automated and manual detection methods for detecting bird species richness in audio recordings
    (University of New Brunswick, 2018) Hines, Jeff; Houlahan, Jeff
    Understanding long-term human impacts over large areas of the environment requires large-scale biological monitoring programs and efficient ways to analyze the volume of data that will be generated. Automatic recording devices (ARDs) can be deployed in the field to record sounds over long periods and are useful for monitoring vocal animal populations. The data can be analyzed by searching through portions of the audio species vocalizations (manual detection) or having software perform the search (automatic detection). Here we compare manual and automated techniques for detecting the presence of 43 bird species over 7 sites based on the vocalizations each method detected in field recordings. Results show that an automated method detected significantly more species than the manual method. Automated detection is, therefore, a viable option to efficiently analyze large audio datasets to determine bird species richness.
  • Thumbnail Image
    Item
    Using commercial machine-learning software to conduct bird species inventories
    (University of New Brunswick, 2020) Hines, Jeff; Houlahan, Jeff
    Bird species inventories are an effective way to measure changes in an ecosystem. They can be conducted using automated sound recorders. Birds vocalizing in recordings can be found using manual detection (an observer finds and identifies the sounds), or automatic detection (machine-learning software finds and identifies the sounds, and an observer verifies). Here, I present a method of training software to identify regional birds and a method to efficiently verify its identifications. I then compared the number of species detected by manual and automatic detection using ~625 hours of field recordings/site over 29 sites. Automatic detection found ~45% more species/site (average: 28 vs. 19 species/site, P<0.01), but each method detected species that the other didn't. Automatic detection finds more species when effort constraints limit manual detection to a small portion of the audio, but for the most complete species list I recommend using a combination of manual and automatic detection.